2012 EDUCATIONAL BOOK - American Society of Clinical Oncology
2012 EDUCATIONAL BOOK - American Society of Clinical Oncology
2012 EDUCATIONAL BOOK - American Society of Clinical Oncology
You also want an ePaper? Increase the reach of your titles
YUMPU automatically turns print PDFs into web optimized ePapers that Google loves.
AMERICAN SOCIETY OF CLINICAL ONCOLOGY<br />
<strong>2012</strong> <strong>EDUCATIONAL</strong> <strong>BOOK</strong><br />
48th Annual Meeting | June 1-5, <strong>2012</strong> | Chicago, Illinois
<strong>American</strong> <strong>Society</strong> <strong>of</strong> <strong>Clinical</strong> <strong>Oncology</strong><br />
Educational Book<br />
The <strong>American</strong> <strong>Society</strong> <strong>of</strong> <strong>Clinical</strong> <strong>Oncology</strong> Educational Book (ISSN 1548-8748) is published by the <strong>American</strong> <strong>Society</strong> <strong>of</strong><br />
<strong>Clinical</strong> <strong>Oncology</strong>.<br />
Requests for permission to reprint all or part <strong>of</strong> any article published in this title should be directed to Permissions,<br />
<strong>American</strong> <strong>Society</strong> <strong>of</strong> <strong>Clinical</strong> <strong>Oncology</strong>, 2318 Mill Road, Suite 800, Alexandria, VA 22314. Tel: 571-483-1300; fax:<br />
571-366-9550; or email: permissions@asco.org.<br />
All other questions should be addressed to Managing Editor, <strong>American</strong> <strong>Society</strong> <strong>of</strong> <strong>Clinical</strong> <strong>Oncology</strong>, 2318 Mill Road, Suite<br />
800, Alexandria, VA 22314. Tel: 703-299-1183; email: EdBook@asco.org.<br />
Single issues, both current and back, exist in limited quantities and are <strong>of</strong>fered for sale subject to availability. For further<br />
information, email EdBook@asco.org or call 888-273-3508.<br />
Copyright © <strong>2012</strong> <strong>American</strong> <strong>Society</strong> <strong>of</strong> <strong>Clinical</strong> <strong>Oncology</strong>. All rights reserved. No part <strong>of</strong> this publication may be reproduced<br />
or transmitted in any form or by any means, electronic or mechanical, including photocopy, recording, or any information<br />
storage and retrieval system, without permission in writing from ASCO or the author.<br />
Copies <strong>of</strong> articles in this publication may be made for personal use. This consent is given on the condition, however, that<br />
the copier pay the stated per-copy fee through the Copyright Clearance Center, Inc. (222 Rosewood Drive, Danvers, MA<br />
01923) for copying beyond that permitted by Sections 107 or 108 <strong>of</strong> the U.S. Copyright Law. This consent does not extend to<br />
other kinds <strong>of</strong> copying, such as copying for general distribution, for advertising or promotional purposes, for creating new<br />
collective works, or for resale.<br />
The <strong>American</strong> <strong>Society</strong> <strong>of</strong> <strong>Clinical</strong> <strong>Oncology</strong> assumes no responsibility for errors or omissions in this publication. The reader<br />
is advised to check the appropriate medical literature and the product information currently provided by the manufacturer<br />
<strong>of</strong> each drug to be administered to verify the dosage, the method and duration <strong>of</strong> administration, or contraindications. It is<br />
the responsibility <strong>of</strong> the treating physician or other health care pr<strong>of</strong>essional, relying on the independent experience and<br />
knowledge <strong>of</strong> the patient, to determine drug dosages and the best treatment for the patient.
<strong>American</strong> <strong>Society</strong> <strong>of</strong> <strong>Clinical</strong><br />
<strong>Oncology</strong><br />
Educational Book<br />
Editor: Ramaswamy Govindan, MD<br />
Associate Editor: Natasha Leighl, MD<br />
Managing Editor: Christine Smith<br />
Editorial Coordinator: Lauren Burke<br />
Production Manager: Donna Dottellis<br />
Executive Editor: Lisa Greaves<br />
© <strong>2012</strong> <strong>American</strong> <strong>Society</strong> <strong>of</strong> <strong>Clinical</strong> <strong>Oncology</strong><br />
Alexandria, VA
ASCO gratefully acknowledges Amgen<br />
for their support <strong>of</strong> the 48th Annual Meeting Educational Book.
<strong>American</strong> <strong>Society</strong> <strong>of</strong> <strong>Clinical</strong><br />
<strong>Oncology</strong><br />
Educational Book<br />
TABLE OF CONTENTS<br />
ASCO Annual Meeting Disclosure. ...................................................................... xiii<br />
2011–<strong>2012</strong> Cancer Education Committee ................................................................. xiv<br />
<strong>2012</strong> Annual Meeting Faculty List ....................................................................... xv<br />
<strong>2012</strong> Educational Book Expert Panel .................................................................... xx<br />
<strong>2012</strong> Annual Meeting Supporters ....................................................................... xxii<br />
Letter from the Editor. .................................................................................. 1<br />
The <strong>2012</strong> ASCO Educational Book is published online at asco.org/edbook. Articles can also be accessed<br />
from the Attendee Resource Center at chicago<strong>2012</strong>.asco.org/attendee. Articles that are included in this<br />
print edition include a page number reference in the table <strong>of</strong> contents below.<br />
Breast Cancer<br />
Controversies in the Adjuvant Treatment <strong>of</strong> Breast Cancer<br />
Adjuvant Therapy for Older Women with Early-Stage Breast Cancer: Treatment Selection in a Complex Population<br />
Cynthia Owusu, Arti Hurria, and Hyman Muss. .............................................................. 3<br />
Management <strong>of</strong> Small T1a/b N0 Breast Cancers<br />
Anthony D. Elias ...................................................................................... 10<br />
Controversies in Adjuvant Endocrine Therapy for Breast Cancer<br />
Stephen R. Johnston. ................................................................................... 20<br />
A Dickens Tale <strong>of</strong> the Treatment <strong>of</strong> Advanced Breast Cancer: The Past, the Present, and the Future<br />
A Dickens Tale <strong>of</strong> the Treatment <strong>of</strong> Advanced Breast Cancer: The Past, the Present, and the Future<br />
George W. Sledge Jr., Fatima Cardoso, Eric P. Winer, and Martine J. Piccart. ................................... 28<br />
A Fresh Look at Ductal Carcinoma In Situ: Basic Science to <strong>Clinical</strong> Management<br />
Ductal Carcinoma In Situ: Challenges, Opportunities, and Uncharted Water<br />
Abigail W. H<strong>of</strong>fman, Catherine Ibarra-Drendall, Virginia Espina, Lance Liotta, and Victoria Seewaldt ................ 40<br />
Ductal Carcinoma In Situ, and the Influence <strong>of</strong> the Mode <strong>of</strong> Detection, Population Characteristics, and Other Risk<br />
Factors<br />
Beth A. Virnig, Shi-Yi Wang, and Todd M. Tuttle ............................................................ 45<br />
Key Questions in the Loco-regional Treatment <strong>of</strong> Breast Cancer<br />
Postmastectomy Radiation and Partial Breast Irradiation<br />
Richard C. Zellars ..................................................................................... 50<br />
The Appropriate Extent <strong>of</strong> Surgery for Early-Stage Breast Cancer<br />
Monica Morrow ....................................................................................... 53
How Does Biology Affect Local Therapy Decisions?<br />
Thomas A. Buchholz ................................................................................... 56<br />
The Many Options for Breast Imaging: What to Use When<br />
<strong>Clinical</strong> and Imaging Surveillance Following Breast Cancer Diagnosis<br />
Chris I. Flowers, Blaise P. Mooney, and Jennifer S. Drukteinis ................................................ 59<br />
Advanced Imaging Techniques for the Detection <strong>of</strong> Breast Cancer<br />
Maxine Jochelson. ..................................................................................... 65<br />
Update <strong>of</strong> the Oxford Overview: New Insights and Perspectives in the Era <strong>of</strong> Personalized Medicine<br />
Update <strong>of</strong> the Oxford Overview: New Insight and Perspectives in the Era <strong>of</strong> Personalized Medicine<br />
Kathleen I. Pritchard, Jonas Bergh, and Harold J. Burstein ................................................... 71<br />
Cancer Genetics<br />
Gene Patenting: Effects on Biotechnology and <strong>Oncology</strong><br />
Gene Patents and Personalized Medicine<br />
Roger D. Klein. ....................................................................................... 81<br />
Cancer Prevention/Epidemiology<br />
Breast Cancer Chemoprevention: If Not Now, When?<br />
Chemoprevention for Breast Cancer: Overcoming Barriers to Treatment<br />
Abenaa M. Brewster, Nancy E. Davidson, and Worta McCaskill-Stevens ......................................... 85<br />
Controversies in Prostate Cancer: PSA Screening, Chemoprevention, and Treatment <strong>of</strong> Early Disease<br />
The Case for Prostate Cancer Risk Reduction by 5-Alpha Reductase Inhibitors<br />
Youssef S. Tanagho and Gerald L. Andriole ................................................................ 92<br />
Screening for Prostate Cancer with Prostate-Specific Antigen: What’s the Evidence?<br />
Pamela M. Marcus and Barnett S. Kramer ................................................................. 96<br />
Central Nervous System Tumors<br />
Glioblastoma: Taking the Standard <strong>of</strong> Care to the Cutting Edge<br />
Glioblastoma: Biology, Genetics, and Behavior<br />
Daniel J. Brat ....................................................................................... 102<br />
Future Directions in Glioblastoma Therapy<br />
Howard Colman. ..................................................................................... 108<br />
Establishing the Standard <strong>of</strong> Care for Patients with Newly Diagnosed and Recurrent Glioblastoma<br />
Mark R. Gilbert ...................................................................................... 112<br />
Imaging in Neuro-oncology: Practical Primer for the Practicing Physician<br />
Current Concepts in Brain Tumor Imaging<br />
Andrew D. Norden, Whitney B. Pope, and Susan M. Chang .................................................. 119<br />
Perspectives on Headline-Making News in Neuro-oncology<br />
Noninvasive Application <strong>of</strong> Alternating Electric Fields in Glioblastoma: A Fourth Cancer<br />
Treatment Modality<br />
Philip H. Gutin and Eric T. Wong ....................................................................... 126<br />
<strong>Clinical</strong> Trials<br />
Adaptive <strong>Clinical</strong> Trial Designs: What Are They and Should They Be Used?<br />
Limitations <strong>of</strong> Adaptive <strong>Clinical</strong> Trials<br />
Marc Buyse ......................................................................................... 133<br />
Endpoints for Cancer Trials in <strong>2012</strong>: Statistical, Regulatory, and <strong>Clinical</strong> Perspectives (eQ&A)<br />
Capturing the Patient Perspective: Patient-Reported Outcomes as <strong>Clinical</strong> Trial Endpoints<br />
Deborah Watkins Bruner, Benjamin Movsas, and Ethan Basch ................................................ 139
New Looks and Challenges for Cooperative Groups<br />
The New National Cancer Institute National <strong>Clinical</strong> Trials Network<br />
Stephen S. Grubbs .................................................................................... 146<br />
Successful Integration <strong>of</strong> Cooperative Groups: The Origin <strong>of</strong> the Children’s <strong>Oncology</strong> Group<br />
Gregory H. Reaman. .................................................................................. 149<br />
Overcoming Disparities in <strong>Clinical</strong> Trial Accrual among African <strong>American</strong>s<br />
A Critical Review <strong>of</strong> the Enrollment <strong>of</strong> Black Patients in Cancer <strong>Clinical</strong> Trials<br />
Deliya R. Banda, Diane St. Germain, Worta McCaskill-Stevens, Jean G. Ford, and Sandra M. Swain ................ 153<br />
Developmental Therapeutics<br />
Antibody Conjugates: Targeted Therapy Meets Chemotherapy in One Drug<br />
Trastuzumab Emtansine (T-DM1): Hitching a Ride on a Therapeutic Antibody<br />
Howard A. Burris III .................................................................................. 159<br />
Targeting CD30 in Hodgkin Lymphoma: Antibody-Drug Conjugates Make a Difference<br />
Catherine S. M. Diefenbach and John P. Leonard .......................................................... 162<br />
Early Drug Development: Casting a Wide Net versus Preselecting a Narrow Audience<br />
Approval <strong>of</strong> New Agents after Phase II Trials<br />
Bruce Chabner ................................................................................... eArticle<br />
Drug Development in the Era <strong>of</strong> Personalized <strong>Oncology</strong>: From Population-Based Trials to Enrichment and<br />
Prescreening Strategies<br />
Rodrigo Dienstmann, Jordi Rodon, and Josep Tabernero. .................................................... 168<br />
Immunotherapy with the T-cell Checkpoint Antibodies: Implications for the Practitioner<br />
Practical Management <strong>of</strong> Immune-Related Adverse Events from Immune Checkpoint Protein Antibodies for<br />
the Oncologist<br />
Jeffrey S. Weber. ..................................................................................... 174<br />
Targeting Critical Molecular Aberrations Early in the Course <strong>of</strong> Solid Tumors: Is It about Time?<br />
Treatment <strong>of</strong> Chronic Myelogenous Leukemia as a Paradigm for Solid Tumors: How Targeted Agents in Newly Diagnosed<br />
Disease Transformed Outcomes<br />
Jason R. Westin, Hagop Kantarjian, and Razelle Kurzrock ................................................... 179<br />
Targeting Molecular Aberrations in Breast Cancer: Is It about Time?<br />
Laura Esserman, Christopher Benz, and Angela DeMichele .................................................. 186<br />
Ethics<br />
Ethical Challenges <strong>of</strong> Health Care<br />
Core Elements <strong>of</strong> the Patient Protection and Affordable Care Act and Their Relevance to the Delivery <strong>of</strong><br />
High-Quality Cancer Care<br />
Beverly Moy, Amy P. Abernethy, and Jeffrey M. Peppercorn .............................................. eArticle<br />
International Variation in Understanding Oncologists’ Pr<strong>of</strong>essional Duties toward Patients, Families, and Themselves<br />
A Balanced Approach to Physician Responsibilities: Oncologists’ Duties toward Themselves<br />
Amy P. Abernethy ................................................................................. eArticle<br />
Are Oncologists Accountable Only to Patients or Also to Their Families? An International Perspective<br />
Antonella Surbone and Lea Baider ................................................................... eArticle<br />
The Oncologist’s Duty to Provide Hope: Fact or Fiction?<br />
Simon Wein ...................................................................................... eArticle<br />
Medical Errors in Cancer Care: Prevention, Disclosure, and Patient and Family Member Responses<br />
Oncologists’ Difficulties in Facing and Disclosing Medical Errors: Suggestions for the Clinic<br />
Antonella Surbone. ................................................................................ eArticle<br />
Preventing Errors in <strong>Oncology</strong>: Perspective <strong>of</strong> a Physician Who Is Also a Cancer Patient<br />
Itzhak Brook ..................................................................................... eArticle
Gastrointestinal (Colorectal) Cancer<br />
“Personalized” <strong>Oncology</strong> for Colorectal Cancer: Ready for Prime Time or Stop the Train?<br />
Is There Currently an Established Role for the Use <strong>of</strong> Predictive or Prognostic Molecular Markers in the Management<br />
<strong>of</strong> Colorectal Cancer? A Point/Counterpoint<br />
Alan P. Venook, Johanna C. Bendell, and Robert S. Warren .................................................. 193<br />
Curative-Intent Treatment <strong>of</strong> Colorectal Cancer Metastases<br />
The Interventional Radiologist Role in Treating Liver Metastases for Colorectal Cancer<br />
Stephen B. Solomon and Constantinos T. S<strong>of</strong>ocleous ........................................................ 202<br />
Curative-Intent Treatment for Colorectal Liver Metastases: A Medical Oncologist’s Perspective<br />
Leonard B. Saltz ..................................................................................... 205<br />
Surgical Treatment <strong>of</strong> Colorectal Cancer Liver Metastases<br />
Steven A. Curley ..................................................................................... 209<br />
Rectal Cancer: New Paradigms beyond Standard Chemotherapy and Radiation<br />
Minimally Invasive Surgery <strong>of</strong> Rectal Cancer: Current Evidence and Options<br />
Atthaphorn Trakarnsanga and Martin R. Weiser. ........................................................... 214<br />
Radiation in Rectal Cancer: What Are the Options and If/When Can It Be Avoided?<br />
Karyn A. Goodman ................................................................................... 219<br />
Stage III Colon Cancer: What Works, What Doesn’t and Why, and What’s Next (eQ&A)<br />
Stage III Colon Cancer: What Works, What Doesn’t and Why, and What’s Next<br />
Thierry André, Bert H. O’Neil, and Jeffrey A. Meyerhardt. ................................................... 223<br />
Gastrointestinal (Noncolorectal) Cancer<br />
Evolving Paradigms in the Management <strong>of</strong> Unresectable Pancreatic Cancer<br />
A New Direction for Pancreatic Cancer Treatment: FOLFIRINOX in Context<br />
Hedy Lee Kindler. .................................................................................... 232<br />
A Matter <strong>of</strong> Timing: Is There a Role for Radiation in Locally Advanced Pancreatic Cancer, and If So, When?<br />
Theodore S. Hong, Jennifer Y. Wo, and Eunice L. Kwak ..................................................... 238<br />
A Myriad <strong>of</strong> Symptoms: New Approaches to Optimizing Palliative Care <strong>of</strong> Patients with Advanced Pancreatic Cancer<br />
Lauren A. Wiebe ..................................................................................... 243<br />
Global Perspective <strong>of</strong> Locally Advanced Gastric Cancer: Different Treatment Paradigms and Their Rationale<br />
Varying Lymphadenectomies for Gastric Adenocarcinoma in the East Compared with the West: Effect on Outcomes<br />
Benjamin Schmidt and Sam S. Yoon. ..................................................................... 250<br />
Will Disease Heterogeneity Help Define Treatment Paradigms for Gastroesophageal Adenocarcinoma? A Global<br />
Perspective<br />
Manish A. Shah ...................................................................................... 256<br />
Adjuvant Treatments for Localized Advanced Gastric Cancer: Differences among Geographic Regions<br />
Yoon-Koo Kang and Changhoon Yoo ................................................................. eArticle<br />
Liver-Directed Therapeutic Options for Hepatocellular Carcinoma: Patient Selection and <strong>Clinical</strong> Outcomes<br />
Stereotactic Body Radiation Therapy for Hepatocellular Carcinoma<br />
Laura A. Dawson, Sameh Hashem, and Alexis Bujold ....................................................... 261<br />
Patient Selection, Resection, and Outcomes for Hepatocellular Carcinoma<br />
Claudius Conrad and Kenneth K. Tanabe ................................................................. 265<br />
The Management <strong>of</strong> Less Common but Complex Upper Gastrointestinal Malignancies: Hepatocellular Carcinoma,<br />
Pancreatic Neuroendocrine Tumors, and Biliary Tract Tumors<br />
A Renaissance in Therapeutic Options for Pancreatic Neuroendocrine Tumors<br />
Pamela L. Kunz ...................................................................................... 271<br />
A Review <strong>of</strong> the Management <strong>of</strong> Hepatocellular Carcinoma: Standard Therapy and a Look to New Targets<br />
Andrew X. Zhu ....................................................................................... 275<br />
Advanced Biliary Tract Cancers<br />
Laura Williams G<strong>of</strong>f and Jordan D. Berlin ................................................................ 281
Genitourinary Cancer<br />
Best Use <strong>of</strong> Imaging Techniques, Old and New, for Genitourinary Cancers in <strong>Clinical</strong> Practice<br />
Optimal Use <strong>of</strong> Imaging to Guide Treatment Decisions for Kidney Cancer<br />
Walter M. Stadler .................................................................................... 284<br />
Castration-Resistant Prostate Cancer: New Insights into Biology and Treatment (eQ&A)<br />
State-<strong>of</strong>-the-Art Management for the Patient with Castration-Resistant Prostate Cancer in <strong>2012</strong><br />
Oliver Sartor ........................................................................................ 289<br />
Prognostic, Predictive, and Surrogate Factors for Individualizing Treatment for Men with Castration-Resistant Prostate<br />
Cancer<br />
Rhonda L. Bitting and Andrew J. Armstrong ............................................................... 292<br />
Kidney Cancer Biology and Therapeutics: VEGFR, mTOR, and Beyond<br />
The Evolving Landscape <strong>of</strong> Metastatic Renal Cell Carcinoma<br />
Daniel Y. C. Heng and Toni K. Choueiri. ................................................................. 299<br />
Recent Innovations in the Management <strong>of</strong> Genitourinary Cancers (eQ&A)<br />
New Developments in Urothelial Cancer<br />
Matthew D. Galsky ................................................................................... 304<br />
New Developments in Prostate Cancer Therapy<br />
Madhuri Bajaj and Elisabeth I. Heath .................................................................... 309<br />
Geriatric <strong>Oncology</strong><br />
Adjuvant Therapy for Older Patients<br />
Adjuvant Treatment <strong>of</strong> Older Patients with Lung Cancer<br />
Jeffrey Crawford ..................................................................................... 315<br />
Treatment <strong>of</strong> Older Patients with Advanced Cancer: Balancing Efficacy with Toxicity (eQ&A)<br />
Considerations and Controversies in the Management <strong>of</strong> Older Patients with Advanced Cancer<br />
Supriya Gupta Mohile, Heidi D. Klepin, and Arati V. Rao. ................................................... 321<br />
Gynecologic Cancer<br />
Recent <strong>Clinical</strong> Highlights in Gynecologic <strong>Oncology</strong><br />
International Perspective on the Global Advances in Gynecologic <strong>Oncology</strong><br />
Lynette Denny ....................................................................................... 330<br />
The European <strong>Society</strong> <strong>of</strong> Gynaecological <strong>Oncology</strong>: Update on Objectives and Educational and Research Activities<br />
Renata Brantnerova, Ranjit Manchanda, and Nicoletta Colombo .............................................. 335<br />
Upfront Treatment <strong>of</strong> Ovarian Cancer: International Consensus and Variation<br />
Biologicals in the Upfront Treatment <strong>of</strong> Ovarian Cancer: Focus on Bevacizumab and Poly (ADP-Ribose) Polymerase<br />
Inhibitors<br />
Rene Roux, Martin Zweifel, and Gordon J. S. Rustin ........................................................ 340<br />
Intraperitoneal Treatment in Ovarian Cancer: The Gynecologic <strong>Oncology</strong> Group Perspective in <strong>2012</strong><br />
Deborah K. Armstrong, Keiichi Fujiwara, and Danijela Jelovac. .............................................. 345<br />
Dose-Dense Chemotherapy and Neoadjuvant Chemotherapy for Ovarian Cancer<br />
Keiichi Fujiwara, Noriyuki Katsumata, and Takashi Onda. ................................................... 349<br />
Uterine Sarcoma: Challenging Cases for the Interdisciplinary Team<br />
Uterine Sarcomas: Histology and Its Implications on Therapy<br />
Martee L. Hensley .................................................................................... 356<br />
Surgical Options for Recurrent Uterine Sarcomas<br />
Sharmilee B. Korets and John P. Curtin .................................................................. 362
Head and Neck Cancer<br />
Patients with HPV-Positive Oropharynx Tumors: Communicating Diagnosis to Family and Treatment Options<br />
Management <strong>of</strong> Human Papillomavirus-Induced Oropharynx Cancer<br />
David Brizel ......................................................................................... 368<br />
Translating Biologic Discoveries into <strong>Clinical</strong> Trials<br />
Important Early Advances in Squamous Cell Carcinoma <strong>of</strong> the Head and Neck<br />
Eric Bissada, Irene Brana, and Lillian L. Siu .............................................................. 373<br />
Application <strong>of</strong> Genomic and Proteomic Technologies in Biomarker Discovery<br />
Elana J. Fertig, Robbert Slebos, and Christine H. Chung .................................................... 377<br />
Treatment <strong>of</strong> Thyroid Cancers: New Information for Medical Oncologists<br />
Evaluation <strong>of</strong> Patients with Disseminated or Locoregionally Advanced Thyroid Cancer: A Primer for<br />
Medical Oncologists<br />
A. Dimitrios Colevas and Manisha H. Shah ............................................................... 384<br />
Systemic Therapeutic Approaches to Advanced Thyroid Cancers<br />
Michael E. Menefee, Robert C. Smallridge, and Keith C. Bible, and the Mayo Clinic Endocrine Malignancies<br />
Disease-Oriented Group ............................................................................... 389<br />
Health Services Research<br />
Avoiding Overdiagnosis and Overtreatment <strong>of</strong> Common Cancers<br />
Overdiagnosis and Overtreatment <strong>of</strong> Prostate Cancer<br />
Ian M. Thompson Jr. .............................................................................. eArticle<br />
Overdiagnosis and Overtreatment <strong>of</strong> Breast Cancer<br />
Michael Alvarado, Elissa Ozanne, and Laura Esserman .................................................. eArticle<br />
Costs <strong>of</strong> Cancer Care: Affordability, Access, and Policy<br />
Reducing the Cost <strong>of</strong> Cancer Care: How to Bend the Curve Downward<br />
Thomas J. Smith, Bruce E. Hillner, and Ronan J. Kelly .................................................. eArticle<br />
New Diagnostics and Devices in the Era <strong>of</strong> Comparative Effectiveness<br />
Why Hasn’t Genomic Testing Changed the Landscape in <strong>Clinical</strong> <strong>Oncology</strong>?<br />
Daniel F. Hayes, Muin J. Khoury, and David Ransoh<strong>of</strong>f. ................................................. eArticle<br />
Leukemia, Myelodysplasia, and Transplantation<br />
Management <strong>of</strong> Chronic Lymphocytic Leukemia<br />
Predicting <strong>Clinical</strong> Outcome in B-Chronic Lymphocytic Leukemia<br />
Neil E. Kay. ......................................................................................... 394<br />
Transplant in Chronic Lymphocytic Leukemia: To Do It or Not and If So, When and How?<br />
John G. Gribben ..................................................................................... 399<br />
New Developments in Myeloproliferative Neoplasms<br />
Therapeutic Advances in Myeloproliferative Neoplasms: The Role <strong>of</strong> New-Small Molecule Inhibitors<br />
Srdan Verstovsek ..................................................................................... 406<br />
Insights into the Molecular Genetics <strong>of</strong> Myeloproliferative Neoplasms<br />
Huong (Marie) Nguyen and Jason Gotlib ................................................................. 411<br />
Classification <strong>of</strong> Myeloproliferative Neoplasms and Prognostic Factors<br />
Francesco Passamonti ................................................................................. 419<br />
Lung Cancer<br />
Around the World in Almost 80 Minutes: Lung Cancer Care and Research<br />
Lung Cancer in Brazil<br />
Gilberto Schwartsmann ................................................................................ 426<br />
Research and Standard Care: Lung Cancer in China<br />
Tony S. Mok, Qing Zhou, and Yi-Long Wu ................................................................ 432
Research and Standard <strong>of</strong> Care: Lung Cancer in Romania<br />
Tudor E. Ciuleanu .................................................................................... 437<br />
Leveraging Virtual Patient Communities for Optimal <strong>Clinical</strong> Care and Research<br />
A New Model: Physician-Patient Collaboration in Online Communities and the <strong>Clinical</strong> Practice <strong>of</strong> <strong>Oncology</strong><br />
Howard J. West, Dave deBronkart, and George D. Demetri .................................................. 443<br />
Lung Cancer Screening 101<br />
Lung Cancer Screening: Promise and Pitfalls<br />
Christine D. Berg, Denise R. Aberle, and Douglas E. Wood .................................................. 450<br />
Personalized Medicine in Lung Cancer in <strong>2012</strong><br />
Molecular Testing <strong>of</strong> Non–Small Cell Lung Carcinoma Biopsy and Cytology Specimens<br />
Ignacio I. Wistuba .................................................................................... 459<br />
Thymoma and Thymic Carcinoma: Update on Management<br />
Multidisciplinary Management <strong>of</strong> Thymic Carcinoma<br />
Gregory J. Riely, James Huang, and Andreas Rimner ....................................................... 466<br />
The Creation <strong>of</strong> the International Thymic Malignancies Interest Group as a Model for Rare Diseases<br />
Frank C. Detterbeck .................................................................................. 471<br />
Thymoma: From Chemotherapy to Targeted Therapy<br />
Nicolas Girard ....................................................................................... 475<br />
Lymphoma and Plasma Cell Disorders<br />
Controversies in Indolent Lymphoma<br />
What Is the Best Strategy for Incorporating New Agents into the Current Treatment <strong>of</strong> Follicular Lymphoma?<br />
Sonali M. Smith ...................................................................................... 481<br />
What Is the Best First-Line Treatment Strategy for Patients with Indolent Lymphomas?<br />
Gilles Salles, Hervé Ghesquières, and Emmanuel Bachy ..................................................... 488<br />
What Is the Role <strong>of</strong> Transplantation for Indolent Lymphoma?<br />
Ryan D. Cassaday and Ajay K. Gopal .................................................................... 494<br />
Controversies in Myeloma: Induction, Transplant, and Maintenance<br />
Stem Cell Transplantation for Multiple Myeloma: Who, When, and What Type?<br />
Amrita Krishnan. ..................................................................................... 502<br />
Upfront Therapy for Myeloma: Tailoring Therapy across the Disease Spectrum<br />
S. Vincent Rajkumar .................................................................................. 508<br />
Maintenance Therapy for Myeloma: How Much, How Long, and at What Cost?<br />
Michel Attal and Murielle Roussel ....................................................................... 515<br />
Melanoma/Skin Cancers<br />
New Options, New Questions: How to Select and Sequence Therapies for Metastatic Melanoma<br />
New Options and New Questions: How to Select and Sequence Therapies for Patients with Metastatic Melanoma<br />
Keith T. Flaherty, Jeff A. Sosman, and Michael B. Atkins .................................................... 524<br />
Patient and Survivor Care<br />
Antiemetics: Current Standards, Emerging Approaches, and Persistent Gaps<br />
Antiemetic Use in <strong>Oncology</strong>: Updated Guideline Recommendations from ASCO<br />
Ethan Basch, Ann Alexis Prestrud, Paul J. Hesketh, Mark G. Kris, Mark R. Somerfield, and Gary H. Lyman .......... 532<br />
Chemotherapy-Induced Nausea and Vomiting Incidence and Prevalence<br />
Steven M. Grunberg. .................................................................................. 541<br />
Mucosal Injury in Patients with Cancer: Targeting the Biology<br />
New Frontiers in Mucositis<br />
Douglas E. Peterson, Dorothy M. Keefe, and Stephen T. Sonis ................................................ 545
Short- and Long-term Cardiovascular Complications <strong>of</strong> Cancer Treatment: Overview for the Practicing Oncologist<br />
Short- and Long-term Cardiovascular Complications <strong>of</strong> Cancer Treatment: Overview for the Practicing Oncologist<br />
Chetan Shenoy and Gretchen Kimmick ................................................................... 553<br />
Recent Advances in Cardiotoxicity <strong>of</strong> Anticancer Therapies<br />
Marianne Ryberg ..................................................................................... 555<br />
When Cancer Becomes Personal: The Physician’s View <strong>of</strong> Patient Care<br />
The Oncologist as the Patient with Cancer or Relative<br />
Teresa Gilewski, Martin Raber, and George W. Sledge Jr. ................................................... 561<br />
Pediatric <strong>Oncology</strong><br />
Advances in Infection Management in Pediatric <strong>Oncology</strong><br />
Prolonged Febrile Neutropenia in the Pediatric Patient with Cancer<br />
Andrew Y. Koh. ...................................................................................... 565<br />
Initial Management <strong>of</strong> Low-Risk Pediatric Fever and Neutropenia: Efficacy and Safety, Costs, Quality-<strong>of</strong>-Life<br />
Considerations, and Preferences<br />
Lillian Sung ......................................................................................... 570<br />
Genetic Counseling <strong>of</strong> the Patient with Pediatric Cancer<br />
Identification, Management, and Evaluation <strong>of</strong> Children with Cancer-Predisposition Syndromes<br />
Sara Knapke, Kristin Zelley, Kim E. Nichols, Wendy Kohlmann, and Joshua D. Schiffman ......................... 576<br />
How to Manage Very Rare Pediatric Cancers<br />
Treating Rare Cancer in Children: The Importance <strong>of</strong> Evidence<br />
Alberto S. Pappo ..................................................................................... 586<br />
Genetic Alterations in Childhood Melanoma<br />
Fariba Navid ........................................................................................ 589<br />
Desmoid-Type Fibromatosis in Children: A Step Forward in the Cooperative Group Setting<br />
Natalie Pounds and Stephen X. Skapek ................................................................... 593<br />
Molecular Pathways for the Practicing Pediatric Oncologist<br />
Targeting the Insulin-Like Growth Factor (IGF) System Is Not as Simple as Just Targeting the Type 1 IGF Receptor<br />
Katia Scotlandi and Antonino Belfiore .................................................................... 599<br />
Hedgehog Pathway in Pediatric Cancers: They’re Not Just for Brain Tumors Anymore<br />
Tobey J. MacDonald .................................................................................. 605<br />
Pediatric <strong>Oncology</strong> Educational Session in Honor <strong>of</strong> Dr. James Nachman<br />
Development and Refinement <strong>of</strong> Augmented Treatment Regimens for Pediatric High-Risk Acute Lymphoblastic Leukemia<br />
Stephen P. Hunger. ................................................................................... 611<br />
Refining the Role <strong>of</strong> Radiation Therapy in Pediatric Hodgkin Lymphoma<br />
Melissa M. Hudson and Louis S. Constine. ................................................................ 616<br />
Role <strong>of</strong> Doxorubicin in Rhabdomyosarcoma: Is the Answer Knowable?<br />
Carola A. S. Arndt .................................................................................... 621<br />
Pontine Gliomas in Children: To Biopsy or Not to Biopsy<br />
Identification <strong>of</strong> Novel Biologic Targets in the Treatment <strong>of</strong> Newly Diagnosed Diffuse Intrinsic Pontine Glioma<br />
Nathan J. Robison and Mark W. Kieran .................................................................. 625<br />
Is Biopsy Safe in Children with Newly Diagnosed Diffuse Intrinsic Pontine Glioma?<br />
Stephanie Puget, Thomas Blauwblomme, and Jacques Grill. .................................................. 629<br />
Ethics <strong>of</strong> Biopsy in Children with Newly Diagnosed Diffuse Intrinsic Pontine Glioma<br />
Nicholas K. Foreman. ................................................................................. 634<br />
Transition and Decision Making for Palliative Care <strong>of</strong> Patients with Pediatric Cancer<br />
Communication and Decision Support for Children with Advanced Cancer and Their Families<br />
Jennifer W. Mack, Chris Feudtner, and Pamela S. Hinds. .................................................... 637
Practice Management and Information Technology<br />
Addressing the Imbalance <strong>of</strong> Supply and Demand: Integrating Advanced Practice Providers into Survivorship Care<br />
Planning for the Future: The Role <strong>of</strong> Nurse Practitioners and Physician Assistants in Survivorship Care<br />
Mary S. McCabe and Todd Alan Pickard .............................................................. eArticle<br />
Doing It Right, and for Less: Implementing Practice Changes to Manage the Growing Complexities, Inefficiencies,<br />
and Costs <strong>of</strong> Cancer Care<br />
Driving Evidence-Based Standardization <strong>of</strong> Care within a Framework <strong>of</strong> Personalized Medicine<br />
Adam Brufsky, Kathleen Lokay, and Melissa McDonald .................................................. eArticle<br />
Patient Portals in <strong>Oncology</strong><br />
The Future <strong>of</strong> <strong>Oncology</strong> Care with Personal Health Records<br />
Henry Feldman and Elizabeth S. Rodriguez ............................................................ eArticle<br />
The ASCO Quality <strong>Oncology</strong> Practice Initiative (QOPI) and Beyond<br />
Measuring and Improving Value <strong>of</strong> Care in <strong>Oncology</strong> Practices: ASCO Programs from Quality <strong>Oncology</strong> Practice<br />
Initiative to the Rapid Learning System<br />
Joseph O. Jacobson, Michael N. Neuss, and Robert Hauser ............................................... eArticle<br />
Pr<strong>of</strong>essional Development<br />
Physician Wellness: Coping with Repetitive Loss and Work-Related Stress<br />
Buoyancy: A Model for Self-Reflection about Stress and Burnout in <strong>Oncology</strong> Providers<br />
Michael J. Fisch .................................................................................. eArticle<br />
Encountering Grief in Patient Care<br />
Teresa Gilewski. .................................................................................. eArticle<br />
Talking in Sync with Patients across Cultural Barriers<br />
New Insights in Cross-Cultural Communication<br />
Lidia Schapira. ................................................................................... eArticle<br />
The Oncologist, the Patient, and the Media<br />
Patients with Cancer, Internet Information, and the <strong>Clinical</strong> Encounter: A Taxonomy <strong>of</strong> Patient Users<br />
Paul R. Helft ..................................................................................... eArticle<br />
Sarcoma<br />
Global Variations in Access to New Therapies for Sarcomas and Gastrointestinal Stromal Tumor<br />
How to Decide Whether to Offer and Use “Nonstandard” Therapies in Patients with Advanced Sarcomas and<br />
Gastrointestinal Stromal Tumors: Global Variations in <strong>Clinical</strong> Practice, Assessment, and Access to Therapies in<br />
Diseases with Limited Incidence and Data<br />
Stefan Sleijfer, Ian Judson, and George D. Demetri ......................................................... 645<br />
Targeted Therapies in Targeted or Untargeted Sarcomas<br />
Targeted Therapy in Sarcoma: Should We Be Lumpers or Splitters?<br />
Richard F. Riedel, Robert G. Maki, and Andrew J. Wagner .................................................. 652<br />
What the Busy Oncologist Needs to Know about Gastrointestinal Stromal Tumor<br />
Adjuvant Treatment <strong>of</strong> Gastrointestinal Stromal Tumor: The Pro<strong>of</strong>, the Pro, and the Practice<br />
Jaap Verweij ........................................................................................ 659<br />
Management <strong>of</strong> Tyrosine Kinase Inhibitor–Resistant Gastrointestinal Stromal Tumors<br />
Christine M. Barnett and Michael C. Heinrich ............................................................. 663<br />
Tumor Biology<br />
Biologic Principles <strong>of</strong> Targeted Combination Therapy<br />
Opportunities and Pitfalls <strong>of</strong> Targeted Therapeutic Combinations in Solid Tumors<br />
Joaquin Mateo, Michael Ong, Timothy A. Yap, and Johann S. de Bono ......................................... 670
Combination Therapies Building on the Efficacy <strong>of</strong> CTLA4 and BRAF Inhibitors for Metastatic Melanoma<br />
Antoni Ribas. ........................................................................................ 675<br />
Mechanisms <strong>of</strong> Resistance to Targeted Anticancer Agents<br />
Mechanisms <strong>of</strong> Resistance to Mitogen-Activated Protein Kinase Pathway Inhibition in BRAF-Mutant Melanoma<br />
Eva M. Goetz and Levi A. Garraway ..................................................................... 680<br />
Mechanisms <strong>of</strong> Resistance to Targeted Therapies in Acute Myeloid Leukemia and Chronic Myeloid Leukemia<br />
Catherine C. Smith and Neil P. Shah ..................................................................... 685<br />
Toward Successful Targeting <strong>of</strong> the PI3 Kinase Pathway in Cancer<br />
Translating PI3K-Delta Inhibitors to the Clinic in Chronic Lymphocytic Leukemia: The Story <strong>of</strong> CAL-101 (GS1101)<br />
John C. Byrd, Jennifer A. Woyach, and Amy J. Johnson ..................................................... 691<br />
PI3 Kinase in Cancer: From Biology to Clinic<br />
Paul Workman and Paul Clarke ..................................................................... eArticle
<strong>2012</strong> ASCO Annual Meeting Disclosure<br />
In compliance with the ASCO Conflict <strong>of</strong> Interest Policy and the Accreditation Council for Continuing Medical<br />
Education (ACCME) Standards for Commercial Support, authors and faculty who participate in any ASCOsponsored<br />
meeting must disclose financial interests and other relationships with entities that have an<br />
investment, licensing, or other commercial interest in the subject matter under consideration in their paper<br />
or presentation. Author and faculty disclosures may be published in ASCO publications where appropriate.<br />
Definition <strong>of</strong> notations: L – Leadership Position; I–Immediate Family Member; B – Myself and immediate<br />
FamilyMember;U–Uncompensated.<br />
Authors and faculty will be asked to disclose financial interests and other relationships in the following<br />
categories:<br />
Employment or Leadership Positions<br />
Any full- or part-time employment or service as an <strong>of</strong>ficer or board member for an entity having an investment,<br />
licensing, or other commercial interest in the subject matter under consideration must be disclosed.<br />
Advisory Role<br />
Consultant or advisory arrangements with an entity having an investment, licensing, or other commercial<br />
interest in the subject matter under consideration must be disclosed if consultation was performed or payments<br />
made for such consultation within 2 years <strong>of</strong> the activity or subject matter in question.<br />
Stock Ownership<br />
Any ownership interest (except when invested in a diversified fund not controlled by the covered individual) in<br />
a start-up company, the stock <strong>of</strong> which is not publicly traded, or in any publicly traded company must be disclosed<br />
if the company is an entity having an investment, licensing, or other commercial interest in the subject matter<br />
under consideration.<br />
Honoraria<br />
Honoraria are reasonable payments for specific speeches, seminar presentations, or appearances. Disclosure <strong>of</strong><br />
honoraria is required when paid directly to the covered individual by an entity having an investment, licensing,<br />
or other commercial interest in the subject matter under consideration and when provided within 2 years <strong>of</strong> the<br />
activity or subject matter in question.<br />
Research Funding<br />
All payments associated with the conduct <strong>of</strong> the clinical research project in question must be disclosed if provided<br />
by the trial sponsor or agents employed by the sponsor.<br />
Expert Testimony<br />
Provision <strong>of</strong> expert testimony must be disclosed when the testimony relates to the subject matter under<br />
consideration.<br />
Other Remuneration<br />
The value <strong>of</strong> trips, travel, gifts, or other in-kind payments not directly related to research activities must be<br />
disclosed if received from an entity having an investment, licensing, or other commercial interest in the subject<br />
matter under consideration and when received within 2 years <strong>of</strong> the activity or subject matter in question. These<br />
payments exclude research-related costs and travel.<br />
Continuing Medical Education Information<br />
This activity has been approved for AMA PRA Category 1 Credit. For information detailing the<br />
Continuing Medical Education components <strong>of</strong> the 48th ASCO Annual Meeting, please visit<br />
chicago<strong>2012</strong>.asco.org.
Harold J. Burstein MD, PhD, Chair<br />
Antoinette R. Tan, MD, Chair-Elect<br />
Charles Blanke, MD, Immediate Past<br />
Chair<br />
Gary I. Cohen, MD, Board Liaison<br />
Breast Cancer<br />
Anna Maria Storniolo, MD (Track<br />
Leader)<br />
Angelo Di Leo, MD, PhD<br />
Stephen R. Grobmyer, MD<br />
Erica L. Mayer, MD, MPH<br />
Karen A. Gelmon, MD<br />
Jonas Bergh, MD, PhD<br />
Barbara A. Parker, MD<br />
Richard C. Zellars, MD<br />
Cancer Genetics<br />
Sancy Ann Leachman, MD (Track<br />
Leader)<br />
James M. Ford, MD<br />
Kenneth Offit, MD, MPH<br />
Michael J. Hall, MD<br />
Zs<strong>of</strong>ia Kinga Stadler, MD<br />
Cancer Prevention/Epidemiology<br />
Pamela Jean Goodwin, MD (Track<br />
Leader)<br />
Christopher S. Lathan, MD, MS, MPH<br />
Abenaa M. Brewster, MD, MHS<br />
James Roger Marshall, PhD<br />
Melanie R. Palomares, MD<br />
Central Nervous System Tumors<br />
Joon H. Uhm, MD (Track Leader)<br />
Jay Steven Loeffler, MD<br />
Timothy Francis Cloughesy, MD<br />
<strong>Clinical</strong> Trials<br />
David M. Dilts, PhD, MBA (Track<br />
Leader)<br />
Donna S. Neuberg, ScD<br />
Tatiana Michelle Prowell, MD<br />
Philip Jordan Gold, MD<br />
Mary W. Redman, PhD<br />
Tim Eisen, PhD<br />
Developmental Therapeutics<br />
Keith T. Flaherty, MD (Track Leader)<br />
Howard A. Burris III, MD<br />
Razelle Kurzrock, MD<br />
Timothy W. Synold, PharmD<br />
Ethics<br />
Antonella Surbone, MD, PhD (Track<br />
Leader)<br />
Amy Pickar Abernethy, MD<br />
Eileen P. Smith, MD<br />
Rudolph M. Navari, MD, PhD<br />
2011–<strong>2012</strong> Cancer Education Committee<br />
Gastrointestinal (Colorectal) Cancer<br />
Jeffrey A. Meyerhardt, MD, MPH<br />
(Track Leader)<br />
Johanna C. Bendell, MD<br />
Alan Paul Venook, MD<br />
David P. Ryan, MD<br />
Gastrointestinal (Noncolorectal)<br />
Cancer<br />
Jordan Berlin, MD (Track Leader)<br />
Hedy Lee Kindler, MD<br />
Matthew Kulke, MD<br />
Pamela L. Kunz, MD<br />
Manish A. Shah, MD<br />
Cliff P. Connery, MD<br />
Genitourinary Cancer<br />
Kim N. Chi, MD (Track Leader)<br />
Toni K. Choueiri, MD<br />
Johann Sebastian De Bono, MD, PhD,<br />
MSc<br />
Matt D. Galsky, MD<br />
Primo Lara Jr., MD<br />
Andrew J. Stephenson, MD<br />
Geriatric <strong>Oncology</strong><br />
Supriya Gupta Mohile, MD, MS (Track<br />
Leader)<br />
Arti Hurria, MD<br />
Arati Rao, MD<br />
Homayoon Sanati, MD<br />
Gynecologic Cancer<br />
Don S. Dizon, MD (Track Leader)<br />
Gini F. Fleming, MD<br />
David E. Cohn, MD<br />
Head and Neck Cancer<br />
David Sidransky, MD (Track Leader)<br />
Ezra E.W. Cohen, MD<br />
Vassiliki Papadimitrakopoulou, MD<br />
Health Services Research<br />
Jennifer J. Griggs, MD, MPH (Track<br />
Leader)<br />
Dawn L. Hershman, MD<br />
Jennifer Malin, MD<br />
Rinaa S. Punglia, MD<br />
Steven J. Katz, MD<br />
Leukemia, Myelodysplasia, and<br />
Transplantation<br />
Martin S. Tallman, MD (Track Leader)<br />
Daniel J. DeAngelo, MD, PhD<br />
Koen Van Besien, MD<br />
Lung Cancer<br />
Renato Martins, MD, MPH (Track<br />
Leader)<br />
David E. Gerber, MD<br />
Pasi A. Janne, MD, PhD<br />
Jyoti D. Patel, MD<br />
Charles Andrew Butts, MD<br />
Natasha B. Leighl, MD<br />
Lymphoma and Plasma Cell<br />
Disorders<br />
Amrita Y. Krishnan, MD (Track Leader)<br />
Ranjana Advani, MD<br />
Nancy Bartlett, MD<br />
Steven M. Horwitz, MD<br />
Jeremy S. Abramson, MD, MSc<br />
Melanoma/Skin Cancers<br />
Leslie Anne Fecher, MD (Track Leader)<br />
F. Stephen Hodi, MD<br />
Vernon K. Sondak, MD<br />
Denise L. Johnson Miller, MD<br />
Hensin Tsao, MD<br />
Patient and Survivor Care<br />
Julia Howe Rowland, PhD (Track<br />
Leader)<br />
Sydney Morss Dy, MD<br />
Teresa Gilewski, MD<br />
Gretchen Genevieve Kimmick, MD<br />
Pediatric <strong>Oncology</strong><br />
Stephen L. Lessnick, MD, PhD (Track<br />
Leader)<br />
Mark W. Kieran MD, PhD<br />
Stephen Skapek, MD<br />
Judith K. Sato, MD<br />
Practice Management and<br />
Information Technology<br />
Craig A. Bunnell, MD (Track Leader)<br />
Robert Stephen Miller, MD<br />
William Charles Penley, MD<br />
Alan M. Keller, MD<br />
David R. Artz, MD<br />
Pr<strong>of</strong>essional Development<br />
Michael Anthony Carducci, MD (Track<br />
Leader)<br />
Emily K. Bergsland, MD<br />
Christine Marie Lovly, MD, PhD<br />
Vikki A. Canfield, MD<br />
Katherine Elizabeth Reeder-Hayes, MD<br />
Robert A. Wolff, MD<br />
Sarcoma<br />
George D. Demetri, MD (Track Leader)<br />
Richard F. Riedel, MD<br />
Warren Allen Chow, MD<br />
Jayesh Desai, MD<br />
R. Lor Randall, MD<br />
Tumor Biology<br />
Levi Garraway, MD, PhD (Track<br />
Leader)<br />
Ross L. Levine, MD<br />
Lee M. Ellis, MD<br />
Josep Tabernero, MD
Matti S. Aapro, MD<br />
Clinique De Genolier<br />
Denise R. Aberle, MD<br />
University <strong>of</strong> California, Los<br />
Angeles<br />
Amy Pickar Abernethy, MD<br />
Duke University Medical Center<br />
Thierry Andre,<br />
Pitie-Salpetriere Hospital<br />
Gerald L. Andriole, MD<br />
Washington University School <strong>of</strong><br />
Medicine<br />
Andrew J. Armstrong, MD, ScM<br />
Duke Cancer Institute<br />
Deborah Kay Armstrong, MD<br />
Sidney Kimmel Comprehensive<br />
Cancer Center at Johns Hopkins<br />
University<br />
Carola A. S. Arndt, MD<br />
Mayo Clinic<br />
Carlos L. Arteaga, MD<br />
Vanderbilt University<br />
Michael B. Atkins, MD<br />
Beth Israel Deaconess Medical<br />
Center<br />
Michel Attal, MD<br />
CHU Purpan<br />
Peter Bach, MD<br />
Memorial Sloan-Kettering Cancer<br />
Center<br />
Deliya Banda, PhD, MPH<br />
Washington Hospital Center<br />
Alberto Bardelli, PhD<br />
Istituto Ricerca Cura Cancro and<br />
FIRC<br />
Ethan M. Basch, MD, MSc<br />
Memorial Sloan-Kettering Cancer<br />
Center<br />
Johanna C. Bendell, MD<br />
The Sarah Cannon Research<br />
Institute<br />
Christine D. Berg, MD<br />
National Cancer Institute<br />
Carl Bergh, MD<br />
Karolinska Institutet<br />
Jordan Berlin, MD<br />
Vanderbilt-Ingram Cancer Center<br />
Donald A. Berry, PhD<br />
University <strong>of</strong> Texas M. D.<br />
Anderson Cancer Center<br />
Keith C. Bible, MD, PhD<br />
Mayo Clinic<br />
Douglas W. Blayney, MD<br />
Stanford University<br />
Diane Blum, MSW<br />
Lymphoma Research Foundation<br />
Daniel Brat, MD, PhD<br />
Emory University<br />
<strong>2012</strong> Annual Meeting Faculty List<br />
Abenaa M. Brewster, MD, MHS<br />
University <strong>of</strong> Texas M. D.<br />
Anderson Cancer Center<br />
David M. Brizel, MD<br />
Duke University Medical Center<br />
Itzhak Brook, MD, MSc<br />
Georgetown University<br />
Adam Brufsky, MD, PhD<br />
University <strong>of</strong> Pittsburgh School<br />
<strong>of</strong> Medicine<br />
Deborah Watkins Bruner, PhD, RN<br />
Abramson Cancer Center,<br />
University <strong>of</strong> Pennsylvania<br />
Thomas A. Buchholz, MD<br />
University <strong>of</strong> Texas M. D.<br />
Anderson Cancer Center<br />
Howard A. Burris, MD<br />
The Sarah Cannon Research<br />
Institute<br />
Harold J. Burstein, MD, PhD<br />
Dana-Farber Cancer Institute<br />
Marc E. Buyse, ScD<br />
International Drug Development<br />
Institute<br />
John C. Byrd, MD<br />
The Ohio State University<br />
Comprehensive Cancer Center<br />
D. Ross Camidge, MD, PhD<br />
University <strong>of</strong> Colorado Cancer<br />
Center<br />
Fatima Cardoso, MD<br />
Breast Unit, Champalimaud<br />
Cancer Center<br />
Bruce Allan Chabner, MD<br />
Massachusetts General Hospital<br />
Susan Marina Chang, MD<br />
University <strong>of</strong> California, San<br />
Francisco<br />
E. Antonio Chiocca, MD, PhD<br />
The Ohio State University<br />
Medical Center<br />
Toni K. Choueiri, MD<br />
Dana-Farber Cancer Institute/<br />
Harvard Medical School<br />
Christine H. Chung, MD<br />
Johns Hopkins University<br />
Tudor-Eliade Ciuleanu, MD<br />
Institute Ion Chiricuta<br />
Rebecca Anne Clark-Snow, RN<br />
University <strong>of</strong> Kansas Cancer<br />
Center<br />
Timothy Francis Cloughesy, MD<br />
University <strong>of</strong> California, Los<br />
Angeles<br />
Kenneth J. Cohen, MD, MBA<br />
Johns Hopkins School <strong>of</strong><br />
Medicine<br />
A. Dimitrios Colevas, MD<br />
Stanford University School <strong>of</strong><br />
Medicine<br />
Howard Colman, MD, PhD<br />
University <strong>of</strong> Utah<br />
Nicoletta Colombo, MD<br />
University <strong>of</strong> Milan-Bicocca<br />
Robert Leo Comis, MD<br />
Coalition <strong>of</strong> Cancer Cooperative<br />
Groups<br />
Carolyn C. Compton, MD, PhD<br />
National Institutes <strong>of</strong> Health<br />
Fergus J. Couch, PhD<br />
Mayo Clinic<br />
Jeffrey Crawford, MD<br />
Duke University Medical Center<br />
Steven A. Curley, MD<br />
University <strong>of</strong> Texas M. D.<br />
Anderson Cancer Center<br />
John Patrick Curtin, MD<br />
New York University School <strong>of</strong><br />
Medicine<br />
Nancy E. Davidson, MD<br />
University <strong>of</strong> Pittsburgh Cancer<br />
Institute<br />
Laura A. Dawson, MD<br />
Princess Margaret Hospital<br />
Johann Sebastian De Bono, MB,<br />
ChB, MSc, PhD<br />
Royal Marsden Hospital<br />
Dave Debronkart<br />
Boston, Massachusetts<br />
John F. Deeken, MD<br />
Lombardi Comprehensive<br />
Cancer Center, Georgetown<br />
University<br />
George D. Demetri, MD<br />
Dana-Farber Cancer Institute/<br />
Harvard Medical School<br />
Lynette Denny, MD, PhD<br />
University <strong>of</strong> Cape Town<br />
Frank C. Detterbeck, MD<br />
Yale School <strong>of</strong> Medicine<br />
Craig Earle, MD<br />
Institute for <strong>Clinical</strong> Evaluative<br />
Sciences<br />
Anthony D. Elias, MD<br />
University <strong>of</strong> Colorado Cancer<br />
Center<br />
Lee M. Ellis, MD<br />
University <strong>of</strong> Texas M. D.<br />
Anderson Cancer Center<br />
William F. Elmquist, PhD,<br />
PharmD<br />
University <strong>of</strong> Minnesota<br />
Laura Esserman, MD, MBA<br />
University <strong>of</strong> California, San<br />
Francisco Helen Diller Family<br />
Comprehensive Cancer Center<br />
Ruth D. Etzioni, PhD<br />
University <strong>of</strong> Washington
Ann T. Farrell, MD<br />
U. S. Food and Drug<br />
Administration<br />
Henry Feldman, MD<br />
Division <strong>of</strong> <strong>Clinical</strong> Informatics<br />
Chris Feudtner, MD, PhD, MPH<br />
Children’s Hospital <strong>of</strong><br />
Philadelphia<br />
Michael Fisch, MD, MPH<br />
University <strong>of</strong> Texas M. D.<br />
Anderson Cancer Center<br />
Keith T. Flaherty, MD<br />
Massachusetts General Hospital<br />
Chris Flowers, MD<br />
H. Lee M<strong>of</strong>fitt Cancer Center &<br />
Research Institute<br />
Jean G. Ford, MD<br />
Johns Hopkins School <strong>of</strong><br />
Medicine<br />
Nicholas K. Foreman, MD<br />
The Children’s Hospital<br />
Josef J. Fox, MD<br />
Memorial Sloan-Kettering Cancer<br />
Center<br />
Keiichi Fujiwara, MD, PhD<br />
Saitama Medical University<br />
International Medical Center<br />
Gwendolyn A. Fyfe, MD<br />
San Francisco, California<br />
Matt D. Galsky, MD<br />
The Tisch Cancer Institute,<br />
Mount Sinai School <strong>of</strong> Medicine<br />
Patricia A. Ganz, MD<br />
University <strong>of</strong> California, Los<br />
Angeles Schools <strong>of</strong> Medicine and<br />
Public Health<br />
Levi A. Garraway, MD, PhD<br />
Dana-Farber Cancer Institute<br />
Karen A. Gelmon, MD<br />
British Columbia Cancer<br />
Agency<br />
Mark R. Gilbert, MD<br />
University <strong>of</strong> Texas M. D.<br />
Anderson Cancer Center<br />
Teresa Gilewski, MD<br />
Memorial Sloan-Kettering Cancer<br />
Center<br />
Maura L. Gillison, MD, PhD<br />
The Ohio State University<br />
Nicolas Girard, MD, MSc<br />
Hopital Louis Pradel<br />
Michael Goldstein, MD<br />
Beth Israel Deaconess Medical<br />
Center and Harvard Medical<br />
School<br />
Paul J. Goodfellow, PhD<br />
Washington University School <strong>of</strong><br />
Medicine<br />
Karyn A. Goodman, MD<br />
Memorial Sloan-Kettering Cancer<br />
Center<br />
Ajay K. Gopal, MD<br />
University <strong>of</strong> Washington<br />
Jason R. Gotlib, MD<br />
Stanford University School <strong>of</strong><br />
Medicine<br />
Bernardo Haddock Lobo Goulart,<br />
MD<br />
Fred Hutchinson Cancer<br />
Research Center<br />
John G. Gribben, DSc, MD<br />
St. Bartholomew’s Hospital<br />
Stephen S. Grubbs, MD<br />
Helen F. Graham Cancer Center<br />
at Christiana Care<br />
Stephen B. Gruber, MD, PhD,<br />
MPH<br />
University <strong>of</strong> Michigan<br />
Steven M. Grunberg, MD<br />
Fletcher Allen Health Care<br />
Eva Grunfeld, MD, DPhil<br />
University <strong>of</strong> Toronto<br />
Philip H. Gutin, MD<br />
Memorial Sloan-Kettering Cancer<br />
Center<br />
Paul Han, MD, MA, MPH<br />
Maine Medical Center Reseach<br />
Institute<br />
Robert Hauser, PhD, PharmD<br />
<strong>American</strong> <strong>Society</strong> <strong>of</strong> <strong>Clinical</strong><br />
<strong>Oncology</strong><br />
Joshua Hauser, MD<br />
Northwestern University<br />
Daniel F. Hayes, MD<br />
University <strong>of</strong> Michigan Medical<br />
Center<br />
Elisabeth I. Heath, MD<br />
Karmanos Cancer Institute<br />
Michael C. Heinrich, MD<br />
Oregon Health & Science<br />
University Knight Cancer<br />
Institute<br />
Paul R. Helft, MD<br />
Indiana University Simon Cancer<br />
Center<br />
Daniel Yick Chin Heng, MD,<br />
MPH<br />
Tom Baker Cancer Centre,<br />
University <strong>of</strong> Calgary<br />
Martee Leigh Hensley, MD, MSc<br />
Memorial Sloan-Kettering Cancer<br />
Center<br />
Dawn L. Hershman, MD, MS<br />
Columbia University Medical<br />
Center<br />
William J. Hicks, MD<br />
The Ohio State University<br />
Peter Hillmen, PhD<br />
Leeds General Infirmary<br />
Bruce E. Hillner, MD<br />
Virginia Commonwealth<br />
University<br />
Pamela S. Hinds, PhD, RN<br />
Children’s National Medical<br />
Center<br />
F. Stephen Hodi, MD<br />
Dana-Farber Cancer Institute<br />
Theodore S. Hong, MD<br />
Massachusetts General Hospital<br />
Melissa M. Hudson, MD<br />
St. Jude Children’s Research<br />
Hospital<br />
Stephen Hunger, MD<br />
University <strong>of</strong> Colorado Denver<br />
School <strong>of</strong> Medicine<br />
Arti Hurria, MD<br />
City <strong>of</strong> Hope<br />
Eun-Sil Shelley Hwang, MD,<br />
MPH<br />
Duke Unversity Medical Center<br />
Nadine Jackson McCleary, MD<br />
Dana-Farber Cancer Institute<br />
Joseph O. Jacobson, MD<br />
North Shore Medical Center<br />
Pasi A. Janne, MD, PhD<br />
Dana-Farber Cancer Institute<br />
Anuja Jhingran, MD<br />
University <strong>of</strong> Texas M. D.<br />
Anderson Cancer Center<br />
Maxine S. Jochelson, MD<br />
Memorial Sloan-Kettering Cancer<br />
Center<br />
Stephen R. D. Johnston, MA, MD,<br />
PhD<br />
Royal Marsden Hospital<br />
Ian Robert Judson, MD<br />
Royal Marsden Hospital and<br />
NHS Foundation Trust<br />
William G. Kaelin, MD<br />
Dana-Farber Cancer Institute<br />
Yoon-Koo Kang, MD, PhD<br />
Asan Medical Center<br />
Hagop Kantarjian, MD<br />
University <strong>of</strong> Texas M. D.<br />
Anderson Cancer Center<br />
Neil E. Kay, MD<br />
Mayo Clinic<br />
Dorothy Mary Kate Keefe, MD,<br />
MBBS<br />
Royal Adelaide Hospital<br />
Karla Kerlikowske, MD<br />
University <strong>of</strong> California, San<br />
Francisco<br />
Muin Khoury, MD, PhD<br />
Office <strong>of</strong> Public Health<br />
Genomics, Centers for Disease<br />
Control and Prevention<br />
Mark W. Kieran, MD, PhD<br />
Dana-Farber Cancer Institute<br />
Gretchen Genevieve Kimmick,<br />
MD, MS<br />
Duke University Medical Center
Hedy Lee Kindler, MD<br />
The University <strong>of</strong> Chicago<br />
Medical Center<br />
Roger David Klein, MD, JD<br />
Bloodcenter <strong>of</strong> Wisconsin<br />
Heidi D. Klepin, MD, MS<br />
Wake Forest University School <strong>of</strong><br />
Medicine<br />
Sara Knapke, MS<br />
Cincinnati Children’s Hospital<br />
Medical Center<br />
Andrew Y. Koh, MD<br />
University <strong>of</strong> Texas Southwestern<br />
Medical Center<br />
Barnett S. Kramer, MD, MPH<br />
National Cancer Institute<br />
Amrita Y. Krishnan, MD<br />
City <strong>of</strong> Hope National Medical<br />
Center<br />
Pamela L. Kunz, MD<br />
Stanford University Medical<br />
Center<br />
Razelle Kurzrock, MD<br />
University <strong>of</strong> Texas M. D.<br />
Anderson Cancer Center<br />
Sancy Ann Leachman, MD<br />
University <strong>of</strong> Utah Hospitals and<br />
Clinics<br />
J. Jack Lee, PhD, DDS<br />
University <strong>of</strong> Texas M. D.<br />
Anderson Cancer Center<br />
Fred T. Lee, MD<br />
University <strong>of</strong> Wisconsin<br />
Natasha B. Leighl, MD<br />
Princess Margaret Hospital<br />
John Leonard, MD<br />
Weill Cornell Medical College<br />
Stephen L. Lessnick, MD, PhD<br />
Huntsman Cancer Institute at the<br />
University <strong>of</strong> Utah<br />
Douglas A. Levine, MD<br />
Memorial Sloan-Kettering Cancer<br />
Center<br />
Lisa F. Licitra, MD<br />
Fondazione IRCCS Istituto<br />
Nazionale dei Tumori<br />
Patricia LoRusso, DO<br />
Karmanos Cancer Institute<br />
Tobey MacDonald, MD<br />
Emory University<br />
Jennifer W. Mack, MD<br />
Dana-Farber Cancer Institute<br />
Michael L. Maitland, MD, PhD<br />
The University <strong>of</strong> Chicago<br />
Robert G. Maki, MD, PhD<br />
Mount Sinai School <strong>of</strong> Medicine<br />
Jeremy Manier<br />
The University <strong>of</strong> Chicago<br />
Elizabeth Mansfield, PhD<br />
U. S. Food and Drug<br />
Administration<br />
John M. Maris, MD<br />
Children’s Hospital <strong>of</strong><br />
Philadelphia<br />
James D. Marks, MD, PhD<br />
University <strong>of</strong> California, San<br />
Francisco<br />
Mary S. McCabe, RN, MA<br />
Memorial Sloan-Kettering Cancer<br />
Center<br />
Worta J. McCaskill-Stevens, MD<br />
Division <strong>of</strong> Cancer Prevention,<br />
National Cancer Institute<br />
D. Scott McMeekin, MD<br />
University <strong>of</strong> Oklahoma Health<br />
Sciences Center<br />
Jerry Menik<strong>of</strong>f, MD, JD<br />
U.S. Department <strong>of</strong> Health and<br />
Human Services<br />
Jeffrey A. Meyerhardt, MD, MPH<br />
Dana-Farber Cancer Institute<br />
Robert Stephen Miller, MD<br />
Sidney Kimmel Comprehensive<br />
Cancer Center at Johns Hopkins<br />
Tetsuya Mitsudomi, MD, PhD<br />
Aichi Cancer Center Hospital<br />
Supriya Gupta Mohile, MD, MS<br />
University <strong>of</strong> Rochester Medical<br />
Center<br />
Tony Mok, MD<br />
Prince <strong>of</strong> Wales Hospital<br />
Monica Morrow, MD<br />
Memorial Sloan-Kettering Cancer<br />
Center<br />
Robert John Motzer, MD<br />
Memorial Sloan-Kettering Cancer<br />
Center<br />
Beverly Moy, MD<br />
Massachusetts General Hospital<br />
Therese M. Mulvey, MD<br />
Southcoast Hospitals Group<br />
Hyman Bernard Muss, MD<br />
University <strong>of</strong> North Carolina<br />
Lineberger Comprehensive<br />
Cancer Center<br />
Fariba Navid, MD<br />
St. Jude Children’s Research<br />
Hospital<br />
Michael N. Neuss, MD<br />
Vanderbilt-Ingram Cancer<br />
Center<br />
Kim Nichols, MD<br />
Children’s Hospital <strong>of</strong><br />
Philadelphia<br />
Andrew David Norden, MD, MPH<br />
Dana-Farber Cancer Institute<br />
Kenneth Offit, MD, MPH<br />
Memorial Sloan-Kettering Cancer<br />
Center<br />
Bert H. O’Neil, MD<br />
University <strong>of</strong> North Carolina at<br />
Chapel Hill<br />
Henry Orlando Otero, MD<br />
Virginia Mason Medical Center<br />
Cynthia Owusu, MD, MS<br />
Case Western Reserve<br />
University<br />
William Pao, MD, PhD<br />
Vanderbilt-Ingram Cancer<br />
Center<br />
Alberto S. Pappo, MD<br />
St. Jude Children’s Research<br />
Hospital<br />
Donald W. Parsons, MD, PhD<br />
Texas Children’s Hospital<br />
Francesco Passamonti, MD<br />
Fondazione IRCCS Policlinico<br />
San Matteo<br />
Jeffrey M. Peppercorn, MD, MPH<br />
Duke University Medical<br />
Center<br />
Douglas E. Peterson, DMD, PhD<br />
University <strong>of</strong> Connecticut Health<br />
Center<br />
Martine J. Piccart-Gebhart, MD,<br />
PhD<br />
Jules Bordet Institute<br />
Todd Alan Pickard, PA-C<br />
University <strong>of</strong> Texas M. D.<br />
Anderson Cancer Center<br />
Whitney Pope, MD, PhD<br />
University <strong>of</strong> California, Los<br />
Angeles<br />
Kathleen I. Pritchard, MD<br />
Sunnybrook Health Sciences<br />
Centre<br />
Stephanie Puget, MD, PhD<br />
Necker Hospital, Universite<br />
Paris-Descartes<br />
Martin N. Raber, MD<br />
University <strong>of</strong> Texas M. D.<br />
Anderson Cancer Center<br />
Derek Raghavan, MD, PhD<br />
Carolinas Medical Center<br />
S. Vincent Rajkumar, MD<br />
Mayo Clinic<br />
David Ransoh<strong>of</strong>f, MD<br />
University <strong>of</strong> North Carolina at<br />
Chapel Hill<br />
Arati Rao, MD<br />
Duke University Medical<br />
Center<br />
Gregory H. Reaman, MD<br />
Children’s National Medical<br />
Center<br />
Mary Weber Redman, PhD<br />
Fred Hutchinson Cancer<br />
Research Center<br />
Robert Evan Reiter, MD<br />
University <strong>of</strong> California, Los<br />
Angeles<br />
Antoni Ribas, MD<br />
University <strong>of</strong> California, Los<br />
Angeles
Richard F. Riedel, MD<br />
Duke Cancer Institute, Duke<br />
University Medical Center<br />
Gregory J. Riely, MD, PhD<br />
Memorial Sloan-Kettering Cancer<br />
Center<br />
Elizabeth Schmidt Rodriguez, RN<br />
Memorial Sloan-Kettering Cancer<br />
Center<br />
Lee S. Rosen, MD<br />
University <strong>of</strong> California, Los<br />
Angeles Division <strong>of</strong> Hematology-<br />
<strong>Oncology</strong><br />
Michael Rowe, PhD<br />
Yale School <strong>of</strong> Medicine<br />
Gordon J. S. Rustin, MD, MBBS<br />
Mount Vernon Cancer Centre<br />
David P. Ryan, MD<br />
Massachusetts General Hospital<br />
Cancer Center<br />
Charles J. Ryan, MD<br />
University <strong>of</strong> California, San<br />
Francisco<br />
Marianne Ryberg, MD<br />
Herlev University Hospital<br />
Gilles A. Salles, MD, PhD<br />
Hospices Civils de Lyon and<br />
Universite de Lyon<br />
Leonard Saltz, MD<br />
Memorial Sloan-Kettering Cancer<br />
Center<br />
Daniel J. Sargent, PhD<br />
Mayo Clinic<br />
Oliver Sartor, MD<br />
Tulane Cancer Center<br />
Hans Sauer, PhD, JD<br />
Biotechnology Industry<br />
Organization<br />
Doug Sawyer, MD, PhD<br />
Vanderbilt University<br />
Lidia Schapira, MD<br />
Massachusetts General Hospital<br />
Joshua David Schiffman, MD<br />
University <strong>of</strong> Utah<br />
Lowell E. Schnipper, MD<br />
Beth Israel Deaconess Medical<br />
Center<br />
Gilberto Schwartsmann, MD, PhD<br />
Federal University <strong>of</strong> Rio Grande<br />
do Sul<br />
Katia Scotlandi, PhD<br />
Istituto Ortopedico Rizzoli<br />
Victoria Louise Seewaldt, MD<br />
Duke University Medical<br />
Center<br />
Lecia V. Sequist, MD, MPH<br />
Massachusetts General Hospital<br />
Manish A. Shah, MD<br />
Weill Cornell Medical College<br />
Manisha H. Shah, MD<br />
The Ohio State University<br />
Comprehensive Cancer Center<br />
Neil P. Shah, MD, PhD<br />
University <strong>of</strong> California, San<br />
Francisco<br />
Ge<strong>of</strong>frey Shapiro, MD, PhD<br />
Dana-Farber Cancer Institute<br />
Alice Tsang Shaw, MD, PhD<br />
Massachusetts General Hospital<br />
Cancer Center<br />
Lillian L. Siu, MD<br />
Princess Margaret Hospital<br />
Stephen Skapek, MD<br />
The University <strong>of</strong> Chicago<br />
Martha L. Slattery, PhD<br />
University <strong>of</strong> Utah<br />
George W. Sledge, MD<br />
Indiana University Simon Cancer<br />
Center<br />
Stefan Sleijfer, MD, PhD<br />
Erasmus University Medical<br />
Center, Daniel den Hoed Cancer<br />
Center<br />
Sonali M. Smith, MD<br />
The University <strong>of</strong> Chicago<br />
Thomas J. Smith, MD<br />
Sidney Kimmel Comprehensive<br />
Cancer Center at Johns Hopkins<br />
Stephen Barnett Solomon, MD<br />
Memorial Sloan-Kettering Cancer<br />
Center<br />
Stephen T. Sonis, DMSc, DDS<br />
Biomodels<br />
Jeffrey Alan Sosman, MD<br />
Vanderbilt Medical Center<br />
Paul Spellman, PhD<br />
Oregon Health & Science<br />
University<br />
Margaret R. Spitz, MD<br />
University <strong>of</strong> Texas M. D.<br />
Anderson Cancer Center<br />
Zs<strong>of</strong>ia Kinga Stadler, MD<br />
Memorial Sloan-Kettering Cancer<br />
Center<br />
Walter Michael Stadler, MD<br />
The University <strong>of</strong> Chicago<br />
Richard Sullivan, PhD, MBBS<br />
Kings Health Partners Integrated<br />
Cancer Centre<br />
Lillian Sung, MD, PhD<br />
The Hospital for Sick Children<br />
Antonella Surbone, MD, PhD<br />
New York University<br />
Sandra M. Swain, MD<br />
Washington Hospital Center<br />
Mario Sznol, MD<br />
Yale Cancer Center<br />
Josep Tabernero, MD<br />
Vall d’Hebron University<br />
Hospital<br />
Kenneth Tanabe, MD<br />
Massachusetts General Hospital<br />
Cancer Center<br />
William D. Tap, MD<br />
University <strong>of</strong> California, Los<br />
Angeles<br />
Ian Murchie Thompson, MD<br />
University <strong>of</strong> Texas Health<br />
Science Center at San Antonio<br />
Michael A. Thompson, MD, PhD<br />
ProHealth Regional Cancer<br />
Center<br />
Anthony W. Tolcher, MD<br />
South Texas Accelerated<br />
Research Therapeutics (START)<br />
Sean R. Tunis, MD, MSc<br />
Center for Medical Technology<br />
Policy<br />
Alan Paul Venook, MD<br />
University <strong>of</strong> California, San<br />
Francisco<br />
Srdan Verstovsek, MD, PhD<br />
University <strong>of</strong> Texas M. D.<br />
Anderson Cancer Center<br />
Jaap Verweij, MD, PhD<br />
Erasmus University Medical<br />
Center<br />
Louise Villejo, MPH<br />
University <strong>of</strong> Texas M. D.<br />
Anderson Cancer Center<br />
Beth A. Virnig, PhD<br />
University <strong>of</strong> Minnesota, School<br />
<strong>of</strong> Public Health<br />
Andrew J. Wagner, MD, PhD<br />
Dana-Farber Cancer Institute<br />
James Ross Waisman, MD<br />
Breastlink Medical Group Inc<br />
Robert S. Warren, MD<br />
University <strong>of</strong> California, San<br />
Francisco<br />
Jeffrey S. Weber, MD, PhD<br />
H. Lee M<strong>of</strong>fitt Cancer Center &<br />
Research Institute<br />
Simon Wein, MD<br />
David<strong>of</strong>f Cancer Center, Rabin<br />
Medical Center<br />
Martin R. Weiser, MD<br />
Memorial Sloan-Kettering Cancer<br />
Center<br />
Howard Jack West, MD<br />
Swedish Cancer Institute<br />
William H. Westra, MD<br />
Johns Hopkins School <strong>of</strong><br />
Medicine<br />
Lauren Allison Wiebe, MD<br />
Medical College <strong>of</strong> Wisconsin<br />
Timothy J. Wilt, MD, MPH<br />
United States Department <strong>of</strong><br />
Veteran’s Affairs<br />
Eric P. Winer, MD<br />
Dana-Farber Cancer Institute<br />
Ignacio I. Wistuba, MD<br />
University <strong>of</strong> Texas M. D.<br />
Anderson Cancer Center
Douglas E. Wood, MD<br />
University <strong>of</strong> Washington<br />
Paul Workman, BSc, PhD, DSc<br />
The Institute <strong>of</strong> Cancer Research<br />
Michael Xing, MD, PhD<br />
Johns Hopkins University<br />
Sam S. Yoon, MD<br />
Massachusetts General Hospital<br />
Anas Younes, MD<br />
University <strong>of</strong> Texas M. D.<br />
Anderson Cancer Center<br />
Theoklis Zaoutis, MD<br />
Children’s Hospital <strong>of</strong><br />
Philadelphia<br />
Richard C. Zellars, MD<br />
Sidney Kimmel Comprehensive<br />
Cancer Center at Johns Hopkins<br />
Andrew X. Zhu, MD, PhD<br />
Massachusetts General Hospital
<strong>2012</strong> Educational Book Expert Panel<br />
The Expert Panel is a group <strong>of</strong> well-recognized physicians and researchers in oncology<br />
and related fields who have served as peer reviewers on the Educational Book articles.<br />
David Adelstein, MD<br />
Cleveland Clinic<br />
Alex Adjei, MD<br />
University <strong>of</strong> Texas M. D.<br />
Anderson Cancer Center<br />
Doug Adkins, MD<br />
Washington University School <strong>of</strong><br />
Medicine<br />
Stefan Anker, MD<br />
Charité, Universitätsmedizin<br />
Berlin<br />
Robert Arceci, MD, PhD<br />
Sidney Kimmel Cancer Center<br />
Voichita Bar-Ad, MD<br />
Thomas Jefferson University<br />
Hospital<br />
Helen Barraclough, MS<br />
Eli Lilly<br />
Leif Bergsagel, MD<br />
Mayo Clinic<br />
Jan Beumer, PharmD, PhD<br />
University <strong>of</strong> Pittsburgh<br />
George Blumenschein, MD<br />
University <strong>of</strong> Texas M. D.<br />
Anderson Cancer Center<br />
Louise Bordeleau, MD, MSc<br />
McMaster University<br />
Alan Bryce, MD<br />
Mayo Clinic<br />
Kevin Camphausen, MD<br />
National Cancer Institute<br />
James Cassidy, MD<br />
Roche<br />
Nelson Chao, MD<br />
Duke University<br />
Carien Creutzberg, MD<br />
Leiden University Medical<br />
Center<br />
Brian Czito, MD<br />
Duke University<br />
Adam Dicker, MD<br />
Thomas Jefferson University<br />
Brenda Diergaarde, PhD<br />
University <strong>of</strong> Pittsburgh<br />
Benjamin Djulbegovic, MD<br />
University <strong>of</strong> South Florida<br />
Laura Dominici, MD<br />
Brigham and Womens’ Hospital<br />
Fritz Eilber, MD<br />
University <strong>of</strong> California,<br />
Los Angeles<br />
Rebecca Elstrom, MD<br />
Weill Cornell Medical College<br />
Edward Garon, MD<br />
University <strong>of</strong> California,<br />
Los Angeles<br />
Julia Glade, MD<br />
Columbia University<br />
Stewart Goldman, MD<br />
Children’s Memorial Hospital<br />
Mary Gospodarowicz, MD<br />
Princess Margaret Hospital<br />
Victor Grann, MD<br />
Columbia University<br />
Parameswaran Hari, MD<br />
Medical College <strong>of</strong> Wisconsin<br />
Michael Hassett, MD<br />
Dana-Farber Cancer Institute<br />
Michael Isak<strong>of</strong>f, MD<br />
Conneticut Children’s Medical<br />
Center<br />
Michael Kelly, MD, PhD<br />
Medical College <strong>of</strong> Wisconsin<br />
Scott Kopetz, MD<br />
University <strong>of</strong> Texas M. D.<br />
Anderson Cancer Center<br />
Mario Lacouture, MD<br />
Northwestern University<br />
Edward Levine, MD<br />
Wake Forest University<br />
Gerry Linette, MD, PhD<br />
Washington University School <strong>of</strong><br />
Medicine<br />
Steven Lipshultz, MD<br />
University <strong>of</strong> Miami<br />
Christophe Louvet, MD<br />
l’Institut Mutualiste Montsouris<br />
Bo Lu, MD, PhD<br />
Vanderbilt University<br />
Mario Malogolowkin, MD<br />
Medical College <strong>of</strong> Wisconsin<br />
Guido Marcucci, MD<br />
The Ohio State University<br />
Medical Center<br />
Erica Mayer, MD<br />
Dana-Farber Cancer Institute<br />
Minesh Mehta, MD<br />
University <strong>of</strong> Wisconsin<br />
Vicki Morrison, MD<br />
University <strong>of</strong> Minnesota<br />
Joanne Mortimer, MD<br />
City <strong>of</strong> Hope<br />
Craig Moskowitz, MD<br />
Memorial Sloan-Kettering Cancer<br />
Center<br />
Timothy Moynihan, MD<br />
Mayo Clinic<br />
Barbara Murphy, MD<br />
Vanderbilt University<br />
P.B. Ottevanger, MD<br />
Radboud University Nijmegen<br />
Medical Centre<br />
Rebecca Pentz, PhD<br />
Emory University<br />
James Pingpank, MD<br />
University <strong>of</strong> Pittsburgh<br />
Katherine Pisters, MD<br />
University <strong>of</strong> Texas M. D.<br />
Anderson Cancer Center<br />
Dennis Priebat, MD<br />
Washington Hospital Center<br />
Rinaa Punglia, MD<br />
Dana-Farber Cancer Institute<br />
David Quinn, MD<br />
University <strong>of</strong> Southern California<br />
Nithya Ramnath, MD<br />
University <strong>of</strong> Michigan<br />
Paul Gerard Guy Richardson, MD<br />
Dana-Farber Cancer Institute<br />
Jonathan Rosenberg, MD<br />
Dana-Farber Cancer Institute<br />
David Rosenthal, MD<br />
Harvard University<br />
Anna Roshal, MD<br />
Washington University School <strong>of</strong><br />
Medicine<br />
John Ruckdeschel, MD<br />
Karmanos Cancer Institute<br />
Hope Rugo, MD<br />
University <strong>of</strong> California, San<br />
Francisco<br />
Rachel Sanborn, MD<br />
Providence Portland Medical<br />
Center<br />
Alan Sandler, MD<br />
Oregon Health & Science<br />
University<br />
Rafael Santana-Davilla, MD<br />
Medical College <strong>of</strong> Wisconsin<br />
Takami Sato, MD, PhD<br />
Jefferson Medical College<br />
Lee Schwartzberg, MD<br />
The West Clinic<br />
Sunil Sharma, MD<br />
University <strong>of</strong> Utah<br />
Wenyin Shi, MD<br />
Thomas Jefferson University<br />
Hospital<br />
Timothy Showalter, MD<br />
Thomas Jefferson University<br />
Hospital
Mary Lou Smith, JD, MBA<br />
Research Advocacy Network,<br />
Chicago<br />
Andrew Stephenson, MD<br />
Cleveland Clinic<br />
Timothy Synold, PharmD<br />
City <strong>of</strong> Hope<br />
Ben Tan, MD<br />
Washington University School <strong>of</strong><br />
Medicine<br />
Meaghan Tenney, MD<br />
University <strong>of</strong> Chicago<br />
Brian Van Tine, MD, PhD<br />
Washington University School <strong>of</strong><br />
Medicine<br />
Gauri Varadhachary, MD<br />
University <strong>of</strong> Texas M. D.<br />
Anderson Cancer Center<br />
Ravi Vij, MD<br />
Washington University School <strong>of</strong><br />
Medicine<br />
Nina Wagner-Johnston, MD<br />
Washington University School <strong>of</strong><br />
Medicine<br />
Heather Wakelee, MD<br />
Stanford University<br />
Ellen Warner, MD<br />
Odette Cancer Center<br />
Theresa Werner, MD<br />
University <strong>of</strong> Utah<br />
Mier Wetzler, MD<br />
Roswell Park Cancer Institute<br />
William Wierda, MD<br />
University <strong>of</strong> Texas M. D.<br />
Anderson Cancer Center<br />
Dennis Wigle, MD, PhD<br />
Mayo Clinic<br />
Tanya Wildes, MD<br />
Washington University School <strong>of</strong><br />
Medicine<br />
Els Witteveen, MD<br />
University Medical Center<br />
Utrecht<br />
Sandra Wong, MD<br />
University <strong>of</strong> Michigan Health<br />
Systems<br />
Seiko Yamada, MD<br />
University <strong>of</strong> Chicago<br />
Darrell Yamashiro, MD, PhD<br />
Columbia University<br />
Jonathan Zager, MD<br />
H. Lee M<strong>of</strong>fitt Cancer Center<br />
Robin Zon, MD<br />
Michiana Hematology-<strong>Oncology</strong>
<strong>2012</strong> Annual Meeting Supporters*<br />
CONQUER CANCER FOUNDATION MISSION ENDOWMENT FOUNDING DONORS<br />
Genentech Bio<strong>Oncology</strong> TM<br />
GlaxoSmithKline <strong>Oncology</strong><br />
Abbott <strong>Oncology</strong><br />
Amgen<br />
Celgene Corporation<br />
Novartis <strong>Oncology</strong><br />
San<strong>of</strong>i <strong>Oncology</strong><br />
CONQUER CANCER FOUNDATION MISSION ENDOWMENT SUSTAINING DONORS<br />
<strong>EDUCATIONAL</strong> SUPPORT<br />
Amgen<br />
Educational Book<br />
Astellas<br />
AVEO Pharmaceuticals, Inc.<br />
Medivation<br />
Genitourinary Cancer Track<br />
Bristol-Myers Squibb<br />
Gastrointestinal (Colorectal) Cancer Track<br />
Gastrointestinal (Noncolorectal) Cancer Track<br />
Head and Neck Cancer Track<br />
Celgene Corporation<br />
Best <strong>of</strong> ASCO ® Meetings<br />
Breast Cancer Track<br />
Gastrointestinal (Noncolorectal) Cancer Track<br />
Leukemia, Myelodysplasia, and Transplantation Track<br />
Lung Cancer Track<br />
Lymphoma and Plasma Cell Disorders Track<br />
Daiichi Sankyo, Inc.<br />
Lung Cancer Track<br />
Eisai Inc.<br />
Best <strong>of</strong> ASCO ® Meetings<br />
Breast Cancer Track<br />
Leukemia, Myelodysplasia, and Transplantation Track<br />
Patient and Survivor Care Track<br />
Genentech Bio<strong>Oncology</strong> TM<br />
Best <strong>of</strong> ASCO ® Meetings<br />
<strong>Clinical</strong> Trials Track<br />
Gastrointestinal (Colorectal) Cancer Track<br />
Gynecologic Cancer Track<br />
Melanoma/Skin Cancer Track<br />
Pr<strong>of</strong>essional Development Track<br />
Tumor Biology Track<br />
Kidney Cancer Association<br />
Genitourinary Cancer Track<br />
Lilly USA, LLC<br />
Best <strong>of</strong> ASCO ® Meetings<br />
Gastrointestinal (Colorectal) Cancer Track<br />
Head and Neck Cancer Track<br />
Lung Cancer Track<br />
* This list reflects commitments as <strong>of</strong> April 16, <strong>2012</strong><br />
Lilly USA, LLC<br />
Millennium: The Takeda <strong>Oncology</strong> Company<br />
San<strong>of</strong>i <strong>Oncology</strong><br />
Merck <strong>Oncology</strong><br />
Melanoma/Skin Cancer Track<br />
Millennium: The Takeda <strong>Oncology</strong> Company<br />
Genitourinary Cancer Track<br />
Pr<strong>of</strong>essional Development Track<br />
Myriad Genetics Laboratories, Inc.<br />
Breast Cancer Track<br />
Cancer Genetics Track<br />
Novartis <strong>Oncology</strong><br />
Breast Cancer Track<br />
Gastrointestinal (Noncolorectal) Cancer Track<br />
Onyx Pharmaceuticals<br />
Lymphoma and Plasma Cell Disorders Track<br />
Pfizer <strong>Oncology</strong><br />
Breast Cancer Track<br />
Gastrointestinal (Noncolorectal) Cancer Track<br />
Genitourinary Cancer Track<br />
Leukemia, Myelodysplasia, and Transplantation Track<br />
Lung Cancer Track<br />
Lymphoma and Plasma Cell Disorders Track<br />
Patient and Survivor Care Track<br />
San<strong>of</strong>i <strong>Oncology</strong><br />
Genitourinary Cancer Track<br />
San<strong>of</strong>i <strong>Oncology</strong> and Regeneron Pharmaceuticals<br />
Best <strong>of</strong> ASCO ® Meetings<br />
Gastrointestinal (Colorectal) Cancer Track<br />
Seattle Genetics<br />
Lymphoma and Plasma Cell Disorders Track<br />
GENERAL SUPPORT<br />
Abbott <strong>Oncology</strong><br />
Young Investigator Award<br />
<strong>American</strong> Cancer <strong>Society</strong><br />
ASCO-<strong>American</strong> Cancer <strong>Society</strong> Award and Lecture<br />
Amgen<br />
Career Development Award<br />
International Development and Education Awards<br />
Long-term International Fellowship (LIFe)<br />
Merit Awards<br />
Young Investigator Award
ASCO and Conquer Cancer Foundation Boards <strong>of</strong><br />
Directors<br />
Jane C. Wright, MD, Young Investigator Award<br />
ASCO Cancer Research Committee<br />
Young Investigator Award<br />
ASCO <strong>Clinical</strong> Practice Committee<br />
Young Investigator Award<br />
Astellas<br />
Young Investigator Award<br />
Avon Foundation for Women<br />
International Development and Education Awards in<br />
Breast Cancer<br />
Bayer Healthcare Pharmaceuticals Inc.<br />
International Development and Education Awards<br />
Boehringer Ingelheim Pharmaceuticals, Inc.<br />
iMeetings Mobile Application<br />
Virtual Meetings<br />
Bradley Stuart Beller Endowed Fund<br />
Bradley Stuart Beller Special Merit Award<br />
The Breast Cancer Research Foundation<br />
Advanced <strong>Clinical</strong> Research Award in Breast Cancer<br />
Career Development Award in Breast Cancer<br />
Comparative Effectiveness Research Pr<strong>of</strong>essorship in<br />
Breast Cancer<br />
Long-term International Award (LIFe) in Breast or<br />
Related Cancers<br />
Young Investigator Award in Breast Cancer<br />
Young Investigator Award in Breast Cancer in honor <strong>of</strong><br />
Susan Hirschhorn and in memory <strong>of</strong> her mother<br />
Bristol-Myers Squibb<br />
Annual Meeting Program<br />
Merit Awards<br />
Young Investigator Award<br />
Celgene Corporation<br />
Career Development Award<br />
Merit Awards<br />
Patient Advocate Scholarship Program<br />
Young Investigator Award<br />
Celltrion Healthcare Co., Ltd.<br />
Annual Meeting Program Mobile Application (iPlanner)<br />
Coalition <strong>of</strong> Cancer Cooperative Groups<br />
<strong>Clinical</strong> Trials Participation Awards<br />
Daiichi Sankyo, Inc.<br />
Annual Meeting Program Mobile Application (iPlanner)<br />
Young Investigator Award<br />
Dendreon Corporation<br />
Annual Meeting ASCO TV<br />
Internet Stations<br />
Laptop Stations (Wi-Fi Zones)<br />
Don Shula Foundation, Inc.<br />
Young Investigator Award in Breast Cancer<br />
Doris Duke Charitable Foundation<br />
Medical Student Rotation for Underrepresented<br />
Populations<br />
Eisai Inc.<br />
Charging Stations<br />
Interactive Locator Maps<br />
International Development and Education Awards<br />
Internet Stations<br />
Patient Advocate Scholarship Program<br />
Laptop Stations (Wi-Fi Zones)<br />
Young Investigator Award (2)<br />
Estate <strong>of</strong> Sandra Syms<br />
Young Investigator Award in Kidney Cancer Research<br />
Genentech Bio<strong>Oncology</strong><br />
Advanced <strong>Clinical</strong> Research Award in Colorectal Cancer<br />
Advanced <strong>Clinical</strong> Research Award in Glioma<br />
Advanced <strong>Clinical</strong> Research Award in Hematologic<br />
Malignancies<br />
Career Development Award (5)<br />
Program Announcement<br />
Registration and Housing Announcement<br />
Translational Research Pr<strong>of</strong>essorship<br />
Young Investigator Award (5)<br />
GlaxoSmithKline <strong>Oncology</strong><br />
Gianni Bonadonna Breast Cancer Award and Fellowship<br />
Translational Research Pr<strong>of</strong>essorship<br />
Incyte Corporation<br />
Annual Meeting ASCO TV<br />
Charging Stations<br />
Interactive Locator Maps<br />
Internet Stations<br />
Jackson Simpson<br />
Merit Awards in Lung Cancer and Lymphoma<br />
James B. Nachman Pediatric <strong>Oncology</strong> Fund<br />
James B. Nachman ASCO Junior Faculty Award in<br />
Pediatric <strong>Oncology</strong><br />
Journal <strong>of</strong> <strong>Clinical</strong> <strong>Oncology</strong><br />
Young Investigator Award<br />
Kidney Cancer Association<br />
Young Investigator Award in Renal Cell Carcinoma<br />
Lilly USA, LLC<br />
Career Development Award<br />
Merit Awards<br />
Patient Advocate Scholarship Program<br />
Young Investigator Award<br />
Merck <strong>Oncology</strong><br />
Young Investigator Award<br />
Millennium: The Takeda <strong>Oncology</strong> Company<br />
International Development and Education Awards<br />
Merit Awards<br />
<strong>Oncology</strong> Trainee Travel Awards<br />
Patient Advocate Scholarship Program<br />
Young Investigator Award (2)
National Cancer Institute<br />
International Development and Education Awards in<br />
Palliative Care<br />
Novartis <strong>Oncology</strong><br />
Career Development Award<br />
Young Investigator Award (2)<br />
Open <strong>Society</strong> Foundations<br />
International Development and Education Awards in<br />
Palliative Care<br />
Palo Alto Medical Foundation<br />
Leaders in Practice—Silver Supporter<br />
Pepper Hamilton LLC<br />
Pro Bono Legal Services: Diversity in <strong>Oncology</strong> Initiative<br />
Pfizer <strong>Oncology</strong><br />
Virtual Meetings<br />
Young Investigator Award<br />
Raj Mantena, RPh<br />
Leaders in Practice—Primary Supporter<br />
Roche<br />
Career Development Award (2)<br />
Annual Meeting Program Mobile Application (iPlanner)<br />
Laptop Stations (Wi-Fi Zones)<br />
Young Investigator Award (5)<br />
San<strong>of</strong>i <strong>Oncology</strong><br />
Career Development Award<br />
Drug Development Research Pr<strong>of</strong>essorship<br />
Sarcoma Foundation <strong>of</strong> America<br />
Career Development Award in Sarcoma Research<br />
Young Investigator Award in Sarcoma Research<br />
Seattle Genetics<br />
iMeetings Mobile Application<br />
Internet Stations<br />
Patient Advocate Scholarship Program<br />
Young Investigator Award<br />
Strike 3 Foundation<br />
Young Investigator Award in Pediatric <strong>Oncology</strong><br />
Susan G. Komen for the Cure ®<br />
The Susan G. Komen for the Cure/Conquer Cancer<br />
Foundation Research Initiative<br />
● Conquer Cancer Foundation <strong>of</strong> ASCO Improving<br />
Cancer Care Grant, funded by Susan G. Komen for the<br />
Cure<br />
The Susan G. Komen for the Cure/ASCO Quality <strong>of</strong> Care<br />
Initiative<br />
● The ASCO Breast Cancer Registry Pilot Program,<br />
funded by Susan G. Komen for the Cure<br />
● The ASCO Diversity in <strong>Oncology</strong> Initiative, funded by<br />
Susan G. Komen for the Cure<br />
● The ASCO Study <strong>of</strong> Collaborative Practice<br />
Arrangements, funded by Susan G. Komen for the<br />
Cure<br />
● The ASCO Study <strong>of</strong> Geographic Access to <strong>Oncology</strong><br />
Care, funded by Susan G. Komen for the Cure<br />
● Assessing and Monitoring Demand for Cancer Care<br />
Services, with funding from Susan G. Komen for the<br />
Cure<br />
William D. Piety Living Trust<br />
Young Investigator Award in Cholangiocarcinoma<br />
Research<br />
WWWW Foundation, Inc. (QuadW) and The Sarcoma<br />
Fund <strong>of</strong> the QuadW Foundation<br />
Young Investigator Award in Sarcoma in memory <strong>of</strong> Willie<br />
Tichenor
Letter from the Editor<br />
The theme <strong>of</strong> the <strong>2012</strong> <strong>American</strong> <strong>Society</strong> <strong>of</strong> <strong>Clinical</strong><br />
<strong>Oncology</strong> Annual Meeting is “Collaborating to Conquer<br />
Cancer,” and the sessions being presented at this year’s<br />
meeting certainly look toward this theme with the goal <strong>of</strong><br />
advancing the education <strong>of</strong> oncologists so that they can<br />
improve the care provided to their patients.<br />
The ASCO Educational Book is meant to serve as an<br />
enduring record <strong>of</strong> the information presented at these sessions.<br />
This year, 92 educational sessions are represented by<br />
the articles published in this book, written by some <strong>of</strong> the<br />
most recognized clinicians and investigators in their field. In<br />
order to ensure the utmost accuracy and clarity, each article<br />
is reviewed by our editorial team, as well as by peer<br />
reviewers from our extensive expert panel. In addition,<br />
Natasha Leighl, MD, an expert in lung cancer and cost-<strong>of</strong>care<br />
issues and the Associate Editor <strong>of</strong> the Educational<br />
Book, has provided valuable insights to the review process.<br />
We are grateful to the authors who, despite their demanding<br />
schedules, took the time to write and, in some cases,<br />
revise their articles according to the peer reviewers’ comments.<br />
We appreciate the efforts <strong>of</strong> our peer reviewers for<br />
their careful, thorough reviews that undoubtedly contributed<br />
to the high quality <strong>of</strong> this book.<br />
Please visit asco.org/edbook for access to all <strong>of</strong> the <strong>2012</strong><br />
Educational Book articles. The following are examples for<br />
citing both print and electronic articles:<br />
Girard N. Thymoma: From Chemotherapy to Targeted<br />
Therapy. In: Govindan R, ed. <strong>2012</strong> ASCO Educational Book.<br />
Alexandria, VA: <strong>American</strong> <strong>Society</strong> <strong>of</strong> <strong>Clinical</strong> <strong>Oncology</strong>;<br />
<strong>2012</strong>;475-479.<br />
Surbone A, Baider L. Are Oncologists Accountable Only to<br />
Patients or Also to Their Families? An International Perspective.<br />
In: Govindan R, ed. <strong>2012</strong> ASCO Educational Book.<br />
Alexandria, VA: <strong>American</strong> <strong>Society</strong> <strong>of</strong> <strong>Clinical</strong> <strong>Oncology</strong>;<br />
<strong>2012</strong>;e15-e19.<br />
The ASCO Educational Book would not exist today but for<br />
the remarkable staff at ASCO. I want to specifically thank<br />
Christine Smith and Lauren Burke, who have worked very<br />
hard year-round to make this publication possible. Dr.<br />
Leighl and I welcome your feedback and suggestions on how<br />
we can improve the content, so please feel free to email me<br />
with your comments at EdBook@asco.org.<br />
Ramaswamy Govindan, MD<br />
Editor<br />
1
CONTROVERSIES IN THE ADJUVANT TREATMENT<br />
OF BREAST CANCER<br />
CHAIR<br />
Stephen R. Johnston, MA, PhD<br />
Royal Marsden Hospital<br />
London, United Kingdom<br />
SPEAKERS<br />
Anthony D. Elias, MD<br />
University <strong>of</strong> Colorado Cancer Center<br />
Aurora, CO<br />
Hyman Muss, MD<br />
University <strong>of</strong> North Carolina Lineberger<br />
Comprehensive Cancer Center<br />
Chapel Hill, NC
Adjuvant Therapy for Older Women with<br />
Early-Stage Breast Cancer: Treatment<br />
Selection in a Complex Population<br />
By Cynthia Owusu, MD, MS, Arti Hurria, MD, and Hyman Muss, MD<br />
Overview: Breast cancer is a disease <strong>of</strong> aging. However,<br />
older women with breast cancer are less likely to participate in<br />
clinical trials or to receive recommended treatment. This<br />
undertreatment has contributed to a lag in breast cancer<br />
survival outcomes for older women compared with that for<br />
their younger counterparts. The principles that govern recommendations<br />
for adjuvant treatment <strong>of</strong> breast cancer are the<br />
same for younger and older women. Systemic adjuvant treatment<br />
recommendations should be <strong>of</strong>fered on the basis <strong>of</strong><br />
tumor characteristics that divide patients into three distinct<br />
subgroups. These include (1) older women with hormone<br />
receptor (HR)-positive and human epidermal growth factor 2<br />
(HER2)-negative breast cancer who should be <strong>of</strong>fered endocrine<br />
therapy; (2) older women with HR-negative and HER2negative<br />
breast cancer who should be <strong>of</strong>fered adjuvant<br />
BREAST CANCER is the most common cancer in <strong>American</strong><br />
women and the second leading cause <strong>of</strong> cancerrelated<br />
deaths among women. In 2011, approximately<br />
230,480 new cases were diagnosed in the United States,<br />
with an expected 39,520 deaths. 1,2 The most important risk<br />
factor for breast cancer is age. The estimated lifetime risk <strong>of</strong><br />
a new breast cancer is 1 in 15, 1 in 29, 1 in 27 and 1 in 207<br />
for women 70 years or older, 60 to 69, 40 to 59, and 39 or<br />
younger, respectively. 1 The median age at the time <strong>of</strong> breast<br />
cancer diagnosis is currently 61 years and an estimated 45%<br />
<strong>of</strong> women are 65 or older at the time <strong>of</strong> initial diagnosis. 2,3<br />
Recent gains in life expectancy, coupled with aging as a risk<br />
factor for breast cancer, makes breast cancer primarily a<br />
disease <strong>of</strong> older women, with increasing public health importance.<br />
In 1980, persons 65 and older represented 11.3%<br />
<strong>of</strong> the total population, but by 2030 this proportion is<br />
expected to increase to 20%. 3 In addition, by 2030, persons<br />
older than 75 will be expected to account for just under 50%<br />
<strong>of</strong> the total cohort older than 65. 4 Given the nonlinear<br />
age-risk relationship and increasing life expectancy, a substantial<br />
proportion <strong>of</strong> older women are expected to be affected<br />
by breast cancer.<br />
Age-Related Cancer Health Disparities<br />
Evaluation <strong>of</strong> the biology <strong>of</strong> breast cancer by patient age<br />
has shown that hormone receptor-positive, low S-phase, low<br />
tumor grade and HER2-negative tumors are more common<br />
among older than younger women, 5 although these differences<br />
are relatively modest. Despite the favorable tumor<br />
pr<strong>of</strong>ile <strong>of</strong> breast cancer in older women, this has not translated<br />
into any major survival advantage for older women<br />
with breast cancer in comparison with their younger counterparts.<br />
In a study that drew data from National Vital<br />
Statistics Reports and the Surveillance Epidemiology and<br />
End Results database <strong>of</strong> the National Cancer Institute,<br />
Smith and colleagues 6 found that although the rate <strong>of</strong> breast<br />
cancer death in the general population and the adjusted risk<br />
<strong>of</strong> death among women with newly diagnosed disease are<br />
declining among all age groups, the least decline has been<br />
among older women. Relative to 1990, the rate <strong>of</strong> breast<br />
chemotherapy; and (3) older women with HER2-positive disease<br />
who should be <strong>of</strong>fered chemotherapy with trastuzumab.<br />
Exceptions to these guidelines may be made for older women<br />
with small node-negative tumors or frail older women with<br />
limited life expectancy, where close surveillance may be a<br />
reasonable alternative. Addressing the current age-related<br />
disparities in breast cancer survival will require that older<br />
women are <strong>of</strong>fered the same state-<strong>of</strong>-the-art-treatment as<br />
their younger counterparts, with a careful weighing <strong>of</strong> the<br />
risks and benefits <strong>of</strong> each treatment in the context <strong>of</strong> the<br />
individual’s preferences. In addition, older women should be<br />
encouraged to participate in breast cancer clinical trials to<br />
generate additional chemotherapy efficacy, toxicity, and quality<br />
<strong>of</strong> life data.<br />
cancer death in the general population decreased 2.5% per<br />
year for women age 20 to 49, 2.1% per year for those age 50<br />
to 64, 2% per year for those age 65 to 74, but 1.1% per year<br />
for those age 75 years and older. Moreover, among women<br />
with newly diagnosed breast cancer between 1980 and 1997,<br />
the adjusted risk <strong>of</strong> death decreased by 3.6% per year among<br />
women younger than age 75 compared with 1.3% per year<br />
among those age 75 and older (p � 0.01). These differences<br />
were even greater for older black women. The age-related<br />
disparity in survival outcomes was hypothesized to be related<br />
to the undertreatment <strong>of</strong> older women with breast<br />
cancer. In an analysis <strong>of</strong> 9,766 patients enrolled in the<br />
TEAM (Tamoxifen Exemestane Adjuvant Multinational)<br />
randomized controlled trial conducted with postmenopausal<br />
women with hormone receptor-positive breast cancer, increasing<br />
age was associated with a higher disease-specific<br />
mortality. 7 Treatments received and tumor characteristics<br />
did not completely explain the age-related differences in<br />
survival outcomes but older patients in this trial were much<br />
less likely to receive chemotherapy. Together, these data<br />
clearly underscore the fact that breast cancer is an important<br />
disease <strong>of</strong> older women who bear a disproportionate<br />
burden <strong>of</strong> the morbidity and mortality associated with the<br />
disease and who should be <strong>of</strong>fered proven treatments that<br />
improve survival outcomes. Given that treatment differences<br />
do not completely explain the age-related disparities<br />
in survival outcomes, additional population and translational<br />
studies are needed to provide further insight into the<br />
reasons for these disparities.<br />
From the Case Western Reserve University School <strong>of</strong> Medicine, Cleveland, OH; City <strong>of</strong><br />
Hope Medical Center and Beckman Research Institute, Duarte, CA; and, Lineberger<br />
Comprehensive Cancer Center, University <strong>of</strong> North Carolina, Chapel Hill, NC.<br />
Authors’ disclosures <strong>of</strong> potential conflicts <strong>of</strong> interest are found at the end <strong>of</strong> this article.<br />
Address reprint requests to Cynthia Owusu, MD, MS, Case Western Reserve University<br />
School <strong>of</strong> Medicine, Division <strong>of</strong> Hematology/<strong>Oncology</strong> and Case Comprehensive Cancer<br />
Center-BHC 5055, 11100 Euclid Avenue, Cleveland, OH 44106; e-mail: cynthia.owusu@<br />
case.edu.<br />
© <strong>2012</strong> by <strong>American</strong> <strong>Society</strong> <strong>of</strong> <strong>Clinical</strong> <strong>Oncology</strong>.<br />
1092-9118/10/1-10<br />
3
Adjuvant Systemic Therapy<br />
Adjuvant systemic therapy refers to the administration <strong>of</strong><br />
anticancer therapy after primary breast surgery for early<br />
stage breast cancer with the goal <strong>of</strong> eradicating occult<br />
micrometastatic disease thought to be responsible for distant<br />
recurrence. Systemic treatment modalities include<br />
endocrine therapy and chemotherapy with or without trastuzumab.<br />
Treatment decision making regarding adjuvant<br />
systemic therapy for older women should involve consideration<br />
<strong>of</strong> such factors as the risk <strong>of</strong> morbidity and mortality<br />
from breast cancer, life expectancy and treatment tolerance,<br />
and patient preference. A geriatric assessment that includes<br />
evaluation <strong>of</strong> functional, cognitive, nutritional, and psychosocial<br />
status, and review <strong>of</strong> comorbidities and concomitant<br />
medications is particularly helpful in estimating the health<br />
status and life expectancy <strong>of</strong> older adults. Each <strong>of</strong> the<br />
components <strong>of</strong> a geriatric assessment can identify older<br />
adults at increased risk for morbidity and mortality. This<br />
assessment can also identify areas <strong>of</strong> vulnerability that may<br />
affect the patient’s ability to participate in a treatment plan<br />
(for example, ability to take medications on one’s own).<br />
Items in a geriatric assessment are able to identify older<br />
adults at risk <strong>of</strong> chemotherapy toxicity. 8,9 In a multisite<br />
study <strong>of</strong> 500 older adults with cancer, conducted by the<br />
Cancer and Aging Research Group, five geriatric assessment<br />
questions were independent predictors <strong>of</strong> the risk <strong>of</strong> chemotherapy<br />
toxicity, in addition to age, tumor, treatment, and<br />
laboratory variables. The geriatric assessment questions<br />
that were predictive <strong>of</strong> chemotherapy toxicity included hearing<br />
impairment (rated at fair or worse), difficulty in walking<br />
one block, need for assistance with taking medications, one<br />
or more falls in the past 6 months, and decrease in social<br />
activities because <strong>of</strong> either physical or emotional health. 8<br />
This predictive model for chemotherapy toxicity is being<br />
validated specifically in older adults receiving adjuvant<br />
chemotherapy for breast cancer (clinicaltrials.gov<br />
NCT01472094). In another study <strong>of</strong> 518 evaluable older<br />
adults with cancer (Chemotherapy Risk Assessment Scale<br />
for High Age Patients), predictors <strong>of</strong> chemotherapy toxicity<br />
included measures <strong>of</strong> functional status (ability to complete<br />
4<br />
KEY POINTS<br />
● The undertreatment <strong>of</strong> older women with breast<br />
cancer has contributed to poorer survival outcomes<br />
for older women than for younger women.<br />
● The principles that guide breast cancer treatment<br />
recommendations for younger and older women are<br />
fundamentally the same.<br />
● Breast cancer treatment decision making should consider<br />
the risk <strong>of</strong> breast cancer relapse, patient health<br />
status, life expectancy, and patient preference.<br />
● The use <strong>of</strong> systemic therapy in the adjuvant setting<br />
should be based on tumor characteristics and include<br />
endocrine therapy, and/or chemotherapy with or<br />
without trastuzumab.<br />
● Omission <strong>of</strong> postoperative radiation therapy is a<br />
reasonable strategy for older women with small<br />
hormone-receptor positive tumors.<br />
instrumental activities <strong>of</strong> daily living), cognition (Mini-<br />
Mental Status score), and nutrition (Mini-Nutritional Assessment<br />
score). 9 This information can be used in the<br />
treatment decision-making process in order to estimate life<br />
expectancy and risk <strong>of</strong> chemotherapy side effects and to<br />
identify areas <strong>of</strong> intervention to assist the patient during her<br />
treatment course.<br />
Endocrine Therapy<br />
OWUSU, HURRIA, AND MUSS<br />
The majority <strong>of</strong> older patients present with hormone<br />
receptor–positive breast cancer. Endocrine therapy is the<br />
mainstay <strong>of</strong> adjuvant therapy for these patients and results<br />
in proportional reductions in the risk <strong>of</strong> relapse and dying<br />
<strong>of</strong> breast cancer that exceed any available chemotherapy<br />
regimen. Current consensus guidelines recommend adjuvant<br />
systemic endocrine therapy for hormone receptorpositive<br />
breast cancer. The National Comprehensive Center<br />
Network (NCCN) guidelines 10 recommend the use <strong>of</strong> adjuvant<br />
endocrine therapy for women with hormone receptorpositive<br />
breast cancer regardless <strong>of</strong> age, menopausal status,<br />
or HER2 status, with the possible exception <strong>of</strong> women<br />
with lymph node-negative cancers 0.5 cm or less, or 0.6 to<br />
1.0 cm in diameter with favorable prognostic features;<br />
benefit from endocrine therapy is likely to be small for<br />
tumors <strong>of</strong> this size. In contrast, the St. Gallen International<br />
Consensus Panel recommends adjuvant systemic endocrine<br />
therapy for all women with endocrine-responsive disease<br />
with no exceptions and has defined endocrine-responsive<br />
tumors as those with as few as 1% <strong>of</strong> cells staining positive<br />
for hormone receptor proteins. 11 The <strong>Society</strong> <strong>of</strong> International<br />
Geriatric-<strong>Oncology</strong> (SIOG) breast cancer task force<br />
recommended that older women with endocrine-responsive<br />
breast cancer be <strong>of</strong>fered systemic endocrine therapy. However,<br />
for women with minimal risk disease, the decision to<br />
<strong>of</strong>fer endocrine therapy should be based on a risk-benefit<br />
analysis. 12<br />
Tamoxifen is the most firmly established adjuvant endocrine<br />
therapy for both premenopausal and postmenopausal<br />
women. Supporting these recommendations are results<br />
from the Early Breast Cancer Trialists’ Collaborative<br />
Group (EBCTCG) overview analysis, which demonstrated<br />
that over a 15-year period, use <strong>of</strong> adjuvant tamoxifen<br />
therapy for women with known estrogen receptor-positive<br />
disease, compared with no tamoxifen, decreased the 15-year<br />
risk <strong>of</strong> recurrence and death by 39% and 31%, respectively,<br />
regardless <strong>of</strong> age. 13 It is clear from these data that<br />
tamoxifen is <strong>of</strong> benefit for older women. In addition to<br />
decreasing the risk <strong>of</strong> disease relapse and death, there are<br />
also potential nonbreast cancer benefits <strong>of</strong> tamoxifen therapy<br />
in postmenopausal women. Tamoxifen may prevent<br />
osteoporosis 14 and reduce the risk <strong>of</strong> cardiovascular disease.<br />
15 Adjuvant tamoxifen therapy is, however, underutilized<br />
in older women. Women 80 years or older are half as<br />
likely to report having had a discussion about tamoxifen<br />
with their doctor compared with women 65 to 79 years, and<br />
women age 85 to 92 years are 25% less likely to receive a<br />
tamoxifen prescription than those 80 to 84 years. 16 Additionally,<br />
older women are more likely to self-discontinue and<br />
to be nonadherent to tamoxifen before the recommended<br />
treatment period <strong>of</strong> 5 years, 17,18 undercutting the treatment<br />
benefit from tamoxifen. Oncologists should always ask their<br />
patients if they are taking their prescribed medications and
TREATMENT FOR OLDER WOMEN WITH BREAST CANCER<br />
should reinforce the importance <strong>of</strong> adherence to maximize<br />
treatment benefit.<br />
The adjuvant use <strong>of</strong> aromatase inhibitors (anastrozole,<br />
letrozole, exemestane) for postmenopausal women with<br />
early breast cancer has been evaluated in several studies.<br />
These studies have involved the use <strong>of</strong> aromatase inhibitors<br />
either as initial adjuvant therapy, 19 as sequential therapy<br />
after 2 to 3 years <strong>of</strong> tamoxifen, 20 or as extended therapy<br />
after 4.5 to 6 years <strong>of</strong> tamoxifen. 21 The findings <strong>of</strong> these<br />
studies are consistent in demonstrating that the use <strong>of</strong> a<br />
third-generation aromatase inhibitor for postmenopausal<br />
women with hormone receptor–positive breast cancer, regardless<br />
<strong>of</strong> patient age, is superior in decreasing the risk <strong>of</strong><br />
disease recurrence, including ipsilateral and contralateral<br />
breast cancer, and distant metastatic disease compared with<br />
tamoxifen. Additionally, sequential use <strong>of</strong> aromatase inhibitors<br />
20 or extended therapy 21 has been shown to provide an<br />
overall survival advantage compared with tamoxifen use for<br />
5 years. No survival advantage has been demonstrated with<br />
the upfront use <strong>of</strong> aromatase inhibitors for 5 years compared<br />
with tamoxifen.<br />
There are differences in the toxicity pr<strong>of</strong>iles <strong>of</strong> aromatase<br />
inhibitors and tamoxifen. The incidence <strong>of</strong> venous thromboembolic<br />
disease, cerebrovascular events, endometrial cancer,<br />
vaginal bleeding, and hot flashes are less likely to be<br />
associated with aromatase inhibitors than tamoxifen,<br />
whereas the incidence <strong>of</strong> musculoskeletal pain, osteoporosis,<br />
and bone fractures have been found to be higher with<br />
aromatase inhibitors. 19 Emerging data also suggest that<br />
aromatase inhibitors may be associated with a small but<br />
higher risk <strong>of</strong> cardiovascular events compared with tamoxifen,<br />
22,23 but not compared with placebo. 24 In a metaanalysis<br />
<strong>of</strong> seven trials in which aromatase inhibitors were<br />
compared with tamoxifen, longer duration <strong>of</strong> aromatase<br />
inhibitor use or aromatase inhibitor use alone for 5 years<br />
was associated with a higher likelihood <strong>of</strong> cardiovascular<br />
events compared with tamoxifen alone (odds ratio [OR] �<br />
1.26, 95% CI 1.10–1.43). 25 Because studies <strong>of</strong> aromatase<br />
inhibitors have not included extensive follow-up, the full<br />
effect <strong>of</strong> aromatase inhibitors on cardiovascular disease and<br />
coronary heart risk remains to be determined.<br />
Although there is little evidence <strong>of</strong> age-related differences<br />
in the benefits <strong>of</strong> aromatase inhibitors for postmenopausal<br />
women, results <strong>of</strong> studies designed to examine age-related<br />
differences in the pattern <strong>of</strong> toxicity have been mixed. In<br />
general, the incidence <strong>of</strong> grade 3–5 nonfracture-related adverse<br />
events is higher among women 75 and older than<br />
among women less than 75 years. 26 However, a comparison<br />
<strong>of</strong> the quality <strong>of</strong> life and the side effect pr<strong>of</strong>ile for women who<br />
participated in MA-17, a study in which 5 years <strong>of</strong> letrozole<br />
was compared with placebo, showed no age-related differences<br />
in side effects. 24 The long-term consequences and<br />
implications <strong>of</strong> these side effects and any-age-related differences<br />
remain to be well characterized.<br />
Based on the results from recent studies that favor aromatase<br />
inhibitors over tamoxifen, current guidelines recommend<br />
that aromatase inhibitors should be <strong>of</strong>fered to all<br />
postmenopausal women with hormone receptor-positive<br />
early stage breast cancer, either alone, as sequential therapy<br />
after 2–3 years <strong>of</strong> tamoxifen. Given the lack <strong>of</strong> overall<br />
survival advantage associated with aromatase inhibitor use<br />
for 5 years, for women with pre-existing heart disease or<br />
bone loss, use <strong>of</strong> tamoxifen for 5 years or a switching<br />
strategy is a reasonable approach. For women with low<br />
grade, node-negative tumors 1 cm or smaller, endocrine<br />
therapy may be optional and observation acceptable, although<br />
the risks and benefits should be discussed with the<br />
patient.<br />
Adjuvant Chemotherapy<br />
Cytotoxic chemotherapy can be considered for older<br />
women with either node-positive or high-risk node-negative<br />
disease, particularly, triple-negative breast cancer. An<br />
abundance <strong>of</strong> literature has demonstrated the benefit <strong>of</strong><br />
adjuvant chemotherapy for younger women, with benefit<br />
decreasing as age increases. The EBCTCG overview analysis,<br />
13 which has 15 years <strong>of</strong> follow-up data, demonstrated<br />
that adjuvant chemotherapy reduced the annual risk <strong>of</strong><br />
recurrence by 37% and 19%, for women younger than 50 and<br />
50 to 69, respectively. The annual risk <strong>of</strong> death was reduced<br />
by 30% and 12%, for women younger than 50 and 50 to 69,<br />
respectively. The benefit <strong>of</strong> adjuvant chemotherapy for<br />
women with early stage breast cancer over age 70 was<br />
difficult to assess in the EBCTCG overview analysis<br />
because <strong>of</strong> the paucity <strong>of</strong> randomized trials that incorporated<br />
this age group. Of 29,000 women included in 60<br />
adjuvant polychemotherapy trials, 4% were 70 and older.<br />
This paucity <strong>of</strong> data has prevented definitive estimates<br />
regarding the magnitude <strong>of</strong> benefit <strong>of</strong> chemotherapy for<br />
women age 70 and older. In addition, with advancing age,<br />
organ function and performance status decline, and comorbidities<br />
increase, making the risks associated with chemotherapy<br />
even greater. Moreover, the risk reductions for<br />
chemotherapy are lower for postmenopausal women than for<br />
premenopausal women, although the reasons are unclear.<br />
Coupled with the apparent decline in the efficacy <strong>of</strong> chemotherapy<br />
with age is the increased risk <strong>of</strong> death from competing<br />
illnesses (comorbidity), leading to additional decline in<br />
the benefit from chemotherapy. As a result <strong>of</strong> all these<br />
factors, older women with early stage breast cancer receive<br />
adjuvant chemotherapy considerably less frequently than do<br />
younger women.<br />
However, a growing body <strong>of</strong> evidence suggests that adjuvant<br />
chemotherapy leads to improved survival outcomes<br />
for older women with breast cancer, particularly older<br />
women with hormone receptor-negative and node-positive<br />
breast cancer. A retrospective analysis <strong>of</strong> four Cancer and<br />
Leukemia Group B (CALGB) randomized clinical trials<br />
showed superior disease-free and overall survival benefits<br />
with the use <strong>of</strong> more aggressive chemotherapy (compared<br />
with less aggressive chemotherapy), among 6,487 women<br />
with node-positive breast cancer in all age groups, including<br />
70 and older. 27 This benefit, however, came at the cost <strong>of</strong><br />
increased risk <strong>of</strong> toxicity in older women, with older women<br />
more likely to discontinue treatment and to have an increased<br />
risk <strong>of</strong> treatment-related mortality. Additionally,<br />
data from large population studies have demonstrated a<br />
survival benefit from adjuvant chemotherapy for older<br />
women with hormone-negative or node-positive early stage<br />
breast cancer. 28,29<br />
Results from randomized controlled clinical trials that<br />
have specifically focused on older women with breast cancer,<br />
though scant, have helped to fill the gap on chemotherapy<br />
benefit for older women with early stage breast cancer.<br />
In the largest study in this population to date, a CALGB and<br />
Breast Cancer Intergroup study, 33 patients 65 and older<br />
5
with early-stage breast cancer were randomly assigned to<br />
either standard treatment (doxorubicin and cyclophosphamide<br />
[AC] for four cycles or cyclophosphamide, methotrexate,<br />
and 5-fluorouracil [CMF] for six cycles) or to an<br />
experimental arm <strong>of</strong> single-agent capecitabine for six cycles.<br />
30 At a median follow-up <strong>of</strong> 2.4 years, the relapse-free<br />
survival for patients receiving single-agent capecitabine was<br />
inferior to that for women receiving standard therapy (hazard<br />
ratio [HR] � 2.09 (1.38–3.17) p � 0 0.0001), as was<br />
overall survival (HR � 1.85 (1.11–3.08) p � 0.02). An<br />
unplanned subset analysis demonstrated that standard<br />
combination chemotherapy was particularly effective in<br />
patients with hormone receptor-negative disease. Overall,<br />
these results demonstrate that standard combination chemotherapy<br />
in the adjuvant setting, provides an overall<br />
survival benefit for older women with breast cancer, particularly<br />
those with hormone-receptor negative disease.<br />
The optimum chemotherapy regimen for treating breast<br />
cancer in the adjuvant setting remains debatable. Regardless,<br />
anthracycline-based regimens have become the norm,<br />
particularly for high-risk disease. However, the long-term<br />
complications associated with anthracycline use include,<br />
but are not limited to, dose-dependent cardiomyopathy. Age<br />
is a risk factor for cardiac disease, including anthracyclinerelated<br />
cardiomyopathy. Often, this precludes the use <strong>of</strong><br />
anthracycline-based regimens in older women with breast<br />
cancer. Jones and colleagues, 31 in a US <strong>Oncology</strong> trial,<br />
compared an anthracycline-based regimen (AC for four cycles)<br />
with a nonanthracycline taxane-based regimen docetaxel<br />
and cyclophosphamide (TC) for four cycles in 1,016<br />
women with node-negative and node-positive disease. At a<br />
median follow-up <strong>of</strong> 7 years, TC use was associated with<br />
superior disease-free survival (HR � 0.74, 95% CI (0.56–<br />
0.98) p � 0 0.03) and overall survival (HR � 0.69, 95%<br />
(0.50–0.97) p � 0 0.03) compared with AC. Sixteen percent<br />
<strong>of</strong> the study population were 65 and older. Unplanned<br />
subgroup analysis showed that TC was associated with<br />
superior disease-free and overall survival in all age groups,<br />
including older ages. TC therefore is a reasonable treatment<br />
regimen in the adjuvant setting for older women, particularly<br />
those with pre-existing heart disease and other contraindications<br />
to anthracycline-based regimens. A recent study<br />
has shown that this regimen is well tolerated in older<br />
women. 32 In another EBCTCG overview analysis in which<br />
different polychemotherapy regimens for early stage breast<br />
cancer were compared, adding a taxane to an anthracyclinebased<br />
chemotherapy regimen or a higher cumulative-dosage<br />
anthracycline-based regimen reduced breast cancer mortality,<br />
on average, by one-third. This benefit was irrespective<br />
<strong>of</strong> node status, tumor size, tumor grade, or age (�70). 33<br />
No definitive conclusion could be drawn regarding women<br />
70 and older because few women in that age group were<br />
included in the meta-analyses.<br />
For older women with node-negative, hormone-positive<br />
breast cancer, gene-expression pr<strong>of</strong>iling analysis can be used<br />
to identify women with high-risk disease who are likely to<br />
benefit from chemotherapy. The most widely used assay for<br />
this purpose is the 21-gene assay, which quantifies the<br />
likelihood <strong>of</strong> breast cancer recurrence in women with nodenegative,<br />
estrogen receptor-positive breast cancer and predicts<br />
the magnitude <strong>of</strong> endocrine therapy and chemotherapy<br />
benefit. 34 To the extent that the 21-gene assay allows for<br />
6<br />
individualization <strong>of</strong> cancer treatment, it is indeed a useful<br />
tool for the management <strong>of</strong> older patients with breast<br />
cancer, who were well represented in the validation cohorts<br />
(NSABP-14 and NSABP-20) for the assay. Moreover, the<br />
predictive ability <strong>of</strong> the 21-gene assay has been found to be<br />
independent <strong>of</strong> age. 34 Older patients with low scores are not<br />
likely to derive substantial benefit from chemotherapy,<br />
whereas those with high scores may derive great benefit.<br />
The benefit <strong>of</strong> adjuvant chemotherapy among patients with<br />
intermediate scores on the assay is being evaluated in the<br />
Tailor Rx study.<br />
Adjuvant! Online is another tool that can assist in decision<br />
making regarding the benefits <strong>of</strong> adjuvant endocrine therapy<br />
and chemotherapy for an individual patient. This online<br />
tool (available at www.adjuvantonline.com) summarizes the<br />
absolute benefit <strong>of</strong> chemotherapy and endocrine therapy,<br />
taking into account the patient’s age, brief assessment <strong>of</strong><br />
comorbid medical illnesses, and tumor characteristics. 35 To<br />
further inform this discussion, the risks associated with<br />
chemotherapy can be calculated with the predictive models<br />
for chemotherapy toxicity 8,9 as described earlier (See “Adjuvant<br />
Systemic Therapy”).<br />
Adjuvant Trastuzumab for HER2-Positive<br />
Breast Cancer<br />
OWUSU, HURRIA, AND MUSS<br />
Amplification or overexpression <strong>of</strong> HER2 is seen in approximately<br />
10% to 15% <strong>of</strong> invasive breast cancers in older<br />
women, 5 and it is associated with an unfavorable prognosis.<br />
A substantial body <strong>of</strong> literature from phase III trials in<br />
the adjuvant setting has demonstrated considerable benefit<br />
in disease-free and overall survival when trastuzumab is<br />
used either concurrently or sequential to chemotherapy<br />
compared with chemotherapy alone. 36-38 The main adverse<br />
effect associated with trastuzumab use is cardiotoxicity. In<br />
five phase III trials <strong>of</strong> adjuvant trastuzumab, the incidence<br />
<strong>of</strong> severe heart failure (New York Heart Association class III<br />
or IV), ranged from 0 to 3.9% among patients receiving<br />
trastuzumab, compared with 0 to 1.3% among patients who<br />
did not receive trastuzumab. 39 In the Breast Cancer International<br />
Research Group (BCIRG) 006 study, 38 two<br />
trastuzumab-containing regimens (AC plus docetaxel and<br />
trastuzumab and a nonanthracycline regimen <strong>of</strong> docetaxel,<br />
carboplatin, and trastuzumab [TCH]) were compared with<br />
standard chemotherapy alone. This study demonstrated<br />
disease-free and overall survival benefits with the use <strong>of</strong><br />
trastuzumab plus chemotherapy compared with chemotherapy<br />
alone. There was no substantial difference between the<br />
two trastuzumab-containing arms. Moreover, the incidence<br />
<strong>of</strong> cardiotoxicity associated with the nonanthracycline-based<br />
trastuzumab regimen was lower than that associated with<br />
the anthracycline-based trastuzumab regimen.<br />
Consistent with the under-representation <strong>of</strong> older women<br />
in breast cancer clinical trials <strong>of</strong> chemotherapy, older women<br />
have also been under-represented in clinical trials <strong>of</strong> trastuzumab<br />
therapy. With the notable exception <strong>of</strong> cardiac<br />
dysfunction, which was found to be associated with increasing<br />
age (older than 50), limitations in data collection precluded<br />
a determination <strong>of</strong> whether the toxicity pr<strong>of</strong>ile <strong>of</strong><br />
trastuzumab in older patients was different from that in<br />
younger patients. The reported clinical experience was also<br />
not adequate to determine whether the efficacy improvements<br />
(overall and disease-free survival) associated with<br />
trastuzumab in older patients was different from that in
TREATMENT FOR OLDER WOMEN WITH BREAST CANCER<br />
Endocrine-positive,<br />
HER2-negative<br />
Endocrine-negative,<br />
HER2-negative<br />
Table 1. Summary <strong>of</strong> Recommendations for Adjuvant Systemic Therapy in Early Stage Breast Cancer in Older Women<br />
patients younger than 65. Regardless, in the absence <strong>of</strong><br />
contraindications, trastuzumab is currently recommended<br />
for the adjuvant treatment <strong>of</strong> HER2-positive breast cancer,<br />
even for older women. In older women, a nonanthracycline,<br />
trastuzumab-containing regimen <strong>of</strong> TCH is <strong>of</strong>ten used because<br />
<strong>of</strong> the increased risk <strong>of</strong> cardiotoxicity associated with<br />
the anthracycline and trastuzumab regimen.<br />
In summary, the principles that govern recommendations<br />
for systemic adjuvant treatment <strong>of</strong> breast cancer are the<br />
same for younger and older women. Older women should<br />
be <strong>of</strong>fered guideline-recommended therapies (Table 1).<br />
Broadly, these recommendations should be <strong>of</strong>fered along<br />
three clinically distinct subgroups based on tumor characteristics.<br />
(1) Older women with hormone receptor-positive<br />
and HER2-negative breast cancer who should be <strong>of</strong>fered<br />
endocrine therapy regardless <strong>of</strong> node status. Gene expression<br />
pr<strong>of</strong>iling assay may be used to determine the added<br />
benefit <strong>of</strong> chemotherapy for those with node-negative<br />
disease; (2) older women with hormone receptor–negative<br />
and HER2-negative breast cancer (triple-negative breast<br />
cancer) who should be <strong>of</strong>fered adjuvant chemotherapy; and<br />
(3) older women with HER2-positive disease who should<br />
be <strong>of</strong>fered chemotherapy with trastuzumab. In the last<br />
group, women with hormone receptor-positive tumors<br />
should also be <strong>of</strong>fered endocrine therapy. Exceptions to<br />
these guidelines may be made for older women in any <strong>of</strong> the<br />
three subgroups who have node-negative disease and a<br />
tumor less than 1 cm or for frail older women with limited<br />
life expectancy, where close surveillance may be a reasonable<br />
alternative.<br />
Adjuvant Radiation Therapy<br />
Node-negative,<br />
Tumor size � 1cm<br />
No adjuvant therapy or<br />
Consider hormonal therapy if tumor size � 0.6 cm,<br />
grade 2, or other high risk features<br />
No adjuvant therapy or<br />
Consider chemotherapy if tumor size � 0.6 cm plus<br />
other high risk features<br />
HER2-positive No adjuvant therapy or<br />
Consider chemotherapy with trastuzumab if tumor<br />
size � 0.6 cm plus high risk features<br />
Until recently, the guideline recommendation, regardless<br />
<strong>of</strong> age, was for all women to receive radiation therapy<br />
after breast-conserving surgery and for postmastectomy<br />
radiation to be <strong>of</strong>fered to women with a high probability <strong>of</strong><br />
local recurrence. A recent meta-analysis <strong>of</strong> the EBCTCG<br />
supports these recommendations, showing that radiation<br />
therapy decreased the 10-year risk <strong>of</strong> any first recurrence<br />
from 35% to 19% and the 15-year risk <strong>of</strong> breast cancer death<br />
from 25% to 21% among women treated with breastconserving<br />
surgery. 40 Although the proportional reductions<br />
in relapse were similar among all women, the absolute<br />
benefits varied substantially by age, grade, estrogen receptor<br />
status, tamoxifen use, and extent <strong>of</strong> surgery. The authors<br />
concluded that radiation therapy after breast-conserving<br />
surgery halves the rate at which the disease recurs and<br />
reduces the breast cancer death rate by about a sixth.<br />
However, older women with small tumors can be spared<br />
radiation therapy. In a landmark randomized controlled<br />
study by Hughes and colleagues 41,42 636 women 70 or older<br />
who had undergone lumpectomy for stage I hormone<br />
receptor-positive breast cancer were randomly assigned to<br />
receive either radiation therapy and adjuvant tamoxifen<br />
for 5 years or to adjuvant tamoxifen for 5 years alone. The<br />
results demonstrated no substantial differences between<br />
the two groups with regard to mastectomy rates for local<br />
recurrence, distant metastases, or overall survival (median<br />
follow-up <strong>of</strong> 12 years). Of the 49% <strong>of</strong> patients who died<br />
during follow-up, 3% died <strong>of</strong> breast cancer. The only statistically<br />
significant difference was found in the rate <strong>of</strong> local<br />
or regional recurrence at 5 years, (2% among women who<br />
had radiation therapy compared with 9% who did not).<br />
Based on results from this study, one may reasonably<br />
consider lumpectomy (with surgically clear margins)<br />
without radiation therapy for women 70 and older with<br />
clinically negative lymph nodes, a tumor 2 cm or smaller,<br />
and hormone receptor-positive breast cancer who agree<br />
to take endocrine therapy. This strategy is limited by the<br />
high rate <strong>of</strong> nonadherence and early discontinuation <strong>of</strong><br />
adjuvant systemic endocrine therapy among older<br />
women. 43,44 Omission <strong>of</strong> postoperative radiation therapy,<br />
coupled with nonadherence to adjuvant systemic endocrine<br />
therapy, may result in earlier recurrences and, ultimately,<br />
poorer survival outcomes for older women. A favorable<br />
outcome from this approach can be achieved only when<br />
older women are adherent to prescribed oral endocrine<br />
therapies. Adherence to prescribed endocrine therapy<br />
should therefore be discussed and encouraged at follow-up<br />
visits.<br />
Conclusion<br />
Node-negative,<br />
Tumor size � 1 cm Node-positive<br />
Hormonal therapy<br />
Consider chemotherapy based on geneexpression<br />
pr<strong>of</strong>iling results<br />
Hormonal therapy<br />
Chemotherapy<br />
Chemotherapy alone Chemotherapy alone<br />
Chemotherapy with trastuzumab<br />
Add hormonal therapy if hormone<br />
positive<br />
Chemotherapy with trastuzumab<br />
Add hormonal therapy<br />
if hormone positive<br />
Breast cancer is a disease <strong>of</strong> aging. With minor differences,<br />
existing data support similar recommendations for<br />
both younger and older women. However, there are agerelated<br />
differences in treatment patterns, with older women<br />
less likely than younger women to receive standard therapies.<br />
Furthermore, survival outcomes lag behind that <strong>of</strong><br />
younger women. Closing the current gap in age-related<br />
disparities in breast cancer survival will require that<br />
older women are <strong>of</strong>fered the same state-<strong>of</strong>-the-art treatment<br />
as younger women, with a careful weighing <strong>of</strong> the risks<br />
and benefits <strong>of</strong> each treatment in the context <strong>of</strong> the individual’s<br />
preferences. Newer tools that estimate life<br />
expectancy and toxicity as well as the potential benefits <strong>of</strong><br />
therapy should make it easier for oncologists to make better<br />
treatment decisions with older patients. In addition, older<br />
women should be encouraged to participate in breast cancer<br />
7
clinical trials to generate additional efficacy and toxicity<br />
data. Such information will provide further knowledge so<br />
Authors’ Disclosures <strong>of</strong> Potential Conflicts <strong>of</strong> Interest<br />
Author<br />
Employment or<br />
Leadership<br />
Positions<br />
Consultant or<br />
Advisory Role<br />
Cynthia Owusu*<br />
Arti Hurria Amgen;<br />
Genentech; GTx<br />
Hyman Muss Boehringer<br />
Ingelheim; Eisai;<br />
Pfizer; Sandoz<br />
*No relevant relationships to disclose.<br />
1. Siegel R, Ward E, Brawley O, et al. Cancer statistics, 2011: the impact <strong>of</strong><br />
eliminating socioeconomic and racial disparities on premature cancer deaths.<br />
CA Cancer J Clin. 2011;61:212-236.<br />
2. DeSantis C, Siegel R, Bandi P, et al. Breast cancer statistics, 2011. CA<br />
Cancer J Clin. 2011;61:409-418.<br />
3. SEER stat fact sheets. http://seer.cancer.gov/csr/1975_2008/results_<br />
single/sect_01_table.11_2pgs.pdf. Accessed March 8, <strong>2012</strong>.<br />
4. Chu KC, Tarone RE, Kessler LG, et al. Recent trends in U.S. breast<br />
cancer incidence, survival, and mortality rates. J Natl Cancer Inst. 1996;88:<br />
1571-1579.<br />
5. Diab SG, Elledge RM, Clark GM. Tumor characteristics and clinical<br />
outcome <strong>of</strong> elderly women with breast cancer. J Natl Cancer Inst. 2000;92:<br />
550-556.<br />
6. Smith BD, Jiang J, McLaughlin SS, et al. Improvement in breast cancer<br />
outcomes over time: are older women missing out? J Clin Oncol. 2011;29:<br />
4647-4653.<br />
7. van de Water W, Markopoulos C, van de Velde CJ, et al. Association<br />
between age at diagnosis and disease-specific mortality among postmenopausal<br />
women with hormone receptor-positive breast cancer. JAMA. <strong>2012</strong>;<br />
307:590-597.<br />
8. Hurria A, Togawa K, Mohile SG, et al. Predicting chemotherapy toxicity<br />
in older adults with cancer: a prospective multicenter study. J Clin Oncol.<br />
2011;29:3457-3465.<br />
9. Extermann M, Boler I, Reich RR, et al. Predicting the risk <strong>of</strong> chemotherapy<br />
toxicity in older patients: The Chemotherapy Risk Assessment Scale for<br />
High-Age Patients (CRASH) score. Cancer. Epub 2011 Nov 9. doi: 10.1002/<br />
cncr.26646<br />
10. National Comprehensive Cancer Network. <strong>Clinical</strong> Practice Guidelines<br />
in Breast Cancer. Version 1. <strong>2012</strong>. www.nccn.org. Accessed February 28,<br />
<strong>2012</strong>.<br />
11. Goldhirsch A, Wood WC, Gelber RD, et al. Meeting highlights: updated<br />
international expert consensus on the primary therapy <strong>of</strong> early breast cancer.<br />
J Clin Oncol. 2003;21:3357-3365.<br />
12. Wildiers H, Kunkler I, Biganzoli L, et al. Management <strong>of</strong> breast cancer<br />
in elderly individuals: recommendations <strong>of</strong> the International <strong>Society</strong> <strong>of</strong><br />
Geriatric <strong>Oncology</strong>. Lancet Oncol. 2007;8:1101-1115.<br />
13. Effects <strong>of</strong> chemotherapy and hormonal therapy for early breast cancer<br />
on recurrence and 15-year survival: an overview <strong>of</strong> the randomised trials.<br />
Lancet. 2005;365:1687-1717.<br />
14. Love RR, Mazess RB, Barden HS, et al. Effects <strong>of</strong> tamoxifen on bone<br />
mineral density in postmenopausal women with breast cancer. N Engl J Med.<br />
1992;326:852-856.<br />
15. Love RR, Wiebe DA, Feyzi JM, et al. Effects <strong>of</strong> tamoxifen on cardiovascular<br />
risk factors in postmenopausal women after 5 years <strong>of</strong> treatment. J Natl<br />
Cancer Inst. 1994;86:1534-1539.<br />
16. Blackman SB, Lash TL, Fink AK, et al. Advanced age and adjuvant<br />
tamoxifen prescription in early-stage breast carcinoma patients. Cancer.<br />
2002;95:2465-2472.<br />
17. Partridge AH, Wang PS, Winer EP, et al. Nonadherence to adjuvant<br />
tamoxifen therapy in women with primary breast cancer. J Clin Oncol.<br />
2003;21:602–606.<br />
18. Owusu C, Buist DS, Field T, et al. Tamoxifen discontinuance among<br />
older women with estrogen-receptor-positive breast cancer. J Clin Oncol.<br />
2006;24: (suppl; abstr 648).<br />
19. Howell A, Cuzick J, Baum M, et al. Results <strong>of</strong> the ATAC (Arimidex,<br />
8<br />
that oncologists can <strong>of</strong>fer older women the best treatment<br />
options.<br />
Stock<br />
Ownership Honoraria<br />
REFERENCES<br />
Research<br />
Funding<br />
Abraxis<br />
BioScience;<br />
Celgene;<br />
GlaxoSmithKline<br />
OWUSU, HURRIA, AND MUSS<br />
Expert<br />
Testimony<br />
Other<br />
Remuneration<br />
Tamoxifen, Alone or in Combination) trial after completion <strong>of</strong> 5 years’<br />
adjuvant treatment for breast cancer. Lancet. 2005;365:60-62.<br />
20. Coombes RC, Hall E, Gibson LJ, et al. A randomized trial <strong>of</strong> exemestane<br />
after two to three years <strong>of</strong> tamoxifen therapy in postmenopausal women<br />
with primary breast cancer. N Engl J Med. 2004;350:1081-1092.<br />
21. Goss PE, Ingle JN, Martino S, et al. A randomized trial <strong>of</strong> letrozole in<br />
postmenopausal women after five years <strong>of</strong> tamoxifen therapy for early-stage<br />
breast cancer. N Engl J Med. 2003;349:1793-1802.<br />
22. Mouridsen H, Keshaviah A, Coates AS, et al. Cardiovascular adverse<br />
events during adjuvant endocrine therapy for early breast cancer using<br />
letrozole or tamoxifen: safety analysis <strong>of</strong> BIG 1-98 trial. J Clin Oncol.<br />
2007;25:5715-5722.<br />
23. Buzdar A, Howell A, et al. Comprehensive side-effect pr<strong>of</strong>ile <strong>of</strong> anastrozole<br />
and tamoxifen as adjuvant treatment for early-stage breast cancer:<br />
long-term safety analysis <strong>of</strong> the ATAC trial. Lancet Oncol. 2006;7:633-643.<br />
24. Muss HB, Tu D, Ingle JN, et al. Efficacy, toxicity, and quality <strong>of</strong> life in<br />
older women with early-stage breast cancer treated with letrozole or placebo<br />
after 5 years <strong>of</strong> tamoxifen: NCIC CTG intergroup trial MA.17. J Clin Oncol.<br />
2008;26:1956-1964.<br />
25. Amir E, Seruga B, Niraula S, et al. Toxicity <strong>of</strong> adjuvant endocrine<br />
therapy in postmenopausal breast cancer patients: a systematic review and<br />
meta-analysis. J Natl Cancer Inst. 2011;103:1299-1309.<br />
26. Crivellari D, Sun Z, Coates AS, et al. Letrozole compared with tamoxifen<br />
for elderly patients with endocrine-responsive early breast cancer: the<br />
BIG 1-98 trial. J Clin Oncol. 2008;26:1972-1979.<br />
27. Muss HB, Woolf S, Berry D, et al. Adjuvant chemotherapy in older and<br />
younger women with lymph node-positive breast cancer. JAMA. 2005;293:<br />
1073-1081.<br />
28. Du XL, Jones DV, Zhang D. Effectiveness <strong>of</strong> adjuvant chemotherapy for<br />
node-positive operable breast cancer in older women. J Gerontol A Biol Sci<br />
Med Sci. 2005;60:1137-1144.<br />
29. Elkin EB, Hurria A, Mitra N, et al. Adjuvant chemotherapy and<br />
survival in older women with hormone receptor-negative breast cancer:<br />
assessing outcome in a population-based, observational cohort. J Clin Oncol.<br />
2006;24:2757-2764.<br />
30. Muss HB, Berry DA, Cirrincione CT, et al. Adjuvant chemotherapy in<br />
older women with early-stage breast cancer. N Engl J Med. 2009;360:2055-<br />
2065.<br />
31. Jones S, Holmes FA, O’Shaughnessy J, et al. Docetaxel with cyclophosphamide<br />
is associated with an overall survival benefit compared with doxorubicin<br />
and cyclophosphamide: 7-year follow-up <strong>of</strong> US <strong>Oncology</strong> Research<br />
Trial 9735. J Clin Oncol. 2009;27:1177-1183.<br />
32. Freyer G, Campone M, Peron J, et al. Adjuvant docetaxel/cyclophosphamide<br />
in breast cancer patients over the age <strong>of</strong> 70: results <strong>of</strong> an observational<br />
study. Crit Rev Oncol Hematol. 2011;80:466-473. Epub 2011 May 11.<br />
33. Peto R, Davies C, et al. Comparisons between different polychemotherapy<br />
regimens for early breast cancer: meta-analyses <strong>of</strong> long-term<br />
outcome among 100,000 women in 123 randomised trials. Lancet. <strong>2012</strong>;379:<br />
432-444.<br />
34. Paik S, Shak S, Tang G, et al. A multigene assay to predict recurrence<br />
<strong>of</strong> tamoxifen-treated, node-negative breast cancer. N Engl J Med. 2004;351:<br />
2817-2826.<br />
35. Ravdin PM, Simin<strong>of</strong>f LA, Davis GJ, et al. Computer program to assist<br />
in making decisions about adjuvant therapy for women with early breast<br />
cancer. J Clin Oncol. 2001;19:980-991.<br />
36. Piccart-Gebhart MJ, Procter M, Leyland-Jones B, et al. Trastuzumab
TREATMENT FOR OLDER WOMEN WITH BREAST CANCER<br />
after adjuvant chemotherapy in HER2-positive breast cancer. N Engl J Med.<br />
2005;353:1659-1672.<br />
37. Romond EH, Perez EA, Bryant J, et al. Trastuzumab plus adjuvant<br />
chemotherapy for operable HER2-positive breast cancer. N Engl J Med.<br />
2005;353:1673-1684.<br />
38. Slamon D, Eiermann W, Robert N, et al. Adjuvant trastuzumab in<br />
HER2-positive breast cancer. N Engl J Med. 2011;365:1273-1283.<br />
39. Ewer SM, Ewer MS. Cardiotoxicity pr<strong>of</strong>ile <strong>of</strong> trastuzumab. Drug Saf.<br />
2008;31:459-467.<br />
40. Darby S, McGale P, et al. Effect <strong>of</strong> radiotherapy after breast-conserving<br />
surgery on 10-year recurrence and 15-year breast cancer death: meta-analysis<br />
<strong>of</strong> individual patient data for 10,801 women in 17 randomised trials. Lancet.<br />
2011;378:1707-1716.<br />
41. Hughes KS, Schnaper LA, Berry D, et al. Lumpectomy plus tamoxifen<br />
with or without irradiation in women 70 years <strong>of</strong> age or older with early<br />
breast cancer. N Engl J Med. 2004;351:971-977.<br />
42. Hughes KS, Schnaper LA, Cirrincione C, et al. Lumpectomy plus<br />
tamoxifen with or without irradiation in women age 70 or older with early<br />
breast cancer. J Clin Oncol. 2010;28:15s (suppl; abstr 507).<br />
43. Owusu C, Buist DS, Field TS, et al. Predictors <strong>of</strong> tamoxifen discontinuation<br />
among older women with estrogen receptor-positive breast cancer.<br />
J Clin Oncol. 2008;26:549-555.<br />
44. Partridge AH, LaFountain A, Mayer E, et al. Adherence to initial<br />
adjuvant anastrozole therapy among women with early-stage breast cancer.<br />
J Clin Oncol. 2008;26:556-562.<br />
9
Management <strong>of</strong> Small T1a/b N0<br />
Breast Cancers<br />
Overview: T1ab N0 breast cancer generally has excellent<br />
prognosis. Adverse prognostic factors include HER2� disease,<br />
ER-negative disease, high-grade histology, T1b, and<br />
young age <strong>of</strong> patient. These patients are largely excluded from<br />
most trials, and to date, no prospective studies for this group<br />
THE INCIDENCE <strong>of</strong> small (� 1 cm) node-negative breast<br />
cancers (BC) is increasing in mammography-screened<br />
populations. 1 About two-thirds <strong>of</strong> all breast cancers (BCs) in<br />
the Surveillance, Epidemiology, and End Results (SEER)<br />
registry are T1; approximately 72% and 28% are T1b and<br />
T1a, respectively. 1<br />
Approximately 50% to 60% <strong>of</strong> T1abN0 human epidermal<br />
growth factor receptor 2 (HER2)-positive tumors are estrogen<br />
receptor (ER) positive. 2-4 Approximately 10% to 15% <strong>of</strong><br />
patients with T1abN0 have HER2-positive disease. 2-9 Several<br />
studies have reported that T1mic and T1a are more<br />
likely to be ER negative and HER2 positive compared with<br />
T1b, although an early study <strong>of</strong> OncotypeDx suggested that<br />
among the ER-positive tumors, a greater percentage <strong>of</strong> the<br />
T1b had low recurrence score (RS). 4,10,11 In the Kwon report,<br />
HER2 was positive in 46%, 23%, and 13%, and ER positive<br />
in 42%, 66%, and 78% <strong>of</strong> T1mic, T1a, and T1b, respectively. 4<br />
Small node-negative BC are generally associated with<br />
excellent outcomes, with disease-free (DFS) and relapse-free<br />
survivals (RFS) in excess <strong>of</strong> 90%, and with survivals exceeding<br />
95% at 5 to 10 years. Older studies have shown 10-year<br />
RFS between 83% and 91%. 12-14 However, these tumors are<br />
quite heterogeneous, and some subsets <strong>of</strong> tumors have a far<br />
higher risk <strong>of</strong> relapse. Because these patients are generally<br />
not included in clinical trials, outcomes are poorly characterized,<br />
management is uncertain, and treatment varies<br />
considerably from one oncologist to another.<br />
This article will review the outcomes literature for T1ab<br />
N0M0 BCs, focusing on articles presented in the last decade,<br />
at a time where molecular testing <strong>of</strong> ER, progesterone<br />
receptor (PR), and HER2 have become routine. For more<br />
details <strong>of</strong> earlier literature, refer to the excellent review by<br />
Hanrahan et al. 15 Because <strong>of</strong> a recent spike in interest in the<br />
oncology community, the majority <strong>of</strong> articles on this topic<br />
have been published in manuscript or abstract form in 2010<br />
and 2011.<br />
T1a tumors are 0.5 cm or smaller, while T1b are larger<br />
than 0.5 to 1 cm or smaller. Typical definitions include RFS:<br />
date <strong>of</strong> diagnosis or initial surgery to date <strong>of</strong> recurrence<br />
(distant or local), death from any cause, or last follow-up.<br />
DFS includes contralateral new BCs as events. Distant DFS<br />
only includes metastatic events.<br />
Literature Review<br />
Studies in Which 10% <strong>of</strong> Patients Received Adjuvant Chemotherapy<br />
and No Trastuzumab<br />
Fisher et al presented the National Surgical Adjuvant<br />
Breast and Bowel Project (NSABP) experience with T1abN0<br />
BC in the B-06, B-13, B-14, B-19, and B-20 studies (Table<br />
1). 16 Median follow-up was 8 years. For the 61 patients who<br />
10<br />
By Anthony D. Elias, MD<br />
yet reported. Treatment guidelines are vague and treatment<br />
inconsistent. As yet, in the HER2� population, little experience<br />
with targeted therapy has been reported. Prospective<br />
trials are needed.<br />
were chemotherapy naive with ER-negative BC, the 8-year<br />
RFS was only 79% with an overall survival (OS) <strong>of</strong> 93%. A<br />
90% 8-year RFS was achieved when chemotherapy was<br />
given. For the 264 patients with ER-positive BC, an 8-year<br />
RFS <strong>of</strong> 86% and an OS <strong>of</strong> 90% were achieved with surgery<br />
alone. For those treated with tamoxifen, a 93% 8-year RFS<br />
was noted.<br />
Wood et al studied the outcomes <strong>of</strong> 282 consecutive women<br />
with T1abN0 BCs at Emory. 17 The mean follow-up was 7<br />
years. Tamoxifen and chemotherapy was administered to 93<br />
and 20 patients, respectively. One patient had local recurrence<br />
and two suffered metastatic spread despite tamoxifen<br />
for an estimated 10-year distant DFS (DDFS) <strong>of</strong> 98.7%.<br />
Joensuu described a cohort <strong>of</strong> 852 Finnish patients with<br />
T1N0 BC, <strong>of</strong> whom 12% had HER2-positive disease and 75%<br />
had ER-positive disease. 18 Median follow-up was 114<br />
months. Three hundred and thirteen patients had T1ab<br />
disease with a 9-year DFS <strong>of</strong> 93%; for T1a 100%. Nine-year<br />
DDFS was much better for HER2-negative compared with<br />
HER2-positive BCs (89% vs. 73%, p � 0.0003). In the T1b<br />
group, this difference was greater (95% vs. 67%). Thirty<br />
patients had TN BC, <strong>of</strong> whom nine received chemotherapy.<br />
No relapses were observed. Adverse prognostic factors included<br />
T size (in mm), HER2 positivity, Ki67 greater than<br />
20%, and possibly grade 2 to 3 disease.<br />
Fisher et al reported on the outcomes <strong>of</strong> the patients with<br />
T1abN0 disease from the NSABP B21 study who were<br />
randomly selected to receive lumpectomy plus tamoxifen,<br />
lumpectomy plus radiation therapy, or all three. 19 Median<br />
follow-up was 87 months. HER2 status was unknown. Twothirds<br />
received endocrine therapy, and the remainder received<br />
observation. Radiation therapy reduced ipsilateral<br />
breast tumor recurrence (IBTR; hazard ratio [HR] � 0.19) as<br />
did tamoxifen (HR � 0.37). Eight-year DFS for the best arm<br />
(preoperative cetuximab and radiation [XRT] plus tamoxifen)<br />
was 84.4%. An update was provided in 2007 with<br />
14-year follow-up. 20 Adverse prognostic factors were grade<br />
3, nontubular histology, and, for survival, lymphovascular<br />
invasion (LVI).<br />
Chia et al reported the outcomes <strong>of</strong> 2,026 patients in<br />
British Columbia with N0 BC treated in 1986 to 1992. 21<br />
Seventy percent had no adjuvant therapy. Ten-year RFS<br />
was 76% (and only 66% in the HER2� population [10% <strong>of</strong><br />
From the University <strong>of</strong> Colorado Cancer Center, Anschutz Medical Campus, Aurora, CO.<br />
Author’s disclosures <strong>of</strong> potential conflicts <strong>of</strong> interest are found at the end <strong>of</strong> this article.<br />
Address reprint requests to Anthony D. Elias, MD, University <strong>of</strong> Colorado Cancer Center,<br />
Anschutz Medical Campus, MS 8117, Research Tower South, 12801 East 17th Avenue,<br />
Room 8111, Aurora, CO 80045; email: anthony.elias@ucdenver.edu.<br />
© <strong>2012</strong> by <strong>American</strong> <strong>Society</strong> <strong>of</strong> <strong>Clinical</strong> <strong>Oncology</strong>.<br />
1092-9118/10/1-10
MANAGEMENT OF T1 BREAST CANCERS<br />
the group]). Of the 307 T1abN0 HER2-negative BCs, 10-year<br />
RFS was approximately 94%. Of the 21 T1abN0 HER2positive<br />
BCs, the 10-year RFS was approximately 85%. T1b<br />
fared a bit worse than T1a.<br />
Gonzalez-Angulo et al analyzed a University <strong>of</strong> Texas<br />
M. D. Anderson Cancer Center (MDACC) dataset (later<br />
reported by Theriault et al 9 ) plus a confirmatory dataset<br />
from the Institutes Bordet and Leoben. 22 Nine hundred and<br />
sixty five patients from MDACC and 350 patients from the<br />
other two institutions with T1abN0 were included. Endocrine<br />
therapy was given to 55% <strong>of</strong> the MDACC patients;<br />
treatment details were sparse from the other group. Overall<br />
5-year RFS was 92% and 96% for the two groups but only<br />
77% and 87% for the HER2-positive subset. Adverse prognostic<br />
factors included HER2-positivity, ER-negativity, age<br />
younger than 50, and grade 3 tumor. Tumor size was not a<br />
factor. The TN group had a 5-year RFS <strong>of</strong> 83% at MDACC,<br />
but it was substantially better (95%) at the other institutions.<br />
It is uncertain whether this was a result <strong>of</strong> differences<br />
in treatment.<br />
Theriault et al summarized the MDACC experience for<br />
1,012 patients with T1abN0 BC in 1990 to 2002. 9 Median<br />
follow-up was roughly 60 months. Patients who had received<br />
chemotherapy or trastuzumab were excluded from this article.<br />
Of the 771 patients with ER-positive BC, 61% received<br />
endocrine therapy. Of the 98 and 143 patients with HER2positive<br />
and TN BC, endocrine therapy was given to 46%<br />
and 21%, respectively. Five-year DDFS was 97.5%, 86.9%,<br />
and 96.3% for the patients with HR-positive, HER2-positive,<br />
and TN disease, respectively. Five-year RFS was 94.5%,<br />
77.2%, and 84.9%, respectively. Adverse prognostic factors<br />
included age younger than 35, HER2 positivity, and less so<br />
TN BC.<br />
Cancello et al reported on the European Institute <strong>of</strong><br />
<strong>Oncology</strong> experience for patients with T1mic/a/bN0 BC<br />
treated in 1997 to 2005. 8 It is not clear how these patients<br />
overlap with a prior report from Curigliano et al. 23 Median<br />
follow-up was 76 months for the 1,691 patients. Overall<br />
treatment consisted <strong>of</strong> endocrine therapy in 35% (93% <strong>of</strong> all<br />
ER�) and chemotherapy in 10% (54% and 71% <strong>of</strong> HER2�<br />
and TN BC). No trastuzumab was administered. Using<br />
cut<strong>of</strong>fs <strong>of</strong> 1% for ER/PR and 14% for Ki67, patients were<br />
categorized by IHC to have BCs that were luminal A (881,<br />
52%), luminal B (532, 31%), luminal B HER2 positive (101,<br />
6%), HER2 positive/ER negative (82, 5%), and TN (95, 6%).<br />
Five-year BC-specific survival was approximately 98%, 97%,<br />
95%, 91%, and 91%, respectively. Adverse prognostic factors<br />
included age younger than 35, HER2 positivity, TN, and<br />
LVI.<br />
KEY POINTS<br />
● T1ab N0 breast cancer generally has excellent prognosis.<br />
● Adverse prognostic factors include HER2� disease,<br />
ER� disease, high-grade histology, T1b, and younger<br />
age.<br />
● Vague treatment guidelines exist.<br />
● No prospective trials have yet been reported.<br />
● For the HER2� population, there is almost no trastuzumab<br />
experience in the absence <strong>of</strong> chemotherapy.<br />
Rouanet et al reported the outcomes <strong>of</strong> 703 French patients<br />
with T1abN0 BC treated in 1999 to 2004. 10 Only 6%<br />
had HER2-positive disease. Of these, 50% had dual positive<br />
disease. Treatment consisted <strong>of</strong> endocrine therapy in 41%,<br />
chemotherapy in 10%, and observation in 55%, which<br />
achieved a 5-year DFS <strong>of</strong> only 74%. For the 661 patients<br />
with HER2-negative disease, 10% had TN disease and 90%<br />
ER-positive disease. Treatment for this group was endocrine<br />
for 80%, chemotherapy for 5%, and observation for 17%,<br />
which achieved a 5-year DFS <strong>of</strong> 95%. Adverse prognostic<br />
factors included HER2-positivity, ER-negativity, age<br />
younger than 50, no adjuvant therapy, and grade 3 disease.<br />
Five-year DFS for the TN cohort was 91%.<br />
Livi et al presented 704 Florentines with T1abN0 BC<br />
treated in 2002 to 2008. 2,24 Median follow-up was 59<br />
months. No chemotherapy or trastuzumab was given; endocrine<br />
therapy was given in 64%. Five hundred and fifty-nine<br />
had ER-positive/HER2-negative disease, 75 had HER2positive<br />
disease—<strong>of</strong> whom 55% had ER-positive disease—<br />
and 70 had TN BC. 24 Five-year DDFS was 97.8%, 92%, and<br />
91.8%, respectively. Overall 5-year DFS was 96.5%. Adverse<br />
prognostic factors included age 50 or younger and ERnegative<br />
and HER2-positive disease.<br />
Studies in Which >10% Patients Received Adjuvant Chemotherapy,<br />
but No Trastuzumab<br />
Colleoni et al reported on a cohort <strong>of</strong> patients with T1mic/<br />
a/b N0 disease from the European Institute <strong>of</strong> <strong>Oncology</strong><br />
treated in 1997 to 2001. 14 Median follow-up was only 43<br />
months. Of the 425 patients, 358 had ER-positive BC (11%<br />
HER2�), and 60 had ER-negative BC (40% HER2�, 60%<br />
TN). Most received adjuvant therapy. Of the patients who<br />
were ER positive, only six relapsed following treatment<br />
(endocrine 88%, chemotherapy 12%, observation 11%). Of<br />
the patients who were ER negative, only three relapsed<br />
following treatment (chemotherapy 60%, observation 40%).<br />
Adverse prognostic factors included Ki67 greater than 20%,<br />
age younger than 35, and PR-negative disease.<br />
Hanrahan et al analyzed the outcomes <strong>of</strong> patients with<br />
T1abN0 disease in the SEER registry from 1988 to 2001. 26<br />
Treatment details are not available. By 10 years, only<br />
approximately 4% died <strong>of</strong> BC, whereas an additional 20%<br />
died <strong>of</strong> other causes, mostly cardiovascular. For OS, prognostic<br />
factors included black race, T1b, and PR-negative<br />
disease. For each mm tumor size, the HR for OS was 1.02.<br />
Prognostic factors for BC-specific mortality were age<br />
younger than 50, grade 3, ER negativity, PR negativity, and<br />
fewer than six lymph nodes in the axillary lymph node<br />
dissection (ALND).<br />
Curigliano et al reviewed 2,130 patients with T1abN0 BC<br />
treated at the European Institute <strong>of</strong> <strong>Oncology</strong> in 1999 to<br />
2006. 23 The extent <strong>of</strong> overlap with Cancello’s presentation is<br />
not clear. Median follow-up was 55 months. A subset had<br />
immunohistochemical (IHC) testing: 79 with HER2-positive/<br />
ER-positive disease and 71 with HER2-positive/ER-negative<br />
disease; 158 patients with ER-positive/HER2-negative BC<br />
were matched by HR status, age, and year <strong>of</strong> surgery.<br />
Treatment for the ER-positive/HER2-negative group was<br />
endocrine in 93%, chemotherapy in 1% and observation in<br />
6% resulting in a 5-year DFS <strong>of</strong> 99%. Treatment for the dual<br />
positive group was endocrine in 66% and chemotherapy in<br />
25%, resulting in a 5-year DFS <strong>of</strong> 92%. Treatment for the<br />
HER2-positive/ER-negative group was endocrine in 3%, che-<br />
11
Table 1. T1ab N0 Breast Cancer Literature Review (Papers Published 2002–2011)<br />
Pt Accrual F/U<br />
Types <strong>of</strong> Treatment<br />
Author Ref(s) Institution Year Yrs (mos) Stage # Pts %ER� %HER2 %OBS %ET %Chemo %anti HER2<br />
Fisher 16 NSABP B21 2002 1989–1998 87 T1abN0 1009 57% (30% unk) nd 33 67 0 0<br />
332 57 XRT only 0<br />
334 54 XRT 100<br />
334 59 no XRT 100<br />
Wood 17 Emory 2002 ? 84 mean T1abN0 282 NR nd NR 33 7 0<br />
Joensuu 18 Finland 2003 1991–1992 114 T1N0 852 75 12 95<br />
T1ab 313<br />
TN 30 70 30<br />
HR�HER2- 396*<br />
*#s don’t add up<br />
93 7 0 0<br />
Fisher 19,20 NSABP 2007 1989–1998 134 T1abN0 638 n/a nd 32 68 0 0<br />
Subset <strong>of</strong><br />
204 XRT only 0<br />
1009<br />
228 XRT 100<br />
in B21<br />
206 no XRT 100<br />
Chia 21 British<br />
Columbia<br />
Gonzalez-<br />
Angulo<br />
2008 1986–1992 N0 2026 10 70 0<br />
T1ab HER2- 307 73% any T 82<br />
T1ab HER2� 21 39% any T 76<br />
9,22 MDACC 2009 1990–2002 74 T1abN0 965 77<br />
86<br />
61<br />
10 45 55 0 0<br />
Bordet<br />
Leoben<br />
350 48 6<br />
Theriault 9 MDACC 2011 1990–2002 �60 T1abN0 1012 76 10 0 0<br />
ER� 771 39 61<br />
HER2� 98 54 46<br />
TN 143 79 21<br />
Cancello 8 Eur Inst Onc 2011 1997–2005 76 T1m/a/bN0 1691 90 11 NR 35 10 0<br />
Luminal A 881 100 0 92 1<br />
Luminal B 532 100 0 94 7<br />
Luminal B HER2� 101 100 100 94 7<br />
HER2� 82 0 100 1 54<br />
TN 95 0 0 3 71<br />
Rouanet 10 ONCO 2011 1999–2004 77 T1abN0 703* 93 6 20 83 6 0<br />
T1a 131 77 14<br />
T1b 572 97 4<br />
424 others* unk<br />
661 90 0 17 80 5 0<br />
42 50 100 55 36 5 0<br />
Livi 2,24 Florence 2011 2002–2008 59 T1a/bN0 704 84 36 64 0 0<br />
T1a/b H� 75 55<br />
T1a/b TN 70 0<br />
Colleoni 16 Eur Inst<br />
<strong>Oncology</strong><br />
ANTHONY D. ELIAS<br />
2004 1997–2001 43 T1mic/a/b N0 425 84 13 16 66 16 0<br />
ER� 358 11 88 12 0<br />
ER- 60 40 0 60 0<br />
Hanrahan 26 SEER 2007 1988–2001 64 T1abN0 51,246 58% (30% unk) nd NR NR NR 0<br />
Curigliano 23 Eur Inst Onc 2009 1999–2006 55 T1abN0 2130 7 0<br />
ER�/HER2� 79 10 89 25 0<br />
matched ER�/H- 158 6 94 1 0<br />
ER-/HER2� 71 54 4 43 0<br />
Matched ER-/H- 71 31 3 66 0<br />
12
MANAGEMENT OF T1 BREAST CANCERS<br />
Table 1. T1ab N0 Breast Cancer Literature Review (Papers Published 2002–2011) (Cont’d)<br />
Time %DFS %RFS<br />
Measure <strong>of</strong> Efficacy<br />
%DDFS %OS %IBTR Adverse Factors Comments HER2 Test<br />
8y �71% T1b HER2 unk<br />
81.6 96.7 94 90.7 IBTR: HR 0.19 for XRT;<br />
84.4 98.5 93.4 97.8<br />
HR 0.37 for Tam<br />
77.8 96.8 94 83.5<br />
7 y 98.7 HER2 unk<br />
4.2 y size, HER2�, Ki67 � 49 T1a 100% DFS IHC 3� 10%<br />
93<br />
100<br />
20%, G � 1<br />
or CISH<br />
DFS: high nuclear IBTR: nuclear grade, HER2 unk<br />
grade, non-tubular DCIS present<br />
60.6 82.1 89.2 OS: LVI<br />
56 77.8 89.8<br />
61.5 82.2 80.5<br />
10 y 66 versus<br />
76 HER2<br />
71 versus<br />
82 HER2<br />
IHC 3� 10%<br />
or FISH<br />
�94 T1b a bit worse SP3 Ab<br />
�85<br />
5 y 92 96 HER2�, ER-, age<br />
� 50, high grade<br />
ER-5y RFS 83.3% IHC 3� 10%<br />
or FISH<br />
94 97 HER2-<br />
77 (HR 2.7) 86 (HR 5.3) HER2�<br />
HR 3.9 HR 2.8 TN<br />
96 5y RFS TN 95% (n � 125);<br />
no diff T1a versus T1b<br />
IHC 3� 10%<br />
or FISH<br />
97 HER2-<br />
87 HER2�<br />
5 y 91.5 96.3 94.6 Age � 35, HER2� worse Fig 1 IHC 3� 10%<br />
94.5 97.5 than TN<br />
excluded CT and trastuzumab or FISH<br />
77.2 (HR 4.98) 86.9 (HR 4.7)<br />
84.9 (HR 2.71) 96.3 (HR 2.1)<br />
more T1a<br />
Age � 35, HER2� then<br />
TN, LVI<br />
ER/PR at 1% threshold;<br />
Ki67 14%<br />
5 y BCS �98 Fig 1<br />
�97<br />
�95<br />
�91 Tmic overrepresented (27%)<br />
�91<br />
5 y 97 HER2�, ER-, � 50 y,<br />
no adj therapy, G3<br />
95<br />
74<br />
IHC 3� 10%<br />
T1a more HER2�, ER- IHC 3� or DISH<br />
50% HER� also ER�<br />
TN 5 y DFS 91%<br />
5 y 96.5 93.7 1 age � 50; ER-, HER2� HR 3.66 for HER2� DFS IHC 10% 3�<br />
92 90.3 1 63% T1b<br />
or FISH<br />
91.8 1<br />
4y Ki67 � 20% HR 12.9; 76% T1b; more G3 in IHC 3� 10%<br />
6 events<br />
3 events<br />
� 35, PR-<br />
T1mic, no events in T1mic<br />
10 y HR 1.02 for<br />
BCSM: � 50, G3, �4% died <strong>of</strong> BC;<br />
HER2 unk<br />
1 mm T size<br />
ER-, PR-, � 6LN in<br />
ALND; OS: � 50,<br />
AA, T1b, PR-<br />
24% all-cause<br />
5 y HR 2.4 IHC 3� 10%<br />
92 (HR 5.2) HR 5.1 for HER2� T1a DFS 88%; T1b 95% or FISH<br />
99 T1a 97%; T1b 99%<br />
91 T1b, multifocal T1a 93%; T1b 85%<br />
92 T1a 87%; T1b 94%<br />
13
Table 1. T1ab N0 Breast Cancer Literature Review (Papers Published 2002–2011) (Cont’d)<br />
ANTHONY D. ELIAS<br />
Pt Accrual F/U<br />
Types <strong>of</strong> Treatment<br />
Author Ref(s) Institution Year Yrs (mos) Stage # Pts %ER� %HER2 %OBS %ET %Chemo %anti HER2<br />
Kwon 4 Seoul 2010 2000–2006 61 T1mic/a/bN0 375 67 26 26 62 16 0<br />
T1mic 120 42 46<br />
T1a 93 66 23<br />
T1b 162 78 13<br />
HER2�HR� 31<br />
HER2�HR- 65 34% <strong>of</strong> ER-<br />
TN 56<br />
Park 27 Seoul 2010 1994–2004 61 T1abN0 370 82 9 (not dual �) NR 74 36 0<br />
T1abN1 57(13%)<br />
289 HR�/HER2 any 94 25<br />
54 ER�/HER2�<br />
31 HER2�/ER- 65<br />
33 TN 64<br />
Lai 28 Taiwan 2011 1995–2006 48 T1ab 377 65 29<br />
N0 or Nx 323<br />
TN 47 <strong>of</strong> 311 36<br />
non-TN 18<br />
Tanaka 5 Japan 2011 2001–2007 46 T1N0 454 17 0<br />
T1N0 H- 376 NR 77 14 0<br />
T1N0 H� 78 NR 44 45 0<br />
T1a/b H� 28<br />
Fehrenbacher 29 KPNC 2010 2000–2006 70 T1abN0 HER2� 237 59 100 NR NR 25 8.4<br />
T1a 116 52 12.9 6.8<br />
T1b 121 65 36.3 9.9<br />
Amar 30 Mayo 2010 2001–2005 34 T1abN0 421 87 7 NR 44 6 1<br />
ER� 364 48 3<br />
HER2� 28 50 36 39 18<br />
TN 29 4 28<br />
Horio 7 Nagoya 2011 2003–2007 52 T1a/b 267 86 16 37 55 9 0<br />
H- 225 36 62 4<br />
T1a 42<br />
T1b 183<br />
H� 42 36 41 24 43 12<br />
T1a 23 48 17 35 9<br />
T1b 19 32 32 53 16<br />
TN 10<br />
Wong 3 Singapore 2011 1989–2009 57 T1N0 519 17<br />
subset T1N0 HER2� 89<br />
subset T1c 315 18<br />
subset T1a/b H- 170 NR 78 8 0<br />
T1a/b H� 34 47 24 9<br />
McArthur 6 MSKCC 2011 2002–2008 T1N0 HER2� 261 NR 0<br />
2002–2004 78 106 60 59 66<br />
subset T1a/b H� 45 42<br />
2005–2008 36 155 63 61 100 100<br />
subset T1a/b H� 54 100 100<br />
Peron 31 AERIO/Unicancer 2011 2000–2010 41 T1abN0 HER2� 205 59 100 54 49 45<br />
HR� 121 100 100 90<br />
HR- 84 0 100<br />
Shao 32 Beth Israel NYC 2011 1995–2008 42 T1abN0 658 75 17 NR NR 0<br />
HR� 494 100 0 NR NR 10 0<br />
HER2� 109 27 100 NR NR 31 17<br />
TN 55 0 0 55 NR 42<br />
14
MANAGEMENT OF T1 BREAST CANCERS<br />
Table 1. T1ab N0 Breast Cancer Literature Review (Papers Published 2002–2011) (Cont’d)<br />
Time %DFS<br />
Measure <strong>of</strong> Efficacy<br />
%RFS %DDFS %OS %IBTR Adverse Factors Comments HER2 Test<br />
5 y 97.2 age � 35, TN 48% MRM IHC 3� (10%?)<br />
96.8<br />
98.5<br />
92.6<br />
5y HER2�, TN IHC 3� 10%<br />
for N1�, multi,<br />
HER2�, TN<br />
or FISH<br />
HR 7.2<br />
�98<br />
�92<br />
�90 (HR 5.7) T1b TN DDFS �84%,<br />
HR 5.5 �90 (HR 6)<br />
HER2� �87%<br />
5 y 85.6 N1, HER2�, TN 3 M1, 54 N1 (15%) IHC 3� 30%<br />
or FISH<br />
95.5 (HR 2.01)<br />
97.4<br />
HR 4.16 HER2, HR 3.37 TN<br />
5y HER2�, ER-, grade 3,<br />
DAKO 3�<br />
97.2 premenopausal, LVI ER- 15%<br />
or FISH<br />
88 ER- 51%<br />
5 y 94.2 96.5 2.9 1 cm tumors have 12.5% 5 y IHC 3� 10%<br />
97.4 99.1<br />
distant recurrence risk<br />
or FISH<br />
91.1 94<br />
3y<br />
99<br />
89<br />
83<br />
IHC 3� 10%<br />
or FISH<br />
5y<br />
97.7<br />
IHC 3� or FISH<br />
5y,<br />
10yOS<br />
90.5 T1b 79%<br />
92.3 95.7 98.6 HER2� (HR 4.1 for DFS);<br />
size for DDFS<br />
5 y DFS HER2� 83.7% DAKO 3�<br />
or FISH<br />
83.7 89.9 83.3 ER- 39%<br />
95.7 100 3.5 HER2 (but not OS)<br />
85.1 99.2 14.9<br />
3y More LVI, T1c in T-treated group IHC 3� or FISH<br />
82 95 97<br />
78 97 98<br />
97 100 99<br />
95 100 98<br />
5y �96 if treated w/chemo/<br />
trastuzumab<br />
25% T1a; 3 pts trastuzumab alone<br />
�86 if not treated Chemo given for ER-, G2-3,<br />
high mitoses<br />
5y TN 35% T1a<br />
97.9<br />
95.6 (HR 1.64)<br />
93.5 (HR 3.13)<br />
Abbreviations: Pt, patient; yrs, years; F/U, follow-up; mos, months; ER, estrogen receptor; HER, human epidermal growth factor receptor; OBS, observation; ET,<br />
endocrine therapy; DFS, disease-free survival; RFS, recurrence-free survival; DDFS, distant disease-free survival; OS, overall survival; IBTR, ipsilateral breast tumor<br />
recurrence; NSABP, National Surgical Breast and Bowel Project; unk, unknown; nd, not done; XRT, radiation therapy; HR, hazard ratio; Tam, tamoxifen; NR, not<br />
reported; IHC, immunohistochemical; CISH, chromagen in situ hybridization; PR, progesterone receptor; LVI, lymphovascular invasion; SEER, Surveillance,<br />
Epidemiology, and End Results; BCSM, breast cancer-specific mortality; LN, lymph node; ALND, axillary lymph node dissection; AA, African <strong>American</strong>; BC, breast<br />
cancer; FISH, fluorescence in situ hybridization; T, tumor; SP3 Ab, SP3 antibody; MDACC, M. D. Anderson Cancer Center; TN, triple-negative; MRM, modified radical<br />
mastectomy; BCS, breast cancer survival; CT, computed tomography; DAKO, DAKO, Inc.; MSKCC, Memorial Sloan-Kettering Cancer Center; KPNC, Kaiser Permanente<br />
Northern California; ONCO, ONCO Languedoc-Roussillon Network.<br />
15
motherapy in 43%, and observation in 54%, resulting in a<br />
5-year DFS <strong>of</strong> 91%. Adverse prognostic factors included T1b<br />
compared with T1a and multifocal BC in the HER2-positive<br />
group.<br />
Kwon et al treated 375 Koreans with T1mic/a/bN0 BC in<br />
2000 to 2006. 4 Median follow-up was 61 months. Treatment<br />
consisted <strong>of</strong> observation in 26%, chemotherapy in 16%, and<br />
endocrine therapy in 62%. No trastuzumab was given.<br />
Five-year RFS was 97.2%. Adverse prognostic factors included<br />
age younger than 35, and TN disease. Five-year RFS<br />
was 96.8%, 98.5%, and 92.5% for the HER2-positive/ERpositive,<br />
HER2-positive/ER-negative, and TN groups.<br />
Park et al reported on 370 Korean patients with T1abN0<br />
BC treated in 1994 to 2004. 27 Median follow-up was 61<br />
months. Treatment consisted <strong>of</strong> endocrine therapy in 74%<br />
and chemotherapy in 36%. No trastuzumab was given.<br />
Five-year DDFS was approximately 98%, 92%, 90%, and<br />
90% for HR-positive, HER2-positive/ER-positive, HER2positive/ER-negative,<br />
and TN disease. Adverse prognostic<br />
factors included T1b compared with T1a, TN, and HER2positive<br />
disease. The 5-year DDFS for the T1b TN and<br />
HER2-positive groups was 84% and 87%, respectively.<br />
Lai et al reported on 377 Taiwanese patients with T1ab<br />
BC treated in 1995 to 2006. 28 Fifty-four patients had node<br />
involvement and three had metastatic disease. The overall<br />
five-year DFS was 85.6% and the prognostic factors were N1,<br />
HER2-positive, and TN disease.<br />
Tanaka et al reported on 454 Japanese patients with<br />
T1N0 BC treated from 2001 to 2007. 5 Median follow-up was<br />
46 months. Treatment followed St. Gallen guidelines, thus<br />
only 14% <strong>of</strong> the T1N0 HER2-negative group and 45% <strong>of</strong> the<br />
HER2-positive group received chemotherapy. No trastuzumab<br />
was given. Five-year RFS was 97.2% and 88% for the<br />
HER2-negative and HER2-positive groups. By univariate<br />
analysis, HER2-positivity, ER-negativity, grade 3 histology,<br />
premenopausal status, and LVI were adverse prognostic<br />
factors. By multivariate analysis, HER2-positivity and LVI<br />
retained significance. Only 28 patients had T1abN0 HER2positive<br />
disease.<br />
Fehrenbach et al reported the outcomes <strong>of</strong> 237 patients<br />
with T1abN0 HER2-positive disease treated at Kaiser Permanente<br />
Northern California (KPNC) in 2000 to 2006. 29<br />
Median follow-up was 70 months. T1a disease was present<br />
in 49% <strong>of</strong> patients and ER-positive disease in 59%. Chemotherapy<br />
was given to 17% (29% in T1b, 4% in T1a) and<br />
trastuzumab to 8% (generally with chemotherapy). Fiveyear<br />
DDFS was excellent (99.1 in T1a and 94% in T1b).<br />
Studies Including the Use <strong>of</strong> Trastuzumab<br />
Amar et al reviewed the experience <strong>of</strong> the Mayo Clinic<br />
with patients with T1abN0 in 2001 to 2005. 30 Median<br />
follow-up was 34 months. Of 421 patients, 87% had ERpositive,<br />
7% HER2-positive, and 7% TN disease. Treatment<br />
consisted <strong>of</strong> hormone therapy in 44%, chemotherapy in 7%,<br />
and trastuzumab in 1%. Of the 28 patients with HER2positive<br />
disease, four had trastuzumab plus chemotherapy<br />
and seven had chemotherapy alone. Endocrine therapy was<br />
given to 48% <strong>of</strong> ER-positive BC and 36% <strong>of</strong> HER2-positive<br />
BC. Chemotherapy was given to 28% <strong>of</strong> patients with TN<br />
disease. Three-year RFS was 99% in ER-positive BC but<br />
only 89% and 83% in patients with HER2-positive and TN<br />
disease.<br />
Horio et al reported on 267 patients from Nagoya with<br />
16<br />
ANTHONY D. ELIAS<br />
T1abN0 BC treated in 2003 to 2007. 9 Median follow-up was<br />
52 months. Overall treatment was observation in 27%,<br />
chemotherapy in 9%, and endocrine therapy in 55%. Of the<br />
42 patients with HER2-positive BC, 12% received trastuzumab<br />
together with chemotherapy. Five-year RFS was<br />
97.7% and 90.5% in the ER-positive/HER2-negative and<br />
HER2-positive groups, respectively. There were no obvious<br />
differences in relapse in the T1a group compared with the<br />
T1b group. Only 10 patients had TN BC.<br />
Wong et al published their institutional outcomes for 519<br />
patients with T1N0 BC treated in Singapore from 1989 to<br />
2009. 3 Treatment details were not provided, although endocrine<br />
therapy was given to most patients with ER-positive<br />
disease, chemotherapy to approximately 10%, and chemotherapy<br />
plus trastuzumab to three patients. Median<br />
follow-up was 57 months. Five-year DFS, DDFS, and OS<br />
were 92.3%, 95.7%, and 98.6%, respectively. Major prognostic<br />
factors were the presence <strong>of</strong> HER2 (HR � 4.1 for DFS)<br />
and tumor size (for DDFS). The 17% who had HER2-positive<br />
disease had a 5-year DFS <strong>of</strong> only 83.7% and an OS <strong>of</strong> 83.3%.<br />
Two hundred and four patients had T1abN0 disease, <strong>of</strong><br />
whom 34 had tumors positive for HER2. The patients with<br />
T1ab HER2-negative disease did well with a 5-year DFS <strong>of</strong><br />
95.7%, OS <strong>of</strong> 100%, and an IBTR <strong>of</strong> only 3.5%. However, the<br />
patients with T1ab HER2-positive disease fared less well<br />
with a 5-year DFS <strong>of</strong> 85.1%, OS <strong>of</strong> 99.2%, and an IBTR <strong>of</strong><br />
14.9%.<br />
McArthur et al summarized the MSKCC experience with<br />
the T1N0 HER2-positive BCs in the pre-trastuzumab era <strong>of</strong><br />
2002 to 2004 and the post-trastuzumab era <strong>of</strong> 2005 to 2008. 6<br />
Median follow-up was 78 and 36 months for the two groups.<br />
In the pre-trastuzumab era, 66% <strong>of</strong> the cohort received<br />
chemotherapy (42% in the T1ab subset), and 59% (essentially<br />
all <strong>of</strong> those ER�) received endocrine therapy. Threeyear<br />
DFS was only 82% (78% for the T1ab group), although<br />
3-year DDFS was 95% to 97%. A high rate <strong>of</strong> local recurrence<br />
was observed. In the post-trastuzumab era, all patients<br />
received chemotherapy plus trastuzumab and 61% (essentially<br />
all <strong>of</strong> those ER�) received endocrine therapy. The<br />
3-year DFS was 97% (95% in the T1ab subset) and the DDFS<br />
was 100%. This trial certainly suggests that aggressive<br />
treatment, including anti-HER2 treatment, can reduce relapses<br />
but at the risk <strong>of</strong> overtreatment in a large proportion.<br />
Peron et al updated the outcomes <strong>of</strong> 205 patients with<br />
T1abN0 HER2-positive BC treated in 2000 to 2010. 31 Median<br />
follow-up was 41 months. Adjuvant therapy consisted<br />
<strong>of</strong> endocrine therapy in 54% (90% <strong>of</strong> the HER2�/ER� group)<br />
and chemotherapy with trastuzumab in 44%, chemotherapy<br />
in 5%, and trastuzumab alone in three patients. Five-year<br />
RFS was 96% if treated with chemotherapy plus trastuzumab<br />
and only 86% if not treated. Chemotherapy was<br />
selectively given for ER-negative disease, grade 2 to 3<br />
disease, and high mitotic count.<br />
Shao et al presented a cohort <strong>of</strong> 658 patients with T1abN0<br />
BC from Beth Israel Deaconess Medical Center treated in<br />
1995 to 2008. 32 Median follow-up was 42 months. Four<br />
hundred and ninety-four patients had ER-positive BC, <strong>of</strong><br />
whom 10% received chemotherapy, and endocrine therapy<br />
was not reported. One hundred and nine patients had<br />
HER2-positive BC (27% ER� as well), <strong>of</strong> whom 14% received<br />
chemotherapy, and 17% received chemotherapy plus trastuzumab.<br />
Fifty-five patients had TN BC, <strong>of</strong> whom 42% received<br />
chemotherapy. Five-year DFS was 97.9%, 95.6%, and
MANAGEMENT OF T1 BREAST CANCERS<br />
93.5%, respectively. Adverse prognostic factors included TN<br />
disease. Hazard function did not reach statistical significance<br />
for HER2-positive disease.<br />
Local-Regional Management<br />
A major trend in BC therapy is the lessening extent <strong>of</strong><br />
axillary surgery. ALN dissection is completed for patients<br />
with clinically node-positive disease but not for patients<br />
with sentinel node (SN)-negative disease and is controversial<br />
even for the patients with clinically negative, SNpositive<br />
disease. SN sampling is occasionally omitted for the<br />
elderly. Since ALN involvement is less common for small<br />
primary tumors, there is a great temptation to avoid this<br />
surgery. In various series <strong>of</strong> small primary tumors, the<br />
likelihood <strong>of</strong> ALN involvement is between 3% and 37%. 33 In<br />
this series <strong>of</strong> 888 T1ab tumors with a 12% ALN involvement<br />
rate, factors associated with ALN involvement in these<br />
small tumors included LVI (HR � 12.63), perineural invasion<br />
(HR � 5.47), grade 3 tumor (HR � 3.07), and ERnegative<br />
disease (HR � 1.84). 33<br />
Rivadeneira et al found that 18% <strong>of</strong> 919 patients (199 T1a)<br />
had ALN metastases following a standard ALND. 34 Although<br />
tumor size was predictive, even the T1a group had a<br />
16% rate <strong>of</strong> ALN involvement. Other factors included grade<br />
3, LVI, and age younger than 50. These risk factors are<br />
consistently observed in other series. 35 HER2 status was not<br />
tested. In the very best group <strong>of</strong> older patients with welldifferentiated<br />
T1a tumors without LVI, 13% still had metastases.<br />
34 The authors recommend routine use <strong>of</strong> SLN biopsy.<br />
Literature Conclusions<br />
T1abN0 BC generally has an excellent prognosis. Consistent<br />
adverse prognostic factors include HER2-positive disease,<br />
ER-negative disease, high grade, T1b, and age younger<br />
than 50 years. Because most <strong>of</strong> the articles treated these<br />
patients variably, these prognostic factors actually are<br />
mixed prognostic/predictive factors demonstrating benefit <strong>of</strong><br />
specific therapies. Grade is problematic because <strong>of</strong> poor<br />
concordance between expert pathologists in its determination.<br />
Ki67—prognostic in many articles—lacks standardization<br />
between pathology laboratories. Age is a complex<br />
variable: low comorbidity rates and high life expectancy<br />
allow extended time for BC events; premenopausal status<br />
with different hormonal milieu; different genetic drivers to<br />
develop BC in the young; a different mix <strong>of</strong> molecular<br />
subtypes; and independent <strong>of</strong> subtypes, a more aggressive<br />
metastasis pattern in the young, perhaps related to the<br />
immunologic and inflammatory signals generated by pregnancy<br />
and weaning. 36<br />
In addition to these factors, which ultimately reflect<br />
molecular subtype, size, and premenopausal status, the fact<br />
remains that most events in these patients with good prognosis<br />
are unrelated to BC. Comorbidities—particularly cardiovascular<br />
health—as competing risks may represent up to<br />
80% <strong>of</strong> all events, particularly in the elderly. 22 Thus, the<br />
best endpoints to help us make decisions may not be OS,<br />
DFS, or RFS but the balancing <strong>of</strong> DDFS (or BCSS) and the<br />
risks <strong>of</strong> treatment.<br />
Most adjuvant trials find a strong correlation between an<br />
intervention and the relative risk reduction independent <strong>of</strong><br />
stage. Thus, we expect approximately 25% risk reduction<br />
with older polychemotherapy, approximately 40% risk re-<br />
duction with docetaxel plus cyclophosphamide (TC), a reduction<br />
<strong>of</strong> approximately 41% with tamoxifen, approximately<br />
50% with aromatase inhibitors, and approximately 50%<br />
reduction with the addition <strong>of</strong> trastuzumab to chemotherapy.<br />
Thus, one can estimate the magnitude <strong>of</strong> absolute risk<br />
reduction with these interventions by estimating the absolute<br />
risk if untreated.<br />
Guidelines<br />
Both major guidelines (National Comprehensive Cancer<br />
Network [NCCN] and St. Gallen’s Expert Panel) require the<br />
oncologist to consider much, but detailed instructions are<br />
nebulous, as paraphrased below.<br />
NCCN 2009. If T1a/T1mic N0, or if T1b and G1, no<br />
adjuvant therapy is recommended, although endocrine therapy<br />
can be considered. If T1b N0 and grade 2 to 3, then<br />
adjuvant endocrine therapy is recommended if ER positive.<br />
If younger than age 60, adjuvant chemotherapy could be<br />
considered. If T1b HER2 positive, endocrine therapy with or<br />
without chemotherapy with or without trastuzumab could<br />
be considered, but no further guidance is provided.<br />
St. Gallen 2011. For luminal T1N0 ER-positive BC, antiestrogen<br />
therapies are recommended but not chemotherapy.<br />
For the HER2-positive subtype, the panel <strong>of</strong> experts was<br />
willing to extrapolate the chemotherapy/trastuzumab studies<br />
to T1bN0 BCs but would not recommend any adjuvant<br />
therapy for T1a tumors. For the TN subtype, chemotherapy<br />
is considered; no specific recommendation was made based<br />
on tumor size. Trastuzumab with or without endocrine<br />
therapy alone without chemotherapy was not considered to<br />
be an acceptable adjuvant treatment for small HER2positive<br />
BC by 78% <strong>of</strong> the experts unless a contraindication<br />
to chemotherapy existed. 37 Chemotherapy would be more<br />
strongly considered for large tumor size, involved nodes, or<br />
bad biology (HER2� or TN, grade 3).<br />
In My Opinion<br />
ER-positive/HER2-negative BC. Adjuvant endocrine therapy<br />
should be considered for all, particularly for the T1b<br />
group, subject to the patient tolerance <strong>of</strong> the endocrine<br />
therapy. DFS/RFS is uniformly greater than 90% and typically<br />
approaches 97%. OncotypeRx is considered valid for<br />
the T1bN0 ER-positive subset but has not been tested in the<br />
T1mic or T1a groups. This test would be potentially useful in<br />
T1b tumors that have adverse features such as higher grade,<br />
low ER positivity, PR negativity, and possibly high Ki67<br />
scores. Size and grade remain weak prognostic factors despite<br />
RS, and a new integrated RS is anticipated to account<br />
for size. 11 Chemotherapy could be considered for T1b disease<br />
with high RS.<br />
HER-positive/ER-positive BC. Treatment for this group is<br />
highly controversial. 38 HER2 amplification increases recurrence<br />
risk. Observation with or without endocrine therapy<br />
alone has resulted in series with 5-year DFS <strong>of</strong> 85% to 92%,<br />
particularly for the T1b group. Combination chemotherapy<br />
plus trastuzumab with or without endocrine therapy has<br />
resulted in extremely few recurrences; however, most patients<br />
are therefore over-treated. This treatment commits<br />
the patient to myelosuppression, acute chemotherapy toxicities,<br />
increased cardiac morbidity, peripheral neuropathy<br />
that can affect balance (particularly in the elderly), and<br />
possible leukemia. Moreover, intravenous therapy is for a<br />
year. Alternative approaches are beginning to be studied. A<br />
17
phase II clinical trial, conducted by the Dana-Farber Cancer<br />
Institute, collaborators, and Genentech, administered<br />
single-agent paclitaxel with or without trastuzumab and<br />
recently completed accrual but is not yet reported. The<br />
Japanese RESPECT study is comparing trastuzumab with<br />
or without chemotherapy in older patients with early-stage<br />
HER2-positive BC.<br />
There is no reported experience with anti-HER2 therapy<br />
alone with or without endocrine therapy, although targeted<br />
therapy alone is an attractive option that should be studied<br />
prospectively. A legitimate concern would be some, but low,<br />
cardiac risk. Another more concerning one would be the<br />
observation that single-agent trastuzumab may not be as<br />
effective in the absence <strong>of</strong> synergy with chemotherapy. For<br />
example, sequential trastuzumab reduced relative relapse<br />
risk by 0%, 14%, and 40% in adjuvant trials compared with<br />
concurrent chemotherapy plus trastuzumab with risk reductions<br />
consistently above 50%. 39-42 More recently, there is<br />
strong data to suggest that dual anti-HER2 therapy in the<br />
absence <strong>of</strong> chemotherapy can be highly effective. 43,44 Thus,<br />
combinations <strong>of</strong> trastuzumab/lapatinib, trastuzumab/pertuzumab,<br />
and T-DM1 with or without pertuzumab would<br />
make sense. A phase III study in this population would<br />
require an international effort. 45<br />
TN BC. At present, TN BC is defined by the absence <strong>of</strong><br />
predictive markers for all but chemotherapy, given its high<br />
proliferative thrust. Thus, chemotherapy, particularly for<br />
the T1b group older than age 50, could be considered. I<br />
would use a regimen such as TC for four cycles, without<br />
anthracyclines. Once predictive biomarkers for specific targeted<br />
treatments are established in the TN subgroup, the<br />
discussion may read just like the previous paragraph, and<br />
Author’s Disclosures <strong>of</strong> Potential Conflicts <strong>of</strong> Interest<br />
Author<br />
Anthony D. Elias*<br />
*No relevant relationships to disclose.<br />
Employment or<br />
Leadership<br />
Positions<br />
Consultant or<br />
Advisory Role<br />
1. Kennedy T, Stewart AK, Bilimoria KY, et al. Treatment trends and<br />
factors associated with survival in T1aN0 and T1bN0 breast cancer patients.<br />
Ann Surg Oncol. 2007;14:2918-2927.<br />
2. Livi L, Meattini I, Saieva C, et al. Prognostic value <strong>of</strong> positive human<br />
epidermal growth factor receptor 2 status and negative hormone status in<br />
patients with T1a/T1b, lymph node-negative breast cancer. Cancer. Epub<br />
2011 October 25.<br />
3. Wong FY, Yip CSP, Chua ET. Implications <strong>of</strong> HER2 amplification in<br />
small, node-negative breast cancers: Do Asians differ? World J Surg. Epub<br />
2011 November 22.<br />
4. Kwon JH, Kim YJ, Lee K-W, et al. Triple negativity and young age as<br />
prognostic factors in lymph node-negative invasive ductal carcinoma <strong>of</strong> 1 cm<br />
or less. BMC Cancer. 2010;10:557.<br />
5. Tanaka K, Kawaguchi H, Nakamura Y, et al. Effect <strong>of</strong> HER2 status on<br />
risk <strong>of</strong> recurrence in women with small, node-negative breast tumours. Br J<br />
Surg. 2011;98:1561-1565.<br />
6. McArthur HL, Mahoney KM, Morris PG, et al. Adjuvant trastuzumab<br />
with chemotherapy is effective in women with small, node-negative, HER2positive<br />
breast cancer. Cancer. 2011;117:5461-5468.<br />
7. Horio A, Fujita T, Hayashi H, et al. High recurrence risk and use <strong>of</strong><br />
adjuvant trastuzumab in patients with small, HER2-positive, node-negative<br />
breast cancers. Int J Clin Oncol. Epub 2011 June 18.<br />
8. Cancello G, Maisonneuve P, Rotmensz N, et al. Prognosis in women with<br />
18<br />
these patients (especially the T1b subset) should be included<br />
in the definitive trials.<br />
The prognosis <strong>of</strong> T1a and T1mic is also variable. Some<br />
articles, but not all, find that size matters. Because T1mic<br />
and T1a tumors frequently represent small areas <strong>of</strong> invasion<br />
within a sea <strong>of</strong> DCIS, careful histologic sampling is required<br />
to avoid missing areas <strong>of</strong> greater invasion. The very young,<br />
especially younger than 35, could be considered for adjuvant<br />
chemotherapy.<br />
Local-regional control is clearly affected by tumor biology.<br />
IBR is higher in the HER2-positive and TN cancers, thus<br />
local-regional treatment recommendations would be the<br />
same as for T1cN0 tumors. 3,6 Sentinel node sampling is still<br />
indicated in these small tumors. The incidence <strong>of</strong> axillary<br />
node involvement, while lower in smaller primary tumors, is<br />
still 5% to 15%. In the future, once the RxSPONDER trial is<br />
reported—which is testing the utility <strong>of</strong> chemotherapy in<br />
node-positive, low to intermediate RS ER-positive BCs—<br />
perhaps nodal stage will become less important. Currently,<br />
nodal involvement still affects absolute risk <strong>of</strong> distant metastasis,<br />
and much <strong>of</strong> the decision making has to do with the<br />
absolute benefit <strong>of</strong> given interventions relative to their risks.<br />
Conclusion<br />
Small BCs are increasing in frequency. Although their<br />
risk <strong>of</strong> disseminated disease is low following local-regional<br />
treatment, certain subtypes are at greater risk. These patients<br />
have generally not been included in prospective clinical<br />
trials; a fact reflected by the heterogeneous approaches<br />
to systemic adjuvant therapy in the community <strong>of</strong> clinical<br />
oncologists. Optimal management remains controversial<br />
and will not be clarified until dedicated clinical trials are<br />
conducted.<br />
Stock<br />
Ownership Honoraria<br />
REFERENCES<br />
Research<br />
Funding<br />
Expert<br />
Testimony<br />
ANTHONY D. ELIAS<br />
Other<br />
Remuneration<br />
small (T1mic, T1a, T1b) node-negative operable breast cancer by immunohistochemically<br />
selected subtypes. Breast Cancer Res Treat. 2011;127:713-720.<br />
9. Theriault RL, Litton JK, Mittendorf EA, et al. Age and survival estimates<br />
in patients who have node-negative T1ab breast cancer by breast<br />
cancer subtype. Clin Breast Cancer. 2011;11:325-331.<br />
10. Rouanet P, Roger P, Daures JP, et al. HER2 expression is the major<br />
risk factor for recurrence in pT1a-b, N0 breast cancer: <strong>Clinical</strong> implications<br />
from a French regional population-based study <strong>of</strong> 703 patients. Proc SABCS.<br />
2011;P2:12-16.<br />
11. Habel LA, Shak S, Jacobs MK, et al. A population-based study <strong>of</strong> tumor<br />
gene expression and risk <strong>of</strong> breast cancer death among lymph node-negative<br />
patients. Breast Cancer Res. Epub 2006 May 31.<br />
12. Rosen PP, Groshen S, Saigo PE, et al. A long-term follow-up study <strong>of</strong><br />
survival in stage I T1N0M0) and stage II (T1N1M0) breast carcinoma. J Clin<br />
Oncol. 1989;7:355-366.<br />
13. Moon TE, Jones SE, Bonadonna G, et al. Development and use <strong>of</strong> a<br />
natural history data base <strong>of</strong> breast cancer studies. Am J Clin Oncol.<br />
1987;10:396-403.<br />
14. Stierer M, Rosen HR, Weber R, et al. Long term analysis <strong>of</strong> factors<br />
influencing the outcome in carcinoma <strong>of</strong> the breast smaller than one centimeter.<br />
Surg Gynecol Obstet. 1992;175:151-160.<br />
15. Hanrahan EO, Valero V, Gonzalez-Angulo AM, et al. Prognosis and<br />
management <strong>of</strong> patients with node-negative invasive breast carcinoma that is
MANAGEMENT OF T1 BREAST CANCERS<br />
1 cm or smaller in size (stage 1; T1a,bN0M0): a review <strong>of</strong> the literature. J Clin<br />
Oncol. 2006;24:2133-2122.<br />
16. Fisher B, Dignam J, Tan-Chiu E, et al. Prognosis and treatment <strong>of</strong><br />
patients with breast tumors <strong>of</strong> one centimeter or less and negative lymph<br />
nodes. J Natl Cancer Inst. 2001;93:112-120.<br />
17. Wood WC, Anderson M, Lyles RH, et al. Can we select which patients<br />
with small breast cancers should receive adjuvant chemotherapy? Ann Surg.<br />
2002;235:859-862.<br />
18. Joensuu H, Isola J, Lundin M, et al. Amplification <strong>of</strong> erbB2 and erbB2<br />
expression are superior to estrogen receptor status as risk factors for distant<br />
recurrence in pT1N0M0 breast cancer: A nationwide population-based study.<br />
Clin Cancer Res. 2003;9:923-930.<br />
19. Fisher B, Bryant J, Dignam JJ, et al. Tamoxifen, radiation therapy, or<br />
both for prevention <strong>of</strong> ipsilateral breast tumor recurrence after lumpectomy in<br />
women with invasive breast cancers <strong>of</strong> one centimeter or less. J Clin Oncol.<br />
2002;20:4141-4149.<br />
20. Fisher ER, Costantino JP, Leon ME, et al. Pathobiology <strong>of</strong> small<br />
invasive breast cancers without metastases (T1a/b, N0, M0). NSABP B-21.<br />
Cancer. 2007;110:1929-1936.<br />
21. Chia S, Norris B, Speers C, et al. Human epidermal growth factor<br />
receptor 2 overexpression as a prognositic factor in a large tissue microarray<br />
series <strong>of</strong> node-negative breast cancers. J Clin Oncol. 2008;26:5697-5704.<br />
22. Gonzalez-Angulo AM, Litton JK, Broglio KR, et al. High risk <strong>of</strong><br />
recurrence for patients with breast cancer who have human epidermal growth<br />
factor receptor 2-positive, node-negative tumors 1 cm or smaller. J Clin Oncol.<br />
2009;27:5700-5706.<br />
23. Curigliano G, Viale G, Bagnardi V, et al. <strong>Clinical</strong> relevance <strong>of</strong> Her2<br />
overexpression/amplification in patients with small tumor size and nodenegative<br />
breast cancer. J Clin Oncol. 2009;27:5693-5699.<br />
24. Meattini I, Livi L, Saieva C, et al. Prognostic value <strong>of</strong> HER2 positivity<br />
and negative hormonal status in patients with small tumor (�1cm) and<br />
node-negative breast cancer. Proc SABCS. 2011;P2-12-14.<br />
25. Colleoni M, Rotmensz N, Peruzzotti G, et al. Minimal and small size<br />
invasive breast cancer with no axillary lymph node involvement: The need for<br />
tailored adjuvant therapies. Ann Oncol. 2004;15:1633-1639.<br />
26. Hanrahan EO, Gonzalez-Angulo AM, Giordano SH, et al. Overall<br />
survival and cause-specific mortality <strong>of</strong> patients with stage T1a,bN0M0<br />
breast carcinoma. J Clin Oncol. 2007;25:4952-4960.<br />
27. Park YH, Kim ST, Cho EY, et al. A risk stratification by hormonal<br />
receptors (ER, PgR) and HER2 status in small (�1cm) invasive breast cancer:<br />
Who might be possible candidates for adjuvant treatment. Breast Cancer Res<br />
Treat. 2010;119:653-661.<br />
28. Lai HW, Kuo SJ, Chen LS, et al. Prognostic significance <strong>of</strong> triple<br />
negative breast cancer at tumor size 1 cm and smaller. Eur J Surg Oncol.<br />
2011;37:18-24.<br />
29. Fehrenbacher L, Shiraz P, Sattavat M, et al. T1abN0M0 HER2�<br />
invasive breast cancer recurrence: Population based cohort <strong>of</strong> 17,000� consecutive<br />
breast cancers 2000-2006 at Kaiser Permanente Northern California,<br />
KPNC. J Clin Oncol. 2011;29: (suppl; abstr 551).<br />
30. Amar S, McCullough AE, Tan W, et al. Prognosis and outcome <strong>of</strong> small<br />
(�1cm), node-negative breast cancer on the basis <strong>of</strong> hormonal and Her2<br />
status. Oncologist. 2010;15:1043-1049.<br />
31. Peron J, Frenel JS, Vano Y, et al. Systemic adjuvant treatment <strong>of</strong> T1a<br />
and T1b N0M0 HER2� breast carcinomas; an AERIO/UNICANCER study.<br />
Proc SABCS. 2011;P2-18-03.<br />
32. Shao T, Boolbol SK, Boachi-Adjei K, Klein P. <strong>Clinical</strong> significance <strong>of</strong><br />
HER2� and triple-negative status in patients with tumor size � 1cmand<br />
node negative breast cancer. Proc SABCS. 2011;4-09-03.<br />
33. Cserni G, Bianchi S, Vezzosi V, et al. Sentinel lymph node biopsy in<br />
staging small (up to 15 mm) breast carcinomas. Results from a European<br />
multi-institutional study. Path Oncol Res. 2007;13:5-14.<br />
34. Rivadeneira DE, Simmons RM, Christos PJ, et al. Predictive factors<br />
associated with axillary lymph node metastases in T1a and T1b breast<br />
carcinomas: Analysis in more than 900 patients. J Am Coll Surg. 2000;191:<br />
1-8.<br />
35. Khair TA, Boolbol SK, Boachi-Adjei K, Klein P. Factors affecting the<br />
development <strong>of</strong> axillary lymph node metastases in T1a-T1b breast cancers.<br />
Proc SABCS. 2011;P4-09-16.<br />
36. Lyons TR, O’Brien J, Borges VF, et al. Postpartum mammary gland<br />
involution drives progression <strong>of</strong> ductal carcinoma in situ through collagen and<br />
COX-2. Nat Med. 2011;17:1109-1115.<br />
37. Goldhirsch A, Wood WC, Coates AS, et al. Strategies for subtypes—<br />
dealing with the diversity <strong>of</strong> breast cancer: Highlights <strong>of</strong> the St. Gallen<br />
International Expert Consensus on the primary therapy <strong>of</strong> early breast<br />
cancer 2011. Ann Oncol. 2011;22:1736-1747.<br />
38. Kelly CM, Pritchard KI, Trudeau M, et al. Coping with uncertainty:<br />
T1a,bN0M0 HER2-positive breast cancer, do we have a treatment threshold?<br />
Ann Oncol. 2011;22:2387-2393.<br />
39. Spielmann M, Roche H, Delozier T, et al. Trastuzumab for patients<br />
with axillary-node positive breast cancer: Results <strong>of</strong> the FNCLCC-PACS04<br />
trial. J Clin Oncol. 2009;27:6129-6134.<br />
40. Perez EA, Romond EH, Suman VJ. Updated results <strong>of</strong> the combined<br />
analysis <strong>of</strong> NCCTG N9831 and NSABP B-31 adjuvant chemotherapy with/<br />
without trastuzumab in patients with HER2-positive breast cancer. J Clin<br />
Oncol. 2007;25 (suppl 18s; abstr 512).<br />
41. Piccart-Gebhart MJ, Procter M, Leyland-Jones B, et al. Trastuzumab<br />
after adjuvant chemotherapy in HER2-positive breast cancer. N Engl J Med.<br />
2005;353:1659-1672.<br />
42. Untch M, Gelber RD, Jackisch C, et al. Estimating the magnitude <strong>of</strong><br />
trastuzumab effects within patient subgroups in the HERA trial. Ann Oncol.<br />
2008;19:1090-1096.<br />
43. Blackwell KL, Burstein HJ, Stroniolo AM, et al. Randomized study <strong>of</strong><br />
Lapatinib alone or in combination with trastuzumab in women with ErbB2positive,<br />
trastuzumab-refractory metastatic breast cancer. J Clin Oncol.<br />
2010;28:1124-1130.<br />
44. Baselga J, Gelmon KA, Verma S, et al. Phase II trial <strong>of</strong> pertuzumab and<br />
trastuzumab in patients with human epidermal growth factor receptor<br />
2-positive metastatic breast cancer that progressed during prior trastuzumab<br />
therapy. J Clin Oncol. 2010;28:1138-1144.<br />
45. Constantinidou A, Smith I. Is there a case for anti-HER2 therapy<br />
without chemotherapy in early breast cancer? Breast. 2011;20:S158-S161.<br />
19
Controversies in Adjuvant Endocrine Therapy<br />
for Breast Cancer<br />
Overview: Adjuvant endocrine therapy for early-stage breast<br />
cancer has had the single biggest impact on improving survival<br />
from the disease—with tamoxifen alone contributing to<br />
saving many thousands <strong>of</strong> lives. In postmenopausal women,<br />
additional progress has been made by the incorporation <strong>of</strong><br />
aromatase inhibitors into the treatment <strong>of</strong> early-stage, estrogen<br />
receptor (ER)–positive breast cancer, as several large<br />
well-conducted trials have established either “up-front” or<br />
“switch” strategies that are now widely used. To date, both<br />
have been shown to be beneficial when compared with tamoxifen<br />
alone, although controversy exists as to which approach<br />
is superior. Increasingly, extended adjuvant therapy is being<br />
considered, as “longer may be better” for some women who<br />
have an ongoing risk <strong>of</strong> recurrence beyond 5 years. However,<br />
controversy remains as to how long adjuvant endocrine therapy<br />
should be given for; in clinical practice, clinicians balance<br />
IT IS well established that for patients with early-stage<br />
ER-positive breast cancer, adjuvant endocrine therapy<br />
given for 5 years after primary surgery delays local and<br />
distant relapse and prolongs overall survival. 1 In addition,<br />
adjuvant therapy substantially reduces the incidence <strong>of</strong><br />
contralateral breast cancer in patients with primary breast<br />
cancer. Endocrine responsiveness, for the most part, is<br />
dependent on the presence <strong>of</strong> a functional estrogen receptor<br />
(ER), a protein that can be detected in approximately 70% to<br />
80% <strong>of</strong> primary breast cancers. Hormonal manipulation is<br />
either achieved at a cellular level by using antiestrogens<br />
such as tamoxifen to compete for ER in the breast tumor,<br />
or by lowering systemic estrogen levels in premenopausal<br />
women with the use <strong>of</strong> LHRH agonists, and in postmenopausal<br />
women by the use <strong>of</strong> aromatase inhibitors that block<br />
estrogen biosynthesis in nonovarian tissues. All <strong>of</strong> these<br />
approaches have been extensively investigated within the<br />
context <strong>of</strong> large-scale international trials <strong>of</strong> adjuvant endocrine<br />
therapy over the past two decades. However, several<br />
key issues and unanswered questions remain, namely the<br />
best choice <strong>of</strong> endocrine agents and the optimal strategies<br />
in the postmenopausal setting, the appropriate duration<br />
<strong>of</strong> endocrine therapy, and the true role <strong>of</strong> ovarian suppression<br />
in premenopausal women. In addition, the molecular<br />
classification <strong>of</strong> breast cancer has now allowed us to better<br />
estimate endocrine responsiveness in ER-positive breast<br />
cancer, introducing another tool in addition to established<br />
guidelines for decision making in the adjuvant setting. In<br />
particular, this has allowed us to address the issue <strong>of</strong><br />
whether endocrine therapy alone is enough for many patients.<br />
The Benefit <strong>of</strong> Tamoxifen<br />
In early-stage breast cancer, tamoxifen has been the gold<br />
standard <strong>of</strong> adjuvant endocrine therapy for both premenopausal<br />
and postmenopausal breast cancer for more than<br />
three decades. In an overview <strong>of</strong> the effects <strong>of</strong> chemotherapy<br />
and hormone therapy for early-stage breast cancer by<br />
the Early Breast Cancer Trialists’ Collaborative Group<br />
(EBCTG), a 50% reduction in the risk <strong>of</strong> relapse and a 31%<br />
reduction in the annual breast cancer death rate was re-<br />
20<br />
By Stephen R. Johnston, PhD, MA<br />
the level <strong>of</strong> risk for individual patients versus any ongoing<br />
toxicity concerns. For premenopausal women, with ERpositive<br />
breast cancer, tamoxifen remains the gold standard<br />
with uncertainty in the added overall benefit <strong>of</strong> ovarian suppression.<br />
Important clinical trials have recently been completed<br />
that may help answers this question, including whether<br />
complete estrogen deprivation using a luteinizing hormone<br />
releasing hormone (LHRH) agonist plus aromatase inhibitors<br />
(AIs) is <strong>of</strong> added benefit. In recent years, molecular pr<strong>of</strong>iling<br />
<strong>of</strong> ER-positive breast cancer has started to distinguish those<br />
women with a low risk <strong>of</strong> recurrence on endocrine therapy who<br />
may not need chemotherapy. Thus, with more therapy options<br />
and greater tumour stratification, modern, adjuvant endocrine<br />
therapy is becoming increasingly personalised to suit each<br />
individual patient’s risk.<br />
ported for 5 years <strong>of</strong> adjuvant tamoxifen at 15 years <strong>of</strong><br />
follow-up. 1 Although the risk <strong>of</strong> distant recurrence is greatest<br />
during the first decade, it can still continue through the<br />
second decade, 2 raising the question about the magnitude<br />
<strong>of</strong> carry-over effect from initial therapy, or the need for<br />
extended adjuvant therapy beyond 5 years in those women<br />
at greater risk. In 2011, the EBCTG updated their metaanalysis<br />
<strong>of</strong> long-term outcomes in 21,457 women with earlystage<br />
breast cancer from 20 randomized trials <strong>of</strong> 5 years <strong>of</strong><br />
tamoxifen compared with observation or placebo. 3 They<br />
showed that 5 years <strong>of</strong> tamoxifen reduced the recurrence<br />
rate substantially in years 0–4 during therapy (rate ratio<br />
[RR] 0.53), and also in years 5–9 (RR 0.68) and throughout<br />
the first 15 years (RR 0.70, p � 0.00001). Most importantly,<br />
the relapse curves do not converge after year 10, with a<br />
continued annual absolute gain from the annual reduction<br />
in breast cancer mortality through to year 15. Furthermore,<br />
the benefit only occurs in those with ER-positive tumors,<br />
being maximal in those with rich expression <strong>of</strong> the receptor.<br />
As such, 5 years <strong>of</strong> tamoxifen probably cures many patients<br />
rather than simply delays an inevitable recurrence.<br />
These mature data from the meta-analysis regarding a<br />
well-established treatment such as tamoxifen give both<br />
clinicians and patients confidence in the beneficial effects<br />
<strong>of</strong> endocrine therapy on improving overall survival from<br />
breast cancer—so how long do patients need to take tamoxifen?<br />
Studies <strong>of</strong> duration such as NSABP-14 have compared<br />
5 to 10 years <strong>of</strong> tamoxifen and shown no advantage beyond<br />
5 years, indeed perhaps a slight disadvantage in terms <strong>of</strong><br />
disease-free survival. 4 However, this study only examined<br />
node-negative patients, and in two further trials (aTTOm<br />
and ATLAS), preliminary results have suggested a small<br />
From the Department <strong>of</strong> Medicine, Royal Marsden NHS Foundation Trust, Chelsea,<br />
London, United Kingdom.<br />
Author’s disclosure <strong>of</strong> potential conflicts <strong>of</strong> interest are found at the end <strong>of</strong> this article.<br />
Address reprint requests to Stephen R.D. Johnston, MA, PhD, Department <strong>of</strong> Medicine,<br />
Royal Marsden NHS Foundation Trust, Fulham Road, Chelsea, London, SW3 6JJ, UK;<br />
email: stephen.johnston@rmh.nhs.uk.<br />
© <strong>2012</strong> by <strong>American</strong> <strong>Society</strong> <strong>of</strong> <strong>Clinical</strong> <strong>Oncology</strong>.<br />
1092-9118/10/1-10
ADJUVANT ENDOCRINE THERAPY<br />
reduction in breast cancer recurrence for those randomized<br />
to continue tamoxifen, although no significant difference<br />
was observed for breast cancer specific or overall mortality.<br />
5,6 Further follow-up <strong>of</strong> these studies is required to<br />
reliably assess the longer term effects on recurrence and<br />
overall effects, if any, on mortality.<br />
Although tamoxifen remains the most appropriate sole<br />
treatment option for many premenopausal and perimenopausal<br />
women, the issue for postmenopausal women remains<br />
whether there are any better options than tamoxifen,<br />
especially given that some women will still relapse despite<br />
therapy, and that for others, tamoxifen has either unacceptable<br />
short-term vasomotor toxicities or increased long-term<br />
risks such as venous thromboses and endometrial cancer.<br />
The incorporation <strong>of</strong> AIs into the treatment <strong>of</strong> postmenopausal<br />
early-stage ER-positive breast cancer has now led<br />
to further improvements in outcomes over tamoxifen. The<br />
key issue is how AIs should be incorporated as adjuvant<br />
therapy in early-stage breast cancer, and whether a role for<br />
tamoxifen still remains in postmenopausal breast cancer.<br />
AIs: How Should We Use Them?<br />
In contrast to tamoxifen, which antagonizes estrogen at<br />
the ER, the oral AIs such as anastrozole, letrozole and<br />
exemestane all reduce serum estrogen levels in postmenopausal<br />
women by preventing the conversion <strong>of</strong> adrenal<br />
androgens (androstenedione and testosterone) into estradiol<br />
(E1) and estrone (E2) by the cytochrome P450 enzyme<br />
KEY POINTS<br />
● Aromatase inhibitors are recommended to be part <strong>of</strong><br />
adjuvant endocrine therapy for most postmenopausal<br />
women with estrogen receptor (ER)-positive breast<br />
cancer, either as up-front therapy or as a switch<br />
strategy following initial tamoxifen therapy. Both<br />
approaches are equally effective, and both are superior<br />
to tamoxifen alone for 5 years.<br />
● The optimal duration for aromatase inhibitors given<br />
as adjuvant therapy is 5 years based on current<br />
safety and efficacy data.<br />
● Extended adjuvant therapy with aromatase inhibitors<br />
after 5 years <strong>of</strong> tamoxifen is an additional accepted<br />
strategy for reducing ongoing risk <strong>of</strong><br />
recurrence, especially for those at greater risk.<br />
● The added benefit <strong>of</strong> ovarian suppression in premenopausal<br />
women with ER-positive early-stage breast<br />
cancer over and above tamoxifen remains unknown.<br />
In women younger than 40 years, added benefit may<br />
exist, but must be weighed up against both short and<br />
long-term tolerability.<br />
● Molecular pr<strong>of</strong>iling <strong>of</strong> ER-positive breast cancer can<br />
identify those subtypes at low risk <strong>of</strong> recurrence for<br />
whom the addition <strong>of</strong> adjuvant chemotherapy over<br />
and above endocrine therapy is not indicated. To date<br />
biomarkers, although clearly <strong>of</strong> prognostic value, have<br />
not been shown to identify those more likely to benefit<br />
from aromatase inhibitors rather than tamoxifen.<br />
Table 1. Comparative Efficacy <strong>of</strong> Up-Front 5 Years Aromatase<br />
Inhibitors Versus 5 Years Tamoxifen in Early Breast Cancer<br />
Study (reference) ATAC 11<br />
BIG-198 16<br />
Number <strong>of</strong> patients 6,241 4,922<br />
Median follow up 100 mo 76 mo<br />
Disease-free survival HR 0.90 p � .025 HR 0.88 p � .03<br />
Five-year disease-free survival difference 2.8% 2.9%<br />
Time to distant recurrence HR 0.86 p � .022 HR 0.85 p � .05<br />
Overall survival HR 1.00 p � .99 HR 0.87 p � .08<br />
Abbreviations: HR, hazard ratio.<br />
p � .05 � significant.<br />
aromatase. 7 Although estrogens are primarily synthesized<br />
in the ovary in premenopausal women under the control <strong>of</strong><br />
stimulatory effects <strong>of</strong> luteinizing hormone (LH) and follicle<br />
stimulating hormone (FSH), following menopause, mean<br />
plasma E2 levels fall from about 400–600 pmol/L to around<br />
25–50 pmol/L. These residual estrogens come solely from<br />
peripheral aromatase conversion, particularly in muscle and<br />
subcutaneous fat.<br />
The establishment <strong>of</strong> the efficacy and tolerability <strong>of</strong> AIs in<br />
advanced breast cancer encouraged the development <strong>of</strong> a<br />
number <strong>of</strong> trials examining their use in the adjuvant setting.<br />
7 In general, these therapeutic studies have either<br />
compared primary “up-front” AI therapy for 5 years with 5<br />
years <strong>of</strong> tamoxifen, or a so-called “switch” strategy <strong>of</strong> initial<br />
tamoxifen for 2–3 years followed either by an AI for 2–3<br />
years <strong>of</strong> continued tamoxifen through to year 5. The absolute<br />
benefits <strong>of</strong> these approaches have been assessed in a recent<br />
meta-analysis <strong>of</strong> all the adjuvant trials <strong>of</strong> AIs compared with<br />
the previous standard <strong>of</strong> care, namely, 5 years <strong>of</strong> tamoxifen. 8<br />
The key message is that AIs produce significantly lower<br />
recurrence rates compared with tamoxifen, either as initial<br />
monotherapy or following 2–3 years <strong>of</strong> tamoxifen, although<br />
the true effect on long-term survival is less clear at this<br />
stage. The recent 2010 <strong>American</strong> <strong>Society</strong> <strong>of</strong> <strong>Clinical</strong> <strong>Oncology</strong><br />
(ASCO) guidelines on adjuvant endocrine therapy for<br />
women with ER-positive breast cancer recommend incorporating<br />
AI therapy at some point during adjuvant treatment, 9<br />
while recognizing that the optimal timing and duration <strong>of</strong><br />
therapy remain unresolved questions. So what are the<br />
benefits, and what has been learned from all the trials <strong>of</strong><br />
adjuvant AIs to date?<br />
Up-Front AIs<br />
Two large studies have assessed the efficacy <strong>of</strong> AIs compared<br />
with tamoxifen as “up-front” adjuvant endocrine<br />
therapy in postmenopausal women with early-stage breast<br />
cancer (Table 1). 10-14 The Arimidex, Tamoxifen, Alone or<br />
in Combination (ATAC) trial was the first large study to<br />
investigate the role <strong>of</strong> AIs as adjuvant therapy for earlystage<br />
breast cancer. Over 4 years, 9,366 postmenopausal<br />
women from 21 countries were enrolled. The hypotheses<br />
tested were that anastrozole was noninferior or superior to<br />
tamoxifen, or that the combination <strong>of</strong> anastrozole and tamoxifen<br />
was superior to tamoxifen alone. The combination<br />
treatment was discontinued following the initial analysis<br />
because it showed no efficacy or tolerability benefits over<br />
tamoxifen alone. 10 Following a median follow-up <strong>of</strong> 100<br />
months, 11 anastrozole compared with tamoxifen was associated<br />
with a significantly improved disease-free survival<br />
(DFS) in hormone-receptor-positive patients (hazard ratio<br />
[HR] 0.85, p � 0.003). In the most recent 10-year analysis,<br />
21
the absolute differences in time to recurrence (TTR) in<br />
hormone-receptor-positive patients increased over time,<br />
being 2.7% at 5 years and 4.3% at 10 years. 12 In addition,<br />
recurrence rates remained significantly lower on anastrozole<br />
compared with tamoxifen after treatment completion (HR<br />
0.81, p � 0.03), although the carry-over effect was smaller<br />
after 8 years. However, these differences in preventing/<br />
delaying disease recurrence did not result in a difference in<br />
overall survival between the two treatments (HR 0.95, 95%<br />
CI 0.84–1.06). 12 Differences in toxicity pr<strong>of</strong>ile demonstrated<br />
a higher incidence <strong>of</strong> thromboembolic and cardiovascular<br />
events with tamoxifen, and more musculoskeletal events<br />
and fractures with anastrozole. 13 A bone sub-study <strong>of</strong> the<br />
main trial confirmed that AI therapy was associated with<br />
accelerated bone loss during the 5-year treatment period,<br />
although no patients with normal bone mineral density at<br />
baseline became osteoporotic at 5 years. 14<br />
The BIG 1–98 trial assessed both the efficacy <strong>of</strong> “up-front”<br />
AI therapy with letrozole for 5 years compared with tamoxifen,<br />
as well as switching strategies. It recruited 8,028<br />
women who were randomized to one <strong>of</strong> four treatment arms.<br />
In 2005, the first analysis at a median follow-up <strong>of</strong> 25.8<br />
months reported that 5 years <strong>of</strong> letrozole demonstrated a<br />
significant improvement in DFS over tamoxifen (HR 0.81,<br />
p � 0.003) and distant disease-free survival (DDFS) (HR<br />
0.73 p � 0.001). 15 These results led to the unblinding <strong>of</strong><br />
the tamoxifen alone arm, and 25.2% <strong>of</strong> patients selectively<br />
crossed over to letrozole, which has complicated subsequent<br />
intention to treat (ITT) analyses <strong>of</strong> the monotherapy arms.<br />
The updated report at a median follow-up <strong>of</strong> 76 months 16<br />
included both an ITT analysis and a censored analysis at<br />
the time <strong>of</strong> cross-over, and still demonstrated a statistically<br />
significant improvement in DFS and DDFS in favor <strong>of</strong><br />
letrozole over tamoxifen, with a nonsignificant improvement<br />
in overall survival (HR 0.87, 95% CI 0.75–1.02, p � 0.08). In<br />
terms <strong>of</strong> toxicity, endometrial cancer, vaginal bleeding, and<br />
thrombo-embolism were more common with tamoxifen,<br />
while musculoskeletal events and hypercholesterolemia<br />
were higher with letrozole.<br />
The meta-analysis <strong>of</strong> both <strong>of</strong> these studies showed an<br />
absolute 5-year reduction in recurrence <strong>of</strong> 2.9% (9.6% for AI<br />
vs 12.6% for tamoxifen, p�0.00001), with an absolute 3.9%<br />
reduction at 8 years. 8 This was associated with only a 1.1%<br />
reduction in breast cancer mortality at 5 years, which was<br />
nonsignificant.<br />
Switching from Tamoxifen to AIs Compared with Tamoxifen Alone<br />
In terms <strong>of</strong> a switching strategy, several trials have<br />
evaluated a switch to an aromatase inhibitor after 2–3 <strong>of</strong><br />
tamoxifen compared with 5 years tamoxifen alone, in an<br />
attempt to pre-empt the development <strong>of</strong> resistance to tamoxifen,<br />
and to minimize long-term side effects <strong>of</strong> either therapy<br />
if given alone for 5 years (Table 2). 17-21 The largest <strong>of</strong> these<br />
studies was the Intergroup Exemestane Study (IES), which<br />
compared switching to exemestane after 2–3 years <strong>of</strong> tamoxifen<br />
with tamoxifen for 5 years. This demonstrated not only<br />
a statistically significant improvement in disease-free survival<br />
(HR 0.68, p � 0.001), 17 but also in overall survival in<br />
an updated analysis <strong>of</strong> those with ER-positive disease (HR<br />
0.86, p � 0.04). 18 These findings were also confirmed in a<br />
meta-analysis <strong>of</strong> three separate trials with the tamoxifenanastrozole<br />
switch compared with tamoxifen alone (the<br />
Austrian Breast and Colorectal Cancer Study Group<br />
22<br />
Table 2. Comparative Efficacy <strong>of</strong> 2–3 Years Tamoxifen Followed<br />
by a Switch to 2–3 Years Aromatase Inhibitor Versus 5 Years<br />
Tamoxifen Alone<br />
Study (reference) IES 18<br />
ARNO 95 19<br />
STEPHEN R. JOHNSTON<br />
ITA 20<br />
ABCSG 8 19<br />
Number <strong>of</strong> patients 4,724 979 448 3,714<br />
Median follow up 55.7 mo 30.1 mo 64 mo 72 mo<br />
Disease-free survival HR 0.76 HR 0.66 HR 0.56 HR 0.79<br />
p � .0001 p � .49 p � .01 p � .038<br />
Overall survival HR 0.83* HR 0.53 HR 0.56 HR 0.77<br />
p � .05 p � .045 p � .1 p � .025<br />
Abbreviations: HR, hazard ratio.<br />
* In a subset <strong>of</strong> estrogen-receptor positive patients only p � .05 � significant.<br />
(ABCSG 8), Arimidex-Nolvadex (ARNO 95) and Italian<br />
Tamoxifen Anastrozole (ITA) studies 19,20 ). The metaanalysis<br />
included more than 4,000 patients with a mean<br />
follow-up <strong>of</strong> 30 months, which is shorter follow-up than in<br />
the IES study. Despite this, the combined analysis still<br />
showed that switching to anastrozole resulted in a significant<br />
improvement in disease-free survival (HR 0.59, p �<br />
0.0001), and was also associated with a statistically significant<br />
reduction in breast cancer mortality and death from<br />
any cause (HR 0.71, p � 0.037). 21<br />
These individual studies certainly all seem to suggest<br />
significantly better outcomes in terms <strong>of</strong> disease-free and<br />
overall survival by a switching strategy when compared<br />
with tamoxfen alone—in addition to less long-term toxicities<br />
with tamoxifen, there appears to be a substantial gain in<br />
efficacy, presumably by preventing relapses that were destined<br />
to occur during continued tamoxifen by utilizing a<br />
noncross-resistant therapy. However, it has been questioned<br />
by some whether early relapses on tamoxifen could be<br />
prevented by an up-front AI strategy, and as such it has<br />
been unclear which approach (tamoxifen switch to an AI, or<br />
AI upfront) would yield better efficacy outcomes and which<br />
would be better tolerated overall. Two large trials (BIG 1–98<br />
and TEAM) have now given us direct data on these comparisons.<br />
Switching Tamoxifen to AIs Compared with AI Upfront<br />
The Breast International Group (BIG) 1–98 study was the<br />
first clinical trial to report a direct comparison between the<br />
use <strong>of</strong> an upfront AI and a switching strategy. 16 At a median<br />
follow-up <strong>of</strong> 76 months, there was no significant difference<br />
in disease-free recurrence, overall survival or time to distant<br />
recurrence between the switching (tamoxifen to letrozole)<br />
and the letrozole monotherapy arms. However, there were<br />
fewer early relapses among women who were assigned to<br />
letrozole alone compared with the switch <strong>of</strong> tamoxifen followed<br />
by letrozole—this trend was greatest in node-positive<br />
patients with a 1.4% difference in breast cancer recurrence<br />
in node-negative patients (3.5% vs. 4.9%), and a 2.3% absolute<br />
difference in node-positive patients (12.4% vs. 14.7%) at<br />
5 years. Interestingly, there was no significant difference<br />
in breast cancer recurrence between the letrozole followed<br />
by a reverse-switch to tamoxifen and the letrozole monotherapy<br />
arm. The clinical significance <strong>of</strong> this is that patients<br />
who commence an AI but experience side effects or toxicity<br />
may be safely switched over to tamoxifen to complete their<br />
adjuvant endocrine therapy, without compromising on the<br />
efficacy provided by the AIs.<br />
The second study to provide a direct comparison was the<br />
Tamoxifen Exemestane Adjuvant Multicenter (TEAM) trial,
ADJUVANT ENDOCRINE THERAPY<br />
which randomized 9,779 postmenopausal women with ERpositive<br />
early-stage breast cancer to either exemestane for<br />
5 years, or tamoxifen for 2.5–3 years followed by exemestane.<br />
22 At median follow-up <strong>of</strong> 5.1 years, there was no<br />
difference in either disease-free or overall survival, a result<br />
very similar to that observed in the BIG 1–98 study.<br />
Who Benefits from Which Strategy—Up-Front or Switch?<br />
So what are we to conclude from these large, wellconducted<br />
studies as to which strategy is the best? From the<br />
meta-analysis <strong>of</strong> all these studies the addition <strong>of</strong> an AI in<br />
either strategy (up-front or switch) clearly improves diseasefree<br />
survival compared with 5 years <strong>of</strong> tamoxifen, although<br />
delaying relapse does not appear to translate into a substantial<br />
survival advantage, apart a modest benefit that was<br />
seen in the analysis <strong>of</strong> all the switching studies, which<br />
yielded an absolute gain in overall survival <strong>of</strong> 1.7% at 8<br />
years. 8 Biomarker studies in many <strong>of</strong> these trials have<br />
attempted to see whether conventional indicators <strong>of</strong> hormone<br />
responsiveness, such as absolute ER and PgR<br />
levels, 23-25 coexpression <strong>of</strong> HER2, 23 proliferation as assessed<br />
by Ki-67 labeling index, 26 or the 21-gene recurrence score, 27<br />
can identify those who benefit more from an AI than tamoxifen.<br />
Although confirming the prognostic role <strong>of</strong> all these<br />
different biomarkers in each individual study, none have<br />
been shown to identify patients with a differential response<br />
between tamoxifen and AI therapy. At the present time, the<br />
two direct comparisons between up-front AI or switch strategy<br />
(BIG 1–98 and TEAM) have shown no significant differences<br />
between these approaches; although in the letrozole<br />
study, for higher-risk node-positive patients, fewer recurrences<br />
were seen with AI up-front compared with switching.<br />
16 As such, tumor burden in addition to other inherent<br />
risk factors related to the biology <strong>of</strong> the disease (higher<br />
grade or proliferation score, lower levels <strong>of</strong> ER or PgR,<br />
coexpression <strong>of</strong> HER2) might be reasonable factors to use in<br />
clinical practice to favor an up-front use <strong>of</strong> AIs as opposed to<br />
a switch strategy.<br />
In addition, patient-related factors and toxicity concerns<br />
should always be considered when making a decision with a<br />
patient between either up-front AI use or a switch strategy.<br />
In general, the third generation AIs are well tolerated, and<br />
in the clinical trial setting, the different side effect pr<strong>of</strong>iles <strong>of</strong><br />
tamoxifen and aromatase inhibitors do not appear to affect<br />
patient’s quality <strong>of</strong> life. 28 Assessing an individual’s risk or<br />
history <strong>of</strong> venous thrombosis may influence whether tamoxifen<br />
is indicated, and assessing joint arthropathies or the<br />
risk <strong>of</strong> osteoporosis may determine the optimal timing <strong>of</strong><br />
starting an AI. Current ASCO guidelines recommend that<br />
postmenopausal women who receive an AI should have their<br />
bone mineral density evaluated, with calcium and vitamin D<br />
supplementation or bisphosphonate use dependent on the<br />
result. 29 As such, clinicians and patients have a choice on<br />
when and how to use an AI, dependent on the level <strong>of</strong> risk<br />
for the cancer and the general health issues for each individual<br />
woman–with <strong>of</strong>ten only marginal differences in efficacy<br />
outcomes (in particular survival), consideration <strong>of</strong> all<br />
these issues is important.<br />
Which AI Is Best?<br />
The only reported comparison to date is the MA.17 study,<br />
where 5 years <strong>of</strong> adjuvant exemestane yielded similar results<br />
to 5 years <strong>of</strong> adjuvant anastrozole. 30 A trial compar-<br />
Table 3. Comparative Efficacy <strong>of</strong> Extended Adjuvant Therapy <strong>of</strong><br />
5 Years Tamoxifen Followed by 3–5 Years Aromatase Inhibitor<br />
Versus 5 Years Tamoxifen Alone<br />
Study (reference) MA-17 32<br />
NSABP B-33 36<br />
ing the two nonsteroidal aromatase inhibitors letrozole and<br />
anastrozole (FACE) has completed recruitment, and results<br />
are awaited. Tolerability pr<strong>of</strong>iles between the different AIs<br />
are very similar, so, to date, there is no evidence to suggest<br />
that one agent is substantially better than another.<br />
Adjuvant Endocrine Therapy—Is There an<br />
Optimal Duration?<br />
ABCSG-6a 35<br />
Number <strong>of</strong> patients 5,170 1,562 852<br />
Median follow up 64 mo 30 mo 62 mo<br />
Disease-free survival HR 0.68 HR 0.68 HR 0.62<br />
p � .0001 p � .07 p � .031<br />
Overall survival HR 0.98 NR HR 0.89<br />
p � .853 p � .57<br />
Abbreviations: HR, hazard ratio; NR, not recorded.<br />
ER-positive breast cancer has a chronic relapsing nature,<br />
with the risk <strong>of</strong> recurrence continuing indefinitely. In the<br />
tamoxifen overview, approximately half <strong>of</strong> all recurrences<br />
occurred between 5 and 15 years after surgery despite 5<br />
years <strong>of</strong> adjuvant tamoxifen treatment. 1,3 There is, therefore,<br />
a clear rationale for considering extended adjuvant<br />
endocrine therapy beyond 5 years in an attempt to reduce<br />
long-term risk <strong>of</strong> recurrence further. The MA.27 study 31 was<br />
a double-blind, placebo-controlled trial designed to test<br />
whether 5 years <strong>of</strong> letrozole therapy in more than 5,000<br />
postmenopausal women who had completed 5 years <strong>of</strong> adjuvant<br />
tamoxifen could lead to a further improvement in<br />
disease-free survival between years 5 and 10 from diagnosis.<br />
The trial demonstrated a significant improvement in<br />
disease-free survival in all patients who continued onto<br />
letrozole after completion <strong>of</strong> 5 years <strong>of</strong> tamoxifen, with a<br />
4-year disease-free survival rate (i.e., year 9 since initial<br />
diagnosis) <strong>of</strong> 93% compared with 87%. 31 Furthermore, after<br />
a median follow-up <strong>of</strong> 30 months among higher risk lymph<br />
node-positive patients, overall survival was statistically significantly<br />
improved with letrozole (HR � 0.61, 95% CI,<br />
0.38–0.98; p � 0.04). 32 Subsequent analyses <strong>of</strong> this trial<br />
have suggested that this benefit was greatest in those with<br />
ER-positive PgR-positive tumors, 33 and that benefit may<br />
also have occurred if the switch to extended adjuvant letrozole<br />
occurred late following a period <strong>of</strong> a few years since<br />
completion <strong>of</strong> tamoxifen. 34 Two additional, smaller separate<br />
studies with either anastrozole 35 or exemestane 36 as extended<br />
adjuvant therapy after completion <strong>of</strong> 5 years <strong>of</strong><br />
tamoxifen have shown similar quantitative results in terms<br />
<strong>of</strong> reducing risk <strong>of</strong> recurrence, with hazard ratios <strong>of</strong> 0.62 and<br />
0.68, respectively (Table 3).<br />
These data confirm that risk <strong>of</strong> recurrence in hormonesensitive<br />
breast cancer does indeed persist on an ongoing<br />
basis well beyond completion <strong>of</strong> 5 years <strong>of</strong> adjuvant tamoxifen.<br />
As such, careful consideration should be given to<br />
<strong>of</strong>fering longer durations <strong>of</strong> adjuvant therapy with AIs for<br />
those deemed to be at greatest risk. Although all <strong>of</strong> these<br />
studies looked at treatment following 5 years <strong>of</strong> tamoxifen,<br />
increasingly, clinical practice is to either use AI therapy<br />
up-front in higher risk node-positive patients, or a switch<br />
approach <strong>of</strong> AIs following an initial 2–3 years <strong>of</strong> tamoxifen.<br />
23
So how do we translate the data from the extended adjuvant<br />
therapy studies into these settings <strong>of</strong> how AIs are being<br />
used today? Can we safely use AIs beyond a total duration<br />
<strong>of</strong> 5 years without, for example, compromising bone health?<br />
The answer at present is unknown, although all the safety<br />
data relating to the optimal duration <strong>of</strong> AIs indicate that<br />
therapy should be given for a maximum <strong>of</strong> 5 years. The<br />
ASCO guidelines recommend that AI therapy should not<br />
exceed 5 years outside the setting <strong>of</strong> clinical trials and that<br />
in the sequential setting, patients should receive an AI after<br />
2–3 years <strong>of</strong> tamoxifen for a total <strong>of</strong> 5 years <strong>of</strong> adjuvant<br />
endocrine therapy. The guidelines recognize that this may<br />
yield “an unfamiliar pattern <strong>of</strong> different durations <strong>of</strong> adjuvant<br />
treatment based on the treatment strategy used,” and<br />
that neither 5 years up-front AI or sequential switch strategies<br />
(2–3 years tamoxifen followed by 2–3 years <strong>of</strong> an AI)<br />
have been compared against the longer overall duration <strong>of</strong><br />
the extended adjuvant therapy regimens where treatment<br />
duration was up to 8–10 years. Two trials (MA.17R and<br />
NSABP B-42) are addressing whether longer durations <strong>of</strong> AI<br />
therapy improve outcomes without compromising safety, but<br />
results are not yet available.<br />
Ovarian Suppression—Is It Necessary for<br />
Premenopausal Women?<br />
LHRH agonists, which initially stimulate and then exhaust<br />
the LHRH receptors in the pituitary, cause reversible<br />
suppression <strong>of</strong> ovarian function, and are currently used as<br />
an alternative to ovarian ablation for treatment <strong>of</strong> advanced<br />
breast cancer in premenopausal women. The Early Breast<br />
Cancer Trialists’ Collaborative Group (EBCTG) 1 reviewed<br />
trials involving almost 8,000 women with ER-positive or<br />
ER-unknown early-stage breast cancer who were randomized<br />
to ovarian ablation by surgery or irradiation or ovarian<br />
suppression with an LHRH agonist. Overall, there was a<br />
definite beneficial effect <strong>of</strong> ovarian ablation/suppression<br />
both on recurrence and breast cancer mortality. However,<br />
the effects <strong>of</strong> ovarian treatment appear smaller in the trials<br />
where both groups got chemotherapy than in the trials<br />
where neither did, probably because chemotherapy-induced<br />
amenorrhea attenuated any additional effect <strong>of</strong> ovarian<br />
suppression. This was best demonstrated in the Intergroup<br />
0101 trial, where in premenopausal women with ER-positive<br />
node-positive early-stage breast cancer, the addition <strong>of</strong> the<br />
LHRH agonist goserelin to CAF chemotherapy appeared<br />
to have a greater effect in improving disease-free survival<br />
in women younger than 40 compared with those older than<br />
40 years. 37 Likewise, the IBCSG trial VIII randomized<br />
premenopausal women with node-negative ER-positive<br />
early-stage breast cancer to either adjuvant CMF chemotherapy,<br />
goserelin for 2 years, or CMF followed by goserelin<br />
for 18 months. 38 The addition <strong>of</strong> goserelin overall only had<br />
a minimal effect on 5-year disease-free survival, and again<br />
the benefit was maximal in women younger than 40 years<br />
with a hazard ratio <strong>of</strong> 0.34.<br />
A more recent meta-analysis looked at only trials where<br />
ER status was known and that used LHRH agonists as the<br />
method <strong>of</strong> ovarian suppression. 39 The primary endpoints<br />
were any recurrence or death after recurrence, with a<br />
median follow-up <strong>of</strong> 6.8 years. In particular, a benefit was<br />
observed when LHRH agonists were used after chemotherapy<br />
(either alone or with tamoxifen) in women younger than<br />
40 years in whom chemotherapy is less likely to induce<br />
24<br />
STEPHEN R. JOHNSTON<br />
permanent amenorrhea. Optimum duration <strong>of</strong> the use <strong>of</strong><br />
reversible ovarian suppression is unknown, although studies<br />
in general have utilized 2–3 years <strong>of</strong> LHRH agonist with<br />
5 years <strong>of</strong> tamoxifen.<br />
Although the standard <strong>of</strong> care for ER-positive premenopausal<br />
breast cancer remains tamoxifen alone for 5 years,<br />
prospective trials to address the added role (if any) <strong>of</strong><br />
ovarian suppression compared with tamoxifen alone have<br />
been undertaken. The SOFT (Suppression <strong>of</strong> Ovarian Function<br />
Trial) randomized trial will assess the role <strong>of</strong> ovarian<br />
suppression/ablation in combination with the aromatase<br />
inhibitor exemestane, compared with either ovarian suppression<br />
plus tamoxifen or tamoxifen alone. More than 3,000<br />
women were randomized into this study, which completed<br />
accrual in January 2011. The TEXT trial (Tamoxifen and<br />
EXemestane Trial) assesses an LHRH agonist with the<br />
addition <strong>of</strong> either tamoxifen or exemestane for 5 years<br />
(chemotherapy is optional), and accrual <strong>of</strong> more than 2,600<br />
women was completed in March 2011. It is hoped that both<br />
these trials will provide important additional information<br />
about the optimal endocrine therapy for premenopausal<br />
women with ER-positive early-stage breast cancer.<br />
Adjuvant Endocrine Therapy—Is It Enough on Its Own?<br />
Perhaps the hottest topic for current debate in the management<br />
<strong>of</strong> ER-positive early-stage breast cancer is not the<br />
choice <strong>of</strong> agent (tamoxifen or aromatase inhibitor, or the role<br />
<strong>of</strong> ovarian suppression), or indeed the optimal duration <strong>of</strong><br />
therapy, but rather the threshold for using adjuvant chemotherapy<br />
in addition to adjuvant endocrine therapy. As we<br />
have come to better understand the various intrinsic subtypes<br />
<strong>of</strong> breast cancer and their differential prognoses, 40<br />
oncologists have come to debate the relative benefit <strong>of</strong><br />
adjuvant chemotherapy compared with endocrine therapy,<br />
and ask whether for many patients with endocrine responsive<br />
breast cancer, a hormonal approach alone will be<br />
sufficient to maximize a patient’s chance <strong>of</strong> cure. This has<br />
been reflected in the recent 2011 St. Galen Consensus<br />
guidelines on the management <strong>of</strong> early-stage breast cancer<br />
41 —previous guidelines in 2007 and 2009 had started to<br />
characterize what might be considered truly endocrineresponsive<br />
breast cancer based on levels <strong>of</strong> receptor expression<br />
for both ER and PgR with absence <strong>of</strong> proliferation<br />
markers. In these so-called endocrine responsive tumors,<br />
the addition <strong>of</strong> adjuvant chemotherapy to endocrine therapy<br />
was only recommended if other risk features related to<br />
tumor burden were present (i.e., tumor size above 5 cm,<br />
greater than 4 nodes, extensive vascular invasion).<br />
The ability to sub-classify ER-positive breast cancer into<br />
Luminal A or Luminal B subtypes based on either clinicopathologic<br />
determination (i.e., ER, PgR, HER2 and Ki-67)<br />
or gene expression pr<strong>of</strong>iling is likely to identify ER-positive<br />
cancers with different prognoses and altered levels <strong>of</strong> endocrine<br />
responsiveness. In addition, the 21-gene signature<br />
(Oncotype DX) may also be used to predict chemotherapy<br />
benefit in a patient with an ER-positive breast cancer based<br />
on similar biologic and pathologic features, and in clinical<br />
practice is increasingly being used to identify good prognosis<br />
ER-positive breast cancer with a low recurrence score,<br />
where the addition <strong>of</strong> chemotherapy to endocrine treatment<br />
may yield no additional benefit. 42 This is now being tested<br />
prospectively in the Trial Assigning IndividuaLized Options<br />
for Treatment (Rx) (TAILORx) where the Oncotype DX
ADJUVANT ENDOCRINE THERAPY<br />
assay is being used in more than 10,000 patients to guide<br />
treatment decisions. As we come to understand breast cancer<br />
subtypes better, it is hoped that these molecular tools<br />
will allow us to stratify women with ER-positive early-stage<br />
breast cancer much more effectively into those where chemotherapy<br />
has a real role to play, and those with truly<br />
endocrine-sensitive disease where the optimal strategy for<br />
endocrine therapy alone discussed above is all that is required<br />
to cure the disease.<br />
Conclusion<br />
Endocrine therapy for early-stage breast cancer has had<br />
the biggest single effect on enhancing survival from the<br />
disease, with tamoxifen alone contributing to saving many<br />
thousands <strong>of</strong> lives. In postmenopausal women, enormous<br />
Author’s Disclosure <strong>of</strong> Potential Conflicts <strong>of</strong> Interest<br />
Author<br />
Employment or<br />
Leadership<br />
Positions<br />
Consultant or<br />
Advisory Role<br />
progress has been made by the incorporation <strong>of</strong> aromatase<br />
inhibitors into the treatment <strong>of</strong> early-stage ER-positive<br />
breast cancer, and large well-conducted trials have established<br />
“up-front” or “switch” strategies that are now widely<br />
used in clinical practice. Increasingly, extended adjuvant<br />
therapy is being considered, as “longer may be better” for<br />
some women who have an ongoing risk <strong>of</strong> recurrence beyond<br />
year 5. For others “less may be more” in terms <strong>of</strong> molecular<br />
pr<strong>of</strong>iling informing us <strong>of</strong> those women who do not need<br />
chemotherapy. As such, we are refining how to use our<br />
therapies beyond the “one strategy fits all” approach <strong>of</strong> old.<br />
Endocrine therapy will continue to evolve, and current<br />
research is now exploring novel approaches to enhance<br />
endocrine responsiveness even further through combination<br />
approaches with novel targeted therapies.<br />
Stock<br />
Ownership Honoraria<br />
Stephen R. Johnston AstraZeneca;<br />
GlaxoSmithKline<br />
1. Early Breast Cancer Trialists’ Collaborative Group (EBCTCG). Effects <strong>of</strong><br />
chemotherapy and hormonal therapy for early breast cancer on recurrence<br />
and 15-year survival: An overview <strong>of</strong> the randomised trials. Lancet. 2005;365:<br />
1687-1717.<br />
2. Dignam JJ, Dukic V, Anderson SJ, et al. Hazard <strong>of</strong> recurrence and<br />
adjuvant treatment effects over time in lymph node negative breast cancer.<br />
Breast Cancer Res Treat. 2009;116:595-602.<br />
3. Early Breast Cancer Trialists’ Collaborative Group (EBCTCG). Relevance<br />
<strong>of</strong> breast cancer hormone receptors and other factors to the efficacy <strong>of</strong><br />
adjuvant tamoxifen: Patient-level meta-analysis <strong>of</strong> randomised trials. Lancet.<br />
2011;378:771-784.<br />
4. Fisher B, Dignam J, Bryant J, Wolmark N. Five versus more than five<br />
years <strong>of</strong> tamoxifen for lymph node-negative breast cancer: Updated findings<br />
from the National Surgical Adjuvant Breast and Bowel Project B-14 randomized<br />
trial. J Natl Cancer Inst. 2001;93:684-690.<br />
5. Peto R. ATLAS (Adjuvant Tamoxifen, Longer Against Shorter); international<br />
rndomised trial <strong>of</strong> 10 versus 5 years <strong>of</strong> adjuvant tamoxifen among 11<br />
500 women - preliminary results. Presented at: 2007 San Antonio Breast<br />
Cancer Symposium; Late breaking abstract 48.<br />
6. Gray RG, Handley A, et al. aTTom (adjuvant tamoxifen- to <strong>of</strong>fer more?)<br />
: Randomised trial <strong>of</strong> 10 versus 5 years <strong>of</strong> adjuvant tamoxifen among 6,934<br />
women with etrogen receptor positive (ER�) or ER untested breast cancerpreliminary<br />
results. J Clin Oncol. 2008;26:15S (suppl; abstr 513).<br />
7. Smith IE, Dowsett M. Aromatase inhibitors in breast cancer. N Engl<br />
J Med. 2003;348:2431-2442.<br />
8. Dowsett M, Cuzick J, Ingle J, et al. Meta-analysis <strong>of</strong> breast cancer<br />
outcomes in adjuvant trials or aromatase inhibitors vs tamoxifen. J Clin<br />
Oncol. 2010;28:509-518.<br />
9. Burstein HJ, Prestrud AA, Seidenfeld J, et al. <strong>American</strong> <strong>Society</strong> <strong>of</strong><br />
<strong>Clinical</strong> <strong>Oncology</strong> <strong>Clinical</strong> Practice Guideline: Update on adjuvant endocrine<br />
therapy for women with hormone-receptor-positive breast cancer. J Clin<br />
Oncol. 2010;28:3784-3796.<br />
10. The ATAC Trialists Group. Anastrozole alone or in combination with<br />
tamoxifen versus tamoxifen alone for adjuvant treatmnet <strong>of</strong> postmenopausal<br />
women with early breast cancer: First results <strong>of</strong> the ATAC randomised trial.<br />
Lancet. 2002;359:2131-2139.<br />
11. The Arimidex, Tamoxifen, Alone or in combination (ATAC) Trialists’<br />
Group. Effect <strong>of</strong> anastrozole and tamoxifen as adjuvant treatment for earlystage<br />
breast cancer: 100-month analysis <strong>of</strong> the ATAC trial. Lancet Oncol.<br />
2008;9:45-53.<br />
12. Cuzick J, Sestak I, Baum M, et al. Effect <strong>of</strong> anastrozole and tamoxifen<br />
as adjuvant treatment for early stage breast cancer: 10-year analysis <strong>of</strong> the<br />
ATAC trial. Lancet Oncol. 2010;11:1135-1141.<br />
13. The ATAC (Arimidex, Tamoxifen, Alone or in Combination) Trialists’<br />
Group. Anastrozole alone or in combination with Tamoxifen versus Tamox-<br />
REFERENCES<br />
Research<br />
Funding<br />
Expert<br />
Testimony<br />
Other<br />
Remuneration<br />
ifen alone for adjuvant treatment <strong>of</strong> postmenopausal women with early stage<br />
breast cancer. Cancer. 2003;98:1802-1810.<br />
14. Eastell R, Adams JE, Coleman RE, et al. Effect <strong>of</strong> anastrozole on bone<br />
mineral density; 5-year resulst from the anastrozole, tamoxifen, alone or in<br />
combination trial 18233230. J Clin Oncol.2008;26:1051-1058.<br />
15. The Breast International Group (BIG) 1-98 Collaborative Group. A<br />
comparison <strong>of</strong> letrozole and tamoxifen in postmenopausal women with early<br />
breast cancer. N Engl J Med. 2005;353:2747-2757.<br />
16. The BIG 1-98 Collaborative Group. Letrozole therapy alone or in<br />
sequence with tamoxifen in women with breast cancer. N Engl J Med.<br />
2009;361:766-776.<br />
17. Coombes RC, Hall E, Gibson LJ, et al. A randomised trial <strong>of</strong> exemestane<br />
after two to three years <strong>of</strong> tamoxifen therapy in postmenopausal women<br />
with primary breast cancer. N Engl J Med. 2004;350:1081-1092.<br />
18. Coombes RC, Kilburn LS, Snowdon CF, et al. Survival and safety <strong>of</strong><br />
exemestane versus tamoxifen after 2-3 years’ tamoxifen treatment (Intergroup<br />
Exemestane Study): a randomised controlled trial. Lancet. 2007;369:<br />
559-570.<br />
19. Jackesz R, Jonat W, Gnant M, et al. Switching <strong>of</strong> postmenopausal<br />
women with endocrine-responsive early breast cancer to anastrozole after 2<br />
years adjuvant tamoxifen; combined resulst <strong>of</strong> ABCSG trial 8 and ARNO 95<br />
trial. Lancet. 2005;366:455-462.<br />
20. Boccardo F, Rubagotti A, Puntoni M, et al. Switching to anastrozole<br />
versus continued tamoxifen treatment <strong>of</strong> early breast cancer. preliminary<br />
results <strong>of</strong> the Italian Tamoxifen Anastrozole (ITA) trial. J Clin Oncol.<br />
2005;23:5138-5147.<br />
21. Jonat W, Gnant M, Boccardo F, et al. Effectiveness <strong>of</strong> switching form<br />
adjuvant tamoxifen to anastrozole in postmenopausal women with hormonesensitive<br />
early-stage breast cancer: A meta-analysis. Lancet Oncol. 2006;7:<br />
991-996.<br />
22. Van de Velde CJH, Rea D, Seynaeve C, et al. Adjuvant tamoxifen and<br />
exemestane in early breast cancer (TEAM): a randomised phase 3 trial.<br />
Lancet. 2011;377:321-331.<br />
23. Dowsett M, Allred C, Knox J, et al. Relationship between quantitative<br />
estrogen and progesterone receptor expression and human epidermal growth<br />
factor 2 (HER-2) status with recurrence in the Arimidex, Tamoxifen, Alone or<br />
in Combination Trial. J Clin Oncol. 2008; 26:1059-1065.<br />
24. Viale G, Regan MM, Maiorano E, et al. Prognostic and predictive value<br />
<strong>of</strong> centrally reviewed expression <strong>of</strong> estrogen and progesterone receptors in a<br />
randomised trial comparing letrozole and tamoxifen adjuvant therapy for<br />
postmenopausal early breast cancer: BIG 1-98. J Clin Oncol. 2007;25:3846-<br />
3852.<br />
25. Bartlett JMS, Brookes CL, Billingham LJ, et al. A prospectively<br />
planned pathology study within the TEAM trial confirms that progesterone<br />
receptor expression is prognostic, but is not predictive for differential re-<br />
25
sponse to exemestane vs tamoxifen. Presented at: 2008 San Antonio Breast<br />
Cancer Symposium.<br />
26. Viale G, Giobbie-Hurder A, Regan MM, et al. Prognostic and predictive<br />
value <strong>of</strong> centrally reviewed Ki-67 labelling index in postmenopausal women<br />
with endocrine-responsive breast cancer: Results form Breast International<br />
Group trial 1-98 comparing adjuvant tamoxifen with letrozole. J Clin Oncol.<br />
2008;26:5569-5575.<br />
27. Dowsett M, Cuzick J, Wale C, et al. Prediction <strong>of</strong> risk <strong>of</strong> distant<br />
recurrence using the 21-gene recurrence score in node-negative and nodepositive<br />
postmenopausal patients with breast cancer treated with anastrozole<br />
or tamoxifen: A TransATAC study. J Clin Oncol. 2010;28:1829-1834.<br />
28. Buzdar A, Howell A, Cuzick J, et al. Comprehensive side-effect pr<strong>of</strong>ile<br />
<strong>of</strong> anastrozole and tamoxifen as adjuvant treatment for early-stage breast<br />
cancer: Long-term safety analysis <strong>of</strong> the ATAC trial. Lancet Oncol. 2006;7:<br />
633-643.<br />
29. Hillner BE, Ingle JN, Chlebowski RT, et al. <strong>American</strong> <strong>Society</strong> <strong>of</strong><br />
<strong>Clinical</strong> <strong>Oncology</strong> 2003 update on the role <strong>of</strong> bisphosphonates and bone health<br />
issues in women with breast cancer. J Clin Oncol. 2003;21:4042-4057.<br />
30. Goss PE, Ingle JN, Chapman J-AW, et al. Final analysis <strong>of</strong> NCIC<br />
MA.27; a randomised phase III trial <strong>of</strong> exemestane versus anastrozole in<br />
postmenopausal women with hormone receptor positive primary breast<br />
cancer. Presented at San Antonio Breast Cancer Symposium 2010, Abstract<br />
S1-1.<br />
31. Goss PE, Ingle JN, Martino S, et al. A randomized trial <strong>of</strong> letrozole in<br />
postmenopausal women after five years <strong>of</strong> tamoxifen therapy for early-stage<br />
breast cancer. N Engl J Med. 2003;349:1793-1802.<br />
32. Goss PE, Ingle JN, Martino S, et al. Randomised trial <strong>of</strong> letrozole<br />
following tamoxifen as extended adjuvant therapy in receptor-positive breast<br />
cancer: Updated findings from NCIC CTG MA. 17. J Natl Cancer Inst.<br />
2005;97:1262-1271.<br />
33. Goss PE, Ingle JN, Martino S, et al. Efficacy <strong>of</strong> letrozole extended<br />
adjuavnt therapy according to estrogen receptor and progesterone receptor<br />
status <strong>of</strong> the primary tumour: National Cancer Institute <strong>of</strong> Canada Trials<br />
Group MA. 17. J Clin Oncol. 2006;25:2006-2011.<br />
26<br />
STEPHEN R. JOHNSTON<br />
34. Goss PE, Ingle JN, Pater JL, et al. Late extended adjuvant treatment<br />
with letrozole improved outcome in women with early stage breast cancer who<br />
completes 5 years <strong>of</strong> tamoxifen. J Clin Oncol. 2008;26:1948-1955.<br />
35. Jakesz R, Greil R, Gnant M, et al. Extended adjuvant therapy with<br />
anastrozole among postmenopausal breast cancer patients; results from the<br />
randomised Austrian Breast and Colorectal Cancer Study Group Trial 6a.<br />
J Natl Cancer Inst. 2007;99:1845-1853.<br />
36. Mamounas EP, Jeong J-H, Wickerham L, et al. Benefit from exemestane<br />
as extended adjuvant therapy after 5 years <strong>of</strong> adjuvant tamoxifen;<br />
intention-to-treat analysis <strong>of</strong> the National Surgical Adjuvant Breast and<br />
Bowel Project B-33 Trial. J Clin Oncol. 2008;26:1965-1971.<br />
37. Davidson NE, O’Neill AM, Vukov AM, et al. Chemoendocrine therapy<br />
for pre-menopausal women with axillary lymph-node positive, steroid hormone<br />
receptor positive breast cancer: Results from INT 0101 (E5188). J Clin<br />
Oncol. 2005;23:2973-2982<br />
38. International Breast Cancer Study Group. Adjuvant chemotherapy<br />
followed by goserelin versus either modality alone for pre-menopausal lymph<br />
node negative breast cancer—a randomised trial. J Natl Cancer Inst. 2003;<br />
95:1833-1846.<br />
39. Cuzick J, Ambroisine L, Davidson N, et al. Use <strong>of</strong> luteinising-hormonereleasing<br />
hormone agonists as adjuvant treatment in premenopausal patients<br />
with hormone-receptor-positive breast cancer: A meta-analysis <strong>of</strong> individual<br />
patient data from randomised adjuvant trials. Lancet. 2007;369:1711-1723.<br />
40. Sotiriou C and Pusztai L. Gene expression signatures in breast cancer.<br />
N Engl J Med. 2009;360:790-800.<br />
41. Goldhirsch A, Wood WC, Coates RD, et al. Strategies for subtypes -<br />
dealing with the diversity <strong>of</strong> breast cancer; highlights <strong>of</strong> the St Galen<br />
International Expert Consensus on the Primary Therapy <strong>of</strong> Early Breast<br />
Cancer 2011. Ann Oncol. 2011;22:1736-1747.<br />
42. Albain K, Barlow WE, Shak S, et al. Prognostic and predictive value <strong>of</strong><br />
the 21-gene recurrence score assay in postmenopausal women with nodepositive,<br />
oestrogen-receptor positive breast cancer on chemotherapy; a retrospective<br />
analysis <strong>of</strong> a randomised trial. Lancet Oncol. 2010;11:55-65.
A DICKENS TALE OF THE TREATMENT OF<br />
ADVANCED BREAST CANCER:<br />
THE PAST, THE PRESENT, AND THE FUTURE<br />
CHAIR<br />
George W. Sledge Jr., MD<br />
Indiana University Simon Cancer Center<br />
Indianapolis, IN<br />
SPEAKERS<br />
Fatima Cardoso, MD<br />
Champalimaud Cancer Center<br />
Lisbon, Portugal<br />
Martine J. Piccart, MD, PhD<br />
Jules Bordet Institute<br />
Brussels, Belgium<br />
Eric P. Winer, MD<br />
Dana-Farber Cancer Institute<br />
Boston, MA
A Dickens Tale <strong>of</strong> the Treatment <strong>of</strong> Advanced<br />
Breast Cancer: The Past, the Present, and<br />
the Future<br />
By George W. Sledge Jr., MD, Fatima Cardoso, MD, Eric P. Winer, MD,<br />
and Martine J. Piccart, MD<br />
Overview: Metastatic breast cancer (MBC), a usually incurable<br />
disease, continues to vex physicians and patients. Recent<br />
decades have seen great improvements in the treatment <strong>of</strong><br />
MBC, based on the availability <strong>of</strong> novel targeted therapeutics<br />
and more standard chemotherapeutic agents. This article<br />
POTENTIAL GOALS <strong>of</strong> care in MBC include cure, prolongation<br />
<strong>of</strong> survival, palliation <strong>of</strong> symptoms and maintenance<br />
<strong>of</strong> quality <strong>of</strong> life, the development <strong>of</strong> new treatment<br />
options, and what might be termed “a good death” for<br />
patients with advanced disease.<br />
Cure<br />
Metastatic breast cancer is usually incurable. Nevertheless,<br />
there are patients who are long-term survivors <strong>of</strong> the<br />
disease (whether the word “cure” should be applied to such<br />
patients is contentious). Though the overall percentage <strong>of</strong><br />
patients is small (1% to 2% <strong>of</strong> patients in the largest<br />
reported series), their existence is inarguable and suggests<br />
that long-term survival represents a possible goal, if a rare<br />
one. 1<br />
Prolongation <strong>of</strong> Survival<br />
If cure is rare, prolongation <strong>of</strong> survival nevertheless represents<br />
a reasonable goal. How much does therapy improve<br />
survival in metastatic breast cancer? We lack trials comparing<br />
active therapy with best supportive care. Studies suggesting<br />
an improvement over time in median survival 2<br />
imply that this improvement is related to therapy, but are<br />
potentially flawed because <strong>of</strong> earlier detection <strong>of</strong> metastatic<br />
disease. A recent study by the Eastern Cooperative <strong>Oncology</strong><br />
Group has suggested that, adjusted for disease relapse-free<br />
interval, overall survival (OS) <strong>of</strong> patients with metastatic<br />
disease has not improved. 3<br />
The best evidence for improvements in survival comes<br />
from studies comparing active therapies. These include<br />
comparisons <strong>of</strong> anthracycline- with nonanthracycline-based<br />
therapies, <strong>of</strong> chemotherapy with chemotherapy plus HER2targeted<br />
therapy, <strong>of</strong> aromatase inhibitor therapy with tamoxifen<br />
in estrogen receptor-positive disease, and <strong>of</strong><br />
eribulin with doctor’s best choice. 4-7 In each <strong>of</strong> these cases,<br />
the improvement associated with the new intervention is<br />
measurable in months. The cumulative effect <strong>of</strong> such improvements<br />
is hard to quantify.<br />
As current patients are more heavily pretreated (because<br />
<strong>of</strong> increasingly intensive adjuvant therapies) than their<br />
predecessors, it may become increasingly difficult to demonstrate<br />
improvements in OS. The availability <strong>of</strong> multiple<br />
lines <strong>of</strong> systemic therapy in the metastatic setting may also<br />
dilute or hide the benefit <strong>of</strong> individual new agents, whether<br />
in the front-line or refractory disease settings. 8<br />
28<br />
describes the goals <strong>of</strong> therapy for MBC, the progress made<br />
against MBC in recent decades, the current standard <strong>of</strong> care,<br />
and the ongoing efforts <strong>of</strong> basic and translational researchers<br />
to transfer the fruits <strong>of</strong> modern scientific discovery to patients<br />
in the clinic.<br />
Palliation <strong>of</strong> Symptoms and Maintenance <strong>of</strong><br />
Quality <strong>of</strong> Life<br />
Metastatic breast cancer impairs quality <strong>of</strong> life through its<br />
symptom-producing effects on organ-specific function (e.g.,<br />
bone metastasis-related fractures and pain) and general<br />
effects on quality <strong>of</strong> life (both physical and psychologic). An<br />
important part <strong>of</strong> the physician’s role is to maintain patient<br />
quality <strong>of</strong> life and palliate cancer-related symptoms.<br />
The tools available for this task have expanded in recent<br />
decades. These include appropriate pain control, treatment<br />
<strong>of</strong> isolated metastases (e.g., brain metastases and epidural<br />
cord compression), antiemetic therapy, bone maintenance<br />
therapy (with bisphosphonates and denosumab), and psychosocial<br />
and dietary interventions. Appropriate pain control<br />
in particular is critical, and prompt referral to pain<br />
specialists and palliative care experts is valuable in complex<br />
cases.<br />
Systemic therapies obviously play a role in maintenance <strong>of</strong><br />
quality <strong>of</strong> life and palliation <strong>of</strong> symptoms. Measuring the<br />
effects <strong>of</strong> systemic therapy on health-related quality <strong>of</strong> life in<br />
the metastatic setting has been difficult because <strong>of</strong> the<br />
general lack <strong>of</strong> placebo-controlled trials, the inherent difficulty<br />
<strong>of</strong> measuring health-related quality <strong>of</strong> life over time,<br />
and the competing toxicities <strong>of</strong> systemic therapy.<br />
Psychosocial support, <strong>of</strong>ten for the patient and the family,<br />
is equally important, and symptoms <strong>of</strong> insomnia, anxiety,<br />
and depression <strong>of</strong>ten require treatment. In the optimal<br />
situation, palliative and psychosocial care should be seamlessly<br />
integrated into the medical care provided by the<br />
primary oncologist after discussion with a multidisciplinary<br />
team.<br />
Developing New Agents<br />
The metastatic setting has played an important role in the<br />
cure <strong>of</strong> micrometastatic breast cancer through the development<br />
<strong>of</strong> novel agents applicable to early disease. Physicians,<br />
patients, and society in general owe much to the altruism <strong>of</strong><br />
patients with MBC. All patients with MBC should be <strong>of</strong>fered<br />
clinical trials <strong>of</strong> novel agents as an appropriate treatment<br />
From the Indiana University Simon Cancer Center, Indianapolis, IN; Breast Unit,<br />
Champalimaud Cancer Center, Lisbon, Portugal; Dana-Farber Cancer Institute, Boston,<br />
MA; Institut Jules Bordet, Université Libre de Bruxelles, Belgium.<br />
Authors’ disclosures <strong>of</strong> potential conflicts <strong>of</strong> interest are found at the end <strong>of</strong> this article.<br />
Address reprint requests to George Sledge, Jr., MD, Indiana Cancer Pavilion, 535<br />
Barnhill Dr., RT-473, Indianapolis, IN 46202; email: gsledge@iupui.edu.<br />
© <strong>2012</strong> by <strong>American</strong> <strong>Society</strong> <strong>of</strong> <strong>Clinical</strong> <strong>Oncology</strong>.<br />
1092-9118/10/1-10
TREATMENT OF ADVANCED BREAST CANCER<br />
option. New trials represent an important source <strong>of</strong> hope–<br />
and <strong>of</strong>fer potential therapeutic benefit–for patients with<br />
advanced breast cancer. The National Institutes <strong>of</strong> Health’s<br />
clinicaltrials.gov website represents an invaluable resource<br />
in this regard.<br />
A Good Death<br />
Because MBC is generally incurable, it is important for<br />
physicians to be honest with patients regarding their prognosis,<br />
particularly as treatment options dwindle. Advanced<br />
care planning should be part <strong>of</strong> a physician’s discussion with<br />
patients. Appropriate end-<strong>of</strong>-life care includes the discussion<br />
<strong>of</strong> hospice care as a therapeutic option. The <strong>American</strong><br />
<strong>Society</strong> <strong>of</strong> <strong>Clinical</strong> <strong>Oncology</strong>’s Advanced Care Planning booklet,<br />
available freely through its Cancer.net website, is a good<br />
starting point for patients with MBC.<br />
Treatment Past: Lessons Learned, but Let’s Move On<br />
One <strong>of</strong> the major criticisms raised nowadays <strong>of</strong> the “older”<br />
clinical trials is the fact that most <strong>of</strong> them were run in<br />
“all-comers” without a biologic-based patient selection. However,<br />
it was in the metastatic setting where the first targeted<br />
trials were conducted. Our oldest form <strong>of</strong> targeted therapy–<br />
endocrine therapy–was first used in MBC patients. The<br />
understanding <strong>of</strong> the crucial predictive role <strong>of</strong> the estrogen<br />
receptor (ER) led to selection based on ER status. The first<br />
studies <strong>of</strong> the anti-HER2 agent trastuzumab were performed<br />
in patients with MBC selected by HER2 receptor<br />
status. In the trials <strong>of</strong> cytotoxic agents, patient selection<br />
according to biology is only a recent phenomenon.<br />
Multidisciplinary and Locoregional Treatments<br />
An important change in oncology occurred with the understanding<br />
<strong>of</strong> the crucial role <strong>of</strong> a multidisciplinary approach<br />
to cancer treatment, with active cooperation among all<br />
specialists involved in the care <strong>of</strong> these patients. Unfortunately,<br />
this multidisciplinary approach is <strong>of</strong>ten forgotten in<br />
the advanced setting. With the development <strong>of</strong> efficacious<br />
locoregional treatments to several types <strong>of</strong> metastases<br />
(mainly brain, bone, and liver) the multimodal approach has<br />
become more prominent.<br />
Many unanswered questions remain–the role <strong>of</strong> locoregional<br />
therapy <strong>of</strong> the primary cancer for patients diagnosed<br />
with stage IV breast cancer (a survival benefit is suggested<br />
KEY POINTS<br />
● The goals <strong>of</strong> care in metastatic breast cancer (MBC)<br />
include cure, prolongation <strong>of</strong> survival, palliation <strong>of</strong><br />
symptoms, development <strong>of</strong> new agents, and a good<br />
death.<br />
● The roots <strong>of</strong> targeted therapy lie in past decades <strong>of</strong><br />
clinical trials.<br />
● These trials also did much to elucidate the proper role<br />
<strong>of</strong> systemic chemotherapy regimens in MBC.<br />
● Current treatment approaches are based on the<br />
proper use <strong>of</strong> knowledge regarding growth factor<br />
receptors (estrogen receptor and HER2).<br />
● Novel treatments for MBC attack increasing portions<br />
<strong>of</strong> the “hallmarks <strong>of</strong> cancer.”<br />
by retrospective series, 9 and there are three ongoing trials);<br />
the best candidates, timing, and approach (surgery, radi<strong>of</strong>requency,<br />
a combination <strong>of</strong> both) for liver and lung metastases;<br />
and the best sequence <strong>of</strong> therapies for brain<br />
metastases (surgery, radiotherapy, radio-surgery, combination<br />
with a systemic therapy).<br />
Systemic Therapy: Best Endpoints, Survival Benefits<br />
Though many new agents have been developed and incorporated<br />
into the treatment <strong>of</strong> MBC, very few provided a<br />
survival benefit, 10 and when they did, it was almost exclusively<br />
as first-line therapy (doxorubicin � paclitaxel vs.<br />
5-FU � doxorubicin � cyclophosphamide; letrozole vs. tamoxifen)<br />
or when compared with very old and abandoned<br />
drugs (i.e., melphalan or mitomycin plus vinblastine) or in<br />
trials that compared combination regimens with single<br />
agents, although neglecting planned cross-over (docetaxel �<br />
capecitabine vs. docetaxel alone or paclitaxel � gemcitabine<br />
vs. paclitaxel alone).<br />
Although OS benefit is undoubtedly the most desired<br />
outcome, progression-free survival (PFS) has been the most<br />
widely used endpoint. However, it is not a good surrogate for<br />
OS benefit, 11 and no good surrogate has yet been developed.<br />
Heated discussions regarding the merits <strong>of</strong> OS or PFS as the<br />
most adequate endpoint are ongoing, as is the incorporation<br />
<strong>of</strong> validated quality-<strong>of</strong>-life measurements and patientreported<br />
outcomes.<br />
Endocrine Therapy (ET): Optimal Agents and Optimal Sequence<br />
<strong>of</strong> Therapies<br />
For ER-positive premenopausal women, some relatively<br />
small trials and a meta-analysis found statistically significant<br />
survival benefits with third-generation aromatase inhibitors<br />
(AIs) (letrozole, exemestane, and anastrozole)<br />
(relative hazard (RH) � 0.87, CI 95%: 0.82 to 0.93; p �<br />
0.001). In first-line trials, in which AIs were compared with<br />
tamoxifen, the survival benefit was <strong>of</strong> 11% RH reduction<br />
(95% CI: 1% to 19%; p � 0.03). Their benefit in second and<br />
subsequent line trials, in which they were compared with<br />
other treatments, was similar. 12 However, these trials have<br />
been performed in a era when tamoxifen was the gold<br />
standard adjuvant ET, and currently the great majority <strong>of</strong><br />
postmenopausal women receives an AI in the adjuvant<br />
setting. Trials in MBC patients pretreated with AIs have<br />
shown that another AI, fulvestrant, and tamoxifen are<br />
viable options at progression.<br />
For ER-positive premenopausal women, tamoxifen with or<br />
without ovarian suppression/ablation (OS/OA) has remained<br />
the standard adjuvant therapy, and some relatively small<br />
trials and a meta-analysis provide evidence that for first-line<br />
therapy for MBC ER-positive premenopausal patients, the<br />
combination <strong>of</strong> OS/OA with tamoxifen is superior to tamoxifen<br />
alone. Although frequently used in the clinic, the<br />
combination OS/OA with an AI has not been properly<br />
evaluated in the metastatic setting, and, unfortunately,<br />
randomized trials <strong>of</strong> fulvestrant for MBC have not included<br />
premenopausal women.<br />
Chemotherapy: Optimal Drug/Regimen for First, Second, and More<br />
Lines <strong>of</strong> Therapy; Combination Versus Sequential Monotherapy;<br />
Optimal Sequence <strong>of</strong> Drugs<br />
One <strong>of</strong> the most challenging and controversial issues in<br />
MBC treatment relates to chemotherapy use. Almost all<br />
29
available data comes from an era when taxane adjuvant use<br />
was not yet standard, and even anthracycline-based regimens<br />
were not always used. These data may not apply to our<br />
current MBC patient population.<br />
Available trials have produced the following conclusions:<br />
for taxane-naive and anthracycline-naive/minimally exposed<br />
patients, single-agent anthracycline or single-agent<br />
taxane yield similar results; for taxane-naive and anthracycline-resistant/refractory<br />
patients, single-agent taxane led<br />
to better outcomes than single-agent anthracyclines; combinations<br />
<strong>of</strong> anthracyclines plus taxanes in the metastatic<br />
setting consistently led to higher response rates, sometimes<br />
higher time to progression or PFS, higher toxicity, but not<br />
better OS. Because <strong>of</strong> this, sequential single agent therapy<br />
is not only appropriate, but preferred for most patients. A<br />
2008 meta-analysis <strong>of</strong> individual patient data concludes that<br />
taxanes do not improve survival when compared with anthracyclines,<br />
either as single agents or in anthracycline<br />
combinations and that taxanes in combination with anthracyclines<br />
modestly improve response rate (RR) and PFS but<br />
not OS. 13<br />
For patients pretreated with both anthracyclines and<br />
taxanes, the most consistent data concerns capecitabine use.<br />
Vinorelbine has been compared head-to-head with docetaxel<br />
and yielded similar efficacy and significantly less toxicity.<br />
Given that both these agents are available as oral formulations<br />
and are associated with much less toxicity including<br />
alopecia, they are attractive options. Although there is also<br />
some data supporting rechallenging with taxanes as firstline<br />
therapy, the wealth <strong>of</strong> other available options and<br />
toxicity concerns make this a less appealing option.<br />
A common question relates to the use <strong>of</strong> combination <strong>of</strong><br />
cytotoxic agents compared with their sequential use as<br />
monotherapy. Available evidence 14 suggests monotherapy<br />
as the preferred option (even in the presence <strong>of</strong> visceral<br />
metastases) in all cases except those with a rapid progressive<br />
or highly symptomatic disease.<br />
It is important to know that there are no data to support<br />
an optimal sequence <strong>of</strong> therapies and that only eribulin led<br />
to an OS benefit. Therefore, treatment decisions must be<br />
individualized, taking into account many other factors beyond<br />
efficacy.<br />
Chemotherapy: Optimal Duration and Number <strong>of</strong> Lines <strong>of</strong> Therapy<br />
Many studies have looked into the question <strong>of</strong> optimal<br />
duration <strong>of</strong> systemic therapy, mainly chemotherapy, for<br />
which the balance between efficacy and toxicity is perhaps<br />
more difficult to achieve. Recently, a meta-analysis <strong>of</strong> published<br />
trials 15 has looked into this issue and concluded that<br />
longer first-line chemotherapy duration is associated with<br />
marginally longer OS and a substantially longer PFS, and<br />
treatment should be prescribed until progression or unacceptable<br />
toxicity.<br />
Treatment Present: Metastatic Breast Cancer in <strong>2012</strong><br />
Today’s systemic treatment decisions are guided by a<br />
range <strong>of</strong> factors: 1) the subtype <strong>of</strong> the breast cancer (defined<br />
by ER status and HER2) 2) sites <strong>of</strong> disease 3) disease tempo<br />
4) presence and extent <strong>of</strong> symptoms 5) prior therapy 6)<br />
availability <strong>of</strong> effective local therapy and 7) patient preferences.<br />
30<br />
Hormone Receptor-Positive MBC<br />
Hormone therapy is the preferred first-line treatment for<br />
most. Several options are available, including the AIs<br />
tamoxifen and fulvestrant (not U.S. Food and Drug Administration<br />
(FDA)-approved for first-line use). In postmenopausal<br />
patients who present de novo with metastatic disease<br />
or those who have only received tamoxifen in the adjuvant<br />
setting, treatment with an AI is standard <strong>of</strong> care. 16 A<br />
substantial number <strong>of</strong> these patients remain on an AI with<br />
disease control for prolonged periods. In a minimally symptomatic<br />
patient without an extensive disease burden, it is<br />
reasonable to try a second-line agent, even if there has been<br />
no or little benefit with the initial treatment. Among patients<br />
who develop disease progression after an objective<br />
response or prolonged disease stabilization on first-line<br />
treatment, another endocrine agent is generally preferred.<br />
Treatment with fulvestrant, a steroidal AI (assuming a<br />
nonsteroidal agent was used initially) or tamoxifen (unless<br />
there had been prior progression on tamoxifen) are reasonable.<br />
17 Unfortunately, clinical benefit from second-line therapy<br />
in the metastatic setting is less impressive than with<br />
first-line treatment. Some patients will continue to derive<br />
benefit from hormone therapy and can go on to receive thirdand<br />
fourth-line therapy.<br />
Many patients with a new diagnosis <strong>of</strong> MBC will have<br />
already progressed on an adjuvant aromatase inhibitor. For<br />
these, options are more limited. In general, fulvestrant is<br />
used as the first-line metastatic treatment in such patients.<br />
A recent study conducted by the Southwest <strong>Oncology</strong> Group<br />
that compared anastrozole alone with anastrozole plus fulvestrant<br />
in the first-line setting 18 suggested an advantage<br />
both in terms <strong>of</strong> PFS and OS for the combination.<br />
In premenopausal women with MBC, tamoxifen remains<br />
the standard treatment, <strong>of</strong>ten administered in conjunction<br />
with a luteinizing hormone releasing hormone (LHRH) agonist.<br />
19 For premenopausal women who develop metastatic<br />
disease while on tamoxifen, the usual approach is a combination<br />
<strong>of</strong> an LHRH agonist and an AI. AIs are ineffective in<br />
women with functioning ovaries.<br />
Ultimately, women with hormone receptor-positive disease<br />
become refractory to endocrine therapy. At that point,<br />
chemotherapy is administered. A wide variety <strong>of</strong> choices are<br />
available, and treatment options are similar to those that<br />
are available for women with triple-negative disease. Both<br />
the taxanes and capecitabine are commonly given in the<br />
first-line setting. Although capecitabine does not have FDA<br />
approval for first-line therapy, several small studies comparing<br />
capecitabine with other agents have suggested that<br />
capecitabine is a reasonable initial approach. 20<br />
Triple-Negative Breast Cancer<br />
SLEDGE, CARDOSO, WINER, AND PICCART<br />
Options are limited to chemotherapy in patients with<br />
triple-negative breast cancer. Most patients have received<br />
adjuvant chemotherapy, <strong>of</strong>ten with a short disease-free<br />
interval. For these reasons, chemotherapy options in the<br />
metastatic setting are <strong>of</strong>ten limited. Survival from the time<br />
<strong>of</strong> metastatic disease tends to be relatively short, with a<br />
median <strong>of</strong> approximately 1 year. 21<br />
Treatment options for metastatic triple-negative disease<br />
include all <strong>of</strong> the available chemotherapeutic agents and<br />
regimens. Given the widespread use <strong>of</strong> anthracyclines in the<br />
adjuvant setting, these agents are not commonly used in the
TREATMENT OF ADVANCED BREAST CANCER<br />
Table 1. Targeting Breast Cancer Cells<br />
Compound<br />
HER2<br />
Molecular Target <strong>Clinical</strong> Trials Pharmaceutical<br />
Pertuzumab HER2 -Phase I pertuzumab, T-DM1 and paclitaxel in HER2� MBC (NCT00951665)<br />
-Phase I/II T-DM1 with pertuzumab in trastuzumab pretreated HER2� MBC (NCT00875979)<br />
-Phase II Pertuzumab, trastuzuman and weekly paclitaxel in HER2� MBC (NCT01276041)<br />
-Phase III pertuzumab � T-DM1 versus T-DM1�placebo versus trastuzumab � taxane in<br />
HER2� MBC (MARIANNE) (NCT01120184)<br />
-Phase II trastuzumab � capecitabine � pertuzumab in HER2� MBC (PHEREXA)<br />
(NCT01026142)<br />
-Phase II pertuzumab with trastuzumab and AI in HR�HER2� MBC (NCT01491737)<br />
-Phase II <strong>of</strong> pertuzumab and trastuzuman with neoadjuvant chemotherapy in HER2� BC<br />
(NCT00976989)<br />
-Phase III, trastuzumab � chemotherapy � pertuzumab in early stage HER2� BC (APHINITY)<br />
(NCT01358877)<br />
Genentech<br />
Trastuzumab-DM1<br />
HER Family<br />
HER2 -Phase I pertuzumab, T-DM1 and paclitaxel in HER2� MBC (NCT00951665)<br />
-Phase I T-DM1 monotherapy in HER2� MBC in patients with normal or reduced hepatic<br />
function (NCT01513083)<br />
-Phase I with T-DM1 � GDC0941 versus trastuzumab � GDC0941 in trastuzumab pretreated<br />
HER2� MBC (NCT00928330)<br />
-Phase I T-DM1 � docetaxel in HER2�MBC (NCT00934856)<br />
-Phase I/II T-DM1 with pertuzumab in trastuzumab pretreated HER2� MBC (NCT00875979)<br />
-Phase II T-DM1 sequentially with anthracycline-based chemotherapy, as adjuvant or<br />
neoadjuvant therapy in HER2� BC (NCT01196052)<br />
-Phase III pertuzumab � T-DM1 versus T-DM1 � placebo versus trastuzumab � taxane in<br />
HER2� MBC (MARIANNE) (NCT01120184)<br />
-Phase III T-DM1 versus lapatinib � capecitabine in HER2� MBC (EMILIA) (NCT00829166)<br />
Genentech<br />
MM-111 HER2/HER3 bispecific -Phase I with cisplatin, capecitabine, trastuzumab or lapatinib, trastuzumab or paclitaxel,<br />
Merrimack<br />
moAb<br />
trastuzumab in HER2� MBC (NCT01304784)<br />
-Phase I with trastuzumab in HER2� MBC (NCT01097460)<br />
-Phase I monotherapy in HER2� MBC (NCT00911898)<br />
MM-121 HER3 -Phase II <strong>of</strong> MM121� exemestane in HER2- MBC (NCT01151046)<br />
-Phase II with paclitaxel in HER2- MBC (NCT01421472)<br />
-Phase I with paclitaxel in HER2- MBC (NCT01209195)<br />
-Phase I with XL147 (SAR245408) in advanced solid tumors (NCT01436565)<br />
Merrimack<br />
Neratinib EGFR, HER2 and HER4<br />
TKI<br />
-Phase I neratinib with trastuzumab and weekly paclitaxel in HER2� MBC (NCT01423123) Pfizer<br />
-Phase I/II neratinib with temsirolimus in HER2� or TN MBC (NCT01111825)<br />
-Phase II neratinib with paclitaxel in HER2� MBC (NCT00445458)<br />
-Phase I/II neratinib with trastuzumab in HER2� MBC (NCT00398567)<br />
-Phase I/II neratinib with vinorelbine in HER2� MBC (NCT00706030)<br />
-Phase II neratinib with capecitabine in HER2� MBC (NCT00741260)<br />
-Phase II neratinib versus lapatinib with capecitabine in HER2� MBC<br />
-Phase II monotherapy in HER2� MBC pts with CNS mets (NCT01494662)<br />
-Phase II <strong>of</strong> neoadjuvant chemotherapy with neratinib or trastuzumab followed by<br />
postoperative trastuzumab in HER2� BC (NCT01008150)<br />
-Phase II <strong>of</strong> neratinib plus paclitaxel versus trastuzumab plus paclitaxel in HER2� MBC<br />
(NEFERTT) (NCT00915018)<br />
-Phase III <strong>of</strong> neratinib after adjuvant trastuzumab in HER2� BC (ExteNET) (NCT00878709)<br />
Afatinib<br />
HGF/cMET<br />
EGFR, HER2 TKI -Phase I afatinib with vinorelbine in EGFR and/or HER2 overexpressing advanced solid<br />
tumors (NCT00906698)<br />
-Phase II afatinib or afatinib with chemotherapy (paclitaxel or vinorelbine weekly) in HER2�<br />
MBC (LUX-Breast 2) (NCT01271725)<br />
-Phase II Afatinib in HER2� inflammatory breast cancer (NCT01325428)<br />
-Phase II afatinib alone or in combination with vinorelbine in HER2� MBC with CNS mets<br />
(Lux-Breast 3) (NCT01441596)<br />
-Phase III afatinib plus vinorelbine versus trastuzumab plus vinorelbine in HER2� MBC after<br />
trastuzumab failure (Lux-Breast 1) (NCT01125566)<br />
Boehringer Ingelheim<br />
Foretinib VEGFR2/cMET -Phase I/II with lapatinib in HER2�MBC (NCT01138384)<br />
-Phase II monotherapy in met TNBC (NCT01147484)<br />
GSK<br />
Cabozantinib c-Met/VEGFR2/AXL/<br />
KIT/TIE2/FLT3/RET<br />
-Phase II monotherapy in HR� MBC (NCT01441947) Exelixis<br />
Onartuzumab (MetMAb) c-Met -Phase II <strong>of</strong> onartuzumab and/or bevacizumab in combination with paclitaxel in met TNBC<br />
(NCT01186991)<br />
Genentech<br />
IGFR<br />
Ganitumab (AMG 479) IGF-1R -Phase I/II with trastuzumab in HER2� MBC ( NCT01479179) Takeda with Amgen<br />
-Phase II with exemestane or fulvestrant in HR� MBC (NCT00626106)<br />
BMS-754807 IGF-1R and InsR dual<br />
Inhibition<br />
-Phase II � letrozole in HR� MBC (NCT01225172) Bristol Myers Squibb<br />
-Phase I/II with trastuzumab in HER2�MBC (NCT00788333)<br />
OSI-906 IGF-1R -Phase II with OSI-906 � letrozole/goserelin � erlotinib in HR� MBC (NCT01205685) Astellas<br />
Cixutumumab (IMC-A12) IGF-1R -Phase I/II with temsirolimus in advanced or MBC (NCT00699491)<br />
-capecitabine and lapatinib � Cixutumumab in HER2� locally advanced or MBC (stages IIIB–<br />
IV) (NCT00684983)<br />
-Phase II cixutumumab � antiestrogens in HR� MBC (NCT00728949)<br />
Lilly<br />
Dalotuzumab IGF1R -Phase I monotherapy in the neoadjuvant setting (NCT00759785)<br />
-Phase II dalotuzumab � ridaforolimus (MK-8669) in HR� MBC (NCT01234857)<br />
Merck<br />
31
Table 1. Targeting Breast Cancer Cells (Cont’d)<br />
SLEDGE, CARDOSO, WINER, AND PICCART<br />
Compound Molecular Target <strong>Clinical</strong> Trials Pharmaceutical<br />
BMS-754807 IGF-1R/InsR TKI -Phase II � letrozole in HR� MBC (NCT01225172)<br />
-Phase I/II with trastuzumab in HER2� MBC (NCT00788333)<br />
Bristol Myers Squibb<br />
MEDI-573<br />
FGFR<br />
IGF anti-ligand mAb -Phase Ib/II with AI versus AI monotherapy in HR� MBC (NCT01446159) MedImmune<br />
Dovitinib (TKI 258) FGF-R and VEGFR -Phase I with AI in HR� MBC (NCT01484041)<br />
-Phase II salvage monotherapy in inflammatory MBC (NCT01262027)<br />
Novartis<br />
BGJ398 Pan-FGF-R TKI -Phase I trial for advanced solid tumors bearing FGFR1 or FGFR2 amplification or FGFR3<br />
mutation (NCT01004224)<br />
Novartis<br />
PI3K<br />
XL147 (SAR245408) Pan-PI3K inhibition -Phase I/II with HER or HER�Taxol in MBC 2 nd line ( NCT01042925)<br />
-Phase I with letrozole (NCT01082068)<br />
-Phase I with MSC1936369B (MEK inhibition) in advanced solid tumors (NCT01357330)<br />
Exelixis with San<strong>of</strong>i-Aventis<br />
GDC-0941 Pan-PI3K Inhibition -Phase I with paclitaxel and bevacizumab (NCT00960960)<br />
-Phase Ib with trastuzumab or T-DM1 in MBC (NCT00928330)<br />
-Phase II with fulvestrant versus fulvestrant (NCT01437566) in hormone resistant MBC<br />
Genentech/Roche<br />
BKM120 Pan-PI3K Inhibition -Phase I with fulvestrant in HR� MBC ( NCT01339442)<br />
-Phase I with letrozole in HR� MBC (NCT01248494)<br />
-Phase I/II with trastuzumab in TzRes MBC ( NCT01132664)<br />
Novartis<br />
GDC-0032 PI3Ka -Phase I monotherapy in advanced solid tumors ( NCT01296555) Genentech<br />
GDC-0980 PI3K-mTOR Inhibition -Phase Ib with paclitaxel, bevacizumab and trastuzumab in MBC (NCT01254526)<br />
-Phase II with fulvestrant versus fulvestrant (NCT01437566) in hormone resistant MBC<br />
Genentech<br />
BEZ235 PI3K-mTOR Inhibition -Phase I/II in advanced solid tumors enriched for MBC (NCT00620594)<br />
-Phase I with everolimus in advanced solid tumors enriched for MBC (NCT01482156)<br />
-Phase I/II with trastuzumab versus lapatinib� capecitabine in MBC (NCT01471847)<br />
-Phase I with paclitaxel � trastuzumab in HER2� MBC (NCT01285466)<br />
-Phase I with letrozole in MBC (NCT01248494)<br />
-Phase I/II with paclitaxel in HER2- MBC (NCT01495247)<br />
-Phase I/II monotherapy in HER2-/HR� MBC (NCT01288092)<br />
Novartis<br />
XL765 PI3K-mTOR Inhibition -Phase I with letrozole ( NCT01082068) Exelixis/Aventis<br />
PF04691502<br />
mTOR<br />
PI3K-mTOR Inhibition -Phase Ib with letrozole in MBC (NCT01430585) Pfizer<br />
INK128<br />
Akt<br />
Pan-mTOR -Phase I with paclitaxel and trastuzumab in HER2� MBC (NCT01351350) Intellikine<br />
MK-2206<br />
Ras/MEK/ERK<br />
Akt -Phase I with paclitaxel in MBC (NCT01263145)–CTCs analysis–cleaved caspase 3 analysis<br />
-Phase I with paclitaxel and trastuzumab in HER2� advanced solid tumors (NCT01235897)<br />
-Phase II monotherapy in MBC with mutPI3K or mutAkt and/or PTEN loss (NCT01277757)<br />
-Phase I with trastuzumab � lapatinib in HER2� advanced solid tumors<br />
(NCT00963547)–completed<br />
-Phase I with AI or fulvestrant in HR� MBC (NCT01344031)<br />
-Phase I with lapatinib in HER2� MBC (NCT01245205)<br />
Merck<br />
AZD6244 (ARRY-142886) MEK/ERK -Phase I with docetaxel in MBC (NCT00600496)<br />
-Phase II monotherapy in stage Ic-III ER- BC to induce ER expression (NCT01313039)<br />
-Phase II fulvestrant � AZD6244 in HR� MBC ( NCT01160718)<br />
AstraZeneca<br />
GSK1120212 MEK1/2 - Defining the triple negative breast cancer kinome response to GSK1120212 in stage Ic-III<br />
TNBC (NCT01467310)<br />
GlaxoSmithKline<br />
CDK (Cyclin Dependent Kinase)<br />
PD 0332991 CDK4, 6 -Phase I/II letrozole � PD 0332991 in HR� MBC (NCT00721409)<br />
-Phase I PD 0332991 with paclitaxel in MBC (NCT01320592)<br />
Pfizer<br />
Dinaciclib (SCH 727965) CDK1, 2, 5, 9 -Phase I veliparib and dinaciclib � carboplatin in advanced solid tumors (BRCA1,2 mutation<br />
carriers included) (NCT01434316)<br />
Merck<br />
Seliciclib (CYC202)<br />
Aurora Kinases<br />
CDK2, 7, 9 -Phase I seliciclib with sapacitabine in advanced solid tumors (NCT00999401)<br />
-Phase I liposomal doxorubicin with seliciclib in met TNBC (NCT01333423)<br />
Cyclacel<br />
Alisertib Aurora-A Kinase -Phase I/II momotherapy in advanced solid tumors (NCT01045421)<br />
-Phase I alisertib with paclitaxel in metastatic ovarian and breast cancer (NCT01091428)<br />
Millennium Pharmaceuticals<br />
AMG 900<br />
Apoptosis<br />
Aurora-A, -B, -C Kinase -Phase I monotherapy in advanced solid tumors (NCT00858377) Amgen<br />
Navitoclax (ABT-263) Bcl-2, Bcl-XL, and Bcl-w -Phase I with paclitaxel in advanced solid tumors (NCT00891605) Abbott<br />
RG7112<br />
HDAC<br />
MDM2 Antagonist -Phase I monotherapy in advanced solid tumors (NCT00559533) Roche<br />
Vorinostat HDAC -Phase II with lapatinib in HER2� MBC (NCT01118975)<br />
-Phase I with ixabepilone in MBC (NCT01084057)<br />
-Phase II with tamoxifen in stage I-III BC (NCT01194427)<br />
-Phase II with tamoxifen in HR� MBC (NCT00365599)<br />
-Phase II with AI in HR� MBC (NCT01153672)<br />
Merck<br />
32
TREATMENT OF ADVANCED BREAST CANCER<br />
metastatic setting. If not previously used, an anthracycline<br />
is an appropriate option. The taxanes (particularly weekly<br />
paclitaxel) are a mainstay <strong>of</strong> treatment and one <strong>of</strong> the most<br />
commonly employed first-line regimens, particularly as a<br />
backbone for new treatment approaches with combined<br />
biologic agents. Other active agents include capecitabine,<br />
gemcitabine, eribulin, and vinorelbine. The platinum salts,<br />
both carboplatin and cisplatin, are also used in patients with<br />
metastatic triple-negative disease. 22<br />
A variety <strong>of</strong> biologic agents have been evaluated in patients<br />
with triple-negative disease in conjunction with chemotherapy.<br />
Bevacizumab was commonly administered with<br />
a taxane for metastatic triple-negative disease, but the<br />
benefits <strong>of</strong> this approach have been called into question<br />
based on the results <strong>of</strong> recent trials and a meta-analysis, and<br />
the FDA withdrew approval for bevacizumab as a treatment<br />
for MBC. 23,24 Although there is great interest in the PARP<br />
inhibitors, and an initial randomized phase II clinical trial<br />
suggested a substantial benefit when iniparib was combined<br />
with carboplatin-based chemotherapy; 25 a subsequent phase<br />
III trial failed to support these preliminary results. 26<br />
HER2-Positive Breast Cancer<br />
Trastuzumab, a humanized monoclonal antibody directed<br />
against HER2, plus chemotherapy represents the standard<br />
first-line approach for patients who have ER-negative disease<br />
or have had progression on endocrine therapy. 4 In<br />
initial trials, trastuzumab was combined with a taxane, but<br />
more recent studies have suggested that other chemotherapeutic<br />
agents, such as vinorelbine, yield similar results. 27<br />
Recent evidence suggests that continued suppression <strong>of</strong><br />
HER2 with trastuzumab is useful even after disease progression.<br />
28<br />
Lapatinib, a small molecule tyrosine kinase inhibitor<br />
directed against EGFR and HER2, is also approved for<br />
the treatment <strong>of</strong> HER2-positive breast cancer in combination<br />
with capecitabine. 29 Many patients receive a combination<br />
<strong>of</strong> lapatinib and capecitabine at some point in their<br />
treatment course, after the first or second trastuzumabbased<br />
regimen, and subsequently return to trastuzumab. Of<br />
interest, the combination <strong>of</strong> trastuzumab and lapatinib<br />
Table 1. Targeting Breast Cancer Cells (Cont’d)<br />
Compound Molecular Target <strong>Clinical</strong> Trials Pharmaceutical<br />
Entinostat (SNDX-275 or<br />
MS-275)<br />
HSP90<br />
Tanespimycin (17-AAG) HSP90–Geldanamycin<br />
derivative<br />
AUY922 HSP90 (nongeldanamycin<br />
derivative)<br />
HDAC class I -Phase II with lapatinib in trastuzumab resistant HER2� MBC (NCT01434303) Syndax<br />
-Phase II with azacitidine in MBC (NCT01349959)<br />
-Phase II exemestane � entinostat in HR� MBC ( NCT00676663)<br />
-Phase II trial completed (ENCORE 301) <strong>of</strong> continuing AI upon progression <strong>of</strong> HR� MBC with the<br />
addition <strong>of</strong> entinostat (NCT00828854)<br />
-Phase II with trastuzumab in HER2� MBC (NCT00817362) completed (02.01.<strong>2012</strong>) Infinity<br />
-Phase II with trastuzumab in trastuzumab pretreated HER2� MBC (NCT00773344) completed<br />
-Phase I/II trial monotherapy in HER2� or ER� MBC (NCT00526045) Novartis<br />
-Phase I/II with trastuzumab in HER2� MBC (NCT01271920)<br />
-Phase I/II trial with lapatinib and letrozole in HER2� MBC (NCT0136194)<br />
Retaspimycin (IPI-504)<br />
AR<br />
HSP90 -Phase II with trastuzumab in trastuzumab pretreated HER2� MBC (NCT00817362) completed Infinity<br />
Bicalutamide AR -Phase II monotherapy in AR�ER-PR- MBC (NCT00468715) AstraZeneca<br />
Abiraterone<br />
PRL<br />
-Phase II abiraterone � exemestane in ER� MBC after AI failure (NCT01381874) Johnson & Johnson<br />
LFA 102 PRL-R -Phase I monotherapy in PRL� MBC (NCT01338831) Novartis<br />
Abbreviations: MBC, metastatic breast cancer; AI, aromatase inhibitor; BC, breast cancer; MoAb, monoclonal antibody; EGFR, epidermal growth factor receptor; CNS,<br />
central nervous system; GSK, GlaxoSmithKline; PTEN, phosphatase and tensin homolog; HDAC, histone deacetylase; AR, androgen receptor; FGF, fibroblast growth<br />
factor; VEGFR, vascular endothelial growth factor receptor; PI3K, phosphatidylinositol 3-kinase; CDK, cyclin-dependent kinase; ER, estrogen receptor; PR,<br />
progesterone receptor; IGF, insulin-like growth factor.<br />
without chemotherapy has also been shown to be an active<br />
regimen and substantially more effective than lapatinib<br />
alone. 30<br />
The development <strong>of</strong> central nervous system (CNS) disease<br />
is a major challenge that is faced by more than a third <strong>of</strong> all<br />
women with HER2-positive metastatic disease. As women<br />
live longer with HER2-positive metastatic disease, the incidence<br />
<strong>of</strong> CNS disease increases. Treatment options include<br />
whole brain irradiation, stereotactic radiosurgery, surgery,<br />
and a number <strong>of</strong> systemic approaches. Of the available<br />
medical therapies, single-agent lapatinib results in rare<br />
objective responses, but a combination <strong>of</strong> capecitabine plus<br />
lapatinib appears to be more active. 31<br />
The development <strong>of</strong> trastuzumab (and, subsequently,<br />
lapatinib) for MBC is an example <strong>of</strong> the problem <strong>of</strong> leaving<br />
unanswered questions in the metastatic setting once therapy<br />
has moved to the adjuvant setting. It took almost a<br />
decade to gather evidence about the best dose and regimen<br />
(weekly and 3-weekly regimens with a loading dose are<br />
similar), the best timing <strong>of</strong> introduction, the best agent to<br />
combine it with (several options now known to exist), and<br />
the efficacy <strong>of</strong> this treatment beyond progression. For patients<br />
with ER-positive, HER2-positive disease, the combination<br />
<strong>of</strong> aromatase inhibitor therapy with HER2-targeted<br />
therapy is superior to aromatase inhibitor therapy alone.<br />
This combination may be considered as an alternative to a<br />
chemotherapy plus HER2-targeted therapy in selected patients.<br />
Treatment Future: Newer Targeted Agents in<br />
Metastatic Breast Cancer<br />
Gene-expression pr<strong>of</strong>iling analysis studies 32 have fundamentally<br />
changed the way we think about breast cancer,<br />
which is now considered a group <strong>of</strong> diseases with distinct<br />
genetic and epigenetic alterations. The translation <strong>of</strong> this<br />
improved molecular knowledge into effective molecularlytargeted<br />
agents is slower than expected. The advent <strong>of</strong><br />
next-generation sequencing technologies has generated new<br />
enthusiasm, 33 with the promise <strong>of</strong> better identification <strong>of</strong> the<br />
molecular defects responsible for carcinogenesis.<br />
33
Dual PI3KmTor<br />
inhibitors<br />
Targeting Breast Cancer Cells<br />
RTK<br />
Lessons learned from the recent randomized testing <strong>of</strong> new<br />
strategies for treating MBC. Recent months have witnessed<br />
three striking advances and one failure with the investigation<br />
<strong>of</strong> novel strategies for MBC: dual receptor blockage, use<br />
<strong>of</strong> an antireceptor antibody-drug conjugate, use <strong>of</strong> drugs<br />
targeting the PI3K/AKT/mTOR pathway, and exploitation <strong>of</strong><br />
genome instability.<br />
Dual receptor blockage was first explored in heavily pretreated<br />
patients with HER2-positive disease, using the combination<br />
<strong>of</strong> trastuzumab and lapatinib, which was superior<br />
to lapatinib alone in a relatively small, although provocative,<br />
randomized trial. 30 The recently published phase III<br />
CLEOPATRA study 34 testing dual HER2 blockade with<br />
trastuzumab and pertuzumab along with docetaxel compared<br />
with docetaxel plus trastuzumab, represents a major<br />
advance in the treatment <strong>of</strong> HER2-positive MBC. The combined<br />
strategy achieved an impressive PFS prolongation <strong>of</strong><br />
6.1 months and is likely to become a new standard <strong>of</strong> care.<br />
Others are building on this concept to design “cocktails” <strong>of</strong><br />
monoclonal antibodies directed at several HER family members<br />
in pairs, such as two antibodies targeting two different<br />
domains <strong>of</strong> each receptor, 35 or at the design <strong>of</strong> bispecific<br />
monoclonal antibodies, which target two HER family members<br />
simultaneously (e.g., HER2 and HER3 simultaneously)<br />
(Table 1).<br />
The hope is to induce downregulation <strong>of</strong> surface membrane<br />
receptors and reduce activation <strong>of</strong> compensatory pathways.<br />
It is unclear whether this strategy will be tolerable<br />
34<br />
BEZ 235<br />
BGT 226<br />
SF1126<br />
GSK 1059615<br />
GDC-0980<br />
XL765<br />
PF-4691502<br />
mTORC2<br />
mTORC1<br />
mTORC1 inhibitors<br />
Everolimus<br />
PIK3CA gene mutations<br />
(exon 9 or 20)<br />
PIK3CA gene<br />
amplification<br />
PI3K<br />
PTEN<br />
AKT1 gene mutations<br />
AKT2 gene amplification<br />
AKT<br />
PI3K inhibitors<br />
BKM120<br />
GDC-0941<br />
XL147<br />
PKI-587<br />
mTOR CATALYTIC SITE<br />
inhibitors<br />
AZD 8055<br />
OSI 027<br />
INK 128<br />
Fig. 1. Genetic aberrations <strong>of</strong> the PI3K pathway in BC and PI3K pathway inhibitors.<br />
SLEDGE, CARDOSO, WINER, AND PICCART<br />
Truncating mutations<br />
(Deletions)<br />
Epigenetic silencing<br />
PAN-PI3K inhib.<br />
Is<strong>of</strong>orm specific<br />
PI3K inhib.<br />
AKT inhibitors<br />
GSK 690693<br />
MK 2206<br />
and superior to the use <strong>of</strong> small molecules that potently<br />
inhibit the tyrosine kinase <strong>of</strong> HER1-HER2 (afatinib) or<br />
HER1-HER2-HER4 (neratinib). These oral agents are currently<br />
in phase III trials for HER2-positive MBC.<br />
Three receptors are attracting growing interest as therapeutic<br />
targets: HER3, cMET, and FGFR. HER3 is not only a<br />
key player in driving the growth <strong>of</strong> HER2-positive breast<br />
cancer through the formation <strong>of</strong> the potent HER2-HER3<br />
heterodimer (able to stimulate growth in a ligand-dependent<br />
or independent manner), but also as a mediator <strong>of</strong> resistance<br />
to antiestrogen therapies. 36 Pertuzumab does not prevent<br />
dimerization <strong>of</strong> HER3 with HER1 or other potential partners.<br />
Several anti-HER3 monoclonal antibodies in early<br />
clinical development (Table 1) will be tested in HER2positive<br />
and luminal MBC.<br />
C-MET and its stromal ligand–the hepatocyte growth<br />
factor–play a key role in cell survival, mobility, and invasion.<br />
They have been proposed as drivers <strong>of</strong> resistance to<br />
EGFR kinase inhibitors 37 and first-line trastuzumabtreatment.<br />
38 New agents targeting C-MET alone (anti C-MET<br />
MAb) or together with an angiogenic receptor such as<br />
VEGFR2 (TKIs) are currently being explored in MBC (Table<br />
1).<br />
FGFR overexpression is robustly associated with FGFR1<br />
amplification, and the latter occurs in triple-negative 39 and<br />
in luminal B-type 40 breast cancer, where it is observed in<br />
16% to 27% <strong>of</strong> the cases. FGFR1 overexpression promotes<br />
endocrine therapy resistance. Two FGFR inhibitors have<br />
entered the clinic, and a recently reported phase II study <strong>of</strong>
TREATMENT OF ADVANCED BREAST CANCER<br />
dovitinib (a dual FGFR1 and VEGFR TKI) demonstrated<br />
antitumor activity in heavily pretreated MBC patients with<br />
FGFR1 amplified tumors. 41<br />
Finally, there is a strong rationale for blocking IGF1R<br />
signaling, which plays an important role in MBC growth and<br />
is hyperactivated as a result <strong>of</strong> the release <strong>of</strong> a negative<br />
feedback loop observed with mTORC1 inhibition. Agents<br />
blocking IGF1R signaling are in clinical development with<br />
early signs <strong>of</strong> clinical activity seen in a phase I trial combining<br />
ridaforolimus (a rapalog) and dalotuzumab (an anti-<br />
IGF1R monoclonal antibody). A randomized phase II <strong>of</strong> this<br />
combination compared with exemestane is ongoing.<br />
The use <strong>of</strong> an antireceptor antibody-drug conjugate is<br />
another striking advance in strategies for MBC.<br />
Trastuzumab-DM1 (T-DM1) consists <strong>of</strong> trastuzumab covalently<br />
bound via a linker to DM1, a derivative <strong>of</strong> the<br />
antimicrotubule cytotoxic drug maytansine. This elegant<br />
way <strong>of</strong> localizing an active chemotherapy inside HER2positive<br />
tumor cells minimizes harm to normal cells, while<br />
maintaining the full properties <strong>of</strong> the monoclonal antibody<br />
(such as ADCC). This translates not only into improved<br />
Table 2. Targeting Cancer Stem Cells<br />
Compound Molecular Target Preclinical/<strong>Clinical</strong> Data Pharmaceutical<br />
Hedgehog<br />
Sarigedib (IPI-926) Smo Antagonist -Phase I trials in pancreatic cancer and HNSCC Infinity<br />
XL139 (BMS-833923) Smo Antagonist -Phase I monotherapy in advanced solid tumors (NCT01413906) Exelixis<br />
LDE225 Smo Antagonist -Phase I and II trials for pancreatic AdenoCa, medulloblastoma and basal cell Ca Novartis<br />
PF-04449913<br />
Notch<br />
Smo Antagonist -Phase I monotherapy in advanced solid tumors (NCT01286467) Pfizer<br />
MK-0752 Gamma Secretase Inhibition -Phase I with letrozole or tamoxifen in the neoadjuvant setting (NCT00756717)<br />
-Phase I with ridaforolimus in advanced solid tumors (NCT01295632)<br />
-Phase I/II with docetaxel in MBC (NCT00645333)<br />
Merck<br />
RO4929097<br />
Wnt<br />
Gamma Secretase Inhibition -Phase I/II with exemestane in HR� MBC (NCT01149356)<br />
-Phase I with paclitaxel and carboplatin in the neoadjuvant TNBC (NCT01238133)<br />
-Phase I with capecitabine in advanced solid tumors (NCT01158274)<br />
-Phase I with cediranib in advanced solid tumors (NCT01131234)<br />
-Phase II monotherapy in met TNBC (NCT01151449)<br />
Roche<br />
PRI-724 CBP-�-catenin -Phase I monotherapy in advanced solid tumors (NCT01302405) Prism Biolab<br />
Abbreviations: HNSCC, head and neck squamous cell carcinoma; AdenoCa, adenocarcinoma; Ca, carcinoma.<br />
Table 3. Targeting Tumor Stroma<br />
antitumor activity in a randomized phase II trial 42 compared<br />
with the standard combination <strong>of</strong> docetaxel and trastuzumab,<br />
but also into a better toxicity pr<strong>of</strong>ile as well as<br />
improved patient reported outcomes in terms <strong>of</strong> quality <strong>of</strong><br />
life.<br />
The use <strong>of</strong> drugs targeting the PI3K/AKT/mTOR pathway<br />
is a novel strategy for MBC with future ramifications. The<br />
central role <strong>of</strong> PI3K signaling in several cellular processes<br />
critical for cancer progression has been clearly established.<br />
Genetic alterations in several components <strong>of</strong> the PI3K pathway<br />
(Fig. 1) lead to aberrant pathway activation and mediate<br />
resistance to endocrine and anti-HER2 drugs. Among<br />
the rapidly growing number <strong>of</strong> potential therapeutics targeting<br />
the PI3K signaling cascade, mTORC1 inhibitors are<br />
the most advanced.<br />
The BOLERO-2 phase III randomized clinical trial compared<br />
the efficacy <strong>of</strong> exemestane combined with the<br />
mTORC1 inhibitor everolimus with exemestane alone in the<br />
setting <strong>of</strong> hormone receptor-positive MBC in patients refractory<br />
to non-steroidal AIs. 43 The biologic rationale involved<br />
ligand-independent activation <strong>of</strong> ER through a cross-talk<br />
Compound Molecular Target Preclinical/<strong>Clinical</strong> Data Pharmaceutical<br />
Angiogenesis Inhibition<br />
MEGF0444A/RG7414 EGFL7 -Phase I with bevacizumab � paclitaxel in advanced solid tumors (NCT01075464) Genentech<br />
MNRP1685/RG7347 NRP1 -Phase I with bevacizumab � paclitaxel in advanced solid tumors (NCT00954642) Genentech<br />
TB-403/RG7334<br />
Integrin Inhibitors<br />
PlGF -Phase I monotherapy in advanced solid tumors (NCT00702494) completed BioInvent International AB<br />
with Genentech<br />
Cilengitide �v�3 -Phase I with paclitaxel in MBC (NCT01276496) Merck<br />
PF-04605412 �5�1 Integrin -Phase I monotherapy in advanced solid tumors (NCT00915278) Pfizer<br />
IMGN388 av -Phase I monotherapy in advanced solid tumors (NCT00721669) ImmunoGen, Inc.<br />
Adhesion Signalling Molecules<br />
Removab (catumaxomab)<br />
Hypoxia<br />
Ep-CAM and CD3 -Phase I monotherapy in advanced solid tumors (NCT01320020) Trion<br />
Amin<strong>of</strong>lavone (AFP464) HIF-1� mRNA -Phase I monotherapy in advanced solid tumors (NCT00369200) Tigris<br />
EZN-2968 HIF-1� mRNA -Phase I monotherapy in advanced solid tumors with hepatic mets (NCT01120288) Enzon<br />
Abbreviation: mets, metastases; HIF, hypoxia-inducible factor; Ep-CAM, epithelial cell adhesion membrane; NRP1, neuroplin1; PIGF, placental growth factor.<br />
35
with the mTOR pathway. 44 The combined treatment led to<br />
an impressive 6.5 month extension <strong>of</strong> PFS. 43<br />
There are four additional classes <strong>of</strong> PI3K pathway inhibitors<br />
in clinical development: dual PI3K-mTOR inhibitors,<br />
PI3K inhibitors that do not inhibit mTOR, AKT inhibitors,<br />
and mTOR catalytic inhibitors (also called pan-mTOR inhibitors<br />
because they inhibit mTORC2 in addition to mTORC1)<br />
(Fig. 1). The future will tell if these newer agents are<br />
effective and tolerable pathway inhibitors or are able to<br />
overcome feedback inhibition normally observed with<br />
mTORC1 inhibitors.<br />
Of critical importance will be the identification <strong>of</strong> populations<br />
in which the benefits <strong>of</strong> these PI3K pathway inhibitors<br />
will justify their manageable but real toxicity–the current<br />
assumption is that PI3KCA-mutated cancers and PTENdeficient<br />
cancers will be at the top <strong>of</strong> the list. This would<br />
suggest that this family <strong>of</strong> compounds may find a role in all<br />
three types <strong>of</strong> breast cancer.<br />
Recently, there has been growing interest in combining<br />
PI3K with MEK pathway inhibitors. Many cancers sensitive<br />
to PI3K pathway inhibitors show tumor stasis rather than<br />
regression. The MEK pathway can represent an escape<br />
route in the presence <strong>of</strong> effective downregulation <strong>of</strong> the PI3K<br />
pathway. Therapeutic inhibition <strong>of</strong> both pathways is under<br />
clinical investigation with the use <strong>of</strong> MEK and PI3K inhibitors<br />
(Table 1).<br />
Instability <strong>of</strong> the genome is inherent to the great majority<br />
<strong>of</strong> human cancer cells and is instrumental for tumor progression.<br />
It has been proposed as a fundamental hallmark <strong>of</strong><br />
cancer. 45<br />
Although PARP inhibitors have clearly demonstrated<br />
striking antitumor activity in BRCA mutation carriers with<br />
advanced breast cancer, their activity in sporadic triplenegative<br />
breast cancer, alone or in combination with chemotherapy,<br />
has been less consistent. Myelosuppression has<br />
been an obstacle to their development. It remains uncertain<br />
whether PARP inhibitors will find an application beyond the<br />
treatment <strong>of</strong> BRCA-mutated tumors.<br />
Strategies directed at hallmarks <strong>of</strong> cancer in early development<br />
for MBC. Cyclin-dependent kinase inhibitors target<br />
evasion from growth suppressors by directly interfering with<br />
the cell cycle, downstream <strong>of</strong> the receptor tyrosine kinase<br />
signaling cascades. Several CDK inhibitors are in phase I or<br />
II trials, in combinations with either endocrine therapy in<br />
luminal cancers or chemotherapy for basal-like cancers. A<br />
randomized trial <strong>of</strong> letrozole with or without a CDK-4, -6<br />
inhibitor 25 has generated promising results. 46<br />
A similar anticancer strategy is inhibition <strong>of</strong> Aurora<br />
kinases. These serine/threonine kinases play an essential<br />
role in cellular proliferation, through control over chromatid<br />
segregation. Agents targeting Aurora kinases have entered<br />
early clinical trials.<br />
Proapoptotic agents aim at antagonizing “resistance to<br />
cell death,” and a number <strong>of</strong> them are in phase I clinical<br />
trials (Table 1). Triple-negative breast cancer seems to be a<br />
priority given the high prevalence <strong>of</strong> p53 mutations seen in<br />
this subtype.<br />
Strategies directed at epigenetic aberrations or at downregulation<br />
<strong>of</strong> key intracellular targets. As our understanding<br />
<strong>of</strong> the deregulation <strong>of</strong> epigenetic mechanisms evolves, growing<br />
interest in targeting these aberrations has emerged.<br />
Histone deacetylase inhibitors are progressing from phase I<br />
to phase II trials with interesting results seen when these<br />
36<br />
SLEDGE, CARDOSO, WINER, AND PICCART<br />
agents are combined with endocrine agents or anti-HER2<br />
drugs.<br />
Heat shock protein 90 (HSP90) represents another interesting<br />
anticancer target in MBC. 47 A molecular chaperone,<br />
HSP90 plays a vital role in the intact function <strong>of</strong> several<br />
oncogenic proteins and, thus, promotes key molecular events<br />
in malignant progression. Several agents have been tested<br />
with promising results in HER2-positive MBC.<br />
Antihormone strategies. The androgen and prolactin receptors<br />
are being explored in advanced breast cancer. The<br />
androgen receptor (AR) has recently been shown to play a<br />
role in HER2-positive/estrogen receptor-negative breast<br />
cancers, which can get stimulated by testosterone and also<br />
show upregulation <strong>of</strong> the Wnt signaling pathway. 48 Moreover,<br />
AR positivity has been found in approximately 20% <strong>of</strong><br />
triple-negative breast cancers, 49 and a phase II study with<br />
bicalutamide for ER-negative breast cancer expressing AR<br />
has recently been performed.<br />
Targeting Breast Cancer Stem Cells<br />
The identification <strong>of</strong> a cellular subpopulation in breast<br />
cancer with a specific immunophenotype (i.e., CD44 �<br />
CD24 - ) 50 bearing a tumor-initiating capacity was the first<br />
identification <strong>of</strong> cancer stem cells (CSCs) in the setting <strong>of</strong> a<br />
solid tumor. Ever since, a detailed exploration <strong>of</strong> this cellular<br />
compartment <strong>of</strong> breast cancer has been undertaken, and<br />
this expanding knowledge provides novel therapeutic targets<br />
51 for this therapy-resistant cellular population.<br />
One <strong>of</strong> the fundamental aspects <strong>of</strong> CSCs is their capacity<br />
for self-renewal, mediated through well-conserved developmental<br />
signaling pathways 52 (the Notch, Hedgehog, and<br />
Wnt pathways) amenable to therapeutic targeting. The<br />
Notch signaling pathway, frequently deregulated in breast<br />
cancer, 53 can be targeted by gammasecretase inhibitors<br />
(Table 2). These small molecules block the translocation <strong>of</strong><br />
the intracellular part <strong>of</strong> the Notch receptor to the nucleus<br />
and corresponding signaling pathway activation. The<br />
Hedgehog signaling pathway has also been shown to promote<br />
the tumor-initiating ability <strong>of</strong> breast CSCs. 54 A class <strong>of</strong><br />
compounds called Smo-antagonists has entered clinical trials<br />
(Table 2).<br />
A second approach targeting breast CSCs is based on<br />
high-throughput screening methods. A population <strong>of</strong> breast<br />
cancer cells enriched for CSCs are screened either for<br />
druggable targets through shRNA/siRNA technology or<br />
through the testing <strong>of</strong> extended small molecule compound<br />
collections. Based on the latter approach, Gupta and colleagues<br />
55 recently discovered that salinomycin is able to<br />
eradicate breast CSCs.<br />
Targeting the Tumor Microenvironment<br />
Tumor cells are integrated into their microenvironment,<br />
constantly interacting with it. 56 These components are<br />
appealing therapeutic targets, with antiangiogenic factors<br />
being the most clinically advanced. The failure <strong>of</strong> bevacizumab,<br />
a humanized monoclonal antibody targeting<br />
VEGF-A, to extend OS <strong>of</strong> MBC patients 23,57 should not be<br />
regarded as a pro<strong>of</strong> <strong>of</strong> failure <strong>of</strong> the concept <strong>of</strong> antiangiogenic<br />
therapy in breast cancer. The assumption that bevacizumab<br />
administration potentiates the delivery <strong>of</strong> higher intratumor<br />
chemotherapy concentrations proved incorrect. 58 Moreover,<br />
the redundancy <strong>of</strong> different angiogenic mechanisms and<br />
proangiogenic factors could explain this clinical failure, as
TREATMENT OF ADVANCED BREAST CANCER<br />
inhibition <strong>of</strong> VEGF-A rapidly leads to compensatory activation<br />
<strong>of</strong> alternative pathways. 59,60 Other compounds targeting<br />
the tumor microenvironment can be found in Table 3.<br />
Conclusion<br />
MBC represents a continuing challenge to physicians and<br />
patients. An improved understanding <strong>of</strong> breast cancer biology<br />
has improved prognosis, and progress based on biologic<br />
Authors’ Disclosures <strong>of</strong> Potential Conflicts <strong>of</strong> Interest<br />
Author<br />
Employment or<br />
Leadership<br />
Positions<br />
Consultant or<br />
Advisory Role<br />
George W. Sledge Jr.*<br />
Fatima Cardoso Abraxis<br />
BioScience;<br />
AstraZeneca;<br />
Celgene; Eisai;<br />
GlaxoSmithKline;<br />
Johnson &<br />
Johnson;<br />
Novartis; Pfizer;<br />
Roche<br />
Eric P. Winer Novartis; Roche/<br />
Genentech (U)<br />
Martine J. Piccart Amgen; Bayer<br />
Schering<br />
Pharma;<br />
Boehringer<br />
Ingelheim;<br />
Bristol-Myers<br />
Squibb;<br />
GlaxoSmithKline;<br />
PharmaMar;<br />
Roche; San<strong>of</strong>i<br />
*No relevant relationships to disclose.<br />
1. Greenberg P, Hortobagyi G, Smith T, et al. Long-term follow-up <strong>of</strong><br />
patients with complete remission following combination chemotherapy for<br />
metastatic breast cancer. J Clin Oncol. 1996;14:2197-2205.<br />
2. Giordano S, Buzdar A, Smith T, et al. Is breast cancer survival improving?<br />
Cancer. 2004;100(1):44-52.<br />
3. Tevaarwerk A, Gray R, Schneider B, et al. Survival in Metastatic Breast<br />
Cancer (MBC): No Evidence for Improved Survival Following Distant Recurrence<br />
after Adjuvant Chemotherapy. Presented at: San Antonio Breast<br />
Cancer Symposium. 2011; San Antonio, TX.<br />
4. Slamon D, et al. Use <strong>of</strong> chemotherapy plus a monoclonal antibody<br />
against HER2 for metastatic breast cancer that overexpresses HER2. N Engl<br />
J Med. 2001;344:783-792.<br />
5. Cortes J, O’Shaughnessy J, Loesch D, et al. Eribulin monotherapy<br />
versus treatment <strong>of</strong> physician’s choice in patients with metastatic breast<br />
cancer (EMBRACE): a phase 3 open-label randomised study. Lancet. 2011;<br />
377:914-923.<br />
6. A’Hern R, Smith I, Ebbs S. Chemotherapy and survival in advanced<br />
breast cancer: The inclusion <strong>of</strong> doxorubicin in Cooper type regimens. Br J<br />
Cancer. 1993;67:801-805.<br />
7. Gibson L, Lawrence D, Dawson C, et al. Aromatase inhibitors for<br />
treatment <strong>of</strong> advanced breast cancer in postmenopausal women. Cochrane<br />
Database Syst Rev. 2009; CD003370.<br />
8. Broglio K, Berry D. Detecting an overall survival benefit that is derived<br />
from progression-free survival. J Natl Cancer Inst. 2009;101:1642-1649.<br />
9. Pagani O, Senkus E, Wood W, et al. International guidelines for<br />
management <strong>of</strong> metastatic breast cancer (MBC) from the European School <strong>of</strong><br />
<strong>Oncology</strong> (ESO)-MBC Task Force: Can metastatic breast cancer be cured?<br />
J Natl Cancer Inst. 2010;102(7):456-463.<br />
10. Cardoso F, Di Leo A, Lohrisch C, et al. Second and subsequent lines <strong>of</strong><br />
chemotherapy for metastatic breast cancer: What did we learn in the last two<br />
decades? Ann Oncol. 2002;13:197-207.<br />
11. Burzykowski T, Buyse M, Piccart-Gebhart M, et al. Evaluation <strong>of</strong><br />
tumor response, disease control, progression-free survival, and time to pro-<br />
subtyping continues to advance therapy. The twin revolutions<br />
<strong>of</strong> cancer genomics and computational chemistry<br />
should lead to many new treatment options for patients in<br />
coming years.<br />
Acknowledgments<br />
M. Piccart would like to thank Jose Baselga, MD, PhD for<br />
useful advice and Dimitrios Zardavas, MD, for great help in<br />
preparation <strong>of</strong> the tables listing newer targets and drugs.<br />
Stock<br />
Ownership Honoraria<br />
REFERENCES<br />
Abraxis<br />
BioScience;<br />
AstraZeneca;<br />
Celgene; Eisai;<br />
GlaxoSmithKline;<br />
Johnson &<br />
Johnson;<br />
Novartis; Pfizer;<br />
Roche<br />
Amgen; Bayer<br />
Schering<br />
Pharma;<br />
Boehringer<br />
Ingelheim;<br />
Bristol-Myers<br />
Squibb;<br />
GlaxoSmithKline;<br />
PharmaMar;<br />
Roche; San<strong>of</strong>i<br />
Research<br />
Funding<br />
AstraZeneca;<br />
Novartis<br />
Roche/Genentech<br />
Bristol-Myers<br />
Squibb; Pfizer;<br />
Roche<br />
Expert<br />
Testimony<br />
Other<br />
Remuneration<br />
gression as potential surrogate end points in metastatic breast cancer. J Clin<br />
Oncol. 2008;26:1987-1992.<br />
12. Mauri D, Pavlidis N, Polyzos N, et al. Survival with aromatase<br />
inhibitors and inactivators versus standard hormonal therapy in advanced<br />
breast cancer: Meta-analysis. J Natl Cancer Inst. 2006;98:1285-1291.<br />
13. Piccart-Gebhart M, Burzykowski T, Buyse M, et al. Taxanes alone or in<br />
combination with anthracyclines as first-line therapy <strong>of</strong> patients with metastatic<br />
breast cancer. J Clin Oncol. 2008;26:1980-1986.<br />
14. Cardoso F, Bedard P, Winer E, et al. International guidelines for<br />
management <strong>of</strong> metastatic breast cancer: Combination vs. sequential singleagent<br />
chemotherapy. J Natl Cancer Inst. 2009;101:1174-1181.<br />
15. Gennari A, Stockler M, Puntoni M, et al. Duration <strong>of</strong> chemotherapy for<br />
metastatic breast cancer: A systematic review and meta-analysis <strong>of</strong> randomized<br />
clinical trials. J Clin Oncol. 2011;29:2144-2149.<br />
16. Riemsma R, Forbes C, Kessels A, et al. Systematic review <strong>of</strong> aromatase<br />
inhibitors in the first-line treatment for hormone sensitive advanced or<br />
metastatic breast cancer. Breast Cancer Res Treat. 2010;123:9-24.<br />
17. Chia S, Gradishar W, Mauriac L, et al. Double-blind, randomized<br />
placebo controlled trial <strong>of</strong> fulvestrant compared with exemestane after prior<br />
nonsteroidal aromatase inhibitor therapy in postmenopausal women with<br />
hormone receptor-positive, advanced breast cancer: Results from EFECT.<br />
J Clin Oncol. 2008;26:1664-1670.<br />
18. Mehta R, Barlow W, Albain K, et al. A Phase III Randomized Trial <strong>of</strong><br />
Anastrozole Versus Anastrozole and Fulvestrant as First-Line Therapy for<br />
Postmenopausal Women with Metastatic Breast Cancer: SWOG S0226.<br />
Presented at: San Antonio Breast Cancer Symposium; 2011; San Antonio, TX.<br />
19. Klijn J, Blarney R, Boccardo F, et al. Combination LHRH-agonist plus<br />
tamoxifen treatment is superior to medical castration alone in premenopausal<br />
metastatic breast cancer. Breast Cancer Res Treat. 1998;50:227.<br />
20. Talbot D, Moiseyenko V, Van Belle S, et al. Randomised, phase II trial<br />
comparing oral capecitabine (Xeloda) with paclitaxel in patients with metastatic/advanced<br />
breast cancer pretreated with anthracyclines. Br J Cancer.<br />
2002;86:1367-1372.<br />
21. Lin N, Claus E, Sohl J, et al. Sites <strong>of</strong> distant recurrence and clinical<br />
37
outcomes in patients with metastatic triple-negative breast cancer: High<br />
incidence <strong>of</strong> central nervous system metastates. Cancer. 2008;113:2638-2645.<br />
22. Isak<strong>of</strong>f S, Goss P, Mayer E, et al. TBCRC009: A multicenter phase II<br />
study <strong>of</strong> cisplatin or carboplatin for metastatic triple-negative breast cancer<br />
and evaluation <strong>of</strong> p63/p73 as a biomarker <strong>of</strong> response. J Clin Oncol. 2011;29<br />
(suppl; abstr 1025).<br />
23. Miller K, Wang M, Gralow J, et al. Paclitaxel plus bevacizumab versus<br />
paclitaxel alone for metastatic breast cancer. N Engl J Med. 2007;357:2666-<br />
2676.<br />
24. Miles D, Chan A, Dirix L, et al. Phase III study <strong>of</strong> bevacizumab plus<br />
docetaxel compared with placebo plus docetaxel for the first-line treatment <strong>of</strong><br />
human epidermal growth factor receptor 2-negative metastatic breast cancer.<br />
J Clin Oncol. 2010;28:3239-3247.<br />
25. O’Shaughnessy J, Osborne C, Pippen J, et al. Iniparib plus chemotherapy<br />
in metastatic triple negative breast cancer. N Engl J Med. 2011;364:205-<br />
214.<br />
26. O’Shaughnessy J, Schwartzberg L, Danso M, et al. A randomized phase<br />
III study <strong>of</strong> iniparib (BSI-201) in combination with gemcitabine/carboplatin<br />
(G/C) in metastatic triple-negative breast cancer (TNBC). J Clin Oncol.<br />
2011;29:(suppl; Abstr 1007).<br />
27. Andersson M, Lidbrink E, Bjerre K, et al. Phase III randomized study<br />
comparing docetaxel plus trastuzumab with vinorelbine plus trastuzumab as<br />
first-line therapy <strong>of</strong> metastatic or locally advanced human epidermal growth<br />
factor receptor 2-positive breast cancer: The HERNATA study. J Clin Oncol.<br />
2011;29:264-271.<br />
28. Von Minckwitz G, du Bois A, Schmidt M, et al. Trastuzumab beyond<br />
progression in human epidermal growth factor receptor 2-positive advanced<br />
breast cancer: A German Breast Group 26/Breast International Group 03-05<br />
study. J Clin Oncol. 2009;27:1999-2006.<br />
29. Geyer C, Forster J, Lindquis TD, et al. Lapatinib plus capecitabine for<br />
HER2-positive advanced breast cancer. N Engl J Med. 2006;355:2733-2744.<br />
30. Blackwell K, Burstein H, Storniolo A, et al. Randomized study <strong>of</strong><br />
Lapatinib alone or in combination with trastuzumab in women with ErbB2positive,<br />
trastuzumab-refractory metastatic breast cancer. J Clin Oncol.<br />
2010;28:1124-1130.<br />
31. Lin N, Diéras V, Paul D, et al. Multicenter phase II study <strong>of</strong> lapatinib<br />
in patients with brain metastases from HER2� breast cancer. Clin Cancer<br />
Res. 2009;15:1452-1459.<br />
32. Perou C, Sorlie T, Eisen M, et al. Molecular portraits <strong>of</strong> human breast<br />
tumours. Nature. 2000;406:747-752.<br />
33. Metzker M. Sequencing technologies - the next generation. Nat Rev<br />
Genet. 2010;11:31-46.<br />
34. Baselga J, Cortés J, Kim S, et al. Pertuzumab plus trastuzumab plus<br />
docetaxel for metastatic breast cancer. N Engl J Med. <strong>2012</strong>;366:109-119.<br />
35. Pedersen M, et al. Sym004: a novel synergistic anti-epidermal growth<br />
factor receptor amtibody mixture with superior anticancer efficacy. Cancer<br />
Res. 2010;70:588-597.<br />
36. Baselga J, Swain S. Novel anticancer targets: Revisiting ERBB2 and<br />
discovering ERBB3. Nat Rev Cancer. 2009;9:463-475.<br />
37. Engelman JA, et al. MET amplification leads to gefitinib resistance in<br />
lung cancer by activating ERBB3 signaling. Science. 2007;316:1039-1043.<br />
38. Liu L, et al. Synergistic effects <strong>of</strong> foretinib with HER-targeted agents in<br />
MET and HER1- or HER2-coactivated tumor cells. Mol. Cancer Ther. 2011;<br />
10:518-530.<br />
39. Turner N, Lambros M, Horlings H, et al. Integrative molecular pr<strong>of</strong>iling<br />
<strong>of</strong> triple negative breast cancers identifies amplicon drivers and potential<br />
therapeutic targets. Oncogene. 2010;29:2013-2023.<br />
40. Turner N, Pearson A, Sharpe R, et al. FGFR1 amplification drives<br />
endocrine therapy resistance and is a therapeutic target in breast cancer.<br />
Cancer Res. 2010;70:2085-2094.<br />
38<br />
SLEDGE, CARDOSO, WINER, AND PICCART<br />
41. Andre F, Bachelot T, Campone M, et al. A multicenter, open-label phase<br />
II trial <strong>of</strong> dovitinib, an FGFR1 inhibitor, in FGFR1 amplified and nonamplified<br />
metastatic breast cancer. J Clin Oncol. 2011;29(suppl; abstr 508).<br />
42. Bianchi G, Kocsis J, Dirix L, et al. Patient-Reported Outcomes From a<br />
Randomized Phase II Study (TDM4450g/BO21976) <strong>of</strong> Trastuzumab<br />
Emtansine vs Trastuzumab Plus Docetaxel in Previously Untreated Human<br />
Epidermal Growth Factor Receptor 2-Positive Metastatic Breast Cancer.<br />
Presented at: San Antonio Breast Cancer Symposium; 2011; San Antonio, TX.<br />
43. Baselga J, Campone M, Piccart M, et al. Everolimus in Postmenopausal<br />
Hormone-Receptor-Positive Advanced Breast Cancer. N Engl J Med. <strong>2012</strong>;<br />
366:520-529.<br />
44. Yamnik R, Holz M. mTOR/S6K1 and MAPK/RSK signaling pathways<br />
coordinately regulate estrogen receptor alpha serine 167 phosphorylation.<br />
FEBS Lett. 2010;584:124-128.<br />
45. Hanahan D, Weinberg R. Hallmarks <strong>of</strong> cancer: The next generation.<br />
Cell. 2011;144:646-674.<br />
46. Finn R, Boer K, Lang I, et al. A randomized phase II study <strong>of</strong> PD<br />
0332991, cyclin-dependent kinase (CDK) 4/6 inhibitor, in combination with<br />
letrozole for first-line treatment <strong>of</strong> patients with postmenopausal, estrogen<br />
receptor (ER)-positive, human epidermal growth factor receptor 2 (HER2)negative<br />
advanced breast cancer. J Clin Oncol. 2011;29 (suppl; abstr<br />
TPS100).<br />
47. Neckers L, Workman P. Hsp90 Molecular Chaperone Inhibitors: Are<br />
We There Yet? Clin Cancer Res. <strong>2012</strong>;16:64-76.<br />
48. Ni M, Chen Y, Lim E, et al. Targeting androgen receptor in estrogen<br />
receptor-negative breast cancer. Cancer Cell. 2011;20:119-131.<br />
49. Doane A, Danso M, Lal P, et al. An estrogen-receptor negative breast<br />
cancer subset characterized by a hormonally regulated transcriptional program<br />
and response to androgen. Oncogene. 2006;25:3994-4008.<br />
50. Al-Hajj M, Wicha M, Benito-Hernandez A, et al. Prospective identification<br />
<strong>of</strong> tumorigenic breast cancer cells. Proc Natl Acad Sci U S A.<br />
2003;100:3983-3988.<br />
51. McDermott S, Wicha M. Targeting breast cancer stem cells. Mol Oncol.<br />
2010;4:404-419.<br />
52. Izrailit J, Reedijk M. Developmental pathways in breast cancer and<br />
breast tumor-initiating cells: Therapeutic Implications. Cancer Lett. <strong>2012</strong>;<br />
317:115-126.<br />
53. Stylianou S, Clarke R, Brennan K. Aberrant activation <strong>of</strong> notch<br />
signaling in human breast cancer. Cancer Res. 2006;66:1517-1525.<br />
54. Kasper M, Jaks V, Fiaschi M, et al. Hedgehog signaling in breast<br />
cancer. Carcinogenesis. 2009;30:903-911.<br />
55. Gupta P, Onder T, Jiang G, et al. Identification <strong>of</strong> selective inhibitors <strong>of</strong><br />
cancer stem cells by high-throughput screening. Cell. 2009;138:645-659.<br />
56. Joyce J, Pollard J. Microenvironmental regulation <strong>of</strong> metastasis. Nature<br />
Rev Cancer. 2009;9:239-252.<br />
57. Robert N, Diéras V, Glaspy J, et al. RIBBON-1: randomized, doubleblind,<br />
placebo-controlled, phase III trial <strong>of</strong> chemotherapy with or without<br />
bevacizumab for first-line treatment <strong>of</strong> human epidermal growth factor<br />
receptor 2-negative, locally recurrent or metastatic breast cancer. J Clin<br />
Oncol. 2011;29:1252-1260.<br />
58. Van der Veldt A, Lubberink M, Bahce I, et al. Rapid Decrease in<br />
Delivery <strong>of</strong> Chemotherapy to Tumors after Anti-VEGF Therapy: Implications<br />
for Scheduling <strong>of</strong> Anti-Angiogenic Drugs. Cancer Cell. <strong>2012</strong>;21:82-91.<br />
59. Ebos J, Lee C, Kerbel RS. Tumor and host-mediated pathways <strong>of</strong><br />
resistance and disease progression in response to antiangiogenic therapy.<br />
Clin Cancer Res. 2009;15:5020-5025.<br />
60. Ebos J, Lee C, Christensen J, et al. Multiple circulating proangiogenic<br />
factors induced by sunitinib malate are tumor-independent and correlate with<br />
antitumor efficacy. Proc Natl Acad Sci USA.2007;104:17069-17074.
A FRESH LOOK AT DUCTAL CARCINOMA IN SITU:<br />
BASIC SCIENCE TO CLINICAL MANAGEMENT<br />
CHAIR<br />
Eun-Sil S. Hwang, MD, MPH<br />
Duke University Medical Center<br />
Durham, NC<br />
SPEAKERS<br />
Victoria Seewaldt, MD<br />
Duke University Medical Center<br />
Durham, NC<br />
Beth A. Virnig, PhD<br />
University <strong>of</strong> Minnesota, School <strong>of</strong> Public Health<br />
Minneapolis, MN
Ductal Carcinoma In Situ: Challenges,<br />
Opportunities, and Uncharted Waters<br />
By Abigail W. H<strong>of</strong>fman, MD, Catherine Ibarra-Drendall, PhD, Virginia Espina, MS,<br />
Lance Liotta, MD, PhD, and Victoria Seewaldt, MD<br />
Overview: Ductal carcinoma in situ (DCIS) is a heterogeneous<br />
group <strong>of</strong> diseases that differ in biology and clinical<br />
behavior. Until 1980, DCIS represented less than 1% <strong>of</strong> all<br />
breast cancer cases. With the increased utilization <strong>of</strong> mammography,<br />
DCIS now accounts for 15% to 25% <strong>of</strong> newly<br />
diagnosed breast cancer cases in the United States. Although<br />
our ability to detect DCIS has radically improved, our understanding<br />
<strong>of</strong> the pathophysiology and factors involved in its<br />
progression to invasive carcinoma is still poorly defined. In<br />
many patients, DCIS will never progress to invasive breast<br />
cancer and these women are overtreated. In contrast, some<br />
DCIS cases are clinically aggressive and the women may be<br />
undertreated. We are able to define some <strong>of</strong> the predictors <strong>of</strong><br />
aggressive DCIS compared with DCIS <strong>of</strong> low malignant poten-<br />
DUCTAL CARCINOMA in situ is defined as a breast<br />
lesion in which neoplastic mammary epithelial cells<br />
are confined to the mammary ductal-lobular system without<br />
invasion into the surrounding stroma. DCIS accounts for<br />
approximately 15% to 25% <strong>of</strong> all female breast cancers. The<br />
primary clinical goal in the management <strong>of</strong> DCIS is to<br />
prevent the development <strong>of</strong> invasive breast cancer. In order<br />
to individualize treatment <strong>of</strong> DCIS, there is a need for<br />
biomarkers to prospectively identify DCIS with aggressive<br />
biologic potential.<br />
DCIS is highly variable in its clinical presentation, pathology,<br />
genetic/epigenetic alterations, and biologic potential. 1,2<br />
Although DCIS is a heterogeneous disease, until recently,<br />
treatment <strong>of</strong> DCIS was relatively uniform. There is emerging<br />
evidence that DCIS has a wide range <strong>of</strong> biologic phenotypes;<br />
however, we lack biomarkers to definitively identify<br />
DCIS that will progress to invasive disease. As a result,<br />
there is considerable debate regarding how best to manage<br />
patients with DCIS. The current primary management<br />
options for women with DCIS include lumpectomy plus<br />
radiation therapy; total mastectomy, with or without sentinel<br />
node biopsy, with or without reconstruction; or lumpectomy<br />
alone followed by clinical observation. Mastectomy is<br />
an effective therapy for most patients; however, for many<br />
women it represents overtreatment, particularly those with<br />
minimal lesions identified by mammography. Randomized<br />
clinical trials comparing breast-conserving surgery and radiation<br />
versus breast-conserving surgery alone show that<br />
radiation therapy reduces local recurrence by approximately<br />
50%. 3,4 Conversely, it is possible that a subgroup <strong>of</strong> patients<br />
with DCIS may not benefit from radiation following breastconserving<br />
surgery.<br />
Women with biologically aggressive DCIS are at high risk<br />
<strong>of</strong> local-regional recurrence or progression to invasive breast<br />
cancer and are logically best served by mastectomy. However,<br />
it is also clear that there are women with DCIS at a low<br />
risk for local-regional recurrence and/or progression who<br />
might be adequately treated by breast-conserving surgery<br />
alone, without radiation therapy. Emerging evidence suggests<br />
that there are also women whose DCIS is at such low<br />
risk for recurrence or progression to invasive breast cancer<br />
40<br />
tial. However, our ability to risk-stratify DCIS is still in its<br />
infancy. <strong>Clinical</strong> risk factors that predict aggressive disease<br />
and increased risk <strong>of</strong> local recurrence include young age at<br />
diagnosis, large lesion size, high nuclear grade, comedo<br />
necrosis, and involved margins. Treatment factors such as<br />
wider surgical margins and radiation therapy reduce the risk<br />
<strong>of</strong> local recurrence. DCIS represents a key intermediate in the<br />
stepwise progression to malignancy, but not all aggressive<br />
breast cancers appear to have a DCIS intermediate, notably<br />
within triple-negative breast cancer. Ongoing studies <strong>of</strong> the<br />
genetic and epigenetic alterations in precancerous breast<br />
lesions (atypia and DCIS) as well as the breast microenvironment<br />
are important for developing effective early detection<br />
and individualized targeted prevention.<br />
that they might be observed following a diagnostic biopsy<br />
and undergo “watchful waiting.” 5 However, at this time, we<br />
cannot accurately predict the course <strong>of</strong> each patient’s DCIS.<br />
Herein, we review our current understanding <strong>of</strong> 1) clinical<br />
risk factors that predict local recurrence in patients with<br />
DCIS treated with breast-conserving therapy, 2) evidence<br />
for invasive breast cancers that potentially emerge from<br />
intraepithelial neoplasia <strong>of</strong> lower grade than DCIS, and<br />
3) molecular studies that aim to improve our understanding<br />
<strong>of</strong> the biologic potential and diversity <strong>of</strong> DCIS lesions.<br />
<strong>Clinical</strong>, Treatment, and Pathologic Risk<br />
Factors That Predict Local Recurrence and<br />
Invasion in Women with DCIS<br />
Diagnosis and Prognosis<br />
Until 1980, DCIS represented less than 1% <strong>of</strong> all breast<br />
cancer cases. With the increased utilization <strong>of</strong> mammography<br />
during the 1980s, DCIS now accounts for 15% to 25% <strong>of</strong><br />
newly diagnosed breast cancer cases in the United States. 6<br />
Currently over 90% <strong>of</strong> DCIS is diagnosed by mammographic<br />
examination alone; however, mammograms detect calcifications<br />
and frequently do not identify the neoplasm size<br />
accurately. 7 Precise preoperative evaluation is required to<br />
determine the size and extent <strong>of</strong> the disease to optimize<br />
surgical management. The recommended work-up and staging<br />
<strong>of</strong> DCIS includes a history and physical examination,<br />
bilateral diagnostic mammography, pathology review, and<br />
tumor estrogen receptor (ER) determination. Breast Magnetic<br />
Resonance Imaging (MRI) is increasingly being used<br />
to evaluate DCIS; however, the exact use <strong>of</strong> breast MRI in<br />
the preoperative management <strong>of</strong> DCIS remains a matter <strong>of</strong><br />
debate. 8<br />
With a follow-up period <strong>of</strong> 25 years, it is estimated that<br />
14% to 60% <strong>of</strong> women with untreated DCIS will develop<br />
From Duke University, Durham, NC; George Mason University, Manassas, VA.<br />
Authors’ disclosures <strong>of</strong> potential conflicts <strong>of</strong> interest are found at the end <strong>of</strong> this article.<br />
Address reprint requests to Victoria Seewaldt, MD, Duke University, Box 2628, Room<br />
221A MSRB, Durham, NC 27710; email: seewa001@mc.duke.edu.<br />
© <strong>2012</strong> by <strong>American</strong> <strong>Society</strong> <strong>of</strong> <strong>Clinical</strong> <strong>Oncology</strong>.<br />
1092-9118/10/1-10
DCIS: OPPORTUNITIES AND UNCHARTED WATERS<br />
invasive breast cancer. 1 In women who receive standard<br />
treatment for DCIS, women with microinvasion have a<br />
prognosis almost identical to women with DCIS alone, and<br />
mastectomy is curative in 95% to 100% <strong>of</strong> women. 9 Furthermore,<br />
approximately 50% <strong>of</strong> the local recurrences following<br />
initial treatment for a pure DCIS is invasive cancer. 1<br />
<strong>Clinical</strong> and Treatment Risk Factors<br />
Some <strong>of</strong> the clinical manifestations and treatment factors<br />
associated with an increased risk <strong>of</strong> local recurrence following<br />
breast-conserving treatment for DCIS have been identified.<br />
Young women are at increased risk for local-regional<br />
recurrence and invasive disease. 6-10 Women with nonpalpable<br />
mammographically detected DCIS have a better prognosis<br />
than women with palpable DCIS, which is <strong>of</strong>ten<br />
associated with microinvasion and occasionally with axillary<br />
nodal involvement. Women who undergo mastectomy are at<br />
low risk for local-regional recurrence. The use <strong>of</strong> radiation<br />
therapy following breast-conserving surgery is associated<br />
with approximately 50% reduction in the risk <strong>of</strong> local<br />
recurrence. 2-4 In the National Surgical Adjuvant Breast and<br />
Bowel Project (NSABP) B-24 trial, the addition <strong>of</strong> tamoxifen<br />
further reduced the risk <strong>of</strong> local-regional recurrence in<br />
women treated with lumpectomy and radiation therapy. 11<br />
However, tamoxifen did not reduce the risk <strong>of</strong> local-regional<br />
recurrence in a second randomized trial. 4 Currently, tamoxifen<br />
treatment may be considered as a strategy to reduce the<br />
risk <strong>of</strong> breast cancer recurrence in women with ER-positive<br />
DCIS. 1,5 The benefit <strong>of</strong> tamoxifen for ER-negative DCIS is<br />
not known.<br />
Prospective and retrospective analyses have identified<br />
pathologic characteristics <strong>of</strong> DCIS that correlate with an<br />
increased risk <strong>of</strong> local recurrence following breastconserving<br />
therapy. Pathologic features that are associated<br />
with an increased risk <strong>of</strong> local-regional recurrence include<br />
KEY POINTS<br />
● DCIS is a heterogeneous disease with a complex<br />
array <strong>of</strong> biologic potential, clinical behavior, and<br />
prognosis.<br />
● Although the biologic behavior <strong>of</strong> DCIS is heterogeneous,<br />
our clinical management has remained relatively<br />
uniform.<br />
● Biomarkers are needed to tailor clinical treatment <strong>of</strong><br />
DCIS to biologic potential for recurrence and invasion,<br />
but the genomic and pathologic nature <strong>of</strong> the<br />
neoplasia that emerges during local recurrence compared<br />
with the original DCIS lesion is unknown.<br />
● Triple-negative breast cancers may not appear to<br />
have a DCIS intermediate.<br />
● Numerous studies comparing the gene expression,<br />
genetic, and epigenetic pr<strong>of</strong>iles <strong>of</strong> DCIS and invasive<br />
breast carcinomas reported a high degree <strong>of</strong> similarity<br />
between the molecular alterations in the DCIS<br />
and the invasive cancer in the same patient. This<br />
provides evidence that the aggressive phenotype <strong>of</strong><br />
the breast cancer may be determined at the level <strong>of</strong><br />
the preinvasive lesion.<br />
high nuclear grade, comedo necrosis, larger tumor size, and<br />
involved margins <strong>of</strong> excision. 1-3 The effect <strong>of</strong> pathologic<br />
factors on the risk <strong>of</strong> local recurrence in patients with DCIS<br />
varies with treatment as well as length <strong>of</strong> follow-up. 10-13<br />
Studies by Silverstein and colleagues provide evidence that<br />
DCIS size, nuclear grade, and margin status are associated<br />
with the risk <strong>of</strong> local-regional recurrence, but only in cases<br />
with small surgical margin widths. 12,13 For women with<br />
DCIS who undergo surgical excision leaving margin widths<br />
<strong>of</strong> 10 mm or more, the risk <strong>of</strong> local recurrence is unchanged<br />
by size, grade, the presence <strong>of</strong> comedo necrosis, and radiation<br />
therapy. 12 The effect <strong>of</strong> grade on local-regional recurrence<br />
<strong>of</strong> DCIS appears to be related to the length <strong>of</strong> followup.<br />
10 The NSABP B17 trial provides evidence that comedo<br />
necrosis in DCIS is associated with an increased risk <strong>of</strong><br />
local-regional recurrence in women treated with excision<br />
alone. 3 Local recurrence rates in women undergoing excision<br />
alone, after 8 years, was 40% for those with moderate or<br />
marked comedo necrosis compared with 23% for women<br />
without comedo necrosis. In women who underwent surgical<br />
excision and radiation therapy, local recurrence rates were<br />
similar for women with moderate or marked comedo necrosis<br />
relative to women without comedo necrosis (14% compared<br />
with 13%, respectively). 3 In summary, these studies<br />
suggest that there are multifaceted interactions between<br />
pathologic factors and other elements that ultimately determine<br />
the risk <strong>of</strong> local recurrence. 1-3,10-13<br />
Evidence That Triple-Negative Breast Cancers May<br />
Lack a DCIS Precursor<br />
Triple-negative breast cancers are defined as tumors that<br />
lack ER, PR, and HER2 expression. These tumors account<br />
for 10% to 15% <strong>of</strong> all breast carcinomas, depending on the<br />
thresholds used to define ER and PR positivity and the<br />
methods used for HER2 assessment. 14-16 The clinical interest<br />
in these tumors stems from the lack <strong>of</strong> targeted therapies<br />
for patients affected by this breast cancer subtype, which is<br />
more prevalent in younger women, 14 black women, 16 and<br />
BRCA1 mutation carriers, 17 and more aggressive than tumors<br />
<strong>of</strong> other molecular subtypes. 14 Patients with triplenegative<br />
cancers also have a significantly shorter survival<br />
following the first metastatic event when compared with<br />
those with non-triple-negative controls (p � 0.0001). 14 A<br />
predictive rather than prognostic classification system is<br />
required for the success <strong>of</strong> targeted therapy, and therefore,<br />
the origins and developmental mechanisms must also be<br />
clearly defined.<br />
The biologic developments <strong>of</strong> triple-negative breast cancers<br />
have debatable beginnings. Triple-negative breast cancers<br />
are thought to develop either from cells that are<br />
triple-negative from the early phase <strong>of</strong> intraductal proliferation,<br />
leading to invasion, or cells that have lost hormone<br />
receptor expression before invasion. 18 A group <strong>of</strong> high-grade<br />
DCIS lacking ER, PR ,and HER2, and expressing “basal”<br />
markers has been identified. 18 Bryan and colleagues studied<br />
66 cases <strong>of</strong> high-grade DCIS and found a small proportion<br />
(four cases, 6%) that exhibited the triple-negative phenotype<br />
with increased invasion. 18 In a study by Meijnen and colleagues,<br />
only eight out <strong>of</strong> 163 cases (5%) demonstrated<br />
triple-negative lesions. 19 In addition, tumors with BRCA1<br />
germ-line mutations are significantly more likely to be<br />
triple-negative breast cancers (p � 0.005) and have early<br />
high-grade DCIS (p � 0.0045). 17 However, it is important to<br />
41
ecognize that there are large discrepancies in the relative<br />
frequency <strong>of</strong> triple-negative subtype between DCIS and<br />
invasive disease. The prevalence <strong>of</strong> DCIS is substantially<br />
lower than invasive triple-negative breast cancers, suggesting<br />
that the latter <strong>of</strong>ten lack an obvious in situ element.<br />
Without supporting evidence or definitive conclusions, many<br />
studies have attributed this discrepancy as a result <strong>of</strong><br />
triple-negative breast cancers progressing rapidly from<br />
DCIS to invasive cancer or obliterating the DCIS precursor<br />
from which they arose. 17,18<br />
Molecular Studies <strong>of</strong> the Biologic Potential and<br />
Diversity <strong>of</strong> DCIS Lesions<br />
Although the transition from DCIS to invasive cancer is<br />
central to understanding the origins <strong>of</strong> breast cancer, little is<br />
known about the time <strong>of</strong> onset or the triggering mechanisms<br />
that promote invasion <strong>of</strong> DCIS. Some DCIS progresses to<br />
invade the stroma, but other DCIS lies dormant. This raises<br />
the question <strong>of</strong> whether there are subsets <strong>of</strong> precancerous<br />
lesions that are preprogrammed to invade and metastasize<br />
or whether the ability to invade and metastasize is acquired<br />
late in the process <strong>of</strong> progression. 20<br />
Molecular Markers Predicting Invasion <strong>of</strong> DCIS<br />
Numerous studies have compared the gene expression,<br />
genetic, and epigenetic pr<strong>of</strong>iles <strong>of</strong> DCIS and invasive breast<br />
carcinomas, showing a high degree <strong>of</strong> similarity between<br />
DCIS and invasive cancer in the same patient. 21 As expected,<br />
there is a significant difference in gene expression<br />
between normal mammary epithelial cells and DCIS, but<br />
surprisingly, DCIS and invasive breast cancer <strong>of</strong> the same<br />
histologic subtype essentially share the same genetic/epigenetic<br />
alterations and gene expression patterns. In contrast,<br />
the molecular pr<strong>of</strong>iles <strong>of</strong> breast tumors <strong>of</strong> distinct subtypes<br />
(e.g., luminal, HER2�) are significantly different. The expression<br />
and mutation status <strong>of</strong> numerous tumor suppressor<br />
and oncogenes have been analyzed in DCIS and invasive<br />
breast cancer including TP53, PTEN, PIK3CA, ERBB2,<br />
MYC; differences in the frequency <strong>of</strong> these changes were<br />
associated with tumor subtype, but none were associated<br />
with invasion. 20,22 For example, amplification <strong>of</strong> ERBB2 is<br />
specific for the HER2� subtype; however, amplification <strong>of</strong><br />
ERBB2 is observed in both DCIS and invasive breast cancer.<br />
The expression <strong>of</strong> several selected candidate genes based<br />
on their biologic function has been analyzed in DCIS. 23 Two<br />
recent studies identified molecular markers that hold promise<br />
for identifying the risk <strong>of</strong> recurrence <strong>of</strong> DCIS. 24,25 Gauthier<br />
and colleagues demonstrated that high expression <strong>of</strong><br />
COX-2 and Ki67 in DCIS correlates with higher risk <strong>of</strong> local<br />
in situ and invasive recurrence. 24 Lu and colleagues identified<br />
a molecular synergy between ERBB2 and 14–3-3� that<br />
may increase the risk <strong>of</strong> invasive progression via epithelialto-mesenchymal<br />
transition regulation. 25 A major limitation<br />
<strong>of</strong> both <strong>of</strong> these studies was the use <strong>of</strong> small cohorts. Despite<br />
the promise and need <strong>of</strong> DCIS risk biomarkers, prospective<br />
validation is difficult. Validation <strong>of</strong> a predictive biomarker<br />
may require a long waiting period to see if a patient’s DCIS<br />
lesion will progress to invasive cancer. Most patients diagnosed<br />
with DCIS do not even want to experience a delay<br />
before surgical therapy. Therefore, patients judged at low<br />
risk may be unwilling to forego treatment. Since the population<br />
<strong>of</strong> DCIS and other preinvasive proliferative lesions is<br />
42<br />
HOFFMAN ET AL<br />
heterogeneous in the same patient, the portion <strong>of</strong> the DCIS<br />
lesion sampled at the initial biopsy diagnosis, which is used<br />
for the predictive biomarker measurement, may not be the<br />
same region <strong>of</strong> the lesion that has malignant potential.<br />
There is very little data comparing the molecular genetic<br />
characteristics <strong>of</strong> the cancers that recur to the original DCIS<br />
lesion.<br />
Evidence for Acquired Invasive Potential and<br />
Intratumor Heterogeneity<br />
The stepwise model <strong>of</strong> breast tumorigenesis assumes a<br />
stepwise transition from epithelial hyperproliferation, to<br />
atypical hyperplasia, to DCIS, and finally to invasive and<br />
metastatic cancer. 22,23 This progression model is supported<br />
by human clinical and epidemiologic data and molecular<br />
clonality studies addressing the relationships between in<br />
situ and invasive regions <strong>of</strong> the same tumor and between<br />
DCIS and its local invasive recurrence. 21 There is emerging<br />
evidence that neoplastic cells with invasive potential arise<br />
frequently within DCIS lesions, but are held in check and<br />
only invade when further molecular changes occur. 20<br />
Emerging data suggest a critical role for the microenvironment<br />
in promoting invasion <strong>of</strong> DCIS.<br />
Numerous studies have shown that DCIS exhibits intratumoral<br />
heterogeneity. The prevailing model for explaining<br />
this heterogeneity, the clonal evolution model, has recently<br />
been challenged by proponents <strong>of</strong> the cancer stem cell<br />
hypothesis. 26 To investigate this issue, Kornelia Polyak’s<br />
group performed combined analyses <strong>of</strong> markers associated<br />
with cellular differentiation states and genotypic alterations<br />
in DCIS and invasive breast cancer and evaluated diversity<br />
with ecological and evolutionary methods. Such studies<br />
demonstrated a high degree <strong>of</strong> genetic heterogeneity both<br />
within and between distinct tumor cell populations that<br />
were defined based on markers <strong>of</strong> cellular phenotypes including<br />
stem cell-like characteristics. 26 The degree <strong>of</strong> diversity<br />
correlated with clinically relevant breast tumor<br />
subtypes. 26 This supports the concept that molecular alterations<br />
at the level <strong>of</strong> DCIS may precede and determine the<br />
breast tumor subtype.<br />
There is emerging evidence that adaptation to hypoxic<br />
and metabolic stress promotes progression <strong>of</strong> DCIS to invasive<br />
breast cancer. It is hypothesized that premalignant and<br />
malignant cells down-regulate proteins that induce apoptosis<br />
or senescence and upregulate pro-survival pathways that<br />
protect against hypoxia, DNA damage, metabolic and genotoxic<br />
stress. 20 Nevertheless, even if a cell can resist programmed<br />
cell death or senescence it will still not survive in<br />
a hypoxic, nutrient-deprived environment unless it can find<br />
alternative sources <strong>of</strong> energy for cellular functions, such as<br />
through autophagy, anaerobic respiration, or increasing the<br />
efficiency <strong>of</strong> aerobic respiration. 20 To appreciate how DCIS<br />
might progress and circumvent stress-induced death or<br />
senescence, and use alternative sources <strong>of</strong> energy, it is<br />
important to consider the stresses that affect DCIS cells and<br />
molecular signaling pathways that may act to protect<br />
against these stressors. 20 Autophagy promotes survival in<br />
the face <strong>of</strong> hypoxic and nutrient stress and is an emerging<br />
target for cancer therapy. 20 Autophagy is upregulated and<br />
colocalizes with areas <strong>of</strong> hypoxic stress, and immortalized<br />
mammary epithelial cells are more susceptible to cell death<br />
under conditions <strong>of</strong> metabolic stress. 20 DCIS growth or the<br />
growth <strong>of</strong> any intraductal proliferative lesion is limited
DCIS: OPPORTUNITIES AND UNCHARTED WATERS<br />
because <strong>of</strong> confinement within the duct and the absence <strong>of</strong><br />
a blood supply. Espina and colleagues proposed that autophagy<br />
is a major survival mechanism that is used by DCIS<br />
cells to persist and proliferate in the high-stress environment<br />
<strong>of</strong> the intraductal space and can be a main determinant<br />
<strong>of</strong> acquired DCIS cell fate in response to metabolic<br />
stress. 20<br />
Evidence for Preprogramming <strong>of</strong> Precancerous Breast Lesions<br />
Triple-negative breast cancers frequently exhibit aggressive<br />
clinical behavior and are many times metastatic at<br />
diagnosis. In addition to being poorly differentiated, as<br />
outlined above, triple-negative breast cancers are thought to<br />
have a lower proportion <strong>of</strong> associated DCIS compared with<br />
other types <strong>of</strong> breast carcinoma. These observations challenge<br />
whether the stepwise model <strong>of</strong> mammary carcinogenesis<br />
adequately models initiation and progression <strong>of</strong> triplenegative<br />
breast cancer. In contrast to the stepwise model,<br />
aggressive cancers may emerge from low-grade DCIS or<br />
atypical proliferative lesions. There is growing recognition<br />
that the activated (e.g., phosphorylated) state <strong>of</strong> cellular<br />
protein signaling networks can play a key role in breast<br />
cancer initiation and progression 27 and provide information<br />
about the functional state <strong>of</strong> the cancer cell that cannot be<br />
accurately predicted based on genomic sequencing alone.<br />
Likewise, it has been recently shown that single biomarkers<br />
are not adequate to capture the complex changes in signaling<br />
networks activated during breast cancer initiation. 28<br />
Ibarra-Drendall et al. are currently testing the hypothesis<br />
that mammary atypia from high-risk women exhibits activation<br />
<strong>of</strong> phospho-protein signaling networks activated in<br />
aggressive triple-negative breast cancer. 29 We performed<br />
Reverse-Phase Protein Microarray (RPMA) pr<strong>of</strong>iling <strong>of</strong> cyto-<br />
Authors’ Disclosures <strong>of</strong> Potential Conflicts <strong>of</strong> Interest<br />
Employment or<br />
Leadership<br />
Positions<br />
Consultant or<br />
Advisory Role<br />
Author<br />
Abigail W. H<strong>of</strong>fman*<br />
Catherine Ibarra-Drendall*<br />
Virginia Espina Theranostics<br />
Health<br />
Lance Liotta*<br />
Victoria Seewaldt*<br />
*No relevant relationships to disclose.<br />
1. Burstein HJ, Polyak K, Wong JS, et al. Ductal carcinoma in situ <strong>of</strong> the<br />
breast. N Engl J Med. 2004;350:1430-1441.<br />
2. Schnitt SJ. Local outcomes in ductal carcinoma in situ based on patient<br />
and tumor characteristics. J Natl Cancer Inst Monogr. 2010;41:158-161.<br />
3. Fisher ER, Dignam J, Tan-Chiu E, et al. Pathologic findings from the<br />
National Surgical Adjuvant Breast Project (NSABP) eight-year update <strong>of</strong><br />
Protocol B-17: intraductal carcinoma. Cancer. 1999;86:429-438.<br />
4. Houghton J, George WD, Cuzick J, et al. Radiotherapy and tamoxifen in<br />
women with completely excised ductal carcinoma in situ <strong>of</strong> the breast in the<br />
UK, Australia, and New Zealand: randomised controlled trial. Lancet. 2003;<br />
362:95-102.<br />
5. Chen YY, DeVries S, Anderson J, et al. Pathologic and biologic response<br />
to preoperative endocrine therapy in patients with ER-positive ductal carcinoma<br />
in situ. BMC Cancer. 2009;9:285.<br />
6. Brinton LA, Sherman ME, Carreon JD, et al. Recent trends in breast<br />
cancer among younger women in the United States. J Natl Cancer Inst.<br />
2008;100:1643-1648.<br />
7. Evans A, Pinder S, Wilson R, et al. Ductal carcinoma in situ <strong>of</strong> the<br />
logic atypia obtained prospectively from unaffected highrisk<br />
women in our cohort. 30,31 We observed coactivation <strong>of</strong><br />
Akt/mTOR/insulin- and IL6/Stat3/vimentin-network signaling<br />
in cytologic atypia and found a statistically significant<br />
association between body mass index <strong>of</strong> �30 kg/m 2 and<br />
vimentin expression (p � 0.028). 30,31 Limited immunohistochemistry<br />
studies demonstrated vimentin expression in<br />
atypical mammary epithelial cells, 31 which in turn correlates<br />
with activation <strong>of</strong> Stat3 and IL6 levels in patientmatched<br />
mammary fluids. 32 Both Akt/mTOR and IL6/Stat3<br />
signaling have been implicated in the aggressive biology <strong>of</strong><br />
triple-negative breast cancer and epithelial-to-mesenchymal<br />
transition. These studies provide preliminary evidence that<br />
the biologically aggressive atypia may be present before the<br />
development <strong>of</strong> an invasive triple-negative breast cancer.<br />
Conclusion<br />
Currently, there are many unanswered questions regarding<br />
the diagnosis and management <strong>of</strong> DCIS. Despite the<br />
complexity and heterogeneous nature <strong>of</strong> DCIS, all women<br />
are treated with a “one size fits all” scenario that includes<br />
mastectomy versus breast-conserving surgery with radiation<br />
therapy. Most recently, Oncotype DX Breast Cancer<br />
Assay for DCIS 33 has been developed and is being utilized to<br />
obtain an estimate <strong>of</strong> 10-year risk <strong>of</strong> local recurrence, with<br />
the hope <strong>of</strong> providing guidance to treatment decision. But<br />
the success <strong>of</strong> this multigene assay in predicting the fate<br />
<strong>of</strong> DCIS remains to be seen. An essential goal <strong>of</strong> future<br />
research should focus on the development <strong>of</strong> accurate riskstratification<br />
methods based on a comprehensive understanding<br />
<strong>of</strong> the biologic, pathologic, radiologic, and clinical<br />
factors associated with DCIS.<br />
Stock<br />
Ownership Honoraria<br />
REFERENCES<br />
Research<br />
Funding<br />
Expert<br />
Testimony<br />
Other<br />
Remuneration<br />
breast: correlation between mammographic and pathologic findings. AJR<br />
Am J Roentgenol. 1994;162:1307-1311.<br />
8. Kuhl CK, Schrading S, Bieling HB, et al. MRI for diagnosis <strong>of</strong> pure<br />
carcinoma in situ: a prospective observational study. Lancet. 2007;370:485-<br />
492.<br />
9. Silver SA, Tavassoli FA. Mammary ductal carcinoma in situ with<br />
microinvasion. Cancer. 1998;82:2382-2390.<br />
10. Solin LJ, Fourquet A, Vicini FA, et al. Long-term outcome after<br />
breast-conservation treatment with radiation for mammographically detected<br />
ductal carcinoma in situ <strong>of</strong> the breast. Cancer. 2005;103:1137-1146.<br />
11. Fisher B, Dignam J, Wolmark N, et al. Tamoxifen in treatment <strong>of</strong><br />
intraductal breast cancer: National Surgical Adjuvant Breast and Bowel<br />
Project B-24 randomised controlled trial. Lancet. 1999;353:1993-2000.<br />
12. Silverstein MJ, Lagios MD, Craig PH, et al. A prognostic index for<br />
ductal carcinoma in situ <strong>of</strong> the breast. Cancer. 1996;77:2267-2274.<br />
13. Silverstein MJ, Lagios MD, Groshen S, et al. The influence <strong>of</strong> margin<br />
width on local control <strong>of</strong> ductal carcinoma in situ <strong>of</strong> the breast. N Engl J Med.<br />
1999;340:1455-1461.<br />
43
14. Dent R, Trudeau M, Pritchard KI, et al. Triple-negative breast cancer:<br />
clinical features and patterns <strong>of</strong> recurrence. Clin Cancer Res. 2007;13:4429-<br />
4434.<br />
15. Haffty BG, Yang Q, Reiss M, et al. Locoregional relapse and distant<br />
metastasis in conservatively managed triple negative early-stage breast<br />
cancer. J Clin Oncol. 2006;24:5652-5657.<br />
16. Harris LN, Broadwater G, Lin NU, et al. Molecular subtypes <strong>of</strong> breast<br />
cancer in relation to paclitaxel response and outcomes in women with<br />
metastatic disease: results from CALGB 9342. Breast Cancer Res. 2006;8:R66.<br />
17. Arun B, Vogel KJ, Lopez A, et al. High prevalence <strong>of</strong> preinvasive lesions<br />
adjacent to BRCA1/2-associated breast cancers. Cancer Prev Res. 2009;2:122-<br />
127.<br />
18. Bryan BB, Schnitt SJ, Collins LC. Ductal carcinoma in situ with<br />
basal-like phenotype: a possible precursor to invasive basal-like breast<br />
cancer. Mod Pathol. 2006;19:617-621.<br />
19. Meijnen P, Peterse JL, Antonini N, et al. Immunohistochemical categorisation<br />
<strong>of</strong> ductal carcinoma in situ <strong>of</strong> the breast. Br J Cancer. 2008;98:137-<br />
142.<br />
20. Espina V, and Liota LA. What is the malignant nature <strong>of</strong> human ductal<br />
carcinoma in situ? Nature Reviews Cancer. 2011;11:68-75.<br />
21. Polyak K. Molecular markers for the diagnosis and management <strong>of</strong><br />
ductal carcinoma in situ. J Natl Cancer Inst Monogr. 2010;41:210-213.<br />
22. Sgroi DC. Preinvasive breast cancer. Annu Rev Pathol. 2010;5:193-221.<br />
23. Kuerer HM, Albarracin CT, Yang WT, et al. Ductal carcinoma in situ:<br />
state <strong>of</strong> the science and roadmap to advance the field. J Clin Oncol.<br />
2009;27:279-288.<br />
24. Gauthier ML, Berman HK, Miller C, et al. Abrogated response to<br />
44<br />
HOFFMAN ET AL<br />
cellular stress identifies DCIS associated with subsequent tumor events and<br />
defines basal-like breast tumors. Cancer Cell. 2007;12:479-491.<br />
25. Lu J, Guo H, Treekitkarnmongkol W, et al. 14-3-3zeta cooperates with<br />
ErbB2 to promote ductal carcinoma in situ progression to invasive breast<br />
cancer by inducing epithelial-mesenchymal transition. Cancer Cell. 2009;16:<br />
195-207.<br />
26. Polyak K. Breast cancers: origins and evolution. J Clin Invest. 2007;<br />
117:3155-3163.<br />
27. Sachs K, Perez O, Pe’er D, et al. Causal protein-signaling networks<br />
derived from multiparameter single-cell data. Science. 2005;308:523-529.<br />
28. Chen FL, Xia W, and Spector NL. Acquired resistance to small molecule<br />
ErbB2 tyrosine kinase inhibitors. Clin Cancer Res. 2008;14:6730-6734.<br />
29. Ibarra-Drendall C, Seewaldt V. Evidence for the Warburg Effect in<br />
Mammary Atypia from High-Risk Women. Oral presentation at: Gordon<br />
Conference in Mammary Gland Biology; June 2011; Newport, RI.<br />
30. Ibarra-Drendall C, Troch MM, Barry WT, et al. Pilot and feasibility<br />
study: prospective proteomic pr<strong>of</strong>iling <strong>of</strong> mammary epithelial cells from<br />
high-risk women provides evidence <strong>of</strong> activation <strong>of</strong> pro-survival pathways.<br />
Breast Cancer Res Treat. Epub 2011 Jun 7.<br />
31. Pilie PG, Ibarra-Drendall C, Troch MM, et al. Protein microarray<br />
analysis <strong>of</strong> mammary epithelial cells from obese and nonobese women at high<br />
risk for breast cancer: Feasibility data. Cancer Epidemiol Biomarkers Prev.<br />
2011 20:476-482.<br />
32. Seewaldt V. Evidence for the Warburg Effect in Mammary Atypia from<br />
High-Risk Women. Oral presentation at: Gordon Conference in Mammary<br />
Gland Biology; June 2011; Newport, RI.<br />
33. Oncotype DX Breast Cancer Assay for DCIS (www.oncotypedx.com).<br />
Accessed March 15, <strong>2012</strong>.
Ductal Carcinoma In Situ, and the Influence<br />
<strong>of</strong> the Mode <strong>of</strong> Detection, Population<br />
Characteristics, and Other Risk Factors<br />
By Beth A. Virnig, PhD, MPH, Shi-Yi Wang, MD, MS, and Todd M. Tuttle, MD, MS<br />
Overview: Approximately 25% <strong>of</strong> breast cancers in the United<br />
States are diagnosed as ductal carcinoma in situ (DCIS). Rates<br />
<strong>of</strong> DCIS have risen from 5.8 per 100,000 women in the 1970s to<br />
32.5 per 100,000 in 2004. This pattern is generally attributed to<br />
increased use <strong>of</strong> screening mammography. DCIS is a major<br />
risk factor for invasive breast cancer, and considerable controversy<br />
remains about whether DCIS should be considered a<br />
direct precursor <strong>of</strong> invasive breast cancer. There is, however,<br />
a general consensus that DCIS represents an intermediate<br />
step between normal breast tissue and invasive breast cancer.<br />
Although the majority <strong>of</strong> major risk factors are similar for DCIS<br />
and invasive breast cancer, prognostic factors including estrogen<br />
and progesterone receptor status and HER2 positivity<br />
DUCTAL CARCINOMA in situ (DCIS) is noninvasive<br />
breast cancer that encompasses a wide spectrum <strong>of</strong><br />
diseases ranging from low-grade lesions that are not life<br />
threatening to high-grade lesions that may harbor foci <strong>of</strong><br />
invasive breast cancer. DCIS is typically classified according<br />
to architectural pattern (solid, cribriform, papillary, and<br />
micropapillary), tumor grade (high, intermediate, and low),<br />
and the presence or absence <strong>of</strong> comedo necrosis. Approximately<br />
25% <strong>of</strong> breast cancers diagnosed in the United States<br />
are DCIS. 1<br />
The fundamental question underlying treatment and detection<br />
<strong>of</strong> DCIS is whether it should be considered a direct<br />
precursor <strong>of</strong> invasive breast cancer. There is general consensus<br />
that DCIS represents an intermediate step between<br />
normal breast tissue and invasive breast cancer. However,<br />
because <strong>of</strong> general treatment patterns and the recognition<br />
that even a biopsy may remove an important portion <strong>of</strong> the<br />
tumor, the natural history <strong>of</strong> untreated DCIS is unknown<br />
and likely is unknowable. This article provides updates on a<br />
report requested by the National Institutes <strong>of</strong> Health Office<br />
<strong>of</strong> Medical Applications <strong>of</strong> Research, which is available at<br />
www.ahrq.gov//clinic/epcix.htm and summarized by Virnig. 2<br />
The report addresses key questions about the incidence <strong>of</strong><br />
DCIS and the associated risk factors, the effect <strong>of</strong> magnetic<br />
resonance imaging (MRI) and sentinel lymph node biopsy<br />
(SLNB) on patient outcomes, and the effect <strong>of</strong> tumor and<br />
patient characteristics, surgery, radiation, and systemic<br />
treatment on outcomes.<br />
The incidence <strong>of</strong> DCIS has increased dramatically since<br />
the early 1970s from 5.8 per 100,000 in 1975 to 32.5 per<br />
100,000 in 2004. Yet, the rate <strong>of</strong> DCIS is considerably less<br />
than the invasive breast cancer incidence, estimated to be<br />
124.3 per 100,000 in 2004. 1 The incidence <strong>of</strong> the most<br />
aggressive subtype <strong>of</strong> DCIS—comedo—has not increased as<br />
rapidly as the less aggressive noncomedo form. It is not clear<br />
whether these patterns reflect screening picking up more<br />
noncomedo tumors or pathologists using the comedo classification<br />
less <strong>of</strong>ten.<br />
The increased risk <strong>of</strong> invasive breast cancer associated<br />
with hormone replacement therapy (HRT) with estrogen<br />
plus progestin is well established. The reported association<br />
between HRT and DCIS is inconsistent across studies. 3-5<br />
are less well studied but look to have similar value in both<br />
cases. The use <strong>of</strong> postdiagnostic MRI, sentinel lymph node<br />
biopsy, surgery, radiation, and endocrine therapy are all<br />
evolving as evidence from randomized and observational<br />
studies continues to accumulate. Treatment <strong>of</strong> DCIS requires<br />
a balance between risk <strong>of</strong> overtreatment and undertreatment.<br />
Ongoing studies are focusing on whether partial-breast irradiation<br />
is as effective as whole-breast irradiation and whether<br />
treatment with endocrine therapies can reduce the likelihood<br />
<strong>of</strong> either invasive breast cancer or DCIS recurrence. In general,<br />
treatment decisions should take into account the likelihood<br />
that an apparent case <strong>of</strong> DCIS could harbor foci <strong>of</strong><br />
invasive disease.<br />
The strongest evidence that the increased incidence <strong>of</strong><br />
DCIS can be attributed to the use <strong>of</strong> screening mammography<br />
comes from eight population-based trials <strong>of</strong> mammography<br />
screening. Mammographic screening consistently was<br />
more likely to lead to the diagnosis <strong>of</strong> invasive breast cancer<br />
than <strong>of</strong> DCIS—all trials reported that less than 20% <strong>of</strong><br />
screen-detected breast cancers were DCIS. There is considerable<br />
evidence that the detection <strong>of</strong> DCIS is greatest at<br />
baseline screening. For example, the breast cancer surveillance<br />
consortium reported DCIS incidence <strong>of</strong> 150 per 1,000<br />
screening mammograms on first screening and incidence <strong>of</strong><br />
0.83 per 1,000 for subsequent screening mammograms. 6<br />
Although several trials have assessed the value <strong>of</strong> tamoxifen<br />
or raloxifene for preventing DCIS, the trials, in reality,<br />
were designed to assess the value <strong>of</strong> the agents for preventing<br />
invasive breast cancer, with DCIS as a secondary outcome.<br />
Two large, controlled trials showed that tamoxifen<br />
had a protective role on the development <strong>of</strong> DCIS (and<br />
invasive breast cancer), though the magnitude <strong>of</strong> effects<br />
varied somewhat. 7,8 The Study <strong>of</strong> Tamoxifen and Raloxifene<br />
(STAR) trial found 40% higher incidence <strong>of</strong> DCIS for women<br />
treated with raloxifene than with tamoxifen. The study also<br />
found both treatments decreased the risk <strong>of</strong> invasive breast<br />
cancer by half. The women randomly assigned to raloxifene<br />
had 36% fewer uterine cancers and 29% fewer blood clots<br />
than the women assigned to tamoxifen. 9<br />
The Continuing Outcomes Relevant to Evista (CORE)/<br />
Multiple Outcomes <strong>of</strong> Raloxifene Evaluation (MORE) study<br />
found significantly reduced incidence <strong>of</strong> invasive breast<br />
cancer associated with raloxifene (hazard ratio [HR] � 0.41),<br />
but a nonsignificant increase in the incidence <strong>of</strong> DCIS<br />
among the treated women (HR � 1.78). 10 The inconsistent<br />
effect <strong>of</strong> raloxifene and tamoxifen on DCIS and invasive<br />
From the University <strong>of</strong> Minnesota School <strong>of</strong> Public Health and University <strong>of</strong> Minnesota<br />
Medical School, Minneapolis, MN.<br />
Authors’ disclosures <strong>of</strong> potential conflicts <strong>of</strong> interest are found at the end <strong>of</strong> this article.<br />
Address reprint requests to Beth A. Virnig, PhD, MPH, University <strong>of</strong> Minnesota School <strong>of</strong><br />
Public Health, 420 Delaware Street SE, MMC 729, Minneapolis, MN 55455; email:<br />
virni001@umn.edu.<br />
© <strong>2012</strong> by <strong>American</strong> <strong>Society</strong> <strong>of</strong> <strong>Clinical</strong> <strong>Oncology</strong>.<br />
1092-9118/10/1-10<br />
45
east cancer incidence deserves further investigation and<br />
may, ultimately, shed light on the biology <strong>of</strong> DCIS and<br />
invasive breast cancer and the factors that control invasive<br />
progression.<br />
The use <strong>of</strong> breast MRI for patients with DCIS is not yet<br />
established. MRI can influence treatment recommendations<br />
for some patients by identifying occult disease not visualized<br />
with mammography. Among patients with DCIS, three<br />
studies found that the sensitivity <strong>of</strong> detecting multicentric<br />
disease is higher with MRI compared with mammography.<br />
11,12 Breast MRI can potentially influence treatment<br />
decisions by providing more accurate information on the size<br />
and extent <strong>of</strong> the known DCIS. Such findings may determine<br />
the choice <strong>of</strong> breast-conserving surgery (BCS) compared<br />
with mastectomy or the width <strong>of</strong> excision margins. Studies<br />
have been mixed when comparing whether MRI overestimates<br />
or underestimates tumor size relative to mammography.<br />
The potential benefits <strong>of</strong> MRI include fewer re-excisions<br />
after BCS and decreased local recurrence rates after excision.<br />
However, no studies to date have reported that MRI<br />
yields these improved patient outcomes. Breast MRI may<br />
also have potential disadvantages such as increased patient<br />
anxiety, costs, and unnecessary use <strong>of</strong> breast biopsy; for<br />
example, one study <strong>of</strong> MRI for women with DCIS reported<br />
that 47.6% <strong>of</strong> the biopsies stemming from a positive MRI<br />
were negative. 13 If MRI leads to overestimation <strong>of</strong> the extent<br />
and size <strong>of</strong> DCIS in some patients, it points to MRI use<br />
resulting in more mastectomies and, for BCS, wider excisions<br />
and their associated less favorable cosmetic outcomes.<br />
SLNB is recommended for patients with invasive breast<br />
cancer to determine prognosis and to guide adjuvant treatment<br />
decisions. The risk <strong>of</strong> SLN metastasis is higher for<br />
patients with a final diagnosis <strong>of</strong> DCIS with microinvasion<br />
compared with pure DCIS (9.3% vs. 4.8%). 14 In addition,<br />
approximately 15% <strong>of</strong> patients who are initially diagnosed<br />
with DCIS on core needle biopsy have invasive breast cancer<br />
identified in the excision or mastectomy specimen. 15 Thus,<br />
some patients may require axillary lymph node staging<br />
after definitive surgical treatment for presumed DCIS. Although<br />
SLNB is feasible for most patients after excision,<br />
it is not feasible after mastectomy. 16 Thus, some authors<br />
recommend routine SLNB for women with high-risk DCIS<br />
(palpable mass, comedo necrosis) and for those patients<br />
undergoing mastectomy. 17<br />
The 10-year breast cancer mortality rate from DCIS is less<br />
46<br />
KEY POINTS<br />
● Most <strong>of</strong> the risk factors for DCIS and invasive breast<br />
cancer are similar.<br />
● Rates <strong>of</strong> DCIS have risen and are associated with<br />
rising use <strong>of</strong> mammography.<br />
● The value <strong>of</strong> breast MRI for women diagnosed with<br />
DCIS is not clear.<br />
● Sentinel lymph node biopsy is thought to be most<br />
valuable for women with high-risk disease or undergoing<br />
mastectomy.<br />
● Treatment <strong>of</strong> DCIS includes surgery, radiation, and<br />
endocrine therapy. Studies <strong>of</strong> the optimal treatment<br />
are ongoing.<br />
VIRNIG, WANG, AND TUTTLE<br />
than 2%. 18 Therefore, the primary outcomes for DCIS are<br />
ipsilateral and contralateral breast cancer recurrence. Many<br />
<strong>of</strong> the prognostic factors are shared between DCIS and<br />
invasive cancer. The local and recurrence rates for DCIS are<br />
between 10% and 24% after 10 years. Younger age at<br />
diagnosis is a consistent adverse prognostic factor. Women<br />
older than 40 or 50 years consistently have reduced risk <strong>of</strong><br />
DCIS or invasive recurrence than younger women. This increased<br />
risk may reflect the observation that DCIS in younger<br />
women is more <strong>of</strong>ten symptomatic and more extensive.<br />
Studies <strong>of</strong> racial differences in DCIS recurrence point to a<br />
somewhat complex story. Studies <strong>of</strong> single treatments that<br />
only adjust for demographic factors alone, 19,20 report that<br />
black women with DCIS are more likely than white women<br />
to die from breast cancer (response rate [RR] � 1.35) or<br />
experience an invasive recurrence (RR � 1.4). However, the<br />
studies that adjust for a more detailed set <strong>of</strong> tumor factors<br />
find no difference between racial groups and risk <strong>of</strong> DCIS or<br />
invasive recurrence (RR � 1.12). This suggests that there<br />
may be differences in the tumor characteristics between<br />
black and white women. There also may be systematic<br />
differences in aggressiveness <strong>of</strong> DCIS treatment by race.<br />
Positive surgical margins are consistently associated with<br />
increased DCIS and invasive breast cancer recurrence, although<br />
the magnitude <strong>of</strong> excess risk varies considerably. 21<br />
There is considerable debate regarding whether width <strong>of</strong> a<br />
negative margin (width <strong>of</strong> a margin negative for tumor cells)<br />
is associated with a decreased risk <strong>of</strong> recurrence. In general,<br />
larger tumors were associated with higher rates <strong>of</strong> local<br />
DCIS and invasive recurrence than smaller tumors. 18,20<br />
Although somewhat inconsistently labeled, a higher pathologic<br />
or nuclear grade (grade 3) was consistently associated<br />
with a higher probability <strong>of</strong> local DCIS or invasive recurrence<br />
than an intermediate or low grade (grade 2 or 1).<br />
Comedo necrosis, a factor unique to DCIS, is strongly and<br />
consistently associated with poorer outcomes and increased<br />
risk <strong>of</strong> DCIS or invasive recurrence. 20<br />
Few <strong>of</strong> the important markers <strong>of</strong> tumor aggressiveness in<br />
invasive breast cancer are well studied in DCIS. Rates <strong>of</strong><br />
estrogen receptor (ER) testing for women with DCIS are<br />
rising but still lag behind testing for invasive cancer. However,<br />
rates <strong>of</strong> ER positivity as a percentage <strong>of</strong> tumors tested<br />
are similar for DCIS and invasive breast cancer with 81% to<br />
85% positivity. ER positivity has been linked with a decreased<br />
risk <strong>of</strong> recurrence in several small studies. DCIS<br />
is rarely tested for HER2 positivity, but, nonetheless, several<br />
small studies have linked it to increased risk <strong>of</strong> recurrence<br />
(RR � 1.5–3.7) and reduced time to recurrence. 22 The<br />
possibility <strong>of</strong> treating HER2-positive tumors is being studied<br />
in ongoing trials. Ongoing research is also focusing on<br />
developing genetic markers <strong>of</strong> DCIS that has more or less<br />
favorable biology.<br />
In randomized trials including NSABP-17 and the European<br />
Organization for Research and Treatment <strong>of</strong> Cancer<br />
(EORTC) randomized phase III trial 10853, whole-breast<br />
radiation therapy (RT) following BCS was associated with<br />
reduced local DCIS or invasive carcinoma recurrence. Although<br />
statistically significant, the number <strong>of</strong> events prevented<br />
per 1,000 treated women was less than 10%. Despite<br />
the reduced recurrence, RT had no effect on either breast<br />
cancer mortality or total mortality. 23-25 Neither randomized<br />
nor observational studies pointed to compelling evidence<br />
that BCS plus radiation has differing relative effectiveness
DCIS AND INFLUENCE OF RISK FACTORS<br />
in the presence or absence <strong>of</strong> adverse prognostic factors. This<br />
lack <strong>of</strong> differential effect can be seen for the most important<br />
prognostic factors, including tumor grade and size and<br />
comedo necrosis. 18 The key issue then becomes whether the<br />
absolute benefit <strong>of</strong> RT for low-risk women is small enough to<br />
justify use <strong>of</strong> BCS alone.<br />
Although outcomes between mastectomy and BCS or BCS<br />
plus RT were not studied in a randomized fashion, several<br />
observational studies reported that women undergoing mastectomy<br />
were less likely than women undergoing BCS or<br />
BCS plus RT to experience local DCIS or invasive recurrence.<br />
26,27 We found no study showing a mortality reduction<br />
associated with mastectomy over BCS with or without<br />
radiation.<br />
The literature is still evolving regarding the use <strong>of</strong> accelerated<br />
partial-breast irradiation (APBI) delivered via MammoSite<br />
or balloon-based brachytherapy 28-30 and whether<br />
it is associated with similar levels <strong>of</strong> control as whole-breast<br />
irradiation.<br />
The NSABP-24 assessed the value <strong>of</strong> tamoxifen following<br />
DCIS diagnosis and found tamoxifen was associated with<br />
statistically significant reductions in local recurrence (RR �<br />
0.60) and contralateral disease (RR � 0.56). However, the<br />
absolute risk reduction in ipsilateral and contralateral disease<br />
was small. Ongoing studies such as the NSABP-37 are<br />
examining the comparative effectiveness <strong>of</strong> tamoxifen and<br />
aromatase inhibitors, and the NSABP B-43 is examining use<br />
<strong>of</strong> trastuzumab for women with HER2-positive tumors.<br />
Discussion<br />
It is clear that many cases <strong>of</strong> DCIS either would not<br />
develop into invasive disease or would do so much later in<br />
life, perhaps never becoming clinically relevant. This issue<br />
<strong>of</strong> overdiagnosed DCIS because <strong>of</strong> screening is not limited to<br />
Authors’ Disclosures <strong>of</strong> Potential Conflicts <strong>of</strong> Interest<br />
Author<br />
Beth A. Virnig*<br />
Shi-Yi Wang*<br />
Todd M. Tuttle*<br />
*No relevant relationships to disclose.<br />
Employment or<br />
Leadership<br />
Positions<br />
Consultant or<br />
Advisory Role<br />
1. Brinton LA, Sherman ME, Carreon JD, Anderson WF. Recent trends in<br />
breast cancer among younger women in the United States. J Natl Cancer Inst.<br />
2008;100:1643-1648.<br />
2. Virnig BA, Tuttle TM, Shamliyan T, Kane RL. Ductal carcinoma in situ<br />
<strong>of</strong> the breast: A systematic review <strong>of</strong> incidence, treatment, and outcomes.<br />
J Natl Cancer Inst. 2008;102:170-178.<br />
3. Chlebowski RT, Hendrix SL, Langer RD, et al. Influence <strong>of</strong> estrogen plus<br />
progestin on breast cancer and mammography in healthy postmenopausal<br />
women: The Women’s Health Initiative Randomized Trial. JAMA. 2003;289:<br />
3243-3253.<br />
4. Collaborative Group on Hormonal Factors in Breast Cancer. Breast<br />
cancer and hormone replacement therapy: Collaborative reanalysis <strong>of</strong> data<br />
from 51 epidemiological studies <strong>of</strong> 52,705 women with breast cancer and<br />
108,411 women without breast cancer. Lancet. 1997;350:1047-1059.<br />
5. Reeves GK, Pirie K, Green J, et al. For the Million Women Study<br />
Collaborators. Comparison <strong>of</strong> the effects <strong>of</strong> genetic and environmental risk<br />
factors on in situ and invasive ductal breast cancer. Int J Cancer. Epub ahead<br />
2011 Sept 27.<br />
6. Ernster VL, Ballard-Barbash R, Barlow WE, et al. Detection <strong>of</strong> ductal<br />
carcinoma in situ in women undergoing screening mammography. J Natl<br />
Cancer Inst. 2002;94:1546-1554.<br />
DCIS and is part <strong>of</strong> an active policy discussion related to<br />
invasive breast cancer. There is also an aspect <strong>of</strong> underdiagnosis<br />
that must be considered. In some instances, DCIS<br />
may be underdiagnosed invasive cancer for which the pathology<br />
sections simply missed the invasive area. Overall,<br />
the arguments for a close relationship between in situ and<br />
invasive breast cancer can be found in the similarity <strong>of</strong> risk<br />
factors for both the incidence <strong>of</strong> the diseases and their<br />
similar responses to treatment.<br />
From a clinical perspective, it seems prudent to approach<br />
DCIS and invasive breast cancer as being related. Given the<br />
rate <strong>of</strong> sampling error in needle biopsies, presumptive DCIS<br />
should be treated as potential invasive cancer until a more<br />
definitive pathologic sample is available. In clinical settings,<br />
efforts should be made to make full use <strong>of</strong> markers such as<br />
estrogen and progesterone receptor status, and necrosis to<br />
differentiate women at high risk from those at lower risk <strong>of</strong><br />
developing invasive disease. Effort should be undertaken to<br />
evaluate whether HER2 status <strong>of</strong>fers any promise for clinical<br />
decision-making. Ultimately, these markers would allow<br />
for focusing aggressive treatment on those who have the<br />
greatest probability <strong>of</strong> benefit.<br />
Note: Because <strong>of</strong> space limitations, the reference list is<br />
necessarily incomplete. See the article by Virnig and colleagues<br />
2 for a more complete review.<br />
ACKNOWLEDGMENTS<br />
Funding: Agency for Healthcare Research and Quality, US<br />
Department <strong>of</strong> Health and Human Services (contract number<br />
290-02-10064-I). The authors <strong>of</strong> this report are responsible for<br />
its content. Statements in the report should not be construed as<br />
endorsement by the Agency for Healthcare Research and Quality<br />
or the U.S. Department <strong>of</strong> Health and Human Services.<br />
Stock<br />
Ownership Honoraria<br />
REFERENCES<br />
Research<br />
Funding<br />
Expert<br />
Testimony<br />
Other<br />
Remuneration<br />
7. Cuzick J, Forbes JF, Sestak I, et al. Long-term results <strong>of</strong> tamoxifen<br />
prophylaxis for breast cancer—96-month follow-up <strong>of</strong> the randomized IBIS-I<br />
trial. J Natl Cancer Inst. 2007;99:272-282.<br />
8. Powles TJ, Ashley S, Tidy A, et al. Twenty-year follow-up <strong>of</strong> the Royal<br />
Marsden randomized, double-blinded tamoxifen breast cancer prevention<br />
trial. J Natl Cancer Inst. 2007;99:283-290.<br />
9. Vogel VG, Costantino JP, Wickerham DL, et al. Effects <strong>of</strong> tamoxifen vs<br />
raloxifene on the risk <strong>of</strong> developing invasive breast cancer and other disease<br />
outcomes: The NSABP Study <strong>of</strong> Tamoxifen and Raloxifene (STAR) P-2 Trial.<br />
JAMA. 2006;295:2727-2741.<br />
10. Martino S, Cauley JA, Barrett-Connor E, et al. Continuing outcomes<br />
relevant to Evista: Breast cancer incidence in postmenopausal osteoporotic<br />
women in a randomized trial <strong>of</strong> raloxifene. J Natl Cancer Inst. 2004;96:1751-<br />
1761.<br />
11. Hwang ES, Kinkel K, Esserman LJ, et al. Magnetic resonance imaging<br />
in patients diagnosed with ductal carcinoma-in-situ: Value in the diagnosis <strong>of</strong><br />
residual disease, occult invasion, and multicentricity. Ann Surg Oncol.<br />
2003;10:381-388.<br />
12. Menell JH, Morris EA, Dershaw DD, et al. Determination <strong>of</strong> the<br />
presence and extent <strong>of</strong> pure ductal carcinoma in situ by mammography and<br />
magnetic resonance imaging. Breast J. 2005;11:382-390.<br />
47
13. Allen LR, Lago-Toro CE, Hughes JH, et al. Is there a role for MRI in the<br />
pre-operative assessment <strong>of</strong> patients with DCIS? Ann Surg Oncol. 2010;17:<br />
2395-2400.<br />
14. Katz A, Gage I, Evans S, et al. Sentinel lymph node positivity <strong>of</strong><br />
patients with ductal carcinoma in situ or microinvasive breast cancer. Am J<br />
Surg. 2006;191:761-766.<br />
15. Bruening W, Fontarosa J, Tipton K, et al. Systematic review: Comparative<br />
effectiveness <strong>of</strong> core-needle and open surgical biopsy to diagnose breast<br />
lesions. Ann Intern Med. Epub 2009 Dec 15.<br />
16. Wong SL, Edwards MJ, Chao C, et al. The effect <strong>of</strong> prior breast biopsy<br />
method and concurrent definitive breast procedure on success and accuracy <strong>of</strong><br />
sentinel lymph node biopsy. Ann Surg Oncol. 2002;9:272-277.<br />
17. Schneider C, Trocha S, McKinley B, et al. The use <strong>of</strong> sentinel lymph<br />
node biopsy in ductal carcinoma in situ. Am Surg. 2010;76:943-946.<br />
18. Fisher ER, Dignam J, Tan-Chiu E, et al. Pathologic findings from the<br />
National Surgical Adjuvant Breast Project (NSABP) eight-year update <strong>of</strong><br />
Protocol B-17: intraductal carcinoma. Cancer. 1999;86:429-438.<br />
19. Joslyn SA. Ductal carcinoma in situ: Trends in geographic, temporal,<br />
and demographic patterns <strong>of</strong> care and survival. Breast J. 2006;12:20-27.<br />
20. Li CI, Malone KE, Saltzman BS, Daling JR. Risk <strong>of</strong> invasive breast<br />
carcinoma among women diagnosed with ductal carcinoma in situ and lobular<br />
carcinoma in situ, 1988-2001. Cancer. 2006;106:2104-2112.<br />
21. Dick AW, Sorbero MS, Ahrendt GM, et al. Comparative effectiveness <strong>of</strong><br />
ductal carcinoma in situ management and the roles <strong>of</strong> margins and surgeons.<br />
J Natl Cancer Inst. 2011;103:92-104.<br />
22. Holmes P, Lloyd J, Chervoneva I, et al. Prognostic markers and<br />
long-term outcomes in ductal carcinoma in situ <strong>of</strong> the breast treated with<br />
excision alone. Cancer. 2011;117:3650-3657.<br />
23. Bijker N, Meijnen P, Peterse JL, et al. Breast-conserving treatment<br />
with or without radiotherapy in ductal carcinoma-in-situ: Ten-year results <strong>of</strong><br />
48<br />
VIRNIG, WANG, AND TUTTLE<br />
European Organisation for Research and Treatment <strong>of</strong> Cancer randomized<br />
phase III trial 10853—a study by the EORTC Breast Cancer Cooperative<br />
Group and EORTC Radiotherapy Group. J Clin Oncol. 2006;24:3381-3387.<br />
24. Holmberg L, Garmo H, Granstrand B, et al. Absolute risk reductions for<br />
local recurrence after postoperative radiotherapy after sector resection for<br />
ductal carcinoma in situ <strong>of</strong> the breast. J Clin Oncol. 2008;26:1247-1252.<br />
25. Wapnir IL, Dignam JJ, Fisher B, et al. Long-term outcomes <strong>of</strong> invasive<br />
ipsilateral breast tumor recurrences after lumpectomy in NSABP B-17 and B-24<br />
randomized clinical trials for DCIS. J Natl Cancer Inst. 2011;103:478-488.<br />
26. Meijnen P, Oldenburg HS, Peterse JL, et al. <strong>Clinical</strong> outcome after<br />
selective treatment <strong>of</strong> patients diagnosed with ductal carcinoma in situ <strong>of</strong> the<br />
breast. Ann Surg Oncol. 2008;15:235-243.<br />
27. Schouten van der Velden AP, van Vugt R, Van Dijck JA,et al. Local<br />
recurrences after different treatment strategies for ductal carcinoma in situ <strong>of</strong><br />
the breast: A population-based study in the East Netherlands. Int J Radiat<br />
Oncol Biol Phys. 2007;69:703-710.<br />
28. Jeruss JS, Kuerer HM, Beitsch PD, et al. Update on DCIS outcomes<br />
from the <strong>American</strong> <strong>Society</strong> <strong>of</strong> Breast Surgeons accelerated partial breast<br />
irradiation registry trial. Ann Surg Oncol. 2011;18:65-71.<br />
29. Israel PZ, Vicini F, Robbins AB, et al. Ductal carcinoma in situ <strong>of</strong> the<br />
breast treated with accelerated partial breast irradiation using balloon-based<br />
brachytherapy. Ann Surg Oncol. 2010;17:2940-2944.<br />
30. Keisch M, Vicini F, Beitsch P, et al. <strong>American</strong> <strong>Society</strong> <strong>of</strong> Breast<br />
Surgeons MammoSite Radiation Therapy System Registry Trial: Ductal<br />
carcinoma-in-situ subset analysis-4-year data in 194 treated lesions. Am J<br />
Surg. 2009;198:505-507.<br />
31. Fisher B, Land S, Mamounas E, et al. Prevention <strong>of</strong> invasive breast<br />
cancer in women with ductal carcinoma in situ: An update <strong>of</strong> the national<br />
surgical adjuvant breast and bowel project experience. Semin Oncol. 2001;28:<br />
400-418.
KEY QUESTIONS IN THE LOCO-REGIONAL<br />
TREATMENT OF BREAST CANCER<br />
CHAIR<br />
Richard C. Zellars, MD<br />
Sidney Kimmel Comprehensive Cancer Center at Johns Hopkins University<br />
Baltimore, MD<br />
SPEAKERS<br />
Thomas A. Buchholz, MD<br />
University <strong>of</strong> Texas M. D. Anderson Cancer Center<br />
Houston, TX<br />
Monica Morrow, MD<br />
Memorial Sloan-Kettering Cancer Center<br />
New York, NY
Postmastectomy Radiation and Partial<br />
Breast Irradiation<br />
Overview: Between 1997 and 1999, three modern postmastectomy<br />
radiation therapy (PMXRT) trials were published.<br />
These trials showed a significant benefit with respect to<br />
local control and survival in women who received adjuvant<br />
radiation after mastectomy. Despite two decades <strong>of</strong> follow-up,<br />
reanalyses, meta-analyses and vigorous and <strong>of</strong>ten acrimonious<br />
debate, questions about the benefit <strong>of</strong> PMXRT in women<br />
with one to three positive lymph nodes (LNs) remain unanswered<br />
for many. This persistent debate has limited the use<br />
BETWEEN 1997 and 1999, three randomized prospective<br />
trials <strong>of</strong> postmastectomy radiation therapy<br />
(PMXRT) were published, the Danish Premenopausal and<br />
Postmenopausal trials (82b and 82c, respectively) and the<br />
British Columbia trial. 1-3 The Danish trials (1982–1989),<br />
with approximately 1,400 patients each and the smaller<br />
Canadian trial (1978–1985) with 318 patients randomly<br />
assigned women receiving treatment with mastectomy and<br />
adjuvant systemic therapy to receive or not receive PMXRT.<br />
As seen in the earlier studies, these modern trials showed a<br />
significant difference in locoregional recurrence (LRR) rate<br />
in favor <strong>of</strong> PMXRT. However, for the first time, all three<br />
trials showed a significant improvement in overall survival<br />
(OS) also in favor <strong>of</strong> radiation. These trials, in contrast to<br />
their predecessors, benefited from modern standardized radiation<br />
therapy techniques, as well as modern chemotherapy.<br />
There was general agreement with the results <strong>of</strong> the trials<br />
with respect to patients with four or more positive lymph<br />
nodes (LNs). However, the true benefit <strong>of</strong> PMXRT in patients<br />
with one to three positive LNs was called into question.<br />
All three trials had LRR rates <strong>of</strong> 30 to 33% at 10 to 15<br />
years in patients with one to three positive LNs who did not<br />
undergo radiation. This rate is in contrast to a 13% 10-year<br />
LRR rate in similarly staged and treated disease as reported<br />
in retrospective reviews <strong>of</strong> prospective Eastern Cooperative<br />
<strong>Oncology</strong> Group (ECOG) and National Surgical Adjuvant<br />
Breast and Bowel Project (NSABP) trials. 4,5 Commonly<br />
proposed reasons for the LRR discrepancies center on what<br />
some critics would consider the nonstandard systemic and<br />
local therapy administered in these trials.<br />
The criticisms <strong>of</strong> systemic therapy concern the use <strong>of</strong> CMF<br />
(cyclophosphamide, methotrexate, fluorouracil) in two <strong>of</strong> the<br />
trials, and tamoxifen in the third. One may argue that the<br />
results <strong>of</strong> these PMXRT trials are not applicable to today’s<br />
breastpatient with cancer because 1) CMF is no longer the<br />
most common first-line chemotherapy regimen in breast<br />
cancer and 2) tamoxifen was administered without knowledge<br />
<strong>of</strong> the patients’ estrogen-(ER) and progesteronereceptor<br />
(PR) status and was prescribed for only 1 year as<br />
opposed to today’s recommended 5-year course. When one<br />
considers that systemic therapy can influence local control<br />
(e.g., B-06, B-21) the argument that substandard systemic<br />
therapy may affect LRR rate seems wholly reasonable.<br />
With respect to local therapy, some have argued that an<br />
inadequate axillary dissection may be the cause <strong>of</strong> the high<br />
rate <strong>of</strong> LRR (30 to 33%) seen in the patients with one to three<br />
50<br />
By Richard C. Zellars, MD<br />
<strong>of</strong> PMXRT. A treatment concept certain to be as vigorously<br />
debated as that <strong>of</strong> PMXRT is that <strong>of</strong> partial breast irradiation<br />
(PBI). However, unlike PMXRT, the acceptance <strong>of</strong> PBI, by<br />
community physicians, has been nothing short <strong>of</strong> meteoric<br />
and represents the first major shift in the local therapy<br />
paradigm in more than 10 years. Herein, information will be<br />
presented that will suggest we are nearing the end <strong>of</strong> one<br />
debate—PMXRT in patients with one to three positive LNs—<br />
and beginning another—appropriateness <strong>of</strong> PBI.<br />
positive LNs. The median numbers <strong>of</strong> LNs removed were<br />
seven and 11 in these trials, respectively. In contrast, the<br />
median numbers <strong>of</strong> LNs removed in the retrospective analyses<br />
<strong>of</strong> ECOG and NSABP trials, both <strong>of</strong> which reported a<br />
13% LRR rate, were 15 and 16, respectively. 4,5 The extent <strong>of</strong><br />
axillary surgery has been shown to have a local therapy<br />
benefit. 5 With that in mind, some concluded that the<br />
radiation in the modern PMXRT trials compensated for<br />
less-than-ideal surgery, and thus the benefit <strong>of</strong> radiation is<br />
exaggerated for patients with one to three positive LNs. It is<br />
this belief that prevents many physicians from <strong>of</strong>fering<br />
PMXRT to patients with breast cancer who have one to three<br />
positive LNs. 6<br />
Meta-analyses can <strong>of</strong>ten provide insight when there are<br />
apparent discrepancies between studies. The Early Breast<br />
Cancer Trialist’s Cooperative Group (EBCTCG) reviewed<br />
the records <strong>of</strong> more than 3,000 postmastectomy patients<br />
with breast cancer who had one to three positive LNs and<br />
were randomly assigned to receive or not receive adjuvant<br />
radiation. The majority <strong>of</strong> the patients received systemic<br />
therapy. In this meta-analysis the authors reported an<br />
absolute reduction in breast cancer mortality and all-cause<br />
mortality with PMXRT <strong>of</strong> 7.6% (log-rank 2p � 0.002) and<br />
5.3% (log-rank 2p � 0.05), respectively. 7<br />
There is also indirect evidence supporting PMXRT in<br />
patients with one to three positive LNs. In the NCIC trial<br />
MA.20, 1,800 women were randomly assigned after lumpectomy<br />
to whole-breast irradiation (WBI) with or without<br />
regional nodal irradiation (RNI). 8 The results <strong>of</strong> this trial<br />
were recently presented at ASCO’s 47th Annual Meeting<br />
(June 3–7, 2011, Chicago, IL). Of those enrolled onto MA.20,<br />
85% had one to three positive LNs. With 5-year median<br />
follow-up, the authors reported an improved locoregional<br />
relapse-free survival (94.5% vs. 96.8%, p � 0.02), distant<br />
disease-free survival (87% vs. 92.4%, p � 0.002) and OS<br />
(90.7% vs. 92.4%, p � 0.07) with the addition <strong>of</strong> RNI. By<br />
showing that there are substantial benefits with the addition<br />
<strong>of</strong> RNI in women with one to three positive LNs, this<br />
From the Departments <strong>of</strong> Radiation <strong>Oncology</strong> and <strong>Oncology</strong>, Sidney Kimmel Comprehensive<br />
Cancer Center, Johns Hopkins University, Baltimore, MD.<br />
Author’s disclosures <strong>of</strong> potential conflicts <strong>of</strong> interest are found at the end <strong>of</strong> this article.<br />
Address reprint requests to Richard C. Zellars, MD, 401 North Broadway, Suite 1440,<br />
Baltimore, MD 21231-2410; email: zellari@jhmi.edu.<br />
© <strong>2012</strong> by <strong>American</strong> <strong>Society</strong> <strong>of</strong> <strong>Clinical</strong> <strong>Oncology</strong>.<br />
1092-9118/10/1-10
POSTMASTECTOMY RADIATION AND PARTIAL BREAST IRRADIATION<br />
trial indirectly validates the findings <strong>of</strong> the three modern<br />
PMXRT trials.<br />
It is likely that other factors can aid in predicting LRR<br />
risk after mastectomy. For example, researchers have<br />
shown that the 21-gene recurrence score assay (Oncotype<br />
DX; Genomic Health, Inc., Redwood City, CA), commonly<br />
used to determine the risk <strong>of</strong> distant recurrence, may also be<br />
predictive <strong>of</strong> the risk <strong>of</strong> locoregional recurrence. 9 Other<br />
factors reported to be associated with increased locoregional<br />
failure include tumor size, positive margins, extracapsular<br />
extension, lymphovascular invasion, response to neoadjuvant<br />
chemotherapy, age, ER/PR status, p53 overexpression,<br />
and breast cancer subtypes. 5,8,9 While we await evaluation<br />
<strong>of</strong> putative biologic, genetic and clinical factors predictive <strong>of</strong><br />
LRR such that we can judge a patient’s need for PMXRT, it<br />
is perhaps prudent to at least seriously consider PMXRT for<br />
women with one to three positive axillary LNs.<br />
PBI<br />
In-breast failures after breast-conserving therapy (defined<br />
here as lumpectomy and WBI) occur in the vicinity <strong>of</strong> the<br />
original primary tumor approximately 70 to 90% <strong>of</strong> the time.<br />
This phenomenon gave birth to the concept <strong>of</strong> PBI. Benefits<br />
<strong>of</strong> this proposed new treatment paradigm include shorter<br />
overall treatment course, increased efficacy via the larger<br />
biologically effective doses that are possible with PBI, and<br />
decreased toxicity because less tissue is exposed to radiation.<br />
PBI can be divided into two general techniques: brachytherapy<br />
based (radiation close to the target) and teletherapy<br />
based (external beam). Brachytherapy-based PBI normally<br />
occurs postoperatively and may be delivered by interstitial<br />
(needles) or intracavitary (mammosite, contura, etc.) techniques.<br />
Teletherapy-based PBI can be delivered either intraoperatively<br />
or postoperatively. Of the three largest<br />
randomized controlled trials comparing various PBI techniques<br />
versus standard WBI, only one has been published;<br />
another is closed to accrual but the results have not been<br />
KEY POINTS<br />
● Three modern postmastectomy radiation therapy<br />
(PMXRT) trials reported improved overall survival<br />
with adjuvant radiation in patients with one to three<br />
positive lymph nodes.<br />
● Meta-analyses support the conclusion <strong>of</strong> these modern<br />
PMXRT trials.<br />
● A new trial from the National Cancer Institute <strong>of</strong><br />
Canada (NCIC), which reported a survival benefit in<br />
women receiving treatment with whole-breast irradiation<br />
(WBI) and regional nodal irradiation when<br />
compared with WBI alone, indirectly supports<br />
PMXRT.<br />
● Partial breast irradiation (PBI) has been shown to be<br />
comparable with WBI in a single large randomized<br />
controlled trial using intraoperative PBI.<br />
● Although smaller trials and retrospective studies<br />
support PBI, it may be best to await the results <strong>of</strong><br />
other large randomized controlled trials evaluating<br />
various PBI techniques.<br />
presented, and a third remains open to accrual. The largest<br />
trial published to date is the TARGIT trial in which more<br />
than 2,000 women were randomly assigned to receive WBI<br />
or intraoperative radiation therapy (IORT). Patients received<br />
20 Gy (photons) to the surface <strong>of</strong> the lumpectomy<br />
bed. 10 Patients enrolled were generally low risk: More than<br />
80% were 50 years <strong>of</strong> age or older, node negative, ERpositive,<br />
and HER2-negative. Interestingly, as a result <strong>of</strong><br />
various high-risk features, approximately 14% <strong>of</strong> the patients<br />
randomly assigned to receive PBI also received WBI.<br />
Estimates <strong>of</strong> ipsilateral breast tumor recurrence rates at 4<br />
years were 1.2% and 1.0% for PBI and WBI, respectively.<br />
Although the results are promising, skepticism has prevented<br />
wide acceptance. There are concerns that the<br />
follow-up period is too short, the prescribed dose is too low,<br />
and delivery <strong>of</strong> IORT is logistically too difficult. Nonetheless,<br />
the TARGIT trial supports the concept <strong>of</strong> PBI.<br />
The second large randomized controlled trial <strong>of</strong> PBI is<br />
from Milan. Similar to the TARGIT trial, PBI consists <strong>of</strong> 20<br />
Gy (electrons) delivered to the lumpectomy bed intraoperatively.<br />
The accrual goal <strong>of</strong> approximately 1,306 women was<br />
reached some time ago. So far, one <strong>of</strong> the investigators has<br />
stated that the there is no difference in survival between the<br />
two arms at 10 years. 11 We eagerly await the results <strong>of</strong> this<br />
important study.<br />
The largest trial is that <strong>of</strong> NSABP B-39/Radiation Therapy<br />
<strong>Oncology</strong> Group (RTOG) 0413. 12 This trial is unique not<br />
only for its accrual goal <strong>of</strong> 4,300 women but also because<br />
three separate PBI techniques are permitted. Women randomly<br />
assigned to receive PBI may be <strong>of</strong>fered, depending on<br />
the hosting institution, interstitial or intracavitary brachytherapy<br />
or postoperative teletherapy. Interestingly, with<br />
more than 4,100 patients enrolled thus far (and on target to<br />
complete accrual in <strong>2012</strong>), more than 70% <strong>of</strong> the patients<br />
receiving treatment with PBI received it via the postoperative<br />
teletherapy technique. The trial completed accrual <strong>of</strong><br />
low-risk patients, and now remains open to accrue patients<br />
with a higher risk <strong>of</strong> recurrence (e.g., ER-negative, LNpositive,<br />
or age younger than 50 years).<br />
Another potential benefit <strong>of</strong> PBI is that it could be combined<br />
with systemic therapy, thereby not only shortening<br />
the course <strong>of</strong> radiation but shortening the the overall<br />
treatment course. For example in a small feasibility study,<br />
patients received treatment with PBI (2.7 Gy twice daily<br />
for 15 days) and concurrent dose-dense doxorubicin and<br />
cyclophosphamide. 13 The authors reported that systemic<br />
toxicity was minimal, radiation dermatitis was far less than<br />
that seen in standard WBI without concurrent chemotherapy,<br />
and cosmetic outcome was good to excellent in more<br />
than 90% <strong>of</strong> the patients. This small pilot study deserves<br />
validation in larger trials but may hint to a new era <strong>of</strong><br />
combined-modality therapy in the management <strong>of</strong> breast<br />
cancer.<br />
Given the general public interest in PBI, and the fact that<br />
smaller randomized controlled trials and retrospective<br />
studies support its use, the <strong>American</strong> <strong>Society</strong> for Therapeutic<br />
Radiology and <strong>Oncology</strong> (ASTRO) published guidelines<br />
defining patient characteristics most appropriate for this<br />
new treatment option. Many practitioners have accepted<br />
these guidelines and use them in daily practice. However, a<br />
recent presentation <strong>of</strong> a study comparing brachytherapybased<br />
PBI versus WBI may have placed a pall over the<br />
51
passionate pursuit <strong>of</strong> PBI. This study, presented at the<br />
34 th Annual San Antonio Breast Cancer Symposium (December<br />
6–10, San Antonio, TX), reviewed the Medicare<br />
records <strong>of</strong> 130,535 patients who received treatment with<br />
brachytherapy-based PBI or standard WBI between 2001<br />
and 2007. 14 The authors reported that women receiving<br />
treatment with brachytherapy-based PBI were almost twice<br />
as likely to subsequently require a mastectomy, be hospital-<br />
Author’s Disclosures <strong>of</strong> Potential Conflicts <strong>of</strong> Interest<br />
Author<br />
Richard C. Zellars*<br />
*No relevant relationships to disclose.<br />
Employment or<br />
Leadership<br />
Positions<br />
Consultant or<br />
Advisory Role<br />
1. Overgaard M, Hansen PS, Overgaard J, et al. Postoperative radiotherapy<br />
in high-risk premenopausal women with breast cancer who receive<br />
adjuvant chemotherapy. Danish Breast Cancer Cooperative Group 82b Trial.<br />
N Engl J Med. 1997;337:949-955.<br />
2. Overgaard M. Overview <strong>of</strong> randomized trials in high risk breast cancer<br />
patients treated with adjuvant systemic therapy with or without postmastectomy<br />
irradiation. Semin Radiat Oncol. 1999;9:292-299.<br />
3. Ragaz J, Olivotto IA, Spinelli JJ, et al. Locoregional radiation therapy in<br />
patients with high-risk breast cancer receiving adjuvant chemotherapy:<br />
20-year results <strong>of</strong> the British Columbia randomized trial. J Natl Cancer Inst.<br />
2005;97:116-126.<br />
4. Taghian A, Jeong JH, Mamounas E, et al. Patterns <strong>of</strong> locoregional<br />
failure in patients with operable breast cancer treated by mastectomy and<br />
adjuvant chemotherapy with or without tamoxifen and without radiotherapy:<br />
results from five National Surgical Adjuvant Breast and Bowel Project<br />
randomized clinical trials. J Clin Oncol. 2004;22:4247-4254.<br />
5. Recht A, Gray R, Davidson NE, et al. Locoregional failure 10 years after<br />
mastectomy and adjuvant chemotherapy with or without tamoxifen without<br />
irradiation: experience <strong>of</strong> the Eastern Cooperative <strong>Oncology</strong> Group. J Clin<br />
Oncol. 1999;17:1689-1700.<br />
6. Ceilley E, Jagsi R, Goldberg S, et al. RADIOTHERAPY for invasive<br />
breast cancer in north America and Europe: results <strong>of</strong> a survey. Int J Radiat<br />
Oncol Biol Phys. 2005;61:365-373.<br />
7. McGale P. The 2006 worldwide overview <strong>of</strong> the effects <strong>of</strong> local treatments<br />
for early breast cancer on long-term outcome: “meta-analysis <strong>of</strong> the randomized<br />
trials <strong>of</strong> radiotherapy after mastectomy with axillary clearance.” Pre-<br />
52<br />
ized for complications, and acquire a treatment-induced<br />
infection compared with those who received the more standard<br />
WBI. One must note that this study is retrospective<br />
and covers a period in which new PBI brachytherapy techniques<br />
were just being introduced. Nonetheless, it is perhaps<br />
most prudent for practitioners to reserve PBI for the<br />
protocol setting, or to wait for longer-term results from<br />
randomized controlled trials.<br />
Stock<br />
Ownership Honoraria<br />
REFERENCES<br />
Research<br />
Funding<br />
Expert<br />
Testimony<br />
RICHARD C. ZELLARS<br />
Other<br />
Remuneration<br />
sented at: 48 th Annual Meeting <strong>of</strong> the <strong>American</strong> <strong>Society</strong> <strong>of</strong> Therapeutic<br />
Radiology and <strong>Oncology</strong>; November 2006; Philadelphia, PA.<br />
8. Whelan T. NCIC-CTG MA.20: an intergroup trial <strong>of</strong> regional nodal<br />
irradiation in early breast cancer. J Clin Oncol. 2011;29 (suppl; abstr<br />
LBA1003).<br />
9. Voogd AC, Nielsen M, Peterse JL, et al. Differences in risk factors for<br />
local and distant recurrence after breast-conserving therapy or mastectomy<br />
for stage I and II breast cancer: pooled results <strong>of</strong> two large European<br />
randomized trials. J Clin Oncol. 2001;19:1688-1697.<br />
10. Vaidya JS, Joseph DJ, Tobias JS, et al. Targeted intraoperative<br />
radiotherapy versus whole breast radiotherapy for breast cancer (TARGIT-A):<br />
an international, prospective, randomized, noninferiority phase 3 trial. Lancet.<br />
2010;376:91-102.<br />
11. <strong>Oncology</strong> Times UK. June 2011 Volume 8.<br />
12. Julian T. Early toxicity results with 3-D conformal external beam<br />
therapy from the NSABP B-39/RTOG 0413 Accelerated Partial Breast Irradiation<br />
(APBI) Trial. Presented at: 53 rd Annual Meeting <strong>of</strong> the <strong>American</strong><br />
<strong>Society</strong> <strong>of</strong> Therapeutic Radiology and <strong>Oncology</strong>; October 2011; Philadelphia,<br />
PA.<br />
13. Zellars RC, Stearns V, Frassica D, et al. Feasibility trial <strong>of</strong> PBI with<br />
concurrent dose-dense doxorubicin and cyclophosphamide in early-stage<br />
breast cancer. J Clin Oncol 2009;27:2816-2822.<br />
14. Smith GL, Buchholz TA, Giordano SH et al. Partial breast brachytherapy<br />
is associated with inferior effectiveness and increased toxicity compared<br />
with whole breast irradiation in older patients. Presented at: 34 th Annual San<br />
Antonio Breast Cancer Symposium; December 2011; San Antonio, TX.
The Appropriate Extent <strong>of</strong> Surgery for<br />
Early-Stage Breast Cancer<br />
Overview: Attitudes regarding the appropriate extent <strong>of</strong> surgery<br />
for breast cancer and the effect <strong>of</strong> surgery on breast<br />
cancer–specific survival have varied over time. Failure to<br />
maintain local control is associated with decreased survival,<br />
but the extent <strong>of</strong> surgery necessary for local control has<br />
decreased as other treatment modalities, such as radiotherapy<br />
and systemic therapy, have become more widely used.<br />
Both endocrine therapy and chemotherapy considerably re-<br />
PERCEPTIONS OF the role <strong>of</strong> surgery in contributing to<br />
the cure <strong>of</strong> breast cancer have varied over time. Until<br />
the mid-1970s, surgery was performed with the intent <strong>of</strong><br />
curing breast cancer, and “bigger was better.” In the 1980s<br />
and 1990s the popularization <strong>of</strong> the “systemic disease”<br />
hypothesis by Dr. Bernard Fisher opened the door to smaller<br />
surgical procedures, the widespread use <strong>of</strong> systemic therapy,<br />
and the thought that surgery was primarily a staging procedure.<br />
In 2005 the Early Breast Cancer Trialists’ Collaborative<br />
Group (EBCTCG) overview analysis 1 demonstrated that failure<br />
to achieve local control in both node-positive and nodenegative<br />
women at 5 years after either lumpectomy or<br />
mastectomy decreased breast cancer specific survival at 15<br />
years, indicating that local therapy does affect survival. This<br />
observation renewed interest in the appropriate extent <strong>of</strong><br />
surgery to maximize local control in the index breast and the<br />
axilla, as well as the role <strong>of</strong> contralateral prophylactic<br />
mastectomy (CPM) in breast cancer management.<br />
Extent <strong>of</strong> Surgery in the Index Breast: Margins<br />
The selection criteria for breast-conserving surgery (BCS)<br />
developed in the 1980s and in use today are related to the<br />
extent <strong>of</strong> the tumor burden in the breast and the ability to<br />
safely deliver radiotherapy. Biologic tumor features such as<br />
estrogen receptor (ER), histologic grade, and node status are<br />
not used as selection criteria for lumpectomy compared with<br />
mastectomy. 2 The most important determinant <strong>of</strong> tumor<br />
burden in the breast with a unicentric cancer is margin<br />
status. Although it is clear that patients with positive<br />
margins, defined as tumor cells touching ink, have an<br />
increased risk <strong>of</strong> local recurrence compared with those with<br />
negative margins; 3 evidence that more widely clear margins<br />
further reduce local recurrence is lacking.<br />
In the prospective, randomized trials that established the<br />
safety <strong>of</strong> BCS, only the National Surgical Adjuvant Breast<br />
and Bowel Project (NSABP) B06 study required a microscopically<br />
negative margin, defined as tumor cells not touching<br />
ink. 4 After 20 years <strong>of</strong> follow-up, patients undergoing<br />
lumpectomy and radiation therapy (RT) in this study had a<br />
survival outcome equivalent to those having mastectomy,<br />
and only 14% developed a local recurrence (LR).<br />
Since the time <strong>of</strong> this study, the rates <strong>of</strong> LR in NSABP<br />
trials have steadily declined, although the required margin<br />
has not changed. 5 This can largely be attributed to the<br />
widespread use <strong>of</strong> adjuvant systemic therapy, although<br />
improvements in pathologic evaluation and the quality <strong>of</strong><br />
mammography have also contributed to this trend. Data<br />
from prospective randomized trials <strong>of</strong> adjuvant systemic<br />
By Monica Morrow, MD<br />
duce rates <strong>of</strong> local recurrence in the breast, as well as the<br />
incidence <strong>of</strong> contralateral breast cancer, and as efficacy in<br />
reducing metastatic disease increases, so does the benefit in<br />
reducing local recurrence. The excellent rates <strong>of</strong> local control<br />
in the ACOSOG Z11 trial after elimination <strong>of</strong> axillary dissection<br />
in patients with positive sentinel nodes receiving whole-breast<br />
irradiation and systemic therapy are a model for reducing<br />
surgical morbidity in the era <strong>of</strong> multimodality therapy.<br />
chemotherapy and endocrine therapy demonstrate reductions<br />
in LR rates from 13% to 15% in patients randomized to<br />
no adjuvant therapy compared with 3% to 4% in those<br />
receiving adjuvant therapy. 6,7<br />
As the efficacy <strong>of</strong> systemic therapy improves, LR is further<br />
reduced. In the randomized trials <strong>of</strong> the addition <strong>of</strong> trastuzumab<br />
to chemotherapy in patients with HER2 overexpressing<br />
tumors, the addition <strong>of</strong> trastuzumab resulted in an<br />
approximately 40% relative reduction in the risk <strong>of</strong> LR<br />
compared with treatment with chemotherapy alone. 8 Rates<br />
<strong>of</strong> LR less than 5% at 8–10 years are regularly seen in<br />
patients undergoing BCS in more recent time periods. 5,9<br />
Despite these favorable outcomes, there appears to be confusion<br />
regarding what constitutes an adequate margin <strong>of</strong><br />
resection as evidenced by the results <strong>of</strong> a population-based<br />
survey <strong>of</strong> 318 surgeons in which only 11% endorsed tumor<br />
not touching ink as an adequate margin, 42% endorsed an<br />
adequate margin <strong>of</strong> greater than 1–2 mm, 28% endorsed an<br />
adequate margin <strong>of</strong> greater than 5 mm, and 19% endorsed<br />
an adequate margin greater than 1 cm. 10 A similar lack <strong>of</strong><br />
consensus has been documented among radiation oncologists.<br />
11 This lack <strong>of</strong> consensus has made re-excision a<br />
common procedure, 12 and a study <strong>of</strong> 2,206 patients with<br />
invasive breast cancer documented that although 23% <strong>of</strong><br />
patients had re-excision, almost half (47.5%) <strong>of</strong> the reexcisions<br />
were performed for margins where tumor was not<br />
touching ink. 13 A meta-analysis <strong>of</strong> 14,571 patients with<br />
1,026 local recurrences found no difference in rates <strong>of</strong> local<br />
recurrence for any negative margin width greater than<br />
tumor not touching ink after adjusting for factors such as<br />
the use <strong>of</strong> a boost dose <strong>of</strong> RT and endocrine therapy. 3 Other<br />
strategies, such as magnetic resonance imaging, that are<br />
intended to reduce the subclinical tumor burden (the goal <strong>of</strong><br />
more widely clear margins) have also not been shown to<br />
decrease LR. 9,14 In addition, there is an increasing body <strong>of</strong><br />
evidence indicating that patients at high risk <strong>of</strong> LR after<br />
BCS are also at high risk <strong>of</strong> LR after mastectomy. 15,16 Two<br />
studies examining LR after BCS and mastectomy for patients<br />
with triple-negative breast cancers have found no<br />
From the Breast Service, Department <strong>of</strong> Surgery, Memorial Sloan-Kettering Cancer<br />
Center, New York, NY.<br />
Author’s disclosures <strong>of</strong> potential conflicts <strong>of</strong> interest are found at the end <strong>of</strong> this article.<br />
Address reprint requests to Monica Morrow, MD, Breast Service, Department <strong>of</strong> Surgery,<br />
Memorial Sloan-Kettering Cancer Center, 300 East 66th St., New York, NY 10065; email:<br />
morrowm@mskcc.org.<br />
© <strong>2012</strong> by <strong>American</strong> <strong>Society</strong> <strong>of</strong> <strong>Clinical</strong> <strong>Oncology</strong>.<br />
1092-9118/10/1-10<br />
53
advantage for mastectomy 17,18 in this patient subset at high<br />
risk <strong>of</strong> distant relapse. In a retrospective study <strong>of</strong> the relationship<br />
<strong>of</strong> the 21-gene recurrence score (OncotypeDx) to the risk<br />
<strong>of</strong> LR, the use <strong>of</strong> appropriate systemic therapy as predicted<br />
by the 21-gene recurrence score, significantly reduced LR<br />
(p � 0.0001), and in a multivariate model, type <strong>of</strong> surgery<br />
(lumpectomy compared with mastectomy) was not significantly<br />
associated with LR after controlling for biologic variables. 16<br />
In aggregate, these findings indicate that local control is a<br />
complex interaction between disease burden, tumor biology,<br />
and the effectiveness <strong>of</strong> systemic therapy. In the current era<br />
where multimodality treatment is routinely employed, there<br />
is little evidence that larger surgical resections improve<br />
patient outcomes.<br />
Axilla<br />
The multidisciplinary nature <strong>of</strong> breast cancer therapy<br />
today has also affected our approach to the axilla. Axillary<br />
dissection has traditionally been performed for staging and<br />
local control. Its role in contributing to the cure <strong>of</strong> breast<br />
cancer has been controversial since the NSABP B04 trial<br />
demonstrated no difference in survival in patients treated<br />
with and without axillary dissection in the prechemotherapy<br />
era. 19 Sentinel node biopsy is a reliable method <strong>of</strong> identifying<br />
axillary node involvement with a lower morbidity than<br />
axillary dissection, 20 and biologic tumor features, rather<br />
than number <strong>of</strong> axillary lymph nodes involved with tumor,<br />
are increasingly used to select systemic therapy. In patients<br />
undergoing BCS, the tangent field RT, which is a standard<br />
part <strong>of</strong> treatment, covers some <strong>of</strong> the axilla, 21 and systemic<br />
therapy also contributes to locoregional control. 6-8 These<br />
observations resulted in the <strong>American</strong> College <strong>of</strong> Surgeons<br />
Group (ACOSOG) Z11 study in which clinically nodenegative<br />
women with T1 and T2 breast cancers undergoing<br />
BCS and found to have hematoxylin and eosin-detected<br />
metastasis in � 3 sentinel nodes were randomized to completion<br />
axillary dissection or no further axillary treatment.<br />
22,23 All patients received tangent breast RT and<br />
systemic therapy. After a median follow-up <strong>of</strong> 6.3 years, the<br />
54<br />
KEY POINTS<br />
● Rates <strong>of</strong> local recurrence after breast-conserving surgery<br />
and radiotherapy have decreased over time.<br />
● Evidence that lumpectomy margins more widely<br />
clear further reduce local recurrence than tumor not<br />
touching ink is lacking.<br />
● Local control can be achieved with removal <strong>of</strong> the<br />
sentinel nodes and no further axillary treatment for<br />
patients with involvement <strong>of</strong> one or two sentinel<br />
nodes treated with whole-breast irradiation and systemic<br />
therapy.<br />
● The incidence <strong>of</strong> contralateral breast cancer has been<br />
declining since the mid-1980s because <strong>of</strong> the increased<br />
use <strong>of</strong> adjuvant systemic therapy.<br />
● Systemic therapy has a major effect on local control,<br />
<strong>of</strong>fering the opportunity to decrease the extent and<br />
morbidity <strong>of</strong> surgery as the effectiveness <strong>of</strong> systemic<br />
therapy increases.<br />
Table 1. Patient Groups for Whom Axillary Dissection Remains<br />
Standard Management <strong>of</strong> a Positive Sentinel Node<br />
● Palpable Node Disease<br />
● Locally Advanced (Stage III) Breast Cancer<br />
● Treatment with Mastectomy<br />
● Treatment with Partial Breast Irradiation<br />
● Treatment with Neoadjuvant Therapy<br />
● Involvement <strong>of</strong> � 3 Sentinel Nodes with Metastases<br />
● Gross Extranodal Tumor Extension in Sentinel Nodes<br />
regional recurrence rate in the sentinel node only arm was<br />
0.9% compared with 0.4% in the axillary dissection arm,<br />
despite the fact that 27.4% <strong>of</strong> patients treated with axillary<br />
dissection had additional positive nodes removed beyond the<br />
sentinel node. 23 No survival differences were seen between<br />
groups, 22 and morbidity was significantly lower in the sentinel<br />
node-only group. 22,23 This study clearly indicates that<br />
axillary dissection can be eliminated for patients meeting<br />
the ACOSOG Z11 eligibility criteria, but does not apply to a<br />
considerable number <strong>of</strong> women with breast cancer (Table 1).<br />
There have been concerns expressed that because most <strong>of</strong><br />
the patients in Z11 were postmenopausal with ER-positive<br />
tumors, these results cannot be extrapolated to younger<br />
women and those with ER-negative tumors. However, large<br />
studies <strong>of</strong> regional node recurrence do not identify age or ER<br />
status as predictors, 24,25 suggesting that they should not be<br />
used to exclude women from this approach. Implementing<br />
the results <strong>of</strong> Z11 necessitates some changes in practice. For<br />
example, the identification <strong>of</strong> a single metastatic axillary<br />
node with ultrasound and fine-needle aspiration no longer<br />
changes the operative approach because the sentinel node<br />
still must be surgically removed, and it should be abandoned.<br />
Frozen section <strong>of</strong> the sentinel node also is not necessary<br />
or appropriate in these cases because the finding <strong>of</strong> tumor<br />
in a single node will not lead to dissection, and because<br />
examination <strong>of</strong> all nodal tissue is necessary to determine the<br />
number <strong>of</strong> nodes involved and the presence <strong>of</strong> extranodal<br />
extension. Finally, based on the results <strong>of</strong> ACOSOG Z10 and<br />
NSABP B32 indicating that micrometastases in the sentinel<br />
node have minimal prognostic impact, the routine use <strong>of</strong><br />
immunohistochemistry to look for small tumor deposits can<br />
be abandoned.<br />
Contralateral Breast<br />
MONICA MORROW<br />
The use <strong>of</strong> contralateral prophylactic mastectomy (CPM)<br />
is increasing, 26 with the most notable increases seen in<br />
women less than 50 years <strong>of</strong> age and those treated in<br />
hospitals managing a higher volume <strong>of</strong> patients with breast<br />
cancer. 27 This is somewhat surprising, considering that the<br />
clinical incidence <strong>of</strong> contralateral breast cancer has been<br />
declining since 1985 because <strong>of</strong> the widespread use <strong>of</strong><br />
adjuvant therapy. 28 This decrease is most evident among<br />
women with ER-positive cancers, where even for those in<br />
their 20s and 30s at initial diagnosis, 10-year rates <strong>of</strong><br />
contralateral cancer are 2.5%–4.5%. 28 With such a low<br />
incidence rate, it is not surprising that evidence indicating<br />
that CPM improves breast cancer mortality is lacking. 29<br />
Evidence suggests that level <strong>of</strong> anxiety, regardless <strong>of</strong> the<br />
presence or absence <strong>of</strong> risk factors for contralateral breast<br />
cancer, is a major predictor <strong>of</strong> undergoing CPM. 30
SURGERY FOR EARLY-STAGE BREAST CANCER<br />
Conclusion<br />
Current risks for LR with lumpectomy to a margin <strong>of</strong><br />
tumor not touching ink, and in the axilla after management<br />
<strong>of</strong> a positive node with sentinel node biopsy alone in patients<br />
undergoing BCS, emphasize the powerful effect <strong>of</strong> systemic<br />
therapy on local control. As the options for targeted therapy<br />
expand and our ability to tailor the extent <strong>of</strong> systemic<br />
therapy to an individual patient’s level <strong>of</strong> risk improves, not<br />
Author’s Disclosures <strong>of</strong> Potential Conflicts <strong>of</strong> Interest<br />
Author<br />
Monica Morrow*<br />
*No relevant relationships to disclose.<br />
Employment or<br />
Leadership<br />
Positions<br />
Consultant or<br />
Advisory Role<br />
1. Clarke M, Collins R, Darby S, et al. Effects <strong>of</strong> radiotherapy and <strong>of</strong><br />
differences in the extent <strong>of</strong> surgery for early breast cancer on local recurrence<br />
and 15-year survival: An overview <strong>of</strong> the randomised trials. Lancet. 2005;<br />
366(9503):2087-2106.<br />
2. Morrow M, Harris JR. Practice guideline for breast conservation therapy<br />
in the management <strong>of</strong> invasive breast cancer. J Am Coll Surg. 2007;205:362-<br />
376.<br />
3. Houssami N, Macaskill P, Marinovich ML, et al. Meta-analysis <strong>of</strong> the<br />
impact <strong>of</strong> surgical margins on local recurrence in women with early-stage<br />
invasive breast cancer treated with breast-conserving therapy. Eur J Cancer.<br />
2010;46(18):3219-3232.<br />
4. Fisher B, Anderson S, Bryant J, et al. Twenty-year follow-up <strong>of</strong> a<br />
randomized trial comparing total mastectomy, lumpectomy, and lumpectomy<br />
plus irradiation for the treatment <strong>of</strong> invasive breast cancer. N Engl J Med.<br />
2002;347(16):1233-1241.<br />
5. Anderson SJ, Wapnir I, Dignam JJ, et al. Prognosis after ipsilateral<br />
breast tumor recurrence and locoregional recurrences in patients treated by<br />
breast-conserving therapy in five National Surgical Adjuvant Breast and<br />
Bowel Project protocols <strong>of</strong> node-negative breast cancer. J Clin Oncol. 2009;<br />
27(15):2466-2473.<br />
6. Fisher B, Dignam J, Bryant J, et al. Five versus more than five years <strong>of</strong><br />
tamoxifen therapy for breast cancer patients with negative lymph nodes and<br />
estrogen receptor-positive tumors. J Natl Cancer Inst. 1996;88(21):1529-1542.<br />
7. Fisher B, Dignam J, Mamounas EP, et al. Sequential methotrexate and<br />
fluorouracil for the treatment <strong>of</strong> node-negative breast cancer patients with<br />
estrogen receptor-negative tumors: Eight-year results from National Surgical<br />
Adjuvant Breast and Bowel Project (NSABP) B-13 and first report <strong>of</strong> findings<br />
from NSABP B-19 comparing methotrexate and fluorouracil with conventional<br />
cyclophosphamide, methotrexate, and fluorouracil. J Clin Oncol. 1996;<br />
14(7):1982-1992.<br />
8. Romond EH, Perez EA, Bryant J, et al. Trastuzumab plus adjuvant<br />
chemotherapy for operable HER2-positive breast cancer. N Engl J Med.<br />
2005;353(16):1673-1684.<br />
9. Solin LJ, Orel SG, Hwang WT, et al. Relationship <strong>of</strong> breast magnetic<br />
resonance imaging to outcome after breast-conservation treatment with<br />
radiation for women with early-stage invasive breast carcinoma or ductal<br />
carcinoma in situ. J Clin Oncol. 2008;26(3):386-391.<br />
10. Azu M, Abrahamse P, Katz SJ, et al. What is an adequate margin for<br />
breast-conserving surgery? Surgeon attitudes and correlates. Ann Surg Oncol.<br />
2010;17(2):558-563.<br />
11. Taghian A, Mohiuddin M, Jagsi R, et al. Current perceptions regarding<br />
surgical margin status after breast-conserving therapy: Results <strong>of</strong> a survey.<br />
Ann Surg. 2005;241(4):629-639.<br />
12. Morrow M, Jagsi R, Alderman AK, et al. Surgeon recommendations and<br />
receipt <strong>of</strong> mastectomy for treatment <strong>of</strong> breast cancer. JAMA. 2009;302(14):<br />
1551-1556.<br />
13. McCahill LE, Single RM, Aiello Bowles EJ, et al. Variability in<br />
Reexcision Following Breast Conservation Surgery. JAMA. <strong>2012</strong>;307(5):467-<br />
475.<br />
14. Hwang N, Schiller DE, Crystal P, et al. Magnetic resonance imaging in<br />
the planning <strong>of</strong> initial lumpectomy for invasive breast carcinoma: Its effect on<br />
ipsilateral breast tumor recurrence after breast-conservation therapy. Ann<br />
Surg Oncol. 2009;16(11):3000-3009.<br />
15. Kyndi M, Sorensen FB, Knudsen H, et al. Estrogen receptor, progesterone<br />
receptor, HER-2, and response to postmastectomy radiotherapy in<br />
only does survival improve, but local recurrence decreases as<br />
well. These findings have important implications for the<br />
future <strong>of</strong> local therapy. More aggressive surgical strategies<br />
are unlikely to improve local outcomes for poor-prognosis<br />
cancers in the absence <strong>of</strong> improved systemic therapies.<br />
Strategies to reduce the extent <strong>of</strong> surgery in patient subsets<br />
where systemic therapy is available should be sought to<br />
reduce the morbidity <strong>of</strong> treatment.<br />
Stock<br />
Ownership Honoraria<br />
REFERENCES<br />
Research<br />
Funding<br />
Expert<br />
Testimony<br />
Other<br />
Remuneration<br />
high-risk breast cancer: The Danish Breast Cancer Cooperative Group. J Clin<br />
Oncol. 2008;26(9):1419-1426.<br />
16. Mamounas EP, Tang G, Fisher B, et al. Association between the<br />
21-gene recurrence score assay and risk <strong>of</strong> locoregional recurrence in nodenegative,<br />
estrogen receptor-positive breast cancer: Results from NSABP B-14<br />
and NSABP B-20. J Clin Oncol. 2010;28(10):1677-1683.<br />
17. Abdulkarim BS, Cuartero J, Hanson J, et al. Increased risk <strong>of</strong> locoregional<br />
recurrence for women with T1-2N0 triple-negative breast cancer<br />
treated with modified radical mastectomy without adjuvant radiation therapy<br />
compared with breast-conserving therapy. J Clin Oncol. 2011;29(21):2852-<br />
2858.<br />
18. Ho AY, Gupta G, King TA, et al. Favorable Prognosis in Patients with<br />
T1a/bN0 Triple Negative Breast Cancers Treated With Multimodality Therapy.<br />
Cancer. <strong>2012</strong> (In Press).<br />
19. Fisher B, Jeong JH, Anderson S, et al. Twenty-five-year follow-up <strong>of</strong> a<br />
randomized trial comparing radical mastectomy, total mastectomy, and total<br />
mastectomy followed by irradiation. N Engl J Med. 2002;347(8):567-575.<br />
20. Krag DN, Anderson SJ, Julian TB, et al. Technical outcomes <strong>of</strong><br />
sentinel-lymph-node resection and conventional axillary-lymph-node dissection<br />
in patients with clinically node-negative breast cancer: Results from the<br />
NSABP B-32 randomised phase III trial. Lancet Oncol. 2007;8(10):881-888.<br />
21. Reznik J, Cicchetti MG, Degaspe B, Fitzgerald TJ. Analysis <strong>of</strong> axillary<br />
coverage during tangential radiation therapy to the breast. Int J Radiat<br />
Oncol Biol Phys. 2005;61(1):163-168.<br />
22. Giuliano AE, Hunt KK, Ballman KV, et al. Axillary dissection vs no<br />
axillary dissection in women with invasive breast cancer and sentinel node<br />
metastasis: A randomized clinical trial. JAMA. 2011;305(6):569-575.<br />
23. Giuliano AE, McCall L, Beitsch P, et al. Locoregional recurrence after<br />
sentinel lymph node dissection with or without axillary dissection in patients<br />
with sentinel lymph node metastases: The <strong>American</strong> College <strong>of</strong> Surgeons<br />
<strong>Oncology</strong> Group Z0011 randomized trial. Ann Surg. 2010;252(3):426-432;<br />
discussion 32-33.<br />
24. Grills IS, Kestin LL, Goldstein N, et al. Risk factors for regional nodal<br />
failure after breast-conserving therapy: Regional nodal irradiation reduces<br />
rate <strong>of</strong> axillary failure in patients with four or more positive lymph nodes. Int<br />
J Radiat Oncol Biol Phys. 2003;56(3):658-670.<br />
25. Yates L, Kirby A, Crichton S, et al. Risk Factors for Regional Nodal<br />
Relapse in Breast Cancer Patients with One to Three Positive Axillary Nodes.<br />
Int J Radiat Oncol Biol Phys (In Press). 2011.<br />
26. Tuttle TM, Habermann EB, Grund EH, et al. Increasing use <strong>of</strong><br />
contralateral prophylactic mastectomy for breast cancer patients: A trend<br />
toward more aggressive surgical treatment. J Clin Oncol. 2007;25(33):5203-<br />
5209.<br />
27. Yao K, Stewart AK, Winchester DJ, Winchester DP. Trends in contralateral<br />
prophylactic mastectomy for unilateral cancer: A report from the National<br />
Cancer Data Base, 1998-2007. Ann Surg Oncol. 2010;17(10):2554-2562.<br />
28. Nichols HB, Berrington de Gonzalez A, Lacey JV, Jr., et al. Declining<br />
incidence <strong>of</strong> contralateral breast cancer in the United States from 1975 to<br />
2006. J Clin Oncol. 2011;29(12):1564-1569.<br />
29. Lostumbo L, Carbine NE, Wallace J. Prophylactic mastectomy for the<br />
prevention <strong>of</strong> breast cancer. Cochrane Database Syst Rev. (Online). 2010(11):<br />
CD002748.<br />
30. Hawley ST, Jagsi R, Morrow M, Katz SJ. Correlates <strong>of</strong> contralateral<br />
prophylactic mastectomy in a population-based sample. J Clin Oncol. 2011;29<br />
(Abstract 6010).<br />
55
How Does Biology Affect Local<br />
Therapy Decisions?<br />
Overview: Breast cancer represents a biologically diverse<br />
set <strong>of</strong> diseases. Previous data suggest that the estrogen/<br />
progesterone receptor (ER/PR) status and HER2/neu status<br />
are important determinants <strong>of</strong> prognosis and response to<br />
various systemic treatments. Recent data also suggest that<br />
these receptors correlate with outcomes <strong>of</strong> local-regional<br />
therapies. Specifically, patients with ER-positive HER2/neunegative<br />
disease have an excellent outcome with radiation<br />
GENOMIC AND molecular analyses <strong>of</strong> breast cancer<br />
have elucidated that breast cancer represents a biologically<br />
diverse group <strong>of</strong> diseases. Studies indicate that molecular<br />
subtypes <strong>of</strong> breast cancer differ with respect to tumor<br />
biology, prognosis and response to hormone therapy, targeted<br />
therapy, and chemotherapy. In part, these subtypes<br />
are driven by differences in ER signaling and HER2/neu<br />
status. While breast cancer biology has had a major role in<br />
determining systemic treatment approaches, it is less clear<br />
how biologic markers <strong>of</strong> disease should affect surgery and<br />
radiation decisions. Fewer data exist correlating breast<br />
cancer biology and local-regional treatment outcome compared<br />
with the available data correlating molecular markers<br />
with response to systemic treatments, overall recurrence,<br />
distant metastases, and death. This article will highlight<br />
some <strong>of</strong> the emerging data that suggest biology plays an<br />
important role in local treatment outcome <strong>of</strong> breast cancer.<br />
Breast Conservation<br />
The local-regional treatment outcomes after breastconservation<br />
surgery and whole-breast radiation are excellent.<br />
Improvements in imaging techniques, increased<br />
attention to surgical margins, and increased use <strong>of</strong> systemic<br />
therapy have helped reduce recurrence rates to very low<br />
levels. For example, our institutional experience <strong>of</strong> BCT<br />
(lumpectomy plus whole-breast radiation) for 974 patients<br />
treated between 1973 and 1993 noted a 5-year risk <strong>of</strong><br />
in-breast recurrence <strong>of</strong> 5.7%. This compared with a 5-year<br />
recurrence risk <strong>of</strong> only 1.3% noted in the 381 patients<br />
treated between 1994 and 1996. 1<br />
Despite these overall excellent outcomes, recent data have<br />
demonstrated that some subtypes <strong>of</strong> breast cancer have<br />
higher local recurrence rates. For example, investigators<br />
from the Harvard Radiation <strong>Oncology</strong> program analyzed a<br />
cohort <strong>of</strong> 1,223 patients treated with BCT and noted 5-year<br />
breast recurrence rates <strong>of</strong> 4.4% for patients with triplenegative<br />
disease, 9% for those with HER2-positive disease<br />
(predated trastuzumab), and only 1% for patients with<br />
ER-positive disease. 2 These data are similar to a report<br />
from University <strong>of</strong> Texas M. D. Anderson Cancer Center<br />
(MDACC) that analyzed 911 patients treated with BCT for<br />
primaries under 1 cm and negative lymph nodes and found<br />
8-year breast recurrence rates <strong>of</strong> 18% in HER2/neu-positive<br />
disease (predated trastuzumab), 11% for those with ERnegative<br />
disease, and only 4% for those with ER-positive<br />
disease. 3 Finally, an elegant study from British Columbia<br />
that investigated more than 1,400 patients with BCT reported<br />
a higher 10-year local-regional recurrence rate in<br />
56<br />
By Thomas A. Buchholz, MD<br />
treatments, either given as a component <strong>of</strong> breast-conservative<br />
therapy (BCT) or for those with more advanced disease, when<br />
given after mastectomy. For patients with triple-negative<br />
disease, data suggest that the proportional benefits <strong>of</strong>fered<br />
from radiation in reducing local-regional recurrences may be<br />
less. This article will highlight some <strong>of</strong> these data and discuss<br />
strategies for new local-regional research avenues that are<br />
based on breast cancer biologic subtype.<br />
patients with HER2/neu-enriched (21%) and basal-like (16%)<br />
cancers compared with the much more common luminal A<br />
subtype (8%). 4<br />
Some groups have analyzed local recurrence after BCT<br />
using more sophisticated genomic signatures. For example,<br />
a cDNA microarray-based wound signature successfully<br />
classified patients into two groups with either high or low<br />
risk <strong>of</strong> local-regional recurrence. 5 In a different study, investigators<br />
from the National Surgical Adjuvant Breast and<br />
Bowel Project (NSABP) analyzed the 21-gene pr<strong>of</strong>ile (Oncotype<br />
DX) in more than 1,500 patients with node-negative<br />
ER-positive disease and reported that patients with a higher<br />
gene pr<strong>of</strong>ile score had a higher risk <strong>of</strong> local recurrence. 6<br />
It remains a question how data such as these should affect<br />
local-regional treatment decisions. The majority <strong>of</strong> patients<br />
treated with BCT have ER-positive or luminal A type<br />
tumors and these data collectively indicate very low rates <strong>of</strong><br />
breast recurrence. Accordingly, the focus <strong>of</strong> future studies<br />
in this population should be aimed at minimizing the cost,<br />
inconvenience, and toxicity <strong>of</strong> such treatments. For patients<br />
with higher-risk biologic features, improvements in systemic<br />
treatments may help mitigate risks. For example, the<br />
NSABP study <strong>of</strong> oncotype noted only an 8% 10-year risk for<br />
patients with high recurrence scores who received chemotherapy<br />
and tamoxifen compared with a 16% risk for those<br />
who received tamoxifen alone. 6 Systemic treatments also<br />
likely overcome some <strong>of</strong> the adverse effects <strong>of</strong> HER2/neu<br />
overexpression. Data from the B-31/N9831 trials indicated<br />
the use <strong>of</strong> adjuvant trastuzumab reduced the risk <strong>of</strong> localregional<br />
recurrence in HER2/neu-positive cancers by 42%<br />
(crude event rates, 47 <strong>of</strong> 1,679 for no trastuzumab compared<br />
with 27 <strong>of</strong> 1,672 with trastuzumab). 7 In addition, a study<br />
from Memorial Sloan-Kettering Cancer Center noted 3-year<br />
local-regional recurrence rates <strong>of</strong> 10% for patients with<br />
HER2/neu-positive tumors treated before trastuzumab and<br />
only a 1% rate in those receiving adjuvant trastuzumab. 8<br />
The one cohort <strong>of</strong> patients who remain at higher risk <strong>of</strong><br />
recurrence after BCT appears to be those with triplenegative<br />
disease. For such patients, it is reasonable to help<br />
From the Division <strong>of</strong> Radiation <strong>Oncology</strong>, Department <strong>of</strong> Radiation <strong>Oncology</strong>, University<br />
<strong>of</strong> Texas M. D. Anderson Cancer Center, Houston, TX.<br />
Author’s disclosure <strong>of</strong> potential conflicts <strong>of</strong> interest are found at the end <strong>of</strong> this article.<br />
Address reprint requests to Thomas A. Buchholz, MD, 1515 Holcombe Blvd., Division <strong>of</strong><br />
Radiation <strong>Oncology</strong>, Unit 97, University <strong>of</strong> Texas M. D. Anderson Cancer Center, Houston,<br />
TX 77030; email: tbuchhol@mdanderson.org.<br />
© <strong>2012</strong> by <strong>American</strong> <strong>Society</strong> <strong>of</strong> <strong>Clinical</strong> <strong>Oncology</strong>.<br />
1092-9118/10/1-10
BIOLOGY AND LOCAL THERAPY DECISIONS<br />
mitigate this risk through routine use <strong>of</strong> chemotherapy,<br />
achievement <strong>of</strong> negative surgical margin status, and use <strong>of</strong><br />
a tumor bed boost. Patients with triple-negative disease<br />
should not be routinely recommended to undergo mastectomy.<br />
A recent study from Edmonton, Canada, evaluated<br />
768 patients with T1–2N0 triple-negative disease and found<br />
better 5-year local-regional recurrence-free survival after<br />
BCT with radiation compared with mastectomy without<br />
radiation (96% compared with 90%, respectively, p � 0.027;<br />
hazard ratio, 2.53; p � 0.0264). 9<br />
Mastectomy with or without Radiation<br />
Biologic subtypes also appear to affect the risk <strong>of</strong> localregional<br />
recurrence after mastectomy. For example, a British<br />
Columbia study that combined data from patients<br />
treated with mastectomy alone and patients treated with<br />
mastectomy plus radiation demonstrated that luminal A<br />
subtype independently correlated with a lower risk <strong>of</strong> localregional<br />
recurrence after mastectomy compared with the<br />
other subtypes. 4 Interestingly, in an analysis <strong>of</strong> patients<br />
KEY POINTS<br />
● Estrogen receptor (ER)-positive, HER2/neu-negative<br />
breast cancers have very low local-regional recurrence<br />
rates after adjuvant radiation treatments.<br />
● Triple-negative breast cancer has higher rates <strong>of</strong><br />
local-regional recurrence after adjuvant radiation<br />
compared with similar stages <strong>of</strong> ER-positive disease.<br />
● Breast-conservative therapy remains an appropriate<br />
treatment option for patients with early-stage, triplenegative<br />
disease.<br />
Author’s Disclosure <strong>of</strong> Potential Conflicts <strong>of</strong> Interest<br />
Author<br />
Thomas A. Buchholz*<br />
*No relevant relationships to disclose.<br />
Employment or<br />
Leadership<br />
Positions<br />
Consultant or<br />
Advisory Role<br />
1. Cabioglu N, Hunt KK, Buchholz TA, et al. Improving local control with<br />
breast-conserving therapy: A 27-year single-institution experience. Cancer.<br />
2005;104:20-29.<br />
2. Nguyen PL, Taghian AG, Katz MS, et al. Breast cancer subtype approximated<br />
by estrogen receptor, progesterone receptor, and HER-2 is associated<br />
with local and distant recurrence after breast-conserving therapy. J Clin<br />
Oncol. 2008;26:2373-2378.<br />
3. Albert JM, Gonzalez-Angulo AM, Guray M, et al. Estrogen/progesterone<br />
receptor negativity and HER2 positivity predict locoregional recurrence in<br />
patients with T1a,bN0 breast cancer. Int J Radiat Oncol Biol Phys. 2010;77:<br />
1296-1302.<br />
4. Voduc KD, Cheang MC, Tyldesley S, et al. Breast cancer subtypes and<br />
the risk <strong>of</strong> local and regional relapse. J Clin Oncol. 2010;28:1684-1691.<br />
5. Nuyten DS, Kreike B, Hart AA, et al. Predicting a local recurrence after<br />
breast-conserving therapy by gene expression pr<strong>of</strong>iling. Breast Cancer Res.<br />
2006;8:R62.<br />
6. Mamounas EP, Tang G, Fisher B, et al. Association between the 21-gene<br />
recurrence score assay and risk <strong>of</strong> locoregional recurrence in node-negative,<br />
estrogen receptor-positive breast cancer: Results from NSABP B-14 and<br />
NSABP B-20. J Clin Oncol. 2010;28:1677-1683.<br />
7. Romond EH, Perez EA, Bryant J, et al. Trastuzumab plus adjuvant<br />
treated on the Danish postmastectomy radiation trials and<br />
those who were randomly selected to not receive radiation,<br />
the risks <strong>of</strong> local-regional recurrence for patients with ERpositive<br />
disease and those with ER-negative disease was<br />
similar (34% and 31%, respectively). 10 However, in the<br />
patients randomly selected to receive postmastectomy radiation,<br />
the rates <strong>of</strong> local-regional recurrence were lower in<br />
the patients with ER-positive disease compared with those<br />
with ER-negative disease (4% and 14%, respectively). 10<br />
Similarly, a report <strong>of</strong> local-regional treatment outcome <strong>of</strong><br />
patients treated at MDACC with mastectomy, no radiation,<br />
and adjuvant chemotherapy did not find ER status to independently<br />
correlate with local-regional recurrence (15% ER<br />
positive and 12% ER negative). 11 However, in a recursive<br />
partition analysis <strong>of</strong> patients treated with postmastectomy<br />
radiation, ER-negative disease proved to be the most powerful<br />
discriminator <strong>of</strong> local-regional recurrence risk. 12<br />
Radiation Resistance and Future Research<br />
Collectively, the data from breast conservation and postmastectomy<br />
radiation raise the hypothesis that ER negativity<br />
or triple-negative disease may be a more radiationresistant<br />
histology than ER-positive disease. These data<br />
were also supported by the most recent update <strong>of</strong> the Early<br />
Breast Cancer Trialists’ Group meta-analysis <strong>of</strong> trials<br />
comparing use with omission <strong>of</strong> radiation after breastconservation<br />
surgery. In these trials, the proportional reduction<br />
in the risk <strong>of</strong> recurrence with radiation was nearly<br />
60% for ER-positive disease treated with tamoxifen, approximately<br />
50% for ER-positive disease without tamoxifen<br />
treatment, and only 35% for ER-negative disease. These<br />
data indicate that additional research is needed to identify<br />
potential molecular targets in triple-negative disease and to<br />
design new therapeutic strategies to enhance radiation<br />
effects in this subtype.<br />
Stock<br />
Ownership Honoraria<br />
REFERENCES<br />
Research<br />
Funding<br />
Expert<br />
Testimony<br />
Other<br />
Remuneration<br />
chemotherapy for operable HER2-positive breast cancer. N Engl J Med.<br />
2005;353:1673-1684.<br />
8. Kiess AP, McArthur HL, Mahoney K, et al. Adjuvant trastuzumab<br />
reduces locoregional recurrence in women who receive breast-conservation<br />
therapy for lymph node-negative, human epidermal growth factor receptor<br />
2-positive breast cancer. Cancer. Epub 2011 Sep 1.<br />
9. Abdulkarim BS, Cuartero J, Hanson J, et al. Increased risk <strong>of</strong> locoregional<br />
recurrence for women with T1-2N0 triple-negative breast cancer<br />
treated with modified radical mastectomy without adjuvant radiation therapy<br />
compared with breast-conserving therapy. J Clin Oncol. 2011;29:2852-2858.<br />
10. Kyndi M, Sorensen FB, Knudsen H, et al. Estrogen receptor, progesterone<br />
receptor, HER-2, and response to postmastectomy radiotherapy in<br />
high-risk breast cancer: The Danish Breast Cancer Cooperative Group. J Clin<br />
Oncol. 2008;26:1419-1426.<br />
11. Katz A, Strom EA, Buchholz TA, et al. Loco-regional recurrence<br />
patterns following mastectomy and doxorubicin-based chemotherapy: Implications<br />
for postoperative irradiation. J Clin Oncol. 2000;18:2817-2827.<br />
12. Woodward WA, Strom EA, Tucker SL, et al. Locoregional recurrence<br />
after doxorubicin-based chemotherapy and postmastectomy: Implications for<br />
breast cancer patients with early-stage disease and predictors for recurrence<br />
after postmastectomy radiation. Int J Radiat Oncol Biol Phys. 2003;57:336-344.<br />
57
THE MANY OPTIONS FOR BREAST IMAGING:<br />
WHAT TO USE WHEN<br />
CHAIR<br />
Chris I. Flowers, MD<br />
H. Lee M<strong>of</strong>fitt Cancer Center and Research Institute<br />
Tampa, FL<br />
SPEAKERS<br />
Maxine Jochelson, MD<br />
Memorial Sloan-Kettering Cancer Center<br />
New York, NY<br />
Karla Kerlikowske, MD<br />
University <strong>of</strong> California, San Francisco<br />
San Francisco, CA
<strong>Clinical</strong> and Imaging Surveillance Following<br />
Breast Cancer Diagnosis<br />
By Chris I. Flowers, MBBS, Blaise P. Mooney, MD, and Jennifer S. Drukteinis, MD<br />
Overview: Breast cancer is the most common malignancy<br />
affecting women worldwide. Women have a1in8lifetime risk<br />
<strong>of</strong> breast cancer. Breast conservation therapy (BCT) is the<br />
most common method <strong>of</strong> definitive treatment. Patients who<br />
previously have had to undergo mastectomy may be now<br />
eligible for BCT or a multitude <strong>of</strong> options for reconstruction,<br />
either immediate or delayed. Surveillance imaging after a<br />
breast cancer diagnosis is important because there is an<br />
increased risk <strong>of</strong> recurrence developing in patients, and early<br />
detection has been shown to improve survival. There is<br />
currently no consensus on a protocol for imaging the postoperative<br />
breast. In patients who have undergone mastectomy,<br />
LOCAL RECURRENCE, second ipsilateral breast cancer,<br />
and contralateral breast cancer may develop in<br />
women with a personal history <strong>of</strong> breast cancer. 1-3 Lumpectomy<br />
followed by whole breast irradiation or BCT is the most<br />
common treatment following a diagnosis <strong>of</strong> breast cancer.<br />
BCT <strong>of</strong> clinical stage I and stage II breast cancer has become<br />
a standard <strong>of</strong> care, with long-term studies showing no<br />
significant difference (p � 0.001) in survival rates between<br />
those treated with BCT as opposed to mastectomy. 4-10 Early<br />
detection <strong>of</strong> local recurrence <strong>of</strong> breast cancer has been shown<br />
to improve long-term survival (HR 2.44 [95% CI] if without<br />
symptoms); therefore, it is important to optimize the use <strong>of</strong><br />
clinical and imaging tools for the patient with a history <strong>of</strong><br />
breast cancer to diagnose recurrence in its most early stages.<br />
Locoregional recurrence occurs in approximately 5% <strong>of</strong><br />
patients at 5 years with a local failure rate <strong>of</strong> approximately<br />
1% to 2.5% per year. In the immediate postoperative period,<br />
suspicious findings likely represent residual disease, whereas<br />
local recurrence typically occurs 3–7 years after BCT. 11,12<br />
Many factors affect local, regional recurrence and distant<br />
metastasis. These include age at diagnosis, primary tumor<br />
subtype and size, presence <strong>of</strong> local or regional node disease,<br />
surgical treatment, use and type <strong>of</strong> radiation therapy, and<br />
use <strong>of</strong> adjuvant hormonal therapy and chemotherapy.<br />
With the increasing number and variability <strong>of</strong> oncoplastic<br />
procedures, there are many potential ways in which local<br />
recurrence can appear and be detected on imaging. Mammographic<br />
surveillance has shown utility in the detection<br />
<strong>of</strong> recurrent disease, and in screening the contralateral<br />
breast for early breast cancer. As a result, mammographic<br />
follow-up has become the norm for women who have been<br />
treated for breast cancer. There are, however, significant<br />
variations in the way this has been implemented across the<br />
United States and the world.<br />
Most published data compare mammographic surveillance<br />
and physical examination. This combination has been<br />
the standard <strong>of</strong> care and is most frequently used in discussion<br />
<strong>of</strong> follow-up protocols. In contrast, mammography <strong>of</strong> the<br />
treated breast for recurrence is seldom performed routinely<br />
for patients who have had a mastectomy and breast reconstruction,<br />
with physical examination being the primary tool<br />
to detect recurrence. MRI has been shown to be a valuable<br />
tool in evaluating the reconstructed breast following mastectomy.<br />
MRI may also have an increasing role in the<br />
detection <strong>of</strong> recurrence has mostly been via clinical symptoms<br />
and physical exam, <strong>of</strong>ten at a later stage. New imaging<br />
modalities, such as magnetic resonance imaging (MRI), ultrasound<br />
(US), and positron emission mammography (PEM) are<br />
changing the way we image the postsurgical breast. MRI,<br />
coupled with physical exam and mammography, approaches<br />
100% sensitivity and high specificity for the identification <strong>of</strong><br />
recurrent disease. We present a review <strong>of</strong> major academic<br />
institutions’ imaging protocols and discuss the advantages<br />
<strong>of</strong> including MRI in traditional mammographic and clinical<br />
exams.<br />
treatment <strong>of</strong> patients who have undergone BCT or women<br />
with an originally mammographically occult primary breast<br />
cancer.<br />
When discussing the follow-up examination <strong>of</strong> a patient<br />
previously treated for breast cancer, we must remember that<br />
the ultimate goal <strong>of</strong> imaging following a breast cancer<br />
diagnosis is to detect recurrence or new primary cancer<br />
before it is clinically detectable.<br />
Surveillance <strong>of</strong> the contralateral breast is essentially a<br />
type <strong>of</strong> screening exam, except that it is targeted to a<br />
high-risk group. As a planned medical intervention, this<br />
is <strong>of</strong>ten referred to as surveillance. However, in a patient<br />
with breast conservation, the contralateral breast is being<br />
screened.<br />
There is very little data available to make specific guidelines<br />
about surveillance <strong>of</strong> the patient previously treated for<br />
breast cancer. Therefore, a review <strong>of</strong> the limited literature<br />
and surveillance protocols at a few selected academic centers<br />
will be presented.<br />
Effectiveness <strong>of</strong> Mammography<br />
Mammography has been proven to be efficacious as a<br />
screening modality in detecting breast cancer recurrence,<br />
with various papers showing a 50% detection rate <strong>of</strong> recurrent<br />
tumors by mammography and the rest by physical<br />
examination. 13-15 Depending on the type <strong>of</strong> prior surgery,<br />
there may be confounding and distracting features both on<br />
physical examination and on imaging, and the number <strong>of</strong><br />
false-positive examinations may be higher than those for the<br />
general population.<br />
Recurrent tumors detected by surveillance mammography<br />
are in general smaller and less invasive than those found<br />
during clinical examination, as demonstrated by Lu and<br />
colleagues. 16 This systematic review compared isolated locoregional<br />
recurrence or contralateral cancers on survival.<br />
The authors demonstrated better overall survival for mam-<br />
From the H. Lee M<strong>of</strong>fitt Cancer Center and Research Institute, Tampa, FL, and<br />
University <strong>of</strong> South Florida, Tampa, FL.<br />
Authors’ disclosures <strong>of</strong> potential conflicts <strong>of</strong> interest are found at the end <strong>of</strong> this article.<br />
Address reprint requests to Chris Flowers, MBBS, H. Lee M<strong>of</strong>fitt Cancer Center and<br />
Research Institute, 12902 Magnolia Dr., Tampa, FL 33612; email: chris.flowers@<br />
m<strong>of</strong>fitt.org.<br />
© <strong>2012</strong> by <strong>American</strong> <strong>Society</strong> <strong>of</strong> <strong>Clinical</strong> <strong>Oncology</strong>.<br />
1092-9118/10/1-10<br />
59
mographically detected recurrence, or in asymptomatic patients.<br />
They also demonstrated an absolute reduction in<br />
mortality <strong>of</strong> 17% to 28% in all patients with breast cancers<br />
found early.<br />
A recent review in the United Kingdom for the Health<br />
Technology Assessment by Robertson and colleagues 17 also<br />
demonstrated that for screening for local recurrence, surveillance<br />
mammography sensitivity ranged from 64% to 67%<br />
and specificity ranged from 85% to 97%. They also found<br />
that for MRI, sensitivity ranged from 86% to 100% and<br />
specificity was 93%. They concluded that mammography is<br />
associated with a high sensitivity and specificity but that<br />
MRI is the most accurate test for detecting local recurrence.<br />
Developing Role <strong>of</strong> MRI Post-BCT Surveillance<br />
MRI is emerging as a valuable tool in the treatment <strong>of</strong><br />
patients with a diagnosis <strong>of</strong> breast cancer. Conventional<br />
imaging techniques <strong>of</strong> mammography and US can be difficult<br />
to interpret as a result <strong>of</strong> post-therapeutic changes, and<br />
some centers are including breast MRI in the surveillance <strong>of</strong><br />
women who are post-BCT, as surveillance MRI in this<br />
population has been shown to detect malignancy in 12%. 18<br />
In the Robertson review, 17 MRI was found to be the most<br />
accurate test for detecting ipsilateral tumor recurrence and<br />
contralateral breast cancer in women previously treated for<br />
primary breast cancer. MRI in combination with mammography,<br />
US, and clinical examination has been reported to<br />
have a sensitivity as high as 100% and specificity <strong>of</strong> 89% for<br />
detecting contralateral breast cancer in patients who have<br />
undergone BCT. 19 MRI has been shown to be extremely useful<br />
in differentiating scar tissue from tumor recurrence, particularly<br />
in showing that nonenhancing areas have a high<br />
negative predictive value for malignancy (88% to 96%). 20,21<br />
The use <strong>of</strong> MRI in the monitoring <strong>of</strong> the post-BCT breast<br />
is variable and <strong>of</strong>ten at the discretion <strong>of</strong> the ordering<br />
physician or surgeon. The <strong>American</strong> College <strong>of</strong> Radiology<br />
practice guidelines state that breast MRI may be useful for<br />
women with a prior history <strong>of</strong> breast cancer and suspicion <strong>of</strong><br />
recurrence when clinical, mammographic, and/or sonographic<br />
findings are inconclusive. 22 Although women with<br />
a previous diagnosis <strong>of</strong> breast cancer are at increased risk <strong>of</strong><br />
a second diagnosis, an <strong>American</strong> Cancer <strong>Society</strong> panel concluded<br />
that the increased risk due to a personal history <strong>of</strong><br />
breast cancer alone does not justify a recommendation for<br />
overall screening in women post-BCT at the present time. 23<br />
Neither National Comprehensive Cancer Network nor<br />
60<br />
KEY POINTS<br />
● Mammographic surveillance has a high sensitivity<br />
and specificity for recurrent malignancy.<br />
● Magnetic resonance imaging is the most sensitive<br />
and accurate tool for evaluation <strong>of</strong> the postoperative<br />
breast.<br />
● Physical examination has an important role in surveillance.<br />
● Robust conclusions cannot be made because <strong>of</strong> the<br />
limited evidence base.<br />
● Further research comparing surveillance mammography<br />
and newer diagnostic tests is required.<br />
ASCO guidelines consider MRI for surveillance. Changes<br />
occurring after BCT on MRI are similar to those described<br />
on mammography and include architectural distortion,<br />
edema, skin thickening, and occasionally a seroma. These<br />
findings are typical and may stabilize or continue to decrease<br />
over time. Enhancement at the lumpectomy site can<br />
be an expected finding depending on the time interval from<br />
BCT. A minimal or small focal area <strong>of</strong> thin linear enhancement<br />
can be seen for up to 18 months, and in some cases<br />
even longer. 24,25 On follow-up imaging, enhancement should<br />
demonstrate stability or decrease over time. Fat necrosis is<br />
commonly present in the post-BCT breast and can be a<br />
challenging pitfall, <strong>of</strong>ten leading to biopsy. Enhancement is<br />
variable in appearance and kinetic patterns. Biopsy may be<br />
avoided if the MRI signal is similar to that <strong>of</strong> adjacent fat on<br />
unenhanced images and the lesion shows no internal enhancement.<br />
For this reason, nonfat-saturated T1-weighted<br />
sequences should be included on all breast MRI patients.<br />
Mass-like enhancement at the lumpectomy site is always<br />
suspicious and requires biopsy. Similarly, clumped, new, or<br />
increasing areas <strong>of</strong> nonmass-like enhancement (not associated<br />
with fat necrosis) should be considered suspicious. 26<br />
Much <strong>of</strong> the utilization <strong>of</strong> MRI is physician- or surgeondependent.<br />
Many centers stagger MRI and mammograms at<br />
6-month intervals. The reasoning behind this is that the<br />
breast is surveyed in a different modality every 6 months. At<br />
our institution, we prefer that MRI be performed 1 year<br />
following surgery, so that it coincides with the bilateral<br />
mammographic exam. This clarifies indeterminate findings<br />
in the contralateral breast and allows normal postsurgical/<br />
radiation therapy enhancement to decrease. Either approach<br />
is acceptable, as there are no specific data-driven<br />
guidelines or a consensus on recommendations for MRI<br />
surveillance intervals in this population. Despite breast<br />
cancer survivors having a high risk <strong>of</strong> second cancer,<br />
Punglia and Hassett 27 recommend against routine MRI,<br />
suggesting selective use by lifetime risk estimates.<br />
Breast Conservation Surveillance<br />
All agree that surveillance <strong>of</strong> patients who are post-BCT is<br />
necessary. However, there is a lack <strong>of</strong> consensus on how<br />
frequently and for how long we should provide routine<br />
surveillance after treatment. Subsequently, there are several<br />
ways that surveillance can take place and may vary<br />
from institution to institution (Table 1). The majority <strong>of</strong><br />
centers surveyed for this article have a common initial<br />
2-year pathway, changing to annual follow-up at year 3.<br />
Benign Findings Mimicking Malignancy<br />
FLOWERS, MOONEY, AND DRUKTEINIS<br />
There are many benign findings that can mimic malignancy<br />
in the postsurgical breast, but essentially they may be<br />
broken down into two categories: a) developing microcalcification,<br />
and b) variants <strong>of</strong> fat necrosis. The aggressive variant<br />
<strong>of</strong> fat necrosis can produce a very hard, spiculate scar.<br />
The matrix may also calcify without the characteristic<br />
dystrophic calcification that is normally associated with it.<br />
This variant <strong>of</strong>ten produces a fine, lace-like, rapidly developing<br />
calcification, prompting biopsy.<br />
Calcifications may develop in a patient who has been<br />
treated for ductal carcinoma in situ (DCIS), especially if<br />
there was noncalcified disease followed by radiation therapy.<br />
There may be an increase in necrosis due to apoptosis <strong>of</strong>
SURVEILLANCE FOLLOWING BREAST CANCER DIAGNOSIS<br />
cells in the ducts, with subsequent developing ductal microcalcifications.<br />
Some centers advocate a postsurgical, preradiation<br />
therapy mammogram <strong>of</strong> the treated side as a<br />
baseline for just this purpose.<br />
In general, calcifications developing in the first 2 years are<br />
dystrophic and thereafter are more likely as a result <strong>of</strong><br />
recurrent disease.<br />
For palpable findings, targeted breast US is the initial<br />
imaging modality <strong>of</strong> choice. Findings are usually related to<br />
scarring or to a lump caused by development <strong>of</strong> fat necrosis.<br />
Imaging Appearances <strong>of</strong> Recurrent Disease<br />
Table 1. Examples <strong>of</strong> Breast Conservation Surveillance Protocol by Institution<br />
Institution Mammography Physical Examination<br />
M. D. Anderson Cancer Center At 6 mo, then annual diagnostic for 5 yr; screening from year 6 Every 6 mo for 5 yr, then annual<br />
University <strong>of</strong> California, San Francisco Every 6 mo for 5 yr, then annual screening Every 6 mo for 5 yr, then annual<br />
University <strong>of</strong> California, Los Angelas Every 6 mo for 2 yr, then annual diagnostic Left to primary care<br />
M<strong>of</strong>fitt Cancer Center Every 6 mo for 2 yr, then annual diagnostic until yr 10 Every 6 mo for 5 yr, then annual<br />
Brigham and Women’s Hospital and<br />
Dana-Farber Cancer Center<br />
At 6 mo after completion <strong>of</strong> radiotherapy, then annual diagnostic Every 6 mo for 5 yr<br />
Abbreviations: mo, months; yr, years.<br />
In general, a recurrent cancer shows similar appearances<br />
to the original malignancy, such that a patient with DCIS<br />
will have developing microcalcifications in the treated breast,<br />
and high-grade tumors with DCIS may present as a developing<br />
focal asymmetry containing calcifications (Fig. 1).<br />
Fig. 1. Example <strong>of</strong> recurrent DCIS with a sharp margin corresponding<br />
to the edge <strong>of</strong> the radiotherapy boost site.<br />
Abbreviation: DCIS, ductal carcinoma in situ.<br />
We recommend that spot magnification views be part <strong>of</strong><br />
the initial post-lumpectomy mammogram for patients in<br />
whom calcified DCIS is present, whether invasive or in situ.<br />
This will act as the new baseline exam for comparison.<br />
Developing microcalcifications usually ring alarm bells but<br />
can be commonly seen in patients prone to developing fat<br />
necrosis.<br />
Special types <strong>of</strong> tumors, such as angiosarcomas, require<br />
more specialized follow-up, which can be a combination <strong>of</strong><br />
Doppler US and MRI. The permeative vascular nature <strong>of</strong> the<br />
recurrent tumor is best seen on MRI (Fig. 2).<br />
Mastectomy without Reconstruction<br />
A patient who has been treated by mastectomy has a<br />
contralateral breast cancer risk <strong>of</strong> approximately 7%, and so<br />
surveillance <strong>of</strong> the treated breast is almost entirely covered<br />
by physical examination, with targeted US scan <strong>of</strong> any<br />
palpable area. The contralateral breast will continue to be<br />
monitored by a regular annual mammographic examination,<br />
with additional imaging being reserved for patients who are<br />
symptomatic or who have a palpable finding.<br />
Male Survivors <strong>of</strong> Breast Cancer<br />
These are the forgotten survivors <strong>of</strong> cancer. The surveillance<br />
protocol for a male patient should be the same as that<br />
for a woman who has had a mastectomy without reconstruction.<br />
Most recurrences in men should be palpable, hence the<br />
importance <strong>of</strong> a regular physical exam, and possibly “breast<br />
awareness” on the part <strong>of</strong> the man. If the patient has had a<br />
Fig. 2. MRI image showing enhancing nodule adjacent to seroma/<br />
lumpectomy site. Patient is 5 years post-surgery and 2 months<br />
post-completion <strong>of</strong> 5 years <strong>of</strong> tamoxifen.<br />
Abbreviation: MRI, magnetic resonance imaging.<br />
61
lumpectomy rather than mastectomy, then there is still no<br />
reason to vary the protocol from that <strong>of</strong> a female patient.<br />
The Reconstructed Breast<br />
The reconstructed breast has not been the subject <strong>of</strong><br />
imaging surveillance until more recently. Some advise the<br />
use <strong>of</strong> mammographic surveillance <strong>of</strong> transverse rectus<br />
abdominus myocutaneous (TRAM) flap reconstructions,<br />
sometimes called “tramograms,” which can be performed<br />
just like any other surveillance mammogram. 28,29 The findings<br />
<strong>of</strong> recurrent cancer are typical and have been reported<br />
before they become palpable, advancing the stage <strong>of</strong> diagnosis.<br />
Others advocate for only selecting women with a high<br />
risk-<strong>of</strong>-recurrence score to have tramograms (Fig. 3).<br />
The most sensitive test for the reconstructed breast is<br />
MRI, although it is a relatively expensive test that is not<br />
always covered by health insurance in this situation, even<br />
Fig. 3. Tramogram image showing TRAM recurrence in the right<br />
XCCL mammogram. Lesion was not initially not detected until the<br />
mammogram finding showed the lesion.<br />
Abbreviations: TRAM, transverse rectus abdominus myocutaneous;<br />
XCCL, exaggerated craniocaudal view.<br />
62<br />
FLOWERS, MOONEY, AND DRUKTEINIS<br />
Fig. 4. Developing microcalcifications in scar on routine mammogram.<br />
though it is the most effective test. According to Devon, 30<br />
MRI is useful in the detection <strong>of</strong> locally recurrent tumors in<br />
patients who have undergone breast reconstruction with a<br />
TRAM flap. It detects cancer recurrences at an earlier stage<br />
than noncompliant annual surveillance.<br />
Improved access to dedicated breast MRI is making evaluation<br />
<strong>of</strong> the reconstructed breast more affordable and<br />
accessible. Early detection may potentially save health care<br />
dollars by catching recurrence before loco-regional spread,<br />
but there are no data to show that this improves survival.<br />
False-Positive Examinations on Follow-up<br />
There are frequently false-positive clinical findings after<br />
reconstruction, mainly with palpable lumps. Palpable lesions<br />
should be first evaluated with US. The palpable lumps<br />
are generally multiple, appear in the upper outer quadrant<br />
or 6 o’clock positions, and are caused by fat necrosis. This is<br />
more common in women with implant reconstruction than
SURVEILLANCE FOLLOWING BREAST CANCER DIAGNOSIS<br />
Fig. 5. Palpable mass near scar; US shows typical features <strong>of</strong> fat<br />
necrosis.<br />
Abbreviation: US, ultrasound.<br />
TRAM flaps in our practice but may vary according to the<br />
distributions <strong>of</strong> types <strong>of</strong> reconstruction in your group. Mammography<br />
<strong>of</strong>ten demonstrates early dystrophic microcalcifications,<br />
which can simulate an aggressive recurrent DCIS<br />
(Fig. 4 and Fig. 5).<br />
Bilateral Mastectomy with Reconstruction<br />
Little is known about the utility <strong>of</strong> MRI as a diagnostic or<br />
surveillance tool after bilateral mastectomy with reconstruction.<br />
The current role <strong>of</strong> MRI in treating patients who have<br />
had a mastectomy is primarily as a diagnostic tool when<br />
clinical findings need clarification. Most recurrences adjacent<br />
to an implant are clinically palpable, and there is little<br />
evidence to support the use <strong>of</strong> MRI in asymptomatic postmastectomy<br />
with reconstruction patients. Patients who<br />
have undergone reconstruction with TRAM or deep inferior<br />
epigastric perforator flaps can be evaluated with MRI;<br />
however, routine surveillance is not common practice, as the<br />
risk <strong>of</strong> nonpalpable recurrence is extremely low. 30 MRI has<br />
been found useful in differentiating benign from malignant<br />
findings in patients with palpable abnormalities or pain<br />
after TRAM reconstruction (Fig. 6). 30<br />
Possible Developing Methods for<br />
Recurrence Detection<br />
To detect additional tumors or metastatic disease at an<br />
early stage, advanced imaging techniques, including diagnostic<br />
mammographic images, focused breast US, breast<br />
Authors’ Disclosures <strong>of</strong> Potential Conflicts <strong>of</strong> Interest<br />
Author<br />
Chris I. Flowers*<br />
Blaise P. Mooney*<br />
Jennifer S. Drukteinis*<br />
*No relevant relationships to disclose.<br />
Employment or<br />
Leadership<br />
Positions<br />
Consultant or<br />
Advisory Role<br />
1. Voogd AC, van Tienhoven G, Peterse HL, et al. Local recurrence after<br />
breast conservation therapy for early stage breast carcinoma: detection,<br />
treatment, and outcome in 266 patients. Cancer. 1999;85:437-446.<br />
2. Hietanen P, Miettinen M, Makinen J. Survival after first recurrence in<br />
breast cancer. Eur J Cancer Clin Oncol. 1986;22:913-919.<br />
MRI, and positron emission tomography imaging is <strong>of</strong>ten<br />
required. Breast-specific gamma imaging, PEM, and molecular<br />
breast imaging may have a problem-solving role, especially<br />
in the dense breast or where there are complications<br />
from ruptured silicone implants. Radiologists who have<br />
completed advanced training in breast cancer detection<br />
should interpret these imaging techniques. Imaging should<br />
be coupled with clinical exams by specialty-trained clinicians<br />
familiar with signs and symptoms <strong>of</strong> recurrence and<br />
should occur at set time intervals. The downside <strong>of</strong> these<br />
newer techniques is the radiation dose from the radioisotopes<br />
currently used, although new techniques are focused<br />
on reducing the radiation dose.<br />
Advanced reconstruction and radiation techniques can be<br />
difficult to interpret unless communication between members<br />
<strong>of</strong> the health care team is timely and accurate, and<br />
the radiologist is familiar with expected appearances. To<br />
achieve the goal <strong>of</strong> recurrence detection in the early stages,<br />
before it is clinically evident, radiologists and clinicians<br />
must be familiar with the behavior <strong>of</strong> subtypes <strong>of</strong> breast<br />
cancer and different surgical/medical/oncoplastic approaches<br />
to their treatment.<br />
Stock<br />
Ownership Honoraria<br />
REFERENCES<br />
Fig. 6. Vague nodularity on physical exam. US <strong>of</strong> nodularity<br />
showed no correlation. MRI demonstrates obvious local recurrence<br />
and metastases.<br />
Abbreviations: US, ultrasound; MRI, magnetic resonance imaging.<br />
Research<br />
Funding<br />
Expert<br />
Testimony<br />
Other<br />
Remuneration<br />
3. Nomura Y, Tsutsui S, Murakami S, et al. Prognostic impact <strong>of</strong> second<br />
cancer on the survival <strong>of</strong> early breast cancer patients. Int J Oncol. 1999;14:<br />
1103-1109.<br />
4. Fisher B, Anderson S, Bryant J, et al. Twenty-year follow-up <strong>of</strong> a<br />
randomized trial comparing total mastectomy, lumpectomy, and lumpectomy<br />
63
plus irradiation for the treatment <strong>of</strong> invasive breast cancer. N Engl J Med.<br />
2002;347:1233-1241.<br />
5. Fisher B, Jeong JH, Anderson S, et al. Twenty-five-year follow-up <strong>of</strong> a<br />
randomized trial comparing radical mastectomy, total mastectomy, and total<br />
mastectomy followed by irradiation. N Engl J Med. 2002;347:567-575.<br />
6. Fisher B, Redmond C, Poisson R, et al. Eight-year results <strong>of</strong> a randomized<br />
clinical trial comparing total mastectomy and lumpectomy with or<br />
without irradiation in the treatment <strong>of</strong> breast cancer. N Engl J Med.<br />
1989;320:822-828.<br />
7. Jacobson JA, Danforth DN, Cowan KH, et al. Ten-year results <strong>of</strong> a<br />
comparison <strong>of</strong> conservation with mastectomy in the treatment <strong>of</strong> stage I and<br />
II breast cancer. N Engl J Med. 1995;332:907-911.<br />
8. van Dongen JA, Bartelink H, Fentiman IS, et al. Factors influencing<br />
local relapse and survival and results <strong>of</strong> salvage treatment after breastconserving<br />
therapy in operable breast cancer: EORTC trial 10801, breast<br />
conservation compared with mastectomy in TNM stage I and II breast cancer.<br />
Eur J Cancer. 1992;28A:801-805.<br />
9. Veronesi U, Cascinelli N, Mariani L, et al. Twenty-year follow-up <strong>of</strong> a<br />
randomized study comparing breast-conserving surgery with radical mastectomy<br />
for early breast cancer. N Engl J Med. 2002;347:1227-1232.<br />
10. Veronesi U, Luini A, Del Vecchio M, et al. Radiotherapy after breastpreserving<br />
surgery in women with localized cancer <strong>of</strong> the breast. N Engl<br />
J Med. 1993;328:1587-1591.<br />
11. Kurtz JM, Amalric R, Brandone H, et al. Local recurrence after<br />
breast-conserving surgery and radiotherapy. Frequency, time course, and<br />
prognosis. Cancer. 1989;63:1912-1917.<br />
12. Recht A, Silen W, Schnitt SJ, et al. Time-course <strong>of</strong> local recurrence<br />
following conservative surgery and radiotherapy for early stage breast cancer.<br />
Int J Radiat Oncol Biol Phys. 1988;15:255-261.<br />
13. Fowble B, Solin LJ, Schultz DJ, et al. Breast recurrence following<br />
conservative surgery and radiation: patterns <strong>of</strong> failure, prognosis, and pathologic<br />
findings from mastectomy specimens with implications for treatment.<br />
Int J Radiat Oncol Biol Phys. 1990;19:833-842.<br />
14. Joseph E, Hyacinthe M, Lyman GH, et al. Evaluation <strong>of</strong> an intensive<br />
strategy for follow-up and surveillance <strong>of</strong> primary breast cancer. Ann Surg<br />
Oncol. 1998;5:522-528.<br />
15. Orel SG, Troupin RH, Patterson EA, et al. Breast cancer recurrence<br />
after lumpectomy and irradiation: role <strong>of</strong> mammography in detection. Radiology.<br />
1992;183:201-206.<br />
16. Lu WL, Jansen L, Post WJ, et al. Impact on survival <strong>of</strong> early detection<br />
<strong>of</strong> isolated breast recurrences after the primary treatment for breast cancer:<br />
a meta-analysis. Breast Cancer Res Treat. 2009;114:403-412.<br />
17. Robertson C, Ragupathy SK, Boachie C, et al. Surveillance mammography<br />
for detecting ipsilateral breast tumour recurrence and metachronous<br />
64<br />
FLOWERS, MOONEY, AND DRUKTEINIS<br />
contralateral breast cancer: a systematic review. Eur Radiol. 2011;21:2484-<br />
2491.<br />
18. Brennan S, Liberman L, Dershaw DD, et al. Breast MRI screening <strong>of</strong><br />
women with a personal history <strong>of</strong> breast cancer. AJR Am J Roentgenol.<br />
2010;195:510-516.<br />
19. Viehweg P, Rotter K, Laniado M, et al. MR imaging <strong>of</strong> the contralateral<br />
breast in patients after breast-conserving therapy. Eur Radiol. 2004;14:402-<br />
408.<br />
20. Nunes LW, Schnall MD, Orel SG. Update <strong>of</strong> breast MR imaging<br />
architectural interpretation model. Radiology. 2001;219:484-494.<br />
21. Schnall MD, Blume J, Bluemke DA, et al. Diagnostic architectural and<br />
dynamic features at breast MR imaging: multicenter study. Radiology.<br />
2006;238:42-53.<br />
22. <strong>American</strong> College <strong>of</strong> Radiology. ACR Practice Guideline for the Performance<br />
<strong>of</strong> Contrast Enhanced Magnetic Resonance Imaging (MRI) <strong>of</strong> the<br />
Breast (revision). 2008.<br />
23. Saslow D, Boetes C, Burke W, et al. <strong>American</strong> Cancer <strong>Society</strong> guidelines<br />
for breast screening with MRI as an adjunct to mammography. CA<br />
Cancer J Clin. 2007;57:75-89.<br />
24. Heywang-Kobrunner SH, Schlegel A, Beck R, et al. Contrast-enhanced<br />
MRI <strong>of</strong> the breast after limited surgery and radiation therapy. J Comput<br />
Assist Tomogr. 1993;17:891-900.<br />
25. Li J, Dershaw DD, Lee CH, Joo S, Morris EA. Breast MRI after<br />
conservation therapy: usual findings in routine follow-up examinations. AJR<br />
Am J Roentgenol. 2010;195:799-807.<br />
26. Drukteinis JS, Gombos EC, Raza S, et al. MR imaging assessment <strong>of</strong><br />
the breast after breast conservation therapy: distinguishing benign from<br />
malignant lesions. Radiographics. <strong>2012</strong>;32:219-234.<br />
27. Punglia RS, Hassett MJ. Using lifetime risk estimates to recommend<br />
magnetic resonance imaging screening for breast cancer survivors. J Clin<br />
Oncol. 2010;28:4108-4110.<br />
28. Eidelman Y, Liebling RW, Buchbinder S, et al. Mammography in the<br />
evaluation <strong>of</strong> masses in breasts reconstructed with TRAM flaps. Ann Plast<br />
Surg. 1998;41:229-233.<br />
29. Helvie MA, Bailey JE, Roubidoux MA, et al. Mammographic screening<br />
<strong>of</strong> TRAM flap breast reconstructions for detection <strong>of</strong> nonpalpable recurrent<br />
cancer. Radiology. 2002;224:211-216.<br />
30. Devon RK, Rosen MA, Mies C, et al. Breast reconstruction with a<br />
transverse rectus abdominis myocutaneous flap: spectrum <strong>of</strong> normal and<br />
abnormal MR imaging findings. Radiographics. 2004;24:1287-1299.<br />
31. Lee JM, Georgian-Smith D, Gazelle GS, et al. Detecting nonpalpable<br />
recurrent breast cancer: the role <strong>of</strong> routine mammographic screening <strong>of</strong><br />
transverse rectus abdominis myocutaneous flap reconstructions. Radiology.<br />
2008;248:398-405.
Advanced Imaging Techniques for the<br />
Detection <strong>of</strong> Breast Cancer<br />
Overview: Mammography is the only breast imaging examination<br />
that has been shown to reduce breast cancer mortality.<br />
Population-based sensitivity is 75% to 80%, but sensitivity in<br />
high-risk women with dense breasts is only in the range <strong>of</strong><br />
50%. Breast ultrasound and contrast-enhanced breast magnetic<br />
resonance imaging (MRI) have become additional standard<br />
modalities used in the diagnosis <strong>of</strong> breast cancer. In<br />
high-risk women, ultrasound is known to detect approximately<br />
four additional cancers per 1,000 women. MRI is exquisitely<br />
sensitive for the detection <strong>of</strong> breast cancer. In high-risk<br />
women, it finds an additional four to five cancers per 100<br />
women. However, both ultrasound and MRI are also known to<br />
lead to a large number <strong>of</strong> additional benign biopsies and<br />
MAMMOGRAPHY IS the only breast imaging examination<br />
shown to reduce breast cancer mortality, with<br />
a population-based sensitivity <strong>of</strong> 75% to 80%. Current<br />
screening guidelines for normal-risk women (�15% lifetime<br />
risk) recommend starting screening at 40 and continuing<br />
until life expectancy is less than 5 years. Women at higher<br />
risks (intermediate 15% to 20% and high �20%) begin<br />
screening at an earlier age, generally 10 years earlier than<br />
the youngest family member presented with her breast<br />
cancer. Sensitivity <strong>of</strong> mammography in high-risk women<br />
with dense breasts is only in the range <strong>of</strong> 50%. Both<br />
ultrasound and MRI are standard examinations used to<br />
improve on mammographic sensitivity; each has its own<br />
flaws. New technologies have been and continue to be<br />
developed to improve the breast cancer detection rate. These<br />
include improvements on the standard techniques (mammography,<br />
ultrasound, and MRI) as well as new platforms<br />
for breast imaging developed from body imaging tools, which<br />
include CT and radionuclide breast imaging. This article<br />
will review the most salient newer breast imaging technologies,<br />
as well as their potential roles.<br />
Mammography and Mammography-Based Techniques<br />
Digital mammography (FFDM) has almost entirely replaced<br />
analog (film-screen) mammography. It has been<br />
shown to be more sensitive in women younger than 50 with<br />
dense breast tissue. 1 However, overall sensitivity remains<br />
the same as with analog. The advent <strong>of</strong> digital mammography<br />
has enabled more advanced techniques, which are<br />
added to the digital platform. These include digital breast<br />
tomosynthesis (DBT) and contrast-enhanced spectral mammography<br />
(CESM).<br />
Tomosynthesis<br />
In 2011, DBT was called the most exciting new technology<br />
in breast imaging. It is U.S. Food and Drug Administration<br />
(FDA)-approved for performance with a screening digital<br />
mammogram. In 1997, it was stated that DBT allowed “the<br />
radiologist to see through the structured noise <strong>of</strong> normal<br />
breast tissue to improve the detection and characterization<br />
<strong>of</strong> early-stage breast cancer.” 2 By seeing through this<br />
noise, additional lesions may be detected, and better margin<br />
analysis can be performed. Additionally, the radiologist can<br />
By Maxine Jochelson, MD<br />
short-term follow-up examinations. Many new breast imaging<br />
tools have improved and are being developed to improve on<br />
our current ability to diagnose early-stage breast cancer.<br />
These can be divided into two groups. The first group is those<br />
that are advances in current techniques, which include digital<br />
breast tomosynthesis and contrast-enhanced mammography<br />
and ultrasound with elastography or microbubbles. The other<br />
group includes new breast imaging platforms such as breast<br />
computed tomography (CT) scanning and radionuclide breast<br />
imaging. These are exciting advances. However, in this era <strong>of</strong><br />
cost and radiation containment, it is imperative to look at all <strong>of</strong><br />
them objectively to see which will provide clinically relevant<br />
additional information.<br />
determine that a “mass” was merely a confluence <strong>of</strong> shadows.<br />
Specificity is, therefore, theoretically enhanced. To date<br />
no large prospective trials have been reported regarding the<br />
ability <strong>of</strong> DBT to screen for cancer as a solo imaging tool.<br />
What has been well documented is its ability to reduce the<br />
need for callbacks. 3,4 More recently, DBT images were<br />
compared to routine spot films and found to show equivalent<br />
mass characterization. DBT found seven additional cancers<br />
along with five additional false-positive findings. 5 The detection<br />
<strong>of</strong> clustered microcalcifications has been more problematic.<br />
Initially, detection was inferior with DBT because<br />
<strong>of</strong> the very thin (1 mm) slices and the blurring inherent to<br />
tomosynthesis. Spangler and colleagues showed superior<br />
sensitivity (84% vs. 75%) and specificity (71% vs. 64%) with<br />
FFDM, which also detected more cancers with calcifications<br />
but the differences were not significant. 6 More recently,<br />
Destounis and colleagues compared the two techniques in<br />
103 patients and found them equivalent. 7 Gur and colleagues<br />
compared FFDM alone to FFDM combined with<br />
DBT in 125 patients in a retrospective evaluation. There<br />
were both more true-positives and true-negatives with DBT<br />
with an overall 16% performance improvement (p � 0.01). 8<br />
There are several limitations: 1) Radiation dose is twice<br />
that <strong>of</strong> FFDM. The thought is that the reduction in callbacks<br />
will reduce the effect <strong>of</strong> that additional radiation, as spot<br />
films generate a higher dose than DBT. However, most<br />
women who are screened do not get called back for additional<br />
views. That being said, the dose does fall within the<br />
MQSA guidelines; 2) DBT only evaluates anatomy, while<br />
many <strong>of</strong> the other emerging technologies also evaluate<br />
physiology; 3) the interpretation <strong>of</strong> DBT takes much longer<br />
than that <strong>of</strong> FFDM, limiting throughput; 4) reimbursement<br />
is limited or absent.<br />
This is a promising technology without a defined role<br />
(Fig. 1). Possibilities include routine screening for everyone,<br />
high-risk screening, and as a diagnostic tool after abnormal<br />
From the Memorial Sloan-Kettering Cancer Center, New York, NY.<br />
Author’s disclosure <strong>of</strong> potential conflicts <strong>of</strong> interest are found at the end <strong>of</strong> this article.<br />
Address reprint requests to Maxine Jochelson, MD, Memorial Sloan-Kettering Cancer<br />
Center, 300 E. 66 th St., New York, NY 10065; email:jochelsm@mskcc.org.<br />
© <strong>2012</strong> by <strong>American</strong> <strong>Society</strong> <strong>of</strong> <strong>Clinical</strong> <strong>Oncology</strong>.<br />
1092-9118/10/1-10<br />
65
Fig. 1. A clip from the newspaper regarding 3-day as a new<br />
technology for screening.<br />
screening. There are multiple ongoing studies to better<br />
define its strengths and weaknesses and to determine its<br />
ideal role.<br />
Contrast-Enhanced Mammography<br />
Contrast-enhanced breast MRI functions by both defining<br />
anatomic abnormalities and evaluating physiologic changes<br />
related to the development <strong>of</strong> neovascularity, which are<br />
visualized by their rapid contrast enhancement in cancers.<br />
MRI has proven to be exquisitely sensitive for the detection<br />
<strong>of</strong> breast cancer, with a sensitivity range from 79% to<br />
98%. 9,10 A meta-analysis has shown that MRI also detects<br />
mammographically occult multicentric or multifocal disease<br />
in approximately 16% <strong>of</strong> patients with known breast cancer.<br />
9<br />
Although breast MRI is extremely sensitive, specificity is<br />
more limited, leading to additional workups and benign<br />
biopsies. Additionally, it is expensive, and good quality<br />
66<br />
KEY POINTS<br />
● Mammography is the only breast imaging exam<br />
known to reduce breast cancer mortality, but sensitivity<br />
is limited particularly in high-risk women with<br />
dense breast tissue.<br />
● The digital mammography platform has allowed for<br />
developments such as tomosynthesis and contrastenhanced<br />
mammography, which have the potential to<br />
improve sensitivity.<br />
● Ultrasound detects additional cancers but at the cost<br />
<strong>of</strong> many benign biopsies. Potential advances in ultrasound<br />
technology include automated whole-breast<br />
scanning, elastography, and microbubbles.<br />
● Breast MRI remains the most sensitive examination<br />
for the detection <strong>of</strong> early-stage breast cancer in a<br />
high-risk screening population.<br />
● Radionuclide breast imaging is able to detect breast<br />
cancers independent <strong>of</strong> breast density, but its high<br />
extramammary radiation dose precludes use in<br />
screening.<br />
MAXINE JOCHELSON<br />
breast MRI is not universally available. Patients with pacemakers,<br />
certain aneurysm clips or other metallic hardware,<br />
and severe claustrophobia are unable to undergo MRI.<br />
Contrast-enhanced mammography has been developed to<br />
potentially duplicate the capacity <strong>of</strong> MRI to detect and stage<br />
breast cancer with the use <strong>of</strong> both anatomy and physiology.<br />
It has been investigated both as an adjunct to mammography<br />
and an alternative to MRI. 11,12 It uses the same iodinated<br />
contrast used for CT in the same doses.<br />
Hardware and s<strong>of</strong>tware adaptations to digital mammography<br />
units that automate a dual energy technique have<br />
been developed. This technology, known as Contrast-<br />
Enhanced Spectral Mammography (CESM), received FDA<br />
approval in 2011. It provides two images in each view: a<br />
low-energy image, which is below the K-edge <strong>of</strong> iodine<br />
(33.2keV), and a high-energy image, just above. The two<br />
images are combined and processed so that the background<br />
breast tissue is essentially subtracted out, maximizing<br />
the ability to see areas <strong>of</strong> enhancement (Fig. 2). There is a<br />
20% additional radiation dose, which is the equivalent <strong>of</strong><br />
one extra mammographic view. The examination takes approximately<br />
10 minutes to perform and is only slightly<br />
longer than a mammogram to read. It is well tolerated by<br />
patients.<br />
Dromain and colleagues recently reported a comparison<br />
<strong>of</strong> contrast-enhanced digital mammography (CEDM) plus<br />
mammography with mammography alone and mammography<br />
plus ultrasound in 142 lesions in 120 patients. Sensitivity<br />
for CEDM plus mammography was 93% compared with<br />
78% (p � 0.001) for mammography alone. Specificity was<br />
unchanged. There was improvement in sensitivity and specificity<br />
between CEDM plus mammography and ultrasound<br />
plus mammography, but it was not significantly better.<br />
Additionally, all 23 multifocal lesions were detected. 13<br />
Jochelson reported early results <strong>of</strong> a comparison <strong>of</strong> CEDM<br />
with breast MRI in 26 patients with known breast cancer.<br />
The two examinations each detected 25 <strong>of</strong> 26 index tumors<br />
compared with the 22 detected by mammography alone.<br />
MRI was more sensitive than CEDM for the detection <strong>of</strong><br />
additional lesions (7 vs. 5), but there were no false-positive<br />
CEDM examinations and seven false-positive MRIs. 14 Results<br />
on additional patients will be reported later this year.<br />
The early data for CESM are promising but require<br />
confirmation. Screening studies need to be performed. If<br />
larger studies confirm early results, CESM could become<br />
available as a possible “poor woman’s” MRI. Larger multiinstitutional<br />
trials are ongoing.<br />
Ultrasound<br />
Targeted breast ultrasound is <strong>of</strong> value in further defining<br />
mammographic and MRI abnormalities as well as evaluating<br />
palpable lesions. Ultrasound-guided core biopsies are<br />
an effective means <strong>of</strong> obtaining histologic diagnosis. Ultrasound<br />
is inexpensive, utilizes no radiation, and is widely<br />
available.<br />
The use <strong>of</strong> ultrasound for screening is more problematic.<br />
Multiple studies, including a large ACRIN trial evaluating<br />
the utility <strong>of</strong> ultrasound screening in high-risk patients,<br />
have shown that ultrasound will detect additional cancers<br />
in three to four <strong>of</strong> 1,000 patients (as compared with MRI,<br />
which detects additional cancers in four to five <strong>of</strong> 100<br />
patients). 15,16 In doing so, a large number <strong>of</strong> additional<br />
lesions are detected leading to short-interval follow-ups or
IMAGING TECHNIQUES FOR BREAST CANCER<br />
Fig. 2. Three images <strong>of</strong> the left breast<br />
from left to right: normal left mammogram;<br />
contrast-enhanced mammography<br />
showing extensive enhancement<br />
caused by multifocal breast cancer in<br />
the lower half <strong>of</strong> the breast; breast<br />
MRI also showing multifocal cancer.<br />
Contrast-enhanced mammogram corresponds<br />
well to the MRI.<br />
biopsies with a positive predictive value (PPV) <strong>of</strong> only 9%<br />
to 10%. Scans are operator-dependent and, therefore, less<br />
reproducible than other modalities. A lesion not detected is<br />
not documented.<br />
To eradicate operator dependence, automated whole<br />
breast ultrasound (AWBU) machines have been developed.<br />
Images are reproducible, and there is 3D capability. Additionally,<br />
studies need not be interpreted in real time. Kelly<br />
prospectively compared mammography alone to mammography<br />
plus AWBU in 4,419 women and showed an improved<br />
diagnostic yield <strong>of</strong> 7.2 <strong>of</strong> 1,000 compared with 3.6 <strong>of</strong> 1,000<br />
and equivalent PPV <strong>of</strong> 38%. 17 Shin found a 92% detection<br />
rate for lesions larger than 1.2 cm. 18<br />
Other advances have been geared toward improving the<br />
sensitivity, specificity, and PPV. Color Doppler detects tumor<br />
vascularity. This improves specificity <strong>of</strong> B-mode ultrasound.<br />
Microbubbles have been injected as contrast agents<br />
to improve the performance <strong>of</strong> Doppler. Although this can<br />
improve sensitivity, specificity has been more problematic<br />
because <strong>of</strong> its ability to detect smaller and not necessarily<br />
malignant vessels. Liu has shown that there is good histologic<br />
correlation when using this contrast agent. Unfortunately,<br />
this technique is labor intensive and a technically<br />
difficult procedure to perform. It is unlikely to be useful in<br />
every day practice.<br />
Elastography is a tool that has been developed to improve<br />
the specificity <strong>of</strong> breast ultrasound. Because most breast<br />
cancers are stiff, there is theoretically less displacement<br />
on compression. There are two types <strong>of</strong> elastography. There<br />
is more experience with static elastography, in which the<br />
operator compresses the lesion with the special probe and<br />
measures displacement. Unfortunately, this is also operator<br />
dependent, and the results have been disappointing. 19 Shear<br />
strain elastography is being studied. The waves are emitted<br />
perpendicular to the transducer, and the speed <strong>of</strong> their<br />
traversing the tissue is measured with faster speeds occurring<br />
in hard tissue. This is less operator dependent, and<br />
early results indicate that this may be the better way to<br />
measure elasticity.<br />
Although these are interesting attempts at improving<br />
screening ultrasound, a great deal more work is necessary.<br />
Radionuclide Breast Imaging<br />
Radionuclide breast imaging evaluates physiology. Sestamibi<br />
and 18-FDG detect cancers by different mechanisms <strong>of</strong><br />
action. Their ability to detect lesions in the breast is independent<br />
<strong>of</strong> breast density, and with 18-FDG, it is independent<br />
<strong>of</strong> hormone status.<br />
MIBI or Gamma Imaging<br />
In the early 1990s, incidental breast cancers were seen in<br />
patients having sestamibi cardiac scans. At that time, attempts<br />
at dedicated breast imaging had limited sensitivity<br />
for small lesions because <strong>of</strong> the large collimators that were<br />
distant from the breasts. Since then, new technology has<br />
become available, wherein high resolution detectors are<br />
used to mildly compress the breasts, which are positioned as<br />
they are with mammography. There are two different systems:<br />
molecular breast imaging (MBI) based on a semiconductor<br />
base and breast-specific gamma imaging (BSGI),<br />
which uses a scintillating crystal detector. The results are<br />
similar and will be discussed together. Patients are ideally<br />
done between days 2 and 14 <strong>of</strong> the menstrual cycle. They<br />
receive 740–1,110 MBq. There is a 1-minute wait. Images<br />
are then acquired for approximately 10 minutes per view.<br />
Two views <strong>of</strong> each breast are provided. Studies show sensitivities<br />
<strong>of</strong> 91% to 96% overall but somewhat lower for<br />
smaller lesions: 69% for lesions less than 5 mm in one<br />
study 20 and 89% for sub-centimeter lesions. 21 That study<br />
also had a specificity <strong>of</strong> only 60%—no improvement over<br />
MRI. Detection <strong>of</strong> additional foci <strong>of</strong> cancer on staging exams<br />
is 9% to 10%, but in doing so there are a large number <strong>of</strong><br />
false positives. For example, Brem reported on 159 patients<br />
67
in whom there were 46 additional lesions, only 14 <strong>of</strong> which<br />
were malignant. 22<br />
Positron Emission Mammography (PEM)<br />
The sensitivity <strong>of</strong> whole-body 18-FDG PET for detection <strong>of</strong><br />
a primary breast cancer is low—approximately 40%. PEM<br />
was developed to better evaluate the breasts. The scanner<br />
again resembles a mammography machine. Twelve images<br />
are generated for each view <strong>of</strong> the breast, providing a sort <strong>of</strong><br />
3D image. PEM is performed like whole-body PET: 4–6<br />
hours <strong>of</strong> fasting, injection <strong>of</strong> about 10 mCi <strong>of</strong> FDG, followed<br />
by a 1-hour rest period. Each view takes 10 minutes. PEM<br />
can also be performed after a whole-body exam without<br />
additional tracer. Multiple studies have shown sensitivities<br />
<strong>of</strong> more than 90% (including DCIS) even for smaller lesions.<br />
23 Specificity is also reported to be more than 90%. In<br />
large trials comparing PEM to MRI in patients with known<br />
breast cancer, MRI was more sensitive on a lesion level in<br />
the ipsilateral breast, more sensitive in the contralateral<br />
breast, but less specific. 24 PEM can be performed with<br />
different tracers.<br />
Radionuclide imaging is less expensive than MRI, and the<br />
scanners themselves need less space. The limiting factor for<br />
their use in routine screening is the high dose <strong>of</strong> radiation,<br />
particularly to extramammary tissues: 50mGy to lower<br />
large intestine for sestamibi and 59mGy to bladder was for<br />
PEM. Vendors <strong>of</strong> both technologies are reporting promising<br />
preliminary results using half the dose, but even those doses<br />
may be too high to be used in yearly screening. Other<br />
potential uses include preoperative staging, problem solv-<br />
Author’s Disclosure <strong>of</strong> Potential Conflicts <strong>of</strong> Interest<br />
Author<br />
Maxine Jochelson*<br />
*No relevant relationships to disclose.<br />
Employment or<br />
Leadership<br />
Positions<br />
Consultant or<br />
Advisory Role<br />
1. Pisano ED, Gatsonis C, Hendrick E, et al. Diagnostic performance <strong>of</strong><br />
digital versus film mammography for breast-cancer screening. N Engl J Med.<br />
2005;353:1773-1783.<br />
2. Niklason LT, Christian BT, Niklason LE, et al. Digital tomosynthesis in<br />
breast imaging. Radiology. 1997;205:399-406.<br />
3. Kopans D, Moore R. Digital breast tomosynthesis (DBT) NCI 3000women<br />
trial. RSNA. 2009.<br />
4. Poplack SP, Tosteson TD, Kogel CA, Nagy HM. Digital breast tomosynthesis:<br />
Initial experience in 98 women with abnormal digital screening<br />
mammography. AJR Am J Roentgenol. 2007;189:616-623.<br />
5. Noroozian M, Hadjiiski L, Rahnama-Moghadam S, et al. Digital breast<br />
tomosynthesis is comparable to mammographic spot views for mass characterization.<br />
Radiology. <strong>2012</strong>;262:61-68.<br />
6. Spangler ML, Zuley ML, Sumkin JH, et al. Detection and classification<br />
<strong>of</strong> calcifications on digital breast tomosynthesis and 2D digital mammography:<br />
A comparison. AJR Am J Roentgenol. 2011;196:320-324.<br />
7. Destounis S, Murphy P, Seifert P, et al. <strong>Clinical</strong> experience with digital<br />
breast tomosynthesis in the characterization and visualization <strong>of</strong> breast<br />
microcalcifications. Amer J Radiol. 2011;196:A1-A3.<br />
8. Gur D, Bandos AI, Rockette HE, et al. Localized detection and classification<br />
<strong>of</strong> abnormalities on FFDM and tomosynthesis examinations rated<br />
under an FROC paradigm. AJR Am J Roentgenol. 2011;196:737-741.<br />
9. Morris EA, Liberman L, Ballon DJ, et al. MRI <strong>of</strong> occult breast carcinoma<br />
in a high-risk population. AJR Am J Roentgenol. 2003;181:619-626.<br />
10. Berg WA. Rationale for a trial <strong>of</strong> screening breast ultrasound: <strong>American</strong><br />
College <strong>of</strong> Radiology Imaging Network (ACRIN) 6666. AJR Am J<br />
Roentgenol. 2003;180:1225-1228.<br />
11. Dromain C, Thibault F, Adler G. Dual energy contrast-enhanced digital<br />
68<br />
ing, and assessment <strong>of</strong> treatment response in patients receiving<br />
neoadjuvant chemotherapy.<br />
MRI and Conclusions<br />
Breast MRI is the gold standard breast imaging test<br />
against which all other new technologies are compared,<br />
particularly in terms <strong>of</strong> sensitivity, which is more than 95%.<br />
The absence <strong>of</strong> radiation is a major asset. Initial inability to<br />
diagnose DCIS and low specificity has been limiting. Both<br />
have improved over the years, but the perception <strong>of</strong> poor<br />
specificity remains. MRI is clearly the ideal way to screen<br />
women at high risk for breast cancer, but is too expensive to<br />
be used in larger populations, and good quality MRI is not<br />
universally available. As newer technologies become available,<br />
however, there are many lessons from the initial<br />
experience with breast MRI that should be applied: 1) not all<br />
MRI scanners/scans are equal, and although most <strong>of</strong> the<br />
technologies discussed here are easier to perform, equipment<br />
performance standards should be defined. 2) There is<br />
always a learning curve in the interpretation <strong>of</strong> new technologies.<br />
Minimal training requirements should be set. 3) A<br />
vocabulary with standard terminology to describe abnormalities<br />
and what they mean (such as the BIRADS system)<br />
should be developed. 4) In this era <strong>of</strong> needed cost and<br />
radiation containment, we cannot do every test on every<br />
woman. It is imperative that in addition to determining<br />
sensitivity, specificity, and other factors, we also learn in<br />
which situations these new technologies can do the most<br />
good.<br />
Stock<br />
Ownership Honoraria<br />
REFERENCES<br />
Research<br />
Funding<br />
Expert<br />
Testimony<br />
MAXINE JOCHELSON<br />
Other<br />
Remuneration<br />
mammography: Preliminary <strong>Clinical</strong> Results 94th Scientific Assembly <strong>of</strong> the<br />
RSNA 2007 Annual Meeting Program. 2007;326.<br />
12. Diekmann F, Marx C, Jong R, et al. Diagnostic accuracy <strong>of</strong> contrastenhanced<br />
digital mammography as an adjunct to mammography. ECR. 2007.<br />
13. Dromain C, Thibault F, Muller S, et al. Dual-energy contrast-enhanced<br />
digital mammography: Initial clinical results. Eur Radiol. 2011;21:565-574.<br />
14. Jochelson M, Sung J, Dershaw DD, Heerdt A. Bilateral-Dual Energy<br />
Contrast-enhanced Digital Mammography: Initial Experience. Presented at<br />
96 th Annual RSNA Scientific Assembly and Annual Meeting; November 2010;<br />
Chicago, IL.<br />
15. Berg WA, Blume JD, Cormack JB, et al. Combined screening with<br />
ultrasound and mammography vs mammography alone in women at elevated<br />
risk <strong>of</strong> breast cancer. JAMA. 2008;299:2151-2163.<br />
16. Kolb TM, Lichy J, Newhouse JH. Occult cancer in women with dense<br />
breasts: Detection with screening US-diagnostic yield and tumor characteristics.<br />
Radiology. 1998;207:191-199.<br />
17. Kelly KM, Dean J, Comulada WS, Lee SJ. Breast cancer detection<br />
using automated whole breast ultrasound and mammography in radiographically<br />
dense breasts. Eur Radiol. 2010;20:734-742.<br />
18. Shin HJ, Kim HH, Cha JH, et al. Automated ultrasound <strong>of</strong> the breast<br />
for diagnosis: Interobserver agreement on lesion detection and characterization.<br />
AJR Am J Roentgenol. 2011;197:747-754.<br />
19. Kohr J, Sung, J, Malak, S, et al. Impact <strong>of</strong> elastography on assessing<br />
the likelihood <strong>of</strong> malignancy <strong>of</strong> ultrasound lesions. Poster presented at RSNA<br />
97th Annual Meeting; 2011; Chicago, IL.<br />
20. Hruska CB, Boughey JC, Phillips SW, et al. Scientific Impact Recognition<br />
Award: Molecular breast imaging: A review <strong>of</strong> the Mayo Clinic experience.<br />
Am J Surg. 2008;196:470-476.
IMAGING TECHNIQUES FOR BREAST CANCER<br />
21. Brem RF, Floerke AC, Rapelyea JA, et al. Breast-specific gamma<br />
imaging as an adjunct imaging modality for the diagnosis <strong>of</strong> breast cancer.<br />
Radiology. 2008;247:651-657.<br />
22. Brem RF, Shahan C, Rapleyea JA, et al. Detection <strong>of</strong> occult foci <strong>of</strong><br />
breast cancer using breast-specific gamma imaging in women with one<br />
mammographic or clinically suspicious breast lesion. Acad Radiol. 2010;17:<br />
735-743.<br />
23. Berg WA, Weinberg IN, Narayanan D, et al. High-resolution fluorodeoxyglucose<br />
positron emission tomography with compression (“positron emission<br />
mammography”) is highly accurate in depicting primary breast cancer.<br />
Breast J. 2006;12:309-323.<br />
24. Berg WA, Madsen KS, Schilling K, et al. Breast cancer: Comparative<br />
effectiveness <strong>of</strong> positron emission mammography and MR imaging in presurgical<br />
planning for the ipsilateral breast. Radiology. 2011;258:59-72.<br />
69
UPDATE OF THE OXFORD OVERVIEW: NEW<br />
INSIGHTS AND PERSPECTIVES IN THE ERA OF<br />
PERSONALIZED MEDICINE<br />
CHAIR<br />
Harold J. Burstein, MD, PhD<br />
Dana-Farber Cancer Institute<br />
Boston, MA<br />
SPEAKERS<br />
Jonas Bergh, MD, PhD<br />
Karolinska Institutet, University Hospital<br />
Stockholm, Sweden<br />
Kathleen I. Pritchard, MD<br />
Sunnybrook Health Sciences Center<br />
Toronto, ON, Canada
Update <strong>of</strong> the Oxford Overview: New<br />
Insight and Perspectives in the Era <strong>of</strong><br />
Personalized Medicine<br />
By Kathleen I. Pritchard, MD, Jonas Bergh, MD, PhD, and Harold J. Burstein MD, PhD<br />
Overview: There is great appreciation for the heterogeneity<br />
<strong>of</strong> breast cancers, particularly <strong>of</strong> hormone-receptor positive<br />
breast cancers. A goal <strong>of</strong> modern oncology managing such<br />
heterogeneity is to determine how to individualize therapy<br />
based on the specific pathological and biological features <strong>of</strong> a<br />
given tumor. Two distinctive clinical literatures exist to guide<br />
treatment <strong>of</strong> hormone-receptor-positive breast cancer. The<br />
Oxford Overview, a seminal meta-analysis effort, has recently<br />
been updated, and suggests that nearly all patients with<br />
ER-positive tumors benefit from adjuvant endocrine therapy.<br />
THE OXFORD Overview <strong>of</strong> the Early Breast Cancer<br />
Trialists’ Collaborative Group (EBCTCG) dates back to<br />
1984 when investigators responsible for trials <strong>of</strong> systemic<br />
adjuvant systemic therapy from all around the globe met<br />
initially, not in Oxford, but at Heathrow Airport to examine<br />
the first meta-analysis <strong>of</strong> adjuvant systemic therapy trials.<br />
The Overview has always involved collaboration between<br />
the Oxford “Secretariat” led by Sir Richard Peto and a<br />
consortium <strong>of</strong> investigators. A series <strong>of</strong> distinguished past<br />
chairs have included I. Craig Henderson, MD, a prominent<br />
breast cancer medical oncologist initially from Harvard<br />
University and later the University <strong>of</strong> California San Francisco<br />
Helen Diller Family Comprehensive Center, and<br />
William C. Wood, MD, then Chief <strong>of</strong> Surgery at Emory<br />
University. The EBCTCG is currently chaired by Kathy<br />
Pritchard, MD, <strong>of</strong> Toronto and Martine Piccart, MD, PhD,<br />
<strong>of</strong> Brussels. The Steering Committee and Executive <strong>of</strong> the<br />
EBCTCG—both <strong>of</strong> which include members <strong>of</strong> the Oxford<br />
Secretariat and clinical investigators (the Trialists)—extensively<br />
meet, teleconference, and e-mail to bring analyses<br />
and publications to fruition. Between 2005 and 2011,<br />
data collection and analyses have resulted in five major<br />
publications. 1-5<br />
The Overview concept dating back to 1984 has not<br />
changed. Collaboration between physicians and biostatisticians<br />
based in Oxford and around the world was sought,<br />
built, and sustained. Data were initially collected from all<br />
randomized trials <strong>of</strong> systemic adjuvant and, later, localregional<br />
therapy. The Overview methodology involves the<br />
collection <strong>of</strong> individual patient data, which includes a wide<br />
variety <strong>of</strong> items such as dates <strong>of</strong> randomization and treatment<br />
allocation. Patients can be stratified by age, node<br />
status, and other criteria and the log-rank statistics from<br />
each trial are combined to give an overall estimate <strong>of</strong> the<br />
effect <strong>of</strong> different treatments either in the whole patient<br />
population or in various stratified subsets.<br />
Outcomes include recurrence, which can be adjusted to<br />
include or exclude contralateral breast cancers and deaths.<br />
Deaths from unknown causes are usually included with<br />
deaths from breast cancer unless specifically stated otherwise.<br />
Recurrences can be divided into local and/or distant<br />
with and without contralateral breast cancers. The<br />
EBCTCG has also been interested in collecting information<br />
on deaths from cardiac events, stroke, and other cancers.<br />
In addition, the overview finds that nearly all subsets <strong>of</strong><br />
patients with ER-positive tumors also benefit from modern<br />
adjuvant chemotherapy regimens. Meanwhile, retrospective<br />
subset analyses <strong>of</strong> specific trials or populations suggests that<br />
the benefits <strong>of</strong> chemotherapy are not so uniform, and in<br />
particular that molecular diagnostics assays can identify patients<br />
who do not warrant chemotherapy. This article will<br />
highlight recent data and controversies in personalizing adjuvant<br />
breast cancer therapy.<br />
In 1984, the Oxford Overview showed unequivocally for<br />
the first time that tamoxifen improved survival and that<br />
cyclophosphamide, methotrexate, and 5-fluorouracil (CMF)<br />
chemotherapy improved survival. It was also shown that<br />
ovarian ablation, which had been mainly tested in small<br />
underpowered trials with nonsignificant results, did, in and<br />
<strong>of</strong> itself, improve overall survival, particularly in women<br />
whose tumors had positive estrogen receptors (ER). By 1990,<br />
it became clear that 5 years <strong>of</strong> tamoxifen was better than 1<br />
or 2 years and that tamoxifen effects were greater in women<br />
with ER-positive cancers. It was first shown in 1990 that<br />
tamoxifen reduced the rate <strong>of</strong> contralateral breast cancer<br />
and that chemotherapy was effective in both older and<br />
younger women.<br />
By 1995, the huge effect <strong>of</strong> 5 years <strong>of</strong> tamoxifen was<br />
clearly demonstrated, and it was obvious from both direct<br />
and indirect comparisons within the Overview that 5 years<br />
<strong>of</strong> tamoxifen was superior to shorter durations <strong>of</strong> treatment.<br />
It was seen for the first time that tamoxifen prevented<br />
contralateral breast cancers only in women whose initial<br />
tumors were ER positive. That year, the Overview also<br />
demonstrated that anthracycline-containing regimens, at<br />
least when given in higher dosages, were better than CMFtype<br />
chemotherapy.<br />
By 2000, the Overview was able to report on long-term<br />
results, such as 15-year outcomes with chemotherapy, demonstrating<br />
sustained benefit in older and in younger women.<br />
There was great controversy at this time suggesting that,<br />
particularly in the CMF trials, the effects <strong>of</strong> chemotherapy<br />
might be greater in women with ER-negative tumors than in<br />
those with ER-positive tumors, a controversy which continues<br />
to this day. In 2000, it was also clearly seen that the<br />
15-year effects <strong>of</strong> 5 years <strong>of</strong> tamoxifen were sustained and<br />
<strong>of</strong> great magnitude. The ATLAS, ATtOM, and a few other<br />
small trials were combined and opened the door to the<br />
suggestion that 5 years <strong>of</strong> tamoxifen might not be optimal<br />
and that longer tamoxifen might further reduce disease-free<br />
From the Odette Cancer Center, McMaster University, Hamilton, ON, Karolinska<br />
Institutet, Stockholm, Sweden, and the Dana-Farber Cancer Institute, Boston, MA.<br />
Authors’ disclosures <strong>of</strong> potential conflicts <strong>of</strong> interest are found at the end <strong>of</strong> this article.<br />
Address reprint requests to Harold J. Burstein, Dana-Farber Cancer Institute, 450<br />
Brookline Avenue, Boston, MA 02215; email: hburstein@partners.org.<br />
© <strong>2012</strong> by <strong>American</strong> <strong>Society</strong> <strong>of</strong> <strong>Clinical</strong> <strong>Oncology</strong>.<br />
1092-9118/10/1-10<br />
71
ecurrence. This remains an important and unresolved<br />
question. It was also clear that ovarian ablation and/or<br />
suppression was effective but not significantly so when<br />
added to chemotherapy, perhaps because <strong>of</strong> chemotherapyinduced<br />
amenorrhea.<br />
In 2005, structural changes were required in the Overview<br />
process. The Trialists formed a new steering committee and<br />
organized subcommittees, many <strong>of</strong> which continue today to<br />
work very effectively. Many new trials were added and there<br />
were many additional years <strong>of</strong> follow-up for all major questions.<br />
However, data from many major trials, particularly<br />
those <strong>of</strong> taxanes, were missing in 2005.<br />
In 2006, the Trialists and their subcommittees met and<br />
a series <strong>of</strong> priorities were set. These included investigations<br />
<strong>of</strong> the type <strong>of</strong> anthracycline-based regimen; all taxane trials;<br />
aromatase inhibitors; trastuzumab; and chemoendocrine<br />
therapy, particularly in relation to the ER-positive compared<br />
with ER-negative issue in pre- and postmenopausal<br />
women. These meetings led five important publications on<br />
tamoxifen, chemotherapy, and loco-regional therapy. 1-5<br />
Nearly 30 years since its inception, the Overview remains<br />
highly relevant and informative. In comparison to individual<br />
trials—even huge trials tallying thousands <strong>of</strong> patients as<br />
has become common in early-stage breast cancer—the Overview<br />
has several methodologic virtues that make it a unique<br />
data resource. In particular, the Overview is important for<br />
(1) having all the worldwide data; (2) avoiding publication<br />
72<br />
KEY POINTS<br />
● The Oxford Overview suggests benefits for adjuvant<br />
endocrine therapy for all patients with ER positive<br />
breast cancer.<br />
● The Oxford Overview suggests benefits for anthracycline-<br />
and taxane-based adjuvant chemotherapy<br />
regardless <strong>of</strong> nodal, ER, or PR status.<br />
● Molecular diagnostic assays may identify ER-positive<br />
tumors that may not warrant chemotherapy.<br />
● Reconciling the historic “overview” and newer “personalized”<br />
approaches to early breast cancer is a<br />
compelling clinical challenge.<br />
PRITCHARD, BERGH, AND BURSTEIN<br />
Fig. 1. Effects <strong>of</strong> approximately 5 years<br />
<strong>of</strong> tamoxifen on the 15-year probabilities<br />
<strong>of</strong> recurrence and <strong>of</strong> breast cancer mortality<br />
for ER-positive disease.<br />
Abbreviations; ER, estrogen receptor;<br />
RR, recurrence rate; SE, standard error;<br />
(O-E)/V, (observed-expected)/variance.<br />
Reprinted from The Lancet, 378, Early<br />
Breast Cancer Trialists’ Collaborative<br />
Group, Davies C, Godwin J, et al. Relevance<br />
<strong>of</strong> breast cancer hormone receptors and<br />
other factors to the efficacy <strong>of</strong> adjuvant<br />
tamoxifen: Patient-level meta-analysis <strong>of</strong><br />
randomised trials, 771–784, 2011, with<br />
permission from Elsevier.<br />
bias; (3) giving average effect sizes; and (4) clarifying the<br />
timeframes <strong>of</strong> effects through large sample size and longterm<br />
follow-up.<br />
Recent Notable Findings from the Overview<br />
Endocrine Therapy<br />
In the main tamoxifen Overview, more than 54,500<br />
women were studied in trials <strong>of</strong> tamoxifen compared with<br />
no tamoxifen and more than 45,000 in trials <strong>of</strong> longer<br />
compared with shorter tamoxifen. The effects <strong>of</strong> 5 years <strong>of</strong><br />
tamoxifen on breast cancer recurrence and mortality are<br />
shown in Fig. 1.<br />
Fundamentally, tamoxifen has long and strong sustained<br />
effects on local recurrence, contralateral breast cancer, and<br />
distant and multiple recurrences. Tamoxifen benefits as<br />
measured by proportional risk reduction are similar regardless<br />
<strong>of</strong> age, stage, grade, or tumor size and with and without<br />
background chemotherapy.<br />
Figure 2 examines the effect <strong>of</strong> ER and progesterone<br />
receptor (PgR) expression on the benefits with tamoxifen.<br />
Tamoxifen is similarly effective in patients with ER-positive<br />
disease regardless <strong>of</strong> PgR expression. Patients with ERnegative<br />
but PgR-positive tumors do not get benefit from<br />
tamoxifen, nor do patients with ER- and PgR-negative<br />
tumors.<br />
Figure 3 explores the relationship <strong>of</strong> quantitative levels <strong>of</strong><br />
ER with the benefits <strong>of</strong> tamoxifen. It demonstrates that<br />
higher ER levels are associated with stronger effects <strong>of</strong><br />
tamoxifen.<br />
Figure 4 shows the effects <strong>of</strong> tamoxifen over time. They<br />
are large in years 0 to 1, 2 to 4 and 5 to 10 in terms <strong>of</strong><br />
recurrence, but by year 10 and beyond, the effects no longer<br />
increase although the previous gains are not lost. In mortality<br />
however, the benefits come out a little later in years 2 to<br />
4 and then years 5 to 9 and they persist into the 10 to<br />
15–year follow-up. Thus, there is a carryover effect on<br />
recurrence <strong>of</strong> 5 years <strong>of</strong> tamoxifen that goes on to at least 10<br />
years and in mortality that goes on to at least 15 years.<br />
The Overview has been pivotal in demonstrating the<br />
benefits <strong>of</strong> endocrine therapy for ER-positive breast cancer.<br />
Five years <strong>of</strong> tamoxifen in ER-positive disease reduces<br />
recurrences by a relative risk <strong>of</strong> 38%, breast cancer deaths<br />
by approximately 30%, and all deaths by approximately
UPDATE OF THE OXFORD OVERVIEW<br />
Fig. 2. Relevance <strong>of</strong> measured ER and<br />
PR status to the effects <strong>of</strong> tamoxifen on the<br />
10-year probably <strong>of</strong> recurrence.<br />
Abbreviations: ER, estrogen receptor; PR,<br />
progesterone receptor; RR, recurrence rate;<br />
SE, standard error; (O-E)/V, (observedexpected)/variance.<br />
Reprinted from The Lancet, 378, Early<br />
Breast Cancer Trialists’ Collaborative<br />
Group, Davies C, Godwin J, et al. Relevance<br />
<strong>of</strong> breast cancer hormone receptors and<br />
other factors to the efficacy <strong>of</strong> adjuvant<br />
tamoxifen: Patient-level meta-analysis <strong>of</strong><br />
randomised trials, 771–784, 2011, with<br />
permission from Elsevier.<br />
22%. Contralateral breast cancer is reduced by approximately<br />
40%. Tamoxifen for 5 years benefits all women with<br />
ER-positive disease and benefits those with very high levels<br />
<strong>of</strong> ER even more. To date, ER expression is the sole deter-<br />
Fig. 3. Relationship <strong>of</strong> quantitative<br />
levels <strong>of</strong> ER with benefits <strong>of</strong> tamoxifen.<br />
Abbreviations: O-E, observed-expected;<br />
ER, estrogen receptor; SE, standard<br />
error.<br />
Reprinted from The Lancet, 378, Early<br />
Breast Cancer Trialists’ Collaborative<br />
Group, Davies C, Godwin J, et al. Relevance<br />
<strong>of</strong> breast cancer hormone receptors<br />
and other factors to the efficacy <strong>of</strong><br />
adjuvant tamoxifen: Patient-level metaanalysis<br />
<strong>of</strong> randomised trials, 771–784,<br />
2011, with permission from Elsevier.<br />
minant <strong>of</strong> likely benefit from tamoxifen. Endometrial cancer,<br />
however, is increased by 2.3-fold following 5 years <strong>of</strong> tamoxifen.<br />
The rate <strong>of</strong> endometrial cancer increases as much as<br />
4-fold with 10 years <strong>of</strong> tamoxifen.<br />
73
Fig. 4. Abbreviations: ER, estrogen receptor; y, year; SE, standard error; (O-E)/V, (observed-expected)/variance.<br />
Reprinted from The Lancet, 378, Early Breast Cancer Trialists’ Collaborative Group, Davies C, Godwin J, et al. Relevance <strong>of</strong> breast cancer<br />
hormone receptors and other factors to the efficacy <strong>of</strong> adjuvant tamoxifen: Patient-level meta-analysis <strong>of</strong> randomised trials, 771–784, 2011,<br />
with permission from Elsevier.<br />
Radiation Therapy<br />
The Overview has also enabled powerful analyses <strong>of</strong> the<br />
effects <strong>of</strong> radiotherapy for early-stage breast cancer. The<br />
proportional effects <strong>of</strong> radiotherapy after breast-conserving<br />
surgery are substantial for both node-negative and nodepositive<br />
disease as shown in Fig. 5. There are substantial<br />
reductions in any recurrence and improvements in breast<br />
cancer survival with postsurgical radiotherapy after breastconserving<br />
surgery.<br />
An important question regarding radiotherapy for breast<br />
cancer has been whether all-cause mortality would be enhanced<br />
by radiation treatments or whether late (rare) side<br />
effects <strong>of</strong> radiation might negate treatment benefits in<br />
breast cancer–specific mortality. The Overview, with its<br />
large number <strong>of</strong> trials and patients, is uniquely positioned to<br />
address this question. Recent data (Fig. 6) indicate that<br />
radiotherapy after breast-conserving surgery reduces breast<br />
cancer recurrence and all-cause mortality among both nodepositive<br />
and node-negative disease. A one in four rule<br />
applies for patients with pN0 and pN1 disease in that one<br />
death is prevented for every four recurrences that are<br />
prevented. These benefits are not substantially reduced by<br />
side effects and are durable through 15 years <strong>of</strong> follow-up.<br />
74<br />
Chemotherapy<br />
PRITCHARD, BERGH, AND BURSTEIN<br />
The recent overview analysis was published in The Lancet<br />
in December 2011 and was based on the analysis <strong>of</strong> 100,000<br />
randomly selected patients treated in different randomized<br />
studies. 5<br />
In short, the overview contained the quinquennial update<br />
<strong>of</strong> the following:<br />
Nontaxane chemotherapy compared with taxanes in<br />
44,000 patients<br />
Different anthracycline based regimens in 6,000 patients<br />
Anthracyclines in the comparisons compared with CMF<br />
in 18,000 patients<br />
Randomized studies in 32,000 patients comparing no<br />
chemotherapy with polychemotherapy<br />
For this EBCTCG round, the chemotherapy regimens<br />
were, for the first time, grouped based on scheduling and<br />
dose intensity, with particular focus on the CMF- and<br />
anthracycline-based regimens. All outcome analyses were<br />
done based on individual patient tumor data with as long <strong>of</strong><br />
follow-up as possible for each study. Individual patient data<br />
were also included from some unpublished randomized stud-
UPDATE OF THE OXFORD OVERVIEW<br />
Fig. 5. Effect <strong>of</strong> radiotherapy after<br />
breast-conserving surgery on 10-year risk<br />
<strong>of</strong> any (loco-regional or distant) first recurrence<br />
on 15-year risk <strong>of</strong> breast cancer<br />
death in women with pathologically verified<br />
node status.<br />
Abbreviations: SE, standard error; RR, recurrence<br />
rate; BCS, breast-conserving surgery;<br />
RT, radiotherapy.<br />
Reprinted from The Lancet, 378, Early<br />
Breast Cancer Trialists’ Collaborative<br />
Group, Darby S, McGale P, et al. Effect <strong>of</strong><br />
radiotherapy after breast-conserving surgery<br />
on 10-year recurrence and 15-year<br />
breast cancer death: Meta-analysis <strong>of</strong> individual<br />
patient data for 10,801 women in<br />
17 randomized trials, 1707–1716, 2011,<br />
with permission from Elsevier.<br />
ies, thereby reducing the potential risks <strong>of</strong> publication bias<br />
<strong>of</strong> studies with a “positive outcome,” the latter being published<br />
earlier. Data were analyzed for recurrence, breast<br />
cancer–specific survival, and overall mortality.<br />
A summary <strong>of</strong> major findings includes the following results:<br />
The use <strong>of</strong> taxanes added a statistically significant<br />
reduced risk <strong>of</strong> breast cancer death by 14% (2.7% absolute<br />
Fig. 6. Effect <strong>of</strong> radiotherapy after breast-conserving surgery on 10-year risk <strong>of</strong> any (loco-regional or distant) first recurrence on 15-year<br />
risks <strong>of</strong> breast cancer death and death from any cause in 10,801 women (67% with pathologically node-negative disease) in 17 trials. 3<br />
Abbreviations: SE, standard error; RR, recurrence rate; BCS, breast-conserving surgery; RT, radiotherapy.<br />
Reprinted from The Lancet, 378, Early Breast Cancer Trialists’ Collaborative Group, Darby S, McGale P, et al. Effect <strong>of</strong> radiotherapy after<br />
breast-conserving surgery on 10-year recurrence and 15-year breast cancer death: Meta-analysis <strong>of</strong> individual patient data for 10,801 women<br />
in 17 randomized trials, 1707–1716, 2011, with permission from Elsevier.<br />
75
gain, p � 0.0005, 2.2% absolute overall survival gain)<br />
compared with anthracycline-based regimens at 8-years<br />
follow-up. This effect was not statistically significant with<br />
only a 6% relative effect if the anthracycline control arm was<br />
given with a high cumulative anthracycline dose (Fig. 6).<br />
There was no difference in schedule or taxane drug<br />
(paclitaxel or docetaxel).<br />
There was no difference whether endocrine therapy was<br />
given concurrently or in sequence with chemotherapy.<br />
There were no age-related effects by the delivered<br />
chemotherapies, taxane, or anthracycline-based regimens.<br />
Indeed, there was a trend suggesting greater chemotherapy<br />
effect in the age 55 to 69 group than in younger women.<br />
The anthracycline and taxane regimens had similar<br />
antitumoral effects irrespective <strong>of</strong> patient age, tumor ER<br />
status, whether tamoxifen was given, the combination <strong>of</strong> ER<br />
by HER2, or the interaction <strong>of</strong> ER and grade. (Fig. 7).<br />
76<br />
PRITCHARD, BERGH, AND BURSTEIN<br />
Fig. 7. Studies comparing the add-on effect<br />
by taxanes, most AC x 4. Demonstration<br />
<strong>of</strong> the outcome (breast cancer survival and<br />
overall survival) when the comparator arm<br />
contained higher doses <strong>of</strong> the nontaxane<br />
drugs, mostly anthracyclines.<br />
Abbreviations: AC, doxorubicin/cyclophosphamide;<br />
RR, recurrence rate; SE, standard<br />
error; (O-E)/V, (observed-expected)/<br />
variance.<br />
Reprinted from The Lancet, 379, Early<br />
Breast Cancer Trialists’ Collaborative Group,<br />
Peto R, Davies C, et al. Comparisons between<br />
different polychemotherapy regimens<br />
for early breast cancer: Meta-analyses<br />
<strong>of</strong> long-term outcome among 100,000<br />
women in 123 randomized trials, 432–444,<br />
<strong>2012</strong>, with permission from Elsevier.<br />
Four cycles <strong>of</strong> doxorubicin and cyclophosphamide produced<br />
a similar result as six courses <strong>of</strong> standard CMF at 10<br />
years follow-up. Anthracycline regimens with higher doses,<br />
however, reduced the breast cancer mortality by about 4%<br />
at 10 years (a relative improvement <strong>of</strong> 22%, p � 0.004)<br />
compared with CMF. The overall mortality was improved by<br />
the same extent.<br />
In comparison with no chemotherapy, CMF-like regimens<br />
and anthracycline-based regimens reduced overall<br />
mortality by 6% at 10 years. No subgroup could be readily<br />
identified that did not benefit from chemotherapy. The<br />
CMF-based studies revealed a less distinct message; the<br />
effect seemed less obvious for the elderly (irrespective <strong>of</strong><br />
ER status). This could be because <strong>of</strong> a combination <strong>of</strong> factors:<br />
lower compliance in those studies run decades ago; no access<br />
to modern antiemetic regimens; a lack <strong>of</strong> supportive care<br />
measurement; and less patient motivation to accept toxicity
UPDATE OF THE OXFORD OVERVIEW<br />
Fig. 8. Breast cancer mortality for anthracyclines compared with no adjuvant chemotherapy. Outcomes for different cumulative anthracyclines<br />
doses, age groups, node status, ER status. and the combination <strong>of</strong> ER status and histopathologic grade. All forest plots reveal a very similar<br />
pattern <strong>of</strong> the same relative magnitude <strong>of</strong> chemotherapy.<br />
Abbreviations: ER, estrogen receptor; CAF, cyclophosphamide/doxorubicin/fluorouracil; SE, standard error; 4AC/EC, four cycles <strong>of</strong> doxorubicin/cyclophosphamide<br />
or epirubicin/cyclophosphamide; NS, nodal status.<br />
Reprinted from The Lancet, 379, Early Breast Cancer Trialists’ Collaborative Group, Peto R, Davies C, et al. Comparisons between different<br />
polychemotherapy regimens for early breast cancer: Meta-analyses <strong>of</strong> long-term outcome among 100,000 women in 123 randomized trials,<br />
432–444, <strong>2012</strong>, with permission from Elsevier.<br />
77
since an eventual benefit by adjuvant therapies was not<br />
known at the time these studies were done.<br />
The Era <strong>of</strong> Personalized Breast Cancer Treatment<br />
During the last decade, there have been major attempts<br />
to tailor adjuvant chemotherapy and, to a lesser degree,<br />
the choices for adjuvant endocrine therapy, based on consideration<br />
<strong>of</strong> the unique biologic features <strong>of</strong> breast cancers. This<br />
effort has grown from recognition <strong>of</strong> major breast cancer<br />
subsets defined by widely used histopathologic markers<br />
(ER, PgR, and human epidermal growth factor receptor 2<br />
[HER2]) as well as traditional pathology (grade, proliferation)<br />
and molecular diagnostics (intrinsic subsets/OncotypeDX<br />
characteristics). 6 These schemes <strong>of</strong> prognostication and<br />
therapy prediction separate breast cancer into distinct subgroups,<br />
with different management, based principally on<br />
tumor histopathology, abetted by detailed insights from<br />
gene expression pr<strong>of</strong>iling <strong>of</strong> tumors.<br />
A number <strong>of</strong> retrospective analyses from prospective,<br />
randomized trials have suggested that the proportional<br />
benefits <strong>of</strong> chemotherapy are not, in fact, the same for all<br />
kinds <strong>of</strong> breast cancer. 6-9 In addition, previous studies have<br />
also revealed that chemotherapy—in particular CMF-based<br />
therapies—may have no or little value to patients with high<br />
ERs levels and that the major antitumoral effect for the<br />
premenopausal patients may be because <strong>of</strong> induction <strong>of</strong><br />
amenorrhea. 10 These subset analyses have suggested that<br />
the clinical benefit <strong>of</strong> chemotherapy is most dramatic in<br />
tumors with low or no ER expression (including triplenegative<br />
cancers), overexpression <strong>of</strong> HER2, higher grade,<br />
and higher proliferative scores. Conversely, endocrine therapies<br />
are most effective in tumors with higher levels <strong>of</strong> ER<br />
expression, lower levels <strong>of</strong> HER2, lower grade, and lower<br />
proliferative scores. Retrospective subset analyses <strong>of</strong> specific<br />
phase III studies have frequently found that the benefit <strong>of</strong><br />
adding taxane-based therapy is least pronounced among<br />
cancers that are ER positive and HER2 negative (the majority<br />
<strong>of</strong> most breast cancers), although these data must now<br />
be seen in the global context <strong>of</strong> the Overview findings to the<br />
contrary. However, consistent with the findings in single<br />
adjuvant trials, neoadjuvant studies have suggested that<br />
complete pathologic response to chemotherapy is far less<br />
common in tumors that are ER positive. Molecular assays<br />
have emerged that appear to predict the likelihood <strong>of</strong> chemotherapy<br />
benefit, particularly in addition to adjuvant<br />
endocrine therapy. Tumors classified as having a luminal A<br />
intrinsic subtype or a low recurrence score on the OncotypeDX<br />
assay, tend to derive minimal benefit from the<br />
addition <strong>of</strong> chemotherapy to endocrine therapy. It is unclear<br />
whether this is because the relative gain by chemotherapy is<br />
the same, but with an absolute gain that is too small to affect<br />
outcomes among patients with low-risk disease, or whether<br />
there truly is no benefit with chemotherapy. These assays<br />
correlate well if imperfectly with other tumor features such<br />
as grade, proliferation, and quantitative levels <strong>of</strong> ER and<br />
HER2, fitting with subset hypotheses about the importance<br />
<strong>of</strong> those pathologic features.<br />
Finally, the advent <strong>of</strong> anti-HER2 therapy in the adjuvant<br />
setting with trastuzumab has radically altered the natural<br />
history <strong>of</strong> HER2-positive breast cancers. These tumors,<br />
which traditionally had a less favorable prognosis, now<br />
appear to have as good a prognosis as HER2-negative<br />
tumors. The use <strong>of</strong> trastuzumab for HER2-positive cancers<br />
78<br />
PRITCHARD, BERGH, AND BURSTEIN<br />
and the additional use <strong>of</strong> adjuvant endocrine therapy for<br />
ER-positive cancers has meant that issues <strong>of</strong> endocrine<br />
therapy or not or chemotherapy or not are less compelling<br />
now than decades ago. The persistent question has become:<br />
are there sufficient data to tailor treatments based on<br />
specific biomarker subsets, or do all patients need the same<br />
treatments?<br />
Can We Reconcile the Overview and the Individual<br />
Treatment Paradigms?<br />
We now confront the paradox <strong>of</strong> the Overview. In large<br />
measure, the Overview argues for adjuvant chemotherapy<br />
for all women, regardless <strong>of</strong> tumor biology or stage. This<br />
blanket observation seems at odds with the growing literature<br />
that suggests that individual tumors—and thus individual<br />
patients—respond differently to chemotherapy and<br />
endocrine therapy. Can these observations be reconciled?<br />
The reason why the EBCTCG analysis demonstrates relatively<br />
similar antitumoral effects in all the different subgroups<br />
is so far not understood. There are two particular<br />
strengths <strong>of</strong> the Overview that should not be forgotten.<br />
First, the Overview is an enormous data repository that<br />
dwarfs in the number <strong>of</strong> events and the duration <strong>of</strong> follow-up<br />
<strong>of</strong> the available data from any given study. This gives the<br />
Overview power to see effects that might be missed in<br />
smaller, retrospective efforts. Of note, the previously published<br />
reports on selectivity <strong>of</strong> the chemotherapy effects are<br />
the result <strong>of</strong> subgroup analysis on materials with much less<br />
statistical power and in some studies lacking inclusion <strong>of</strong> all<br />
randomly selected patients and overlapping 95% confidence<br />
intervals for some <strong>of</strong> the interesting findings.<br />
Secondly, by including data from almost all adjuvant<br />
trials, the Overview minimizes the temptation to focus on<br />
positive, published studies, a bias particularly notable in the<br />
literature on biomarker subsets where “negative” studies<br />
are all but unpublishable. These are important considerations<br />
that should give clinicians pause before dismissing<br />
the Overview findings.<br />
At the same time, there are features <strong>of</strong> the Overview<br />
design that may overstate the benefits <strong>of</strong> chemotherapy. The<br />
Overview reports on proportional risk reduction and related<br />
absolute differences in outcome. However, for most patients,<br />
the absolute gains are the driving force for decisions on<br />
chemotherapy. Differences in absolute benefit from chemotherapy<br />
clearly vary based on tumor stage and on the<br />
residual degree <strong>of</strong> risk that remains despite adjuvant endocrine<br />
therapy. In addition, the Overview has not been able to<br />
capture adequate information on chemotherapy-induced<br />
amenorrhea so as to tease out the effects <strong>of</strong> chemotherapy on<br />
ovarian function in women with ER-positive tumors, though<br />
chemotherapy benefit is notably similar regardless <strong>of</strong> age<br />
and ER status. 5<br />
More critically, perhaps, is that differences in treatment<br />
effect may hinge on consideration <strong>of</strong> multiple variables that<br />
the Overview has not as yet grappled with in detail. It could<br />
simply be because the EBCTCG data just describes mean<br />
effects for all the different subgroups without enabling the<br />
identification <strong>of</strong> distinct subgroups with claimed different<br />
biology and outcome. The characterization <strong>of</strong> subsets <strong>of</strong><br />
ER-positive cancers, in particular, by molecular assays suggests<br />
that simultaneous consideration <strong>of</strong> ER levels, HER2<br />
expression, and grade/proliferation is important for classifying<br />
cancers and determining treatment benefit. As yet, the
UPDATE OF THE OXFORD OVERVIEW<br />
Overview analyses looking at quantitative hormone receptor<br />
levels or considering “2 � 2” analyses <strong>of</strong> ER by HER2 or ER<br />
by grade may not be sufficiently “multiplexed” to figure out<br />
which patients can avoid chemotherapy. A major limitation<br />
with the present Overview is the lack <strong>of</strong> data on proliferation,<br />
gene expression, modern immunohistochemical measurements<br />
<strong>of</strong> receptors, quality controlled pathology<br />
classification, and central marker review. Information on<br />
these variables would, <strong>of</strong> course, have added value to the<br />
EBCTCG processes and analyses, and some investigators<br />
may claim that these shortcomings may explain some <strong>of</strong> the<br />
results. Despite these shortcomings, the example <strong>of</strong> ER<br />
values in the overall EBCTCG suggest very robust results in<br />
relation to adjuvant tamoxifen, further supporting the validity<br />
<strong>of</strong> the EBCTCG findings. 4<br />
Additional methodologic details may be relevant. Several<br />
studies that have identified selective benefit for chemotherapy<br />
have centrally reanalyzed ER or HER2 data on some<br />
patients, a detailed pathologic step not preformed in the<br />
EBCTCG database. 8,10 Similarly, the Overview used a surrogate<br />
strategy for high proliferation/higher OncotypeDx<br />
scores by using grade as reported by local laboratories (in<br />
particular, poorly differentiated cancers) that have not been<br />
confirmed. Furthermore, the EBCTCG demonstration <strong>of</strong><br />
similar effects <strong>of</strong> chemotherapies, independent <strong>of</strong> studied<br />
subgroups, could merely be a reflection <strong>of</strong> an inherent<br />
common biology <strong>of</strong> breast cancer with a similar type <strong>of</strong><br />
Authors’ Disclosures <strong>of</strong> Potential Conflicts <strong>of</strong> Interest<br />
Author<br />
Employment or<br />
Leadership<br />
Positions<br />
Consultant or<br />
Advisory Role<br />
Kathleen I. Pritchard GlaxoSmithKline;<br />
Novartis; Ortho<br />
Biotech; Pfizer;<br />
Roche; San<strong>of</strong>i;<br />
YM BioSciences<br />
Jonas Bergh Affibody; Amgen;<br />
AstraZeneca;<br />
Bayer;<br />
GlaxoSmithKline;<br />
i3innovus; Onyx;<br />
Pfizer; San<strong>of</strong>i;<br />
Tapestry<br />
Pharmaceuticals<br />
Harold J. Burstein*<br />
*No relevant relationships to disclose.<br />
1. Early Breast Cancer Trialists’ Collaborative Group. Adjuvant polychemotherapy<br />
in estrogen-receptor-poor breast cancer: Meta-analysis <strong>of</strong> individual<br />
patient data from randomized trials. Lancet. 2008;371:29-40.<br />
2. Early Breast Cancer Trialists’ Collaborative Group. Overview <strong>of</strong> the<br />
randomized trials <strong>of</strong> radiotherapy in ductal carcinoma in situ (DCIS) <strong>of</strong> the<br />
breast. J Natl Cancer Inst Monogr. 2010;41:162-177.<br />
3. Early Breast Cancer Trialists’ Collaborative Group. Effect <strong>of</strong> radiotherapy<br />
after breast-conserving surgery on 10-year recurrence and 15-year breast<br />
cancer death: Meta-analysis <strong>of</strong> individual patient data for 10,801 women in 17<br />
randomized trials. Lancet. 2011;378:1707-1716.<br />
4. Early Breast Cancer Trialists’ Collaborative Group. Relevance <strong>of</strong> breast<br />
cancer hormone receptors and other factors to the efficacy <strong>of</strong> adjuvant<br />
tamoxifen: Patient-level meta-analysis <strong>of</strong> randomised trials. Lancet. 2011;<br />
378:771-784.<br />
5. Early Breast Cancer Trialists’ Collaborative Group. Comparisons between<br />
different polychemotherapy regimens for early breast cancer: Metaanalyses<br />
<strong>of</strong> long-term outcome among 100,000 women in 123 randomized<br />
trials. Lancet. <strong>2012</strong>;379:432-444.<br />
sensitivity to the commonly used cytostatics. However, this<br />
seems counterintuitive based on present extensive knowledge<br />
<strong>of</strong> breast cancer biology in relation to outcome and<br />
therapy strategies, but it has happened before in the era <strong>of</strong><br />
science that conclusions made with the best intentions have<br />
required modification.<br />
A possible explanation to the general findings <strong>of</strong> a similar<br />
relative magnitude in anti–breast cancer effects in the<br />
adjuvant setting by taxanes and anthracyclines could be<br />
that all epithelial breast cancers seem to share so far<br />
unidentified gene cassettes associated with a similar sensitivity<br />
to chemotherapy. This could, to some extent, be<br />
potentially substantiated by ongoing studies on human<br />
breast cancer stem cells from primary cultures, which indicate<br />
a common phenotype for most breast cancer stem cells<br />
in relation to a specific receptor status, despite the fact that<br />
cancers per se have different tumor and receptor characteristics<br />
(Johan Hartman, Irma Fredriksson, Jonas Bergh,<br />
verbal communications, March 1, <strong>2012</strong>).<br />
At its heart, the task <strong>of</strong> reconciling the invaluable data<br />
from the ongoing Oxford Overview with the emerging data<br />
from subset studies using novel markers remains the fundamental<br />
challenge for adjuvant therapy for breast cancer.<br />
Ongoing collaboration and research will be critical for helping<br />
clinicians and patients understand these different but<br />
important clinical datasets and solving the riddle <strong>of</strong> the<br />
Overview paradox.<br />
Stock<br />
Ownership Honoraria<br />
REFERENCES<br />
AstraZeneca;<br />
GlaxoSmithKline;<br />
Novartis; Pfizer;<br />
Roche; San<strong>of</strong>i<br />
AstraZeneca;<br />
Pfizer; Roche;<br />
San<strong>of</strong>i<br />
Research<br />
Funding<br />
Merck; Pfizer;<br />
Roche<br />
Expert<br />
Testimony<br />
AstraZeneca;<br />
Novartis; Pfizer;<br />
San<strong>of</strong>i<br />
Other<br />
Remuneration<br />
AstraZeneca;<br />
Roche<br />
6. Goldhirsch A, Wood WC, Coates AS, et al. Strategies for subtypesdealing<br />
with the diversity <strong>of</strong> breast cancer: Highlights <strong>of</strong> the St. Gallen<br />
International Expert Consensus on the Primary Therapy <strong>of</strong> Early Breast<br />
Cancer 2011. Ann Oncol. 2011;22:1736-1747.<br />
7. Paik S, Tang G, Shak S, et al. Gene expression and benefit <strong>of</strong><br />
chemotherapy in women with node-negative, estrogen receptor-positive<br />
breast cancer. J Clin Oncol. 2006;24:3726-3734.<br />
8. Hayes DF, Thor AD, Dressler LG, et al. HER2 and response to paclitaxel<br />
in node-positive breast cancer. N Engl J Med. 2007;357:1496-1506.<br />
9. Berry DA, Cirrincione C, Henderson IC, et al. Estrogen-receptor status<br />
and outcomes <strong>of</strong> modern chemotherapy for patients with node-positive breast<br />
cancer. JAMA. 2006;295:1658-1667.<br />
10. Karlsson P, Sun Z, Braun D, et al. Long-term results <strong>of</strong> International<br />
Breast Cancer Study Group Trial VIII: Adjuvant chemotherapy plus goserelin<br />
compared with either therapy alone for premenopausal patients with nodenegative<br />
breast cancer. Ann Oncol. 2011;22:2216-2226.<br />
79
GENE PATENTING: EFFECTS ON BIOTECHNOLOGY<br />
AND ONCOLOGY<br />
CHAIR<br />
Kenneth Offitt, MD, MPH<br />
Memorial Sloan-Kettering Cancer Center<br />
New York, NY<br />
SPEAKERS<br />
Roger D. Klein, MD, JD<br />
Blood Center <strong>of</strong> Wisconsin<br />
Milwaukee, WI<br />
Hans Sauer, PhD, JD<br />
Biotechnology Industry Organization<br />
Houston, TX
Gene Patents and Personalized Medicine<br />
Overview: The use <strong>of</strong> genetic information to design and guide<br />
therapies and to develop novel diagnostic procedures creates<br />
important patent issues. Patents on human gene sequences<br />
have likely helped stimulate the introduction <strong>of</strong> new biologics;<br />
however, their role and that <strong>of</strong> patents on genotype-phenotype<br />
THE PRACTICE <strong>of</strong> personalized medicine has been<br />
described as giving the right drug, to the right patient,<br />
at the right dose, at the right time. Central to the concept <strong>of</strong><br />
personalized patient management is knowledge and application<br />
<strong>of</strong> somatic and heritable molecular pathologic (molecular<br />
genetic) changes associated with disease susceptibility,<br />
prognosis, therapeutic responsiveness, and susceptibility to<br />
side effects. Although we are far from achieving truly individually<br />
designed care for most patients, in oncology the era<br />
<strong>of</strong> targeted therapies linked to measurable biomarkers has<br />
arrived. Diagnostic tests have long been important in cancer<br />
management, but the high sensitivity, specificity, and direct<br />
relationship to therapy <strong>of</strong> many nucleic acid-based analyses<br />
adds enormously to their utility.<br />
Because they have been rationally designed to directly<br />
influence biochemical pathways implicated in malignant<br />
progression, targeted cancer therapies potentially <strong>of</strong>fer patients<br />
greater benefits with less morbidity than most currently<br />
used chemotherapeutic agents. Moreover, by allowing<br />
for prospective selection <strong>of</strong> individuals most likely to benefit<br />
from a given therapy, molecular pathology-guided approaches<br />
benefit patients by rescuing antineoplastic agents,<br />
the efficacy <strong>of</strong> which is confined to select populations. Without<br />
the ability to specifically identify potential responders,<br />
such drugs may mistakenly be abandoned. Finally, personalized<br />
medicine encourages cost-effective use <strong>of</strong> expensive<br />
therapies by restricting administration to patients most<br />
likely to benefit and least likely to suffer harms from them.<br />
Gene Patents and Personalized Medicine<br />
The use <strong>of</strong> molecular pathologic information to develop<br />
and prescribe therapies raises important new patent issues.<br />
In the United States, patents confer on the patent holder<br />
the right to exclude others from making, using, or selling an<br />
invention for 20 years from the filing date. Under the Patent<br />
Act, patentable inventions must be novel, nonobvious, and<br />
useful. 1 Patent claims define an invention’s features much<br />
as a deed delineates the boundaries <strong>of</strong> a plot <strong>of</strong> land. Patent<br />
applications are submitted to the U.S. Patent and Trademark<br />
Office (USPTO) where they are rejected or allowed and<br />
issued. Importantly, “[p]rocesses, machines, manufactures,<br />
and compositions <strong>of</strong> matter” can be patented, 2 but patents<br />
may not be obtained on products <strong>of</strong> nature or, under the<br />
“natural phenomenon doctrine,” “laws <strong>of</strong> nature, natural<br />
phenomena, and abstract ideas.” 3<br />
Patents have been integral to the discovery and commercialization<br />
<strong>of</strong> drugs and biologics. The product exclusivity<br />
patents confer has been instrumental in ensuring that<br />
manufacturers can raise sufficient capital to identify, test,<br />
obtain regulatory approval for, and bring to market new<br />
therapeutic agents. Absent enforceable patent rights for<br />
new drugs, investment dollars would likely be redeployed to<br />
By Roger D. Klein, MD, JD<br />
correlations in diagnostic testing is highly controversial.<br />
Genotype-phenotype associations are at the heart <strong>of</strong> personalized<br />
medicine. The intellectual property rules by which these<br />
biologic relationships are governed have pr<strong>of</strong>ound implications<br />
for the growth <strong>of</strong> individualized medicine.<br />
areas with a more favorable risk-reward balance. So clear<br />
are the positive incentives patents provide the pharmaceutical<br />
industry, even patent system critics have acknowledged<br />
them. 4 Similarly, the stimulatory effects <strong>of</strong> patent<br />
protection on pharmaceutical development may extend to<br />
patents granted on human genes, when patented genes are<br />
cloned to produce biologic therapies and vaccines. 5<br />
By contrast, patents on relationships between genetic<br />
variants and clinical phenotypes are controversial. Such<br />
patents have been widely criticized and their legitimacy<br />
challenged, whether the patents directly claim genotypephenotype<br />
associations or gene sequences themselves. 5,6<br />
Because genotype-phenotype associations are the center <strong>of</strong><br />
personalized medicine, the intellectual property rules by<br />
which these biologic relationships are governed have pr<strong>of</strong>ound<br />
implications for its growth.<br />
<strong>Clinical</strong> implementation <strong>of</strong> molecular pathology procedures<br />
is substantially less costly than the discovery, development,<br />
and commercialization <strong>of</strong> in vivo pharmaceuticals.<br />
Manufacturers typically invest hundreds <strong>of</strong> millions <strong>of</strong> dollars<br />
over a prolonged period <strong>of</strong> time to bring a drug or<br />
biologic to market. Drug candidates must be extensively<br />
tested in preclinical and clinical studies before U.S. Food<br />
and Drug Administration (FDA) approval. Unlike therapeutics,<br />
most molecular pathology procedures are designed,<br />
developed, and validated by individual diagnostic laboratories<br />
at limited expense. The wide availability <strong>of</strong> automated<br />
nucleic acid extractors, thermal cyclers, and DNA sequencing<br />
instruments combined with broad licensing <strong>of</strong> standard<br />
molecular methods and the free accessibility <strong>of</strong> online gene<br />
sequence databases has rendered development straightforward<br />
for many molecular pathology procedures.<br />
Laboratory quality is ensured by the <strong>Clinical</strong> Laboratory<br />
Improvement Amendments, accreditation by the College <strong>of</strong><br />
<strong>American</strong> Pathologists and other pr<strong>of</strong>essional societies, and<br />
individual state requirements. Most molecular pathology<br />
services do not undergo FDA clearance or approval, although<br />
laboratories serving New York patients submit tests<br />
to the state’s <strong>Clinical</strong> Laboratory Evaluation Program for<br />
pre-implementation review. The small minority <strong>of</strong> molecular<br />
pathology procedures manufactured and sold as kits to<br />
clinical laboratories must be cleared or approved by the<br />
FDA. However, this process is much less burdensome and<br />
orders <strong>of</strong> magnitude less expensive for in vitro diagnostic<br />
kits than for drugs.<br />
From the Medical College <strong>of</strong> Wisconsin, Milwaukee, WI.<br />
Author’s disclosure <strong>of</strong> potential conflicts <strong>of</strong> interest are found at the end <strong>of</strong> this article.<br />
Address reprint requests to Roger D. Klein, MD, JD, 27500 Cedar Road #808, Beachwood,<br />
OH 44122; email: roger.klein@aya.yale.edu.<br />
© <strong>2012</strong> by <strong>American</strong> <strong>Society</strong> <strong>of</strong> <strong>Clinical</strong> <strong>Oncology</strong>.<br />
1092-9118/10/1-10<br />
81
The majority <strong>of</strong> gene-related discoveries have been made<br />
by federally funded academic researchers, for whom patents<br />
are unlikely to have provided primary or even significant<br />
motivation. 7 Further, given the relative ease by which<br />
molecular pathology procedures are introduced into practice,<br />
patents are unnecessary for the clinical implementation<br />
<strong>of</strong> molecular pathologic discoveries. Rather, enforcement <strong>of</strong><br />
gene patents in the diagnostic realm results in the elimination<br />
<strong>of</strong> already existing services and dissuades laboratories<br />
from adding new ones. 8 The resultant elimination <strong>of</strong> competition<br />
is likely to result in higher prices, decreased quality,<br />
reduced innovation, and diminished patient access to these<br />
important medical services. Moreover, loss <strong>of</strong> academic<br />
providers will adversely affect training and related clinical<br />
research.<br />
Greater interest in companion diagnostic development by<br />
the pharmaceutical industry may alter the preceding dynamics.<br />
Control <strong>of</strong> companion diagnostics could allow drug<br />
manufacturers to ensure assay standardization, enforce<br />
specific quality requirements, and provide additional<br />
sources <strong>of</strong> revenue. Companion diagnostics could serve as<br />
vehicles by which companies preserve market share for<br />
linked drugs, assuming they avoid patent misuse or antitrust<br />
violations. Under this model, gene-related patents may<br />
create incentives for marker discovery and in vitro diagnostic<br />
commercialization. However, such patents could prove a<br />
double-edged sword by creating holdout problems for drug<br />
vendors who do not own the underlying molecular pathologic<br />
relationships.<br />
Thousands <strong>of</strong> U.S. patents have been issued on human<br />
gene sequences and genotype-phenotype associations, but<br />
their legal status remains uncertain. Several cases presently<br />
making their way through the courts may clarify and add<br />
reason to the law in this area.<br />
Recent Legal Developments<br />
In Mayo Collaborative Servs. v. Prometheus Labs., Inc., 9<br />
Prometheus sued Mayo Clinic in the District Court for the<br />
82<br />
KEY POINTS<br />
● Personalized medicine in the form <strong>of</strong> targeted drug<br />
therapies is already making important contributions<br />
to oncology care.<br />
● Patents on human gene sequences have likely helped<br />
stimulate the introduction <strong>of</strong> novel biologics and<br />
pharmaceutical agents.<br />
● Conversely, patents on human gene sequences and<br />
genotype-phenotype correlations appear to reduce<br />
the availability <strong>of</strong> and patient access to already<br />
existing diagnostic services, while increasing costs<br />
and decreasing innovation in the development <strong>of</strong><br />
diagnostic methods.<br />
● The intellectual property rules by which patents on<br />
human genes and genotype-phenotype associations<br />
are governed will have a pr<strong>of</strong>ound impact on the<br />
advancement <strong>of</strong> personalized cancer care.<br />
● Several key legal cases now before the courts may add<br />
clarity and guidance to the law governing this area.<br />
ROGER D. KLEIN<br />
Southern District <strong>of</strong> California for infringement <strong>of</strong> a process<br />
patent covering the postadministration correlation between<br />
thiopurine drug activity and side effects, and blood levels <strong>of</strong><br />
the metabolites 6-methyl mercaptopurine (6-MMP) and<br />
6-thioguanine (6-TG). The patent claims at issue include<br />
“administering a drug,” “determining the level” <strong>of</strong> a metabolite<br />
<strong>of</strong> the drug, and correlating this level with therapeutic<br />
efficacy or side effects.<br />
The district court found as a matter <strong>of</strong> law that the patent,<br />
which essentially claims the reference range for the drugs,<br />
covered an unpatentable natural phenomenon. The Court <strong>of</strong><br />
Appeals for the Federal Circuit (CAFC), the national patent<br />
appeals court, reversed the lower court, finding that the<br />
patent covers a treatment method. Further, the CAFC held<br />
that the in vivo metabolism <strong>of</strong> thiopurine agents constituted<br />
a transformation <strong>of</strong> matter, consistent with a patentable<br />
process on an application <strong>of</strong> a natural phenomenon as<br />
opposed to a patent on the phenomenon itself.<br />
The Prometheus Labs case was appealed to and accepted<br />
by the U.S. Supreme Court. Oral arguments were held<br />
before the Court on December 7, 2011. Of note, during oral<br />
arguments, the attorney for Prometheus Labs acknowledged<br />
that it had patented “a fact” they identified, and that<br />
physicians can infringe the patent merely by thinking about<br />
the biologic relationship between metabolite levels and patient<br />
response. However, because the case involved an exogenously<br />
administered drug rather than a genetic variant<br />
intrinsic to a patient, and the CAFC considered the patent<br />
a therapeutic method patent, and Prometheus defended<br />
its validity on this basis, the implications for genotypephenotype<br />
association patents are unclear.<br />
In a case addressing legal standards for patent obviousness,<br />
KSR Int’l Co. v. Teleflex Inc., 550 U.S. 398 (2007),<br />
Teleflex sued KSR for infringement <strong>of</strong> a patent that claimed<br />
the combination <strong>of</strong> an adjustable brake pedal and an electronic<br />
sensor. The district court found the patent obvious,<br />
but the CAFC reversed the lower court decision. The Supreme<br />
Court in turn reversed the CAFC, ruling that the<br />
patent was obvious. The CAFC, the high court stated, had<br />
applied overly restrictive criteria in finding the patent<br />
“nonobvious.” Importantly, the Supreme Court held that<br />
“obviousness to try” a problem-solving approach can render<br />
a patent obvious if there is a demonstrated need for a<br />
discovery and a finite number <strong>of</strong> identified, predictable<br />
solutions.<br />
Many if not most patented genes were initially mapped<br />
to a chromosomal region before discovery. Moreover, many<br />
genes are involved in sequential biochemical pathways in<br />
which disease-related changes were known before specific<br />
genetic variations were identified. Therefore, for many <strong>of</strong><br />
these discoveries, it was arguably obvious to look for variants<br />
in the potentially responsible genes among the finite<br />
number <strong>of</strong> available solutions. As a result, under KSR Int’l<br />
Co. v. Teleflex Inc. (KSR) many gene-related patents may be<br />
subject to challenge on obviousness grounds. In an early<br />
application <strong>of</strong> the KSR ruling, the USPTO refused to award<br />
a patent on the gene sequence <strong>of</strong> the natural killer cell<br />
activation–inducing ligand. Although the sequence had not<br />
been previously described, the USPTO found it obvious in<br />
light <strong>of</strong> the prior art. The CAFC upheld the USPTO’s<br />
decision. 10<br />
Finally, in Association for Molecular Pathology v. United<br />
States Patent and Trademark Office, a lawsuit sponsored
GENE PATENTS AND EFFECTS ON PERSONALIZED CANCER CARE<br />
by the <strong>American</strong> Civil Liberties Union, a number <strong>of</strong> medical<br />
and pr<strong>of</strong>essional societies, health care providers, and patients<br />
with breast cancer sued Myriad Genetics and the<br />
USPTO. The plaintiffs seek to invalidate key claims <strong>of</strong><br />
patents covering wild-type and mutated sequences <strong>of</strong> the<br />
BRCA1 and BRCA2 genes and associations between those<br />
sequences and the predisposition to breast and/or ovarian<br />
cancer. In addition to arguing that the patents claim unpatentable<br />
natural products and natural phenomena, the plaintiffs<br />
asserted that they violate Article I, section 8, clause 8<br />
and the First Amendment to the U.S. Constitution.<br />
On March 29, 2010, in a landmark decision, the court held<br />
that both composition <strong>of</strong> matter claims on the human<br />
BRCA1 and BRCA2 gene sequences, and process claims<br />
covering the correlations between mutations in these genes<br />
and a predisposition to breast cancer are invalid as a matter<br />
<strong>of</strong> law. These patent claims, the court ruled, were directed<br />
toward unpatentable subject matter. 10<br />
Author’s Disclosure <strong>of</strong> Potential Conflicts <strong>of</strong> Interest<br />
Author<br />
Roger D. Klein*<br />
*No relevant relationships to disclose.<br />
Employment or<br />
Leadership<br />
Positions<br />
Consultant or<br />
Advisory Role<br />
1. 35 U.S.C. §§ 101-103 (2006).<br />
2. 35 U.S.C. § 101 (2006).<br />
3. Diamond v. Diehr. 450 U.S. 175, 185 (1981).<br />
4. Bessen J, Meurer MJ. Patent Failure: How Judges, Bureaucrats, and<br />
Lawyers Put Innovators at Risk. Princeton, NJ: Princeton University Press; 2008.<br />
5. Klein RD. Gene patents and genetic testing in the United States. Nat<br />
Biotechnol. 2007;25:989-990.<br />
6. Secretary’s Advisory Committee on Genetics, Health, and <strong>Society</strong>. Revised<br />
Draft Report on Gene Patents and Licensing Practices and Their Impact<br />
on Patient Access to Genetic Tests (2010). http://oba.od.nih.gov/SACGHS/<br />
sacghs_documents.html. Accessed February 10, 2010.<br />
On July 29, 2011, although upholding the district court’s<br />
finding <strong>of</strong> invalidity <strong>of</strong> the genotype-phenotype correlation<br />
claims, the CAFC held that human gene sequences are<br />
patent-eligible subject matter, reversing ruling <strong>of</strong> the lower<br />
court. The plaintiffs have appealed the case to the Supreme<br />
Court.<br />
Conclusion<br />
Genes as chemicals have an important role to play in<br />
therapeutic development. However, ready access to the<br />
information contained within patients’ genes is also critical<br />
to the advancement <strong>of</strong> personalized medicine. Therefore, a<br />
prudent course for the courts when addressing gene-related<br />
patents is to strike a balance, upholding their validity when<br />
the patents protect drug development, while denying their<br />
enforceability when the patents create the potential for<br />
genetic testing monopolies.<br />
Stock<br />
Ownership Honoraria<br />
REFERENCES<br />
Research<br />
Funding<br />
Expert<br />
Testimony<br />
Other<br />
Remuneration<br />
7. Klein RD, Mahoney MJ. LabCorp v Metabolite Laboratories: the Supreme<br />
Court listens, but declines to speak. J Law Med Ethics. 2008;36:141-<br />
149.<br />
8. Cho MK, Illangasekare S, Weaver MA, et al. Effects <strong>of</strong> patents and<br />
licenses on the provision <strong>of</strong> clinical genetic testing services. J Mol. Diagn.<br />
2003;5:3-8.<br />
9. In v. Mayo Collaborative Servs. Prometheus Labs., Inc. 581 F. 3d 1136<br />
(Fed. Circ. 2009), cert. granted (No. 10-1150).<br />
10. In re Kubin, 561 F. 3d 1351 (Fed. Cir. 2009).<br />
11. Ass’n for Molecular Pathology v. U.S. Patent and Trademark Office. 653<br />
F. 3d 1329 (Fed. Cir. 2011), petition for cert. filed.<br />
83
BREAST CANCER CHEMOPREVENTION:<br />
IF NOT NOW, WHEN?<br />
CHAIR<br />
Abenaa M. Brewster, MD, MHS<br />
University <strong>of</strong> Texas M. D. Anderson Cancer Center<br />
Houston, TX<br />
SPEAKERS<br />
Nancy E. Davidson, MD<br />
University <strong>of</strong> Pittsburgh Cancer Center<br />
Pittsburgh, PA<br />
Worta McCaskill-Stevens, MD, MS<br />
National Cancer Institute<br />
Bethesda, MD
Chemoprevention for Breast Cancer:<br />
Overcoming Barriers to Treatment<br />
By Abenaa M. Brewster, MD, MHS, Nancy E. Davidson, MD, and<br />
Worta McCaskill-Stevens, MD, MS<br />
Overview: Evidence from placebo-controlled, randomized<br />
clinical trials supports the use <strong>of</strong> chemoprevention in women<br />
at high risk for developing breast cancer, and two agents—<br />
tamoxifen and raloxifene—are U.S. Food and Drug Administration<br />
(FDA)-approved for the indication. Despite clinical<br />
guidelines that recommend physicians counsel high-risk<br />
women about the use <strong>of</strong> chemoprevention and the estimated<br />
2.4 million women in the United States who meet eligibility<br />
criteria for net benefit, the uptake <strong>of</strong> breast cancer chemoprevention<br />
has been exceedingly low. Assessments <strong>of</strong> the risks<br />
and benefits <strong>of</strong> chemoprevention are aided by the availability<br />
<strong>of</strong> models that can be used to estimate <strong>of</strong> the risk–benefit<br />
ratio. However, many physicians remain unaware <strong>of</strong> these<br />
SEVERAL RANDOMIZED clinical trials have provided<br />
evidence for the chemoprevention <strong>of</strong> invasive breast<br />
cancer by selective estrogen receptor modulators (SERMs).<br />
The Breast Cancer Prevention Trial (BCPT) randomly selected<br />
13,388 women age 35 or older with a 5-year predicted<br />
risk for breast cancer <strong>of</strong> 1.66% or higher to receive either<br />
tamoxifen or placebo. 1 At a mean follow-up <strong>of</strong> 4 years, there<br />
was a 49% reduction in the risk <strong>of</strong> developing invasive breast<br />
carcinoma and a 50% reduction in the risk <strong>of</strong> noninvasive<br />
breast cancer among women who were assigned to the<br />
tamoxifen arm. The absolute risk reduction was 21.4 cases<br />
per 1,000 women over 5 years. The benefit <strong>of</strong> tamoxifen was<br />
limited to estrogen receptor–positive tumors, and the greatest<br />
risk reduction (86%) was observed in women with a<br />
history <strong>of</strong> atypical hyperplasia. However, tamoxifen was also<br />
associated with a 2.5-fold increased risk <strong>of</strong> endometrial<br />
carcinoma and an increased risk <strong>of</strong> stroke, deep venous<br />
thrombosis, pulmonary embolus, and cataracts. 1 The results<br />
<strong>of</strong> the Study <strong>of</strong> Tamoxifen and Raloxifene (STAR) demonstrated<br />
that raloxifene was as effective as tamoxifen in<br />
reducing the risk <strong>of</strong> breast cancer in postmenopausal women<br />
at increased risk <strong>of</strong> breast cancer with less adverse events <strong>of</strong><br />
thromboembolic events, endometrial cancer, and cataracts. 2<br />
Similar risks for ischemic heart disease, fractures, and<br />
stroke were found for both drugs. There was a nonstatistically<br />
significant higher risk <strong>of</strong> noninvasive breast cancer<br />
from raloxifene, which diminished at longer follow-up. 2<br />
Additional evidence for the efficacy <strong>of</strong> SERMs is derived<br />
from the European randomized trials <strong>of</strong> tamoxifen compared<br />
with placebo, which have demonstrated a more modest<br />
benefit <strong>of</strong> tamoxifen for reducing the risk <strong>of</strong> invasive breast<br />
cancer. 3<br />
The significant reduction in contralateral breast cancer<br />
seen in the adjuvant aromatase inhibitor treatment trials<br />
and the potentially serious toxicities associated with SERMs<br />
led to the investigation <strong>of</strong> aromatase inhibitors for the<br />
primary prevention <strong>of</strong> breast cancer. Goss et al published<br />
the first results <strong>of</strong> a randomized placebo-controlled trial<br />
designed to investigate exemestane for the primary prevention<br />
<strong>of</strong> invasive breast cancer. The National Cancer Institute<br />
<strong>of</strong> Canada <strong>Clinical</strong> Trials Group MAP.3 randomly selected<br />
4,560 postmenopausal women to receive exemestane or<br />
resources to determine patient eligibility for chemoprevention<br />
and lack the time to provide informed counseling to their<br />
patients. The barriers for patients’ acceptance <strong>of</strong> chemoprevention<br />
treatment include fear <strong>of</strong> side effects and the perception<br />
that chemoprevention will not substantially lower their<br />
risk <strong>of</strong> developing breast cancer. Despite these challenges,<br />
there are substantial opportunities to increase the utilization<br />
<strong>of</strong> chemoprevention. These strategies include education, dissemination<br />
<strong>of</strong> user-friendly risk–benefit models, and the support<br />
<strong>of</strong> research efforts focused on identifying biomarkers that<br />
can more accurately select women most likely to develop<br />
breast cancer and predict responsiveness <strong>of</strong> treatment.<br />
placebo. The inclusion criteria were similar to the BCPT and<br />
STAR trials and included a 5-year predicted risk for breast<br />
cancer <strong>of</strong> 1.66% or higher, history <strong>of</strong> atypia, or lobular<br />
carcinoma in situ (LCIS). Women with ductal cancer in situ<br />
were also eligible but represented a small portion <strong>of</strong> the<br />
study participants (2.5%). At a median follow-up time <strong>of</strong> 35<br />
months, exemestane was found to significantly reduce the<br />
annual risk <strong>of</strong> invasive breast cancer by 65% (p � 0.002). 4<br />
Arthritis (p � 0.01) and hot flashes (p ��0.001) were more<br />
common in the exemestane treatment group, but there was<br />
no statistically significant difference between the groups<br />
in rates <strong>of</strong> new diagnoses <strong>of</strong> osteoporosis or cardiovascular<br />
disease. Another phase III trial, IBIS-II, has randomly<br />
selected 6,640 high-risk women to receive either placebo or<br />
anastrozole and the results are eagerly awaited.<br />
Evaluating Risk–Benefit Indices for Chemoprevention<br />
The FDA-approved indication to reduce the incidence <strong>of</strong><br />
invasive breast cancer for tamoxifen includes both premenopausal<br />
and postmenopausal women. However, raloxifene—<br />
which was FDA approved in 2007—is indicated for postmenopausal<br />
women with osteoporosis or at high risk <strong>of</strong><br />
breast cancer based on the Gail model 5-year projected risk<br />
<strong>of</strong> 1.66% or higher. Adverse events that are associated with<br />
these agents raise concerns among women and physicians<br />
about the benefits and risks, especially in women at high<br />
risk <strong>of</strong> developing breast cancer who are otherwise healthy.<br />
In addition to chronic diseases that accompany increasing<br />
age, the risks <strong>of</strong> endometrial cancer, venous thromboembolic<br />
events, bone fractures, and cataracts complicate the counseling<br />
for tamoxifen and raloxifene.<br />
In 1998, the National Cancer Institute convened a workshop<br />
to determine the best ways <strong>of</strong> communicating the<br />
results <strong>of</strong> the BCPT. The primary goal <strong>of</strong> this workshop was<br />
From the University <strong>of</strong> Texas M. D. Anderson Cancer Center, Houston, TX; University <strong>of</strong><br />
Pittsburgh Cancer Institute, Pittsburgh, PA; and National Cancer Institute, Bethesda, MD.<br />
Authors’ disclosures <strong>of</strong> potential conflicts <strong>of</strong> interest are found at the end <strong>of</strong> this article.<br />
Address reprint requests to Abenaa Brewster, MD, MHS, University <strong>of</strong> Texas M. D.<br />
Anderson Cancer Center, 1155 Herman P. Pressler, PO Box 301439, Houston, TX 77230;<br />
email: abrewster@mdanderson.org.<br />
© <strong>2012</strong> by <strong>American</strong> <strong>Society</strong> <strong>of</strong> <strong>Clinical</strong> <strong>Oncology</strong>.<br />
1092-9118/10/1-10<br />
85
to develop a tool to identify those women whose benefits<br />
from the use <strong>of</strong> tamoxifen outweighs the risks in reducing<br />
the incidence <strong>of</strong> breast cancer. Of particular interest was the<br />
need to explore the influence <strong>of</strong> tamoxifen on those health<br />
outcomes that disproportionately affect different populations.<br />
Such a tool was developed by Gail et al, which weighed<br />
the benefits and risk <strong>of</strong> other health outcomes in the absence<br />
<strong>of</strong> tamoxifen and the effects <strong>of</strong> tamoxifen on other health<br />
outcomes. 5 The benefits and risks for the use <strong>of</strong> tamoxifen<br />
were described in the format <strong>of</strong> tables by race, age, presence<br />
or absence <strong>of</strong> a uterus, and the projected 5-year risk <strong>of</strong><br />
invasive breast cancer. Net benefits increased with increasing<br />
risk and decreased with age and in postmenopausal<br />
women with an intact uterus. Younger women with elevated<br />
risks <strong>of</strong> breast cancer had the best net benefit for tamoxifen<br />
use. Importantly, this study showed that the adverse events<br />
varied by age and race, preferences by the women, and<br />
varied weights for specific events were critical complements<br />
for weighing risk and benefits <strong>of</strong> tamoxifen.<br />
Limitations <strong>of</strong> the 1999 Gail et al data included the lack<br />
<strong>of</strong> incidence rates for stroke and venous thromboembolic<br />
events from populations other than the non-Hispanic white<br />
population and the uncertainty <strong>of</strong> applying the estimated<br />
benefits and risks to all populations. A comparison <strong>of</strong> the<br />
benefits and risks <strong>of</strong> tamoxifen and raloxifene for postmenopausal<br />
women to assist in communication was identified as<br />
a research and clinical need. In addition, the availability <strong>of</strong><br />
86<br />
KEY POINTS<br />
● Evidence from randomized, placebo-controlled clinical<br />
trials supports the use <strong>of</strong> chemoprevention for<br />
reducing the risk <strong>of</strong> primary invasive breast cancer,<br />
and there are three options: tamoxifen, raloxifene,<br />
and exemestane.<br />
● Risk–benefit models have great potential for decision<br />
making about chemoprevention in high-risk clinical<br />
practices and primary care practices if used in combination<br />
with a full assessment <strong>of</strong> the woman’s health<br />
and preferences.<br />
● Guidelines for the prescription <strong>of</strong> tamoxifen and<br />
raloxifene are provided by the United States Preventive<br />
Services Task Force, <strong>American</strong> <strong>Society</strong> <strong>of</strong> <strong>Clinical</strong><br />
<strong>Oncology</strong>, the Canadian Task Force on Preventive<br />
Health Care, and the National Comprehensive Cancer<br />
Network.<br />
● Only a small portion <strong>of</strong> women who are eligible for<br />
chemoprevention receive treatment, and the barriers<br />
to uptake include lack <strong>of</strong> education about breast<br />
cancer, inadequate physician training in risk assessment,<br />
lack <strong>of</strong> time to provide counseling, concerns<br />
about side effects, cost, and misconceptions about<br />
actual and perceived risks and benefits <strong>of</strong> treatment.<br />
● An active area <strong>of</strong> cancer prevention research is the<br />
identification <strong>of</strong> biomarkers and clinical features to<br />
improve the selection <strong>of</strong> women most likely to benefit<br />
from breast cancer chemoprevention and to predict<br />
and monitor treatment responsiveness.<br />
BREWSTER, DAVIDSON, AND STEVENS<br />
improved incidence rates for invasive cancer for Hispanic<br />
women from the Surveillance, Epidemiology, and End Results<br />
(SEER), modifications to the Gail model prediction tool<br />
for the black population, and baseline incidence rates for<br />
other health outcomes from the Women’s Health Initiative<br />
could enrich a risk–benefit analysis <strong>of</strong> tamoxifen and raloxifene.<br />
Freedman et al evaluated the probability <strong>of</strong> having a<br />
health event in 5 years in the absence and presence <strong>of</strong><br />
tamoxifen and raloxifene. 6 Risks and benefits were defined<br />
as life-threatening (invasive breast cancer, hip fractures,<br />
endometrial cancer, stroke, and pulmonary emboli) and<br />
severe events (in situ breast cancer and deep vein thrombosis).<br />
An index <strong>of</strong> expected number <strong>of</strong> life-threatening events<br />
without chemoprevention in 10,000 women minus the expected<br />
number <strong>of</strong> life-threatening events if tamoxifen and<br />
raloxifene were used was presented in color-coded tables.<br />
The authors used a complex statistical modeling approach to<br />
simulate weighted outcomes to provide levels <strong>of</strong> evidence for<br />
risk estimation. Strong evidence <strong>of</strong> a positive net benefit <strong>of</strong><br />
tamoxifen or raloxifene compared with no treatment was<br />
deemed to exist if the net benefit was positive in 90% or<br />
greater <strong>of</strong> replications in Bayesian boot-strap testing for<br />
variability. Freedman et al showed that raloxifene results in<br />
a better risk–benefit pr<strong>of</strong>ile for black, Hispanic, and non-<br />
Hispanic white postmenopausal women with a uterus when<br />
compared with tamoxifen. Similar benefits for postmenopausal<br />
black and non-Hispanic white women without a<br />
uterus were seen for tamoxifen and raloxifene.<br />
There are several limitations <strong>of</strong> the tables presented by<br />
Freedman et al. Since the Gail model performs poorly at<br />
predicting risk <strong>of</strong> breast cancer among women at higher risk<br />
levels, generalization <strong>of</strong> the results to these women is<br />
limited. In addition, the benefit <strong>of</strong> raloxifene over tamoxifen<br />
may continue to diminish with longer-term follow-up and<br />
the time line <strong>of</strong> the analysis was limited to 5 years. 7 Despite<br />
these limitations, risk–benefit models have great potential<br />
<strong>of</strong> assisting cancer care providers in high-risk clinical practices<br />
and primary care practices if used in combination with<br />
a full assessment <strong>of</strong> the woman’s health and preferences.<br />
Guidelines for Breast Cancer Chemoprevention<br />
In 2002, the United States Preventive Services Task Force<br />
(USPSTF) issued a recommendation statement that women<br />
who are at both high risk <strong>of</strong> developing breast cancer and<br />
low risk for adverse events should be counseled about<br />
chemoprevention for breast carcinoma. 8 Chemoprevention<br />
was not recommended for women at low or average risk for<br />
whom the adverse health events associated with tamoxifen<br />
or raloxifene may outweigh the benefits. Although no set<br />
definition <strong>of</strong> “high risk” was specified, the USPSTF referred<br />
to the entry eligibility for the BCPT, which was a predicted<br />
5-year risk <strong>of</strong> greater than 1.67% defined using the Gail<br />
model. 1 This guideline for the prescription <strong>of</strong> tamoxifen and<br />
raloxifene has been supported by the <strong>American</strong> <strong>Society</strong> <strong>of</strong><br />
<strong>Clinical</strong> <strong>Oncology</strong> (ASCO), the Canadian Task Force on<br />
Preventive Health Care, and the National Comprehensive<br />
Cancer Network (NCCN). 9-11 Based on the results <strong>of</strong> the<br />
MAP.3 trial, exemestane was included as one <strong>of</strong> the choices<br />
for breast cancer chemoprevention by the NCCN panel <strong>of</strong><br />
experts. Their recommendation for exemestane was limited<br />
to postmenopausal women age 35 or older with a Gail model<br />
5-year predicted risk higher than 1.66% or a history <strong>of</strong> LCIS.
CHEMOPREVENTION FOR BREAST CANCER<br />
There are no data directly comparing the benefits and risks<br />
<strong>of</strong> exemestane with those <strong>of</strong> tamoxifen and raloxifene and it<br />
is not currently FDA approved for primary breast cancer<br />
risk reduction.<br />
Uptake <strong>of</strong> Breast Cancer Chemoprevention in the<br />
United States<br />
Despite clinical guidelines recommending tamoxifen or<br />
raloxifene for the risk reduction <strong>of</strong> primary breast cancer,<br />
the majority <strong>of</strong> eligible women elect not to use chemoprevention.<br />
It is estimated that approximately 10 million women<br />
(16%) would be eligible for chemoprevention and approximately<br />
2.4 million would have a positive risk–benefit pr<strong>of</strong>ile<br />
for tamoxifen treatment. 12 Unfortunately the uptake <strong>of</strong><br />
breast chemoprevention has been substantially less than<br />
predicted. Waters et al used data from the National Health<br />
Interview Survey between 2000 and 2005 to estimate the<br />
number <strong>of</strong> women taking tamoxifen for chemoprevention. 13<br />
In 2000, two years after the FDA approved tamoxifen for<br />
breast cancer chemoprevention, only 0.2% <strong>of</strong> women ages 40<br />
to 79 had taken tamoxifen to prevent a diagnosis <strong>of</strong> breast<br />
cancer. In 2005, 0.08% <strong>of</strong> women had taken tamoxifen,<br />
which represented a 50% decrease in the number <strong>of</strong> women<br />
receiving tamoxifen from 120,000 in 2000 to 60,000 in<br />
2005. 13 Armstrong et al surveyed 350 primary care physicians<br />
between 2002 and 2004 and found that only 27% had<br />
prescribed tamoxifen for breast cancer prevention in the<br />
previous year, which was higher than would have been<br />
predicted based on the prevalence <strong>of</strong> women reporting taking<br />
the medication. 14 There are no available data available<br />
on the uptake <strong>of</strong> raloxifene, however, the general perception<br />
is that public acceptance remains low for reasons that<br />
remain unclear given its lower toxicity pr<strong>of</strong>ile compared<br />
with tamoxifen. 15<br />
Physician Barriers to Prescribing Breast Cancer<br />
Chemoprevention<br />
It is well recognized that women are more likely to accept<br />
preventive services if recommended by their physicians. 16-18<br />
The reluctance <strong>of</strong> physicians to prescribe tamoxifen or raloxifene<br />
is therefore likely to be a major reason for the low<br />
uptake <strong>of</strong> chemoprevention. Since primary care physicians<br />
(PCPs) serve as the gatekeepers for implementing screening<br />
recommendations and usually have the first interaction with<br />
women at increased risk <strong>of</strong> developing breast cancer, studies<br />
investigating physicians’ attitudes and acceptance <strong>of</strong> chemoprevention<br />
have focused on this group <strong>of</strong> health care providers.<br />
The PCP role in the preventive care setting is to assess a<br />
woman’s risk <strong>of</strong> developing breast cancer and provide a<br />
recommendation for a risk-reduction option. How a PCP<br />
views a woman’s risk <strong>of</strong> developing breast cancer has been<br />
shown to be an important predictor <strong>of</strong> whether a prescription<br />
<strong>of</strong> tamoxifen will be <strong>of</strong>fered. 19,20 Tchou et al found that<br />
women with a histologic diagnosis <strong>of</strong> atypical hyperplasia<br />
or LCIS and women with a Gail model 5-year predicted risk<br />
<strong>of</strong> 5.0% or greater were more likely to be <strong>of</strong>fered tamoxifen<br />
than women at lower risk. 19 In a random sample <strong>of</strong> 822<br />
Californian PCPs, Haas et al presented hypothetical patient<br />
scenarios to assess how a woman’s breast cancer risk<br />
would influence her physician’s recommendations for riskreduction<br />
options. 20 Although 80% <strong>of</strong> PCPs would recom-<br />
mend mammography screening and counseling for high-risk<br />
women on lifestyle modifications, a minority endorsed the<br />
use <strong>of</strong> tamoxifen. Some <strong>of</strong> the physician-perceived barriers<br />
to risk-reduction counseling included “not sufficiently<br />
trained in counseling techniques” and “not sufficiently informed<br />
about risk-reduction options.” Lack <strong>of</strong> time with<br />
patients was the most frequently cited barrier, and 13% <strong>of</strong><br />
physicians reported insufficient reimbursement as an impediment<br />
to counseling. 21 Interestingly, physicians’ concern<br />
about the side effects <strong>of</strong> chemoprevention has not been<br />
reported to be a significant deterrent to prescribing tamoxifen<br />
or raloxifene. 14 However, more PCPs report prescribing<br />
raloxifene (30%) compared with tamoxifen (10%) for primary<br />
breast cancer risk reduction, which is not surprising since<br />
Freedman et al showed that raloxifene results in a better<br />
risk–benefit pr<strong>of</strong>ile for women with a uterus when compared<br />
with tamoxifen. 6,21 A good understanding <strong>of</strong> how to determine<br />
a woman’s eligibility for tamoxifen and ability to<br />
determine whether the benefits <strong>of</strong> treatment outweigh the<br />
risks is a significant predictor <strong>of</strong> whether a PCP prescribes<br />
tamoxifen. Unfortunately in two surveys <strong>of</strong> PCPs conducted<br />
by Armstrong et al and Sabatino et al, only 15% and 9% <strong>of</strong><br />
PCPs, respectively, felt capable <strong>of</strong> making this clinical assessment,<br />
which highlights the challenges faced by PCPs in<br />
using and interpreting the risk–benefit models that have<br />
been developed. 14,22 Contrary to the experience <strong>of</strong> PCPs, a<br />
survey <strong>of</strong> 851 oncologists revealed that 73% discussed breast<br />
cancer chemoprevention with their patients. 23 The high<br />
percentage <strong>of</strong> oncologists who discuss chemoprevention recommendations<br />
compared with PCPs may reflect differences<br />
in the receipt <strong>of</strong> formal and informal training in cancer<br />
prevention and encounters with women in the high-risk<br />
clinical setting who are highly motivated to reduce their risk<br />
<strong>of</strong> developing breast cancer.<br />
Patient Barriers to Accepting Breast Cancer<br />
Chemoprevention<br />
Few women who meet eligibility criteria based on an<br />
assessment <strong>of</strong> the risks and benefits <strong>of</strong> chemoprevention<br />
treatment elect to initiate treatment. Ropka et al conducted<br />
a meta-analysis <strong>of</strong> studies that evaluated real or hypothetical<br />
decisions made by women about breast cancer chemoprevention.<br />
The mean uptake <strong>of</strong> tamoxifen for the five<br />
studies that reported real decision rates was 14.8% (range,<br />
0.5% to 51.2%) and 24% (range, 5.7% to 60%) for the studies<br />
reporting hypothetical decisions. 24<br />
The most commonly cited reason for not taking tamoxifen<br />
is fear <strong>of</strong> the side effect <strong>of</strong> endometrial cancer, thromboembolic<br />
disease, and menopausal symptoms. 16,25,26 In a study<br />
<strong>of</strong> 43 women who were eligible to take tamoxifen for primary<br />
prevention, the most feared side effects were endometrial<br />
cancer and thromboembolic disease, and the majority <strong>of</strong><br />
women tended to overestimate their risk <strong>of</strong> developing these<br />
side effects even after a presentation <strong>of</strong> the balance <strong>of</strong> risks<br />
and benefits. 25 When asked to estimate the risk <strong>of</strong> developing<br />
a side effect associated with tamoxifen, the majority<br />
perceived that the risk would be 40% or greater. A survey <strong>of</strong><br />
29 women interviewed in a family practice setting revealed<br />
additional concerns about nausea and vomiting, symptoms<br />
mostly attributed to chemotherapy. In this study, the authors<br />
proposed that the name “chemoprevention” might<br />
itself act as a barrier to patient acceptance because it is <strong>of</strong>ten<br />
confused with chemotherapy. 26<br />
87
Previous studies have shown that women tend to overestimate<br />
their risk <strong>of</strong> developing breast cancer and women<br />
with a higher perceived risk <strong>of</strong> breast cancer are more likely<br />
to be accepting <strong>of</strong> chemoprevention. 16,27 Many women, however,<br />
have the perception that chemoprevention will not<br />
substantially lower their risk <strong>of</strong> developing breast cancer<br />
even after receiving information about the 50% risk reduction<br />
associated with tamoxifen treatment. 28,29 Acceptance<br />
<strong>of</strong> chemoprevention treatment has been shown to be significantly<br />
higher among women with a history <strong>of</strong> atypical<br />
hyperplasia or LCIS compared with patients at risk on the<br />
basis <strong>of</strong> other factors (56% vs. 28%, p ��0.0001). 19<br />
Additional barriers that contribute to the reluctance <strong>of</strong><br />
women to use chemoprevention are related to physician<br />
difficulty in accurately selecting individuals most likely to<br />
develop breast cancer and the lack <strong>of</strong> a biomarker that can<br />
be used to monitor the effect <strong>of</strong> the drug on risk. Unlike the<br />
monitoring <strong>of</strong> cardiovascular risk factors (e.g., cholesterol<br />
and blood pressure), which provide a measurable target for<br />
assessing the efficacy <strong>of</strong> cholesterol or hypertension lowering<br />
agents, there is no available reliable biomarker to measure<br />
the preventive effect <strong>of</strong> chemoprevention that may serve to<br />
motivate patients to accept treatment. 15 Salant et al interviewed<br />
women seen at a high-risk clinic, <strong>of</strong> which the<br />
majority were black, and found that the women understood<br />
risk not as a numerical probability or chronic disease state<br />
but as an immediate physical sign or symptom warranting<br />
medical intervention. Therefore, despite meeting criteria for<br />
being high-risk using the Gail model, many women did not<br />
“feel” at high risk and therefore were not interested in<br />
taking breast cancer chemoprevention treatment. 30 This<br />
finding has implications for the support <strong>of</strong> research efforts<br />
aimed at identifying biomarkers and clinical features that<br />
identify high-risk states and individualize the risk prediction<br />
<strong>of</strong> invasive breast cancer as a means <strong>of</strong> improving the<br />
uptake <strong>of</strong> chemoprevention treatment. 31<br />
Among low-income women, acceptance <strong>of</strong> chemoprevention<br />
is dependent also on how much it costs and whether the<br />
cost is covered by health insurance. 28,30 At least three<br />
studies have found that education and lower income are<br />
inversely associated with breast cancer chemoprevention.<br />
27,32,33 These studies did not report on the accuracy <strong>of</strong><br />
the women’s understanding <strong>of</strong> the risks and benefits <strong>of</strong> the<br />
chemoprevention treatments, which is an important consideration<br />
since a greater understanding <strong>of</strong> the risk–benefit<br />
pr<strong>of</strong>ile <strong>of</strong> chemoprevention has been demonstrated to result<br />
in decreased patient acceptance <strong>of</strong> chemoprevention treatment.<br />
24 To overcome the significant patient-related barriers<br />
to the uptake <strong>of</strong> chemoprevention, a comprehensive strategy<br />
is needed that educates women about breast cancer and<br />
their competing health risks and incorporates easily accessible<br />
decision aids that accurately convey the risks and<br />
benefits <strong>of</strong> chemoprevention treatment.<br />
Strategies to Improve the Uptake <strong>of</strong> Breast Cancer<br />
Chemoprevention<br />
Several investigators have explored the use <strong>of</strong> educational<br />
or decision-making aids to increase the uptake <strong>of</strong> breast<br />
cancer chemoprevention. 34,35 The goal <strong>of</strong> the decision aids is<br />
to facilitate a discussion between the patient and provider<br />
about breast cancer chemoprevention and help guide informed<br />
decisions. 36 Key components <strong>of</strong> the decision aids<br />
88<br />
BREWSTER, DAVIDSON, AND STEVENS<br />
have included education about breast cancer, an assessment<br />
<strong>of</strong> the patient’s risk <strong>of</strong> breast cancer using the Gail or Claus<br />
risk assessment models, integration <strong>of</strong> the patient’s risk <strong>of</strong><br />
breast cancer within the context <strong>of</strong> population risk, and an<br />
assessment <strong>of</strong> the change in risks and benefits that would be<br />
expected with tamoxifen or raloxifene treatment. 34-36 In two<br />
studies that have evaluated patient willingness to take<br />
chemoprevention after receiving a decision aid, the uptake<br />
rates have been disappointingly low (range, 4% to 6%). 34,35<br />
In a randomized trial designed to test whether a decision aid<br />
intervention increased a woman’s interest in chemoprevention,<br />
among the 6% <strong>of</strong> high-risk women who received the<br />
decision aid and said that they were likely to take the drug,<br />
less than 1% had initiated therapy after the intervention. 35<br />
The challenges for PCPs considering prescribing breast<br />
cancer chemoprevention are identifying women eligible for<br />
treatment based on an acceptable risk–benefit pr<strong>of</strong>ile and<br />
communicating the net benefit estimate using a balanced<br />
approach that allows the patient to consider her personal<br />
preferences. Increased PCP awareness <strong>of</strong> the online Gail<br />
risk model and access to the risk–benefit tools for tamoxifen<br />
that were published in 1999 and updated in 2011 to include<br />
raloxifene are needed to facilitate this process. It has been<br />
advocated that the risk–benefit models for breast cancer<br />
chemoprevention should be more interactive and available<br />
online, accessible to both physicians and patients. 37 In<br />
addition, the risk–benefit models should incorporate additional<br />
variables, such as 5-year and lifetime mortality,<br />
comorbidities including menopausal symptoms, and competing<br />
health risks that are essential to an informed decisionmaking<br />
discussion about whether to use chemoprevention<br />
treatment. 37,38 To increase awareness and education about<br />
cancer preventive and control strategies, the ASCO Cancer<br />
Prevention Committee is in the process <strong>of</strong> developing curricular<br />
material to disseminate to the educational programs<br />
<strong>of</strong> a spectrum <strong>of</strong> disciplines, including family practice and<br />
internal medicine. A recent breast cancer consensus statement<br />
recommended that to advance efforts for breast cancer<br />
prevention, the term “chemoprevention” should be replaced<br />
with the term “preventive therapy” to remove its inappropriate<br />
association with chemotherapy. 39<br />
A major hurdle in promoting breast cancer chemoprevention<br />
to the general population is the recognition that many<br />
women who receive chemoprevention treatment will never<br />
have developed breast cancer anyway. Because <strong>of</strong> the poor<br />
discriminatory accuracy <strong>of</strong> the Gail risk model in determining<br />
individual level risk <strong>of</strong> breast cancer, many women and<br />
their physicians remain uncertain about the benefits <strong>of</strong><br />
treatment and concerned about the potential serious toxicities<br />
and side effects that may affect quality <strong>of</strong> life. In<br />
addition, advocacy groups including the Breast Cancer National<br />
Coalition are skeptical <strong>of</strong> chemoprevention and express<br />
concern about the lack <strong>of</strong> data on the long term<br />
side-effects <strong>of</strong> the chemoprevention drugs. 40 Strategies to<br />
refine risk and predict responsiveness to chemoprevention<br />
treatment are being explored and include random periareolar<br />
fine-needle aspiration to evaluate for cellular atypia in<br />
otherwise normal breast tissue, breast density as a surrogate<br />
marker for monitoring response to treatment, and<br />
single nucleotide polymorphisms associated with increased<br />
susceptibility. 36,39
CHEMOPREVENTION FOR BREAST CANCER<br />
Conclusion<br />
Breast cancer is the most common malignancy among<br />
women in the United States affecting approximately<br />
225,000 women annually. Evidence from randomized,<br />
placebo-controlled clinical trials supports the use <strong>of</strong> chemoprevention<br />
for reducing the risk <strong>of</strong> primary invasive breast<br />
cancer, and there are three options: tamoxifen, raloxifene,<br />
and exemestane. Risk–benefit models indicate that raloxifene<br />
results in a better risk–benefit pr<strong>of</strong>ile for black, Hispanic,<br />
and non-Hispanic white postmenopausal women with<br />
a uterus when compared with tamoxifen. Similar benefits<br />
for postmenopausal black and non-Hispanic white women<br />
without a uterus are seen for tamoxifen and raloxifene. Only<br />
a small portion <strong>of</strong> women who are eligible for chemopreven-<br />
Authors’ Disclosures <strong>of</strong> Potential Conflicts <strong>of</strong> Interest<br />
Employment or<br />
Leadership<br />
Positions<br />
Consultant or<br />
Advisory Role<br />
Author<br />
Abenaa M. Brewster*<br />
Nancy E. Davidson GlaxoSmithKline<br />
(U)<br />
Worta McCaskill-Stevens*<br />
*No relevant relationships to disclose.<br />
1. Fisher B, Costantino JP, Wickerham DL, et al. Tamoxifen for prevention<br />
<strong>of</strong> breast cancer: report <strong>of</strong> the National Surgical Adjuvant Breast and Bowel<br />
Project P-1 Study. J Natl Cancer Inst. 1998;90:1371-1388.<br />
2. Vogel VG, Costantino JP, Wickerham DL, et al. Update <strong>of</strong> the National<br />
Surgical Adjuvant Breast and Bowel Project Study <strong>of</strong> Tamoxifen and Raloxifene<br />
(STAR) P-2 Trial: Preventing breast cancer. Cancer Prev Res (Phila).<br />
2010;3:696-706.<br />
3. Nelson HD, Fu R, Griffin JC, et al. Systematic review: comparative<br />
effectiveness <strong>of</strong> medications to reduce risk for primary breast cancer. Ann<br />
Intern Med. 2009;151:703-715.<br />
4. Goss PE, Ingle JN, Ales-Martinez JE, et al. Exemestane for breastcancer<br />
prevention in postmenopausal women. N Engl J Med. 2011;364:2381-<br />
2391.<br />
5. Gail MH, Costantino JP, Bryant J, et al. Weighing the risks and benefits<br />
<strong>of</strong> tamoxifen treatment for preventing breast cancer. J Natl Cancer Inst.<br />
1999;91:1829-1846.<br />
6. Freedman AN, Yu B, Gail MH, et al. Benefit/risk assessment for breast<br />
cancer chemoprevention with raloxifene or tamoxifen for women age 50 years<br />
or older. J Clin Oncol. 2011;29:2327-2333.<br />
7. Amir E, Goodwin P. Breast cancer chemoprevention gets personal. J<br />
Clin Oncol. 2011;29:2296-2298.<br />
8. U.S. Preventive Services Task Force. Chemoprevention <strong>of</strong> breast cancer:<br />
recommendations and rationale. Ann Intern Med. 2002;137:56-58.<br />
9. Visvanathan K, Chlebowski R, Hurley P. <strong>American</strong> <strong>Society</strong> <strong>of</strong> <strong>Clinical</strong><br />
<strong>Oncology</strong> practice guideline update on the use <strong>of</strong> pharmacologic interventions<br />
including tamoxifen, raloxifene, and aromatase inhibition for breast cancer<br />
risk reduction. J Clin Oncol. 1999;27:3235-3258.<br />
10. Levine M, Moutquin JM, Walton R, et al. Chemoprevention <strong>of</strong> breast<br />
cancer. A joint guideline from the Canadian Task Force on Preventive Health<br />
Care and the Canadian Breast Cancer Initiative’s Steering Committee on<br />
<strong>Clinical</strong> Practice Guidelines for the Care and Treatment <strong>of</strong> Breast Cancer.<br />
CMAJ. 2001;164:1681-1690.<br />
11. National Comprehensive Cancer Network. NCCN clinical practice<br />
guidelines in oncology: Breast cancer risk reduction, v. 3.2011. http://www.<br />
nccn.org/pr<strong>of</strong>essionals/physician_gls/f_guidelines.asp#detection. Accessed February<br />
13, <strong>2012</strong>.<br />
12. Freedman AN, Graubard BI, Rao SR, et al. Estimates <strong>of</strong> the number <strong>of</strong><br />
US women who could benefit from tamoxifen for breast cancer chemoprevention.<br />
J Natl Cancer Inst. 2003;95:526-532.<br />
13. Waters EA, Cronin KA, Graubard BI, et al. Prevalence <strong>of</strong> tamoxifen use<br />
for breast cancer chemoprevention among U.S. women. Cancer Epidemiol<br />
Biomarkers Prev. 2010;19:443-446.<br />
14. Armstrong K, Quistberg DA, Micco E, et al. Prescription <strong>of</strong> tamoxifen<br />
tion receive treatment, and the barriers to uptake include<br />
lack <strong>of</strong> education about breast cancer, inadequate physician<br />
training in risk assessment, lack <strong>of</strong> time to provide counseling,<br />
concerns about side effects, cost, and misconceptions<br />
about actual and perceived risks and benefits <strong>of</strong> treatment.<br />
Strategies to close this gap have focused on educating<br />
women about breast cancer and their competing health risks<br />
as well as developing easily accessible decision aids and risk<br />
assessment models that accurately convey the risks and<br />
benefits <strong>of</strong> chemoprevention treatment. An active area <strong>of</strong><br />
research is the identification <strong>of</strong> biomarkers and clinical<br />
features to improve the selection <strong>of</strong> women most likely to<br />
benefit from chemoprevention and to predict and monitor<br />
responsiveness <strong>of</strong> treatment.<br />
Stock<br />
Ownership Honoraria<br />
REFERENCES<br />
Research<br />
Funding<br />
Breast Cancer<br />
Research<br />
Foundation; NIH<br />
Expert<br />
Testimony<br />
Other<br />
Remuneration<br />
for breast cancer prevention by primary care physicians. Arch Intern Med.<br />
2006;166:2260-2265.<br />
15. Meyskens FL Jr., Curt GA, Brenner DE, et al. Regulatory approval <strong>of</strong><br />
cancer risk-reducing (chemopreventive) drugs: Moving what we have learned<br />
into the clinic. Cancer Prev Res (Phila). 2011;4:311-323.<br />
16. Bober SL, Hoke LA, Duda RB, et al. Decision-making about tamoxifen<br />
in women at high risk for breast cancer: clinical and psychological factors.<br />
J Clin Oncol. 2004;22:4951-4957.<br />
17. Lerman C, Rimer B, Trock B, et al. Factors associated with repeat<br />
adherence to breast cancer screening. Prev Med. 1990;19:279-290.<br />
18. Kinney AY, Richards C, Vernon SW, et al. The effect <strong>of</strong> physician<br />
recommendation on enrollment in the Breast Cancer Chemoprevention Trial.<br />
Prev Med. 1998;27:713-719.<br />
19. Tchou J, Hou N, Rademaker A, et al. Acceptance <strong>of</strong> tamoxifen chemoprevention<br />
by physicians and women at risk. Cancer. 2004;100:1800-1806.<br />
20. Haas JS, Kaplan CP, Gregorich SE, et al. Do physicians tailor their<br />
recommendations for breast cancer risk reduction based on patient’s risk?<br />
J Gen Intern Med. 2004;19:302-309.<br />
21. Kaplan CP, Haas JS, Perez-Stable EJ, et al. Factors affecting breast<br />
cancer risk reduction practices among California physicians. Prev Med.<br />
2005;41:7-15.<br />
22. Sabatino SA, McCarthy EP, Phillips RS, et al. Breast cancer risk<br />
assessment and management in primary care: provider attitudes, practices,<br />
and barriers. Cancer Detect Prev. 2007;31:375-383.<br />
23. Ganz PA, Kwan L, Somerfield MR, et al. The role <strong>of</strong> prevention in<br />
oncology practice: results from a 2004 survey <strong>of</strong> <strong>American</strong> <strong>Society</strong> <strong>of</strong> <strong>Clinical</strong><br />
<strong>Oncology</strong> members. J Clin Oncol. 2006;24:2948-2957.<br />
24. Ropka ME, Keim J, Philbrick JT. Patient decisions about breast cancer<br />
chemoprevention: a systematic review and meta-analysis. J Clin Oncol.<br />
2010;28:3090-3095.<br />
25. Port ER, Montgomery LL, Heerdt AS, et al. Patient reluctance toward<br />
tamoxifen use for breast cancer primary prevention. Ann Surg Oncol. 2001;<br />
8:580-585.<br />
26. Heisey R, Pimlott N, Clemons M, et al. Women’s views on chemoprevention<br />
<strong>of</strong> breast cancer: qualitative study. Can Fam Physician. 2006;52:624-<br />
625.<br />
27. Melnikow J, Paterniti D, Azari R, et al. Preferences <strong>of</strong> Women Evaluating<br />
Risks <strong>of</strong> Tamoxifen (POWER) study <strong>of</strong> preferences for tamoxifen for<br />
breast cancer risk reduction. Cancer. 2005;103:1996-2005.<br />
28. Cyrus-David MS, Strom SS. Chemoprevention <strong>of</strong> breast cancer with<br />
selective estrogen receptor modulators: views from broadly diverse focus<br />
groups <strong>of</strong> women with elevated risk for breast cancer. Psychooncology.<br />
2001;10:521-533.<br />
89
29. Fagerlin A, Zikmund-Fisher BJ, Nair V, et al. Women’s decisions<br />
regarding tamoxifen for breast cancer prevention: responses to a tailored<br />
decision aid. Breast Cancer Res Treat. 2010;119:613-620.<br />
30. Salant T, Ganschow PS, Olopade OI, et al. “Why take it if you don’t<br />
have anything?” breast cancer risk perceptions and prevention choices at a<br />
public hospital. J Gen Intern Med. 2006;21:779-785.<br />
31. Temple R. Cancer chemoprevention—the cardiovascular model. Cancer<br />
Prev Res. 2011;4:307-310.<br />
32. Fasching PA, von Minckwitz G, Fischer T, et al. The impact <strong>of</strong> breast<br />
cancer awareness and socioeconomic status on willingness to receive breast<br />
cancer prevention drugs. Breast Cancer Res Treat. 2007;101:95-104.<br />
33. Tija J, Micco E, Armstrong K. Interest in breast cancer chemoprevention<br />
among older women. Breast Cancer Res Treat. 2008;108:435-453.<br />
34. McKay A, Martin W, Latosinsky S. How should we inform women at<br />
higher risk <strong>of</strong> breast cancer about tamoxifen? An approach with a decision<br />
guide. Breast Cancer Res Treat. 2005;94:153-159.<br />
90<br />
BREWSTER, DAVIDSON, AND STEVENS<br />
35. Fagerlin A, Dillard AJ, Smith DM, et al. Women’s interest in taking<br />
tamoxifen and raloxifene for breast cancer prevention: response to a tailored<br />
decision aid. Breast Cancer Res Treat. 2011;127:681-688.<br />
36. Ozanne EM, Klemp JR, Esserman LJ. Breast cancer risk assessment<br />
and prevention: a framework for shared decision-making consultations.<br />
Breast J. 2006;12:103-113.<br />
37. Ravdin PM. The lack, need, and opportunities for decision-making and<br />
informational tools to educate primary-care physicians and women about<br />
breast cancer chemoprevention. Cancer Prev Res (Phila). 2010;3:686-688.<br />
38. Mulley AG, Sepucha K. Making good decisions about breast cancer<br />
chemoprevention. Ann Intern Med. 2002;137:52-54.<br />
39. Cuzik J, DeCensi A, Arun B. Preventive therapy for breast cancer: a<br />
consensus statement. Lancet Oncol. 2011;12:496-503.<br />
40. The National Breast Cancer Coalition position statement on chemoprevention.<br />
Updated June 2011. www.breastcancerdeadline2020.org/know/<br />
assets/... /chemoprevention.pdf. Accessed March 7, <strong>2012</strong>.
CONTROVERSIES IN PROSTATE CANCER:<br />
PSA SCREENING, CHEMOPREVENTION, AND<br />
TREATMENT OF EARLY DISEASE<br />
CHAIR<br />
Barnett S. Kramer, MD, MPH<br />
National Cancer Institute<br />
Rockville, MD<br />
SPEAKERS<br />
Gerald L. Andriole, MD<br />
Washington University School <strong>of</strong> Medicine<br />
St. Louis, MO<br />
Timothy J. Wilt, MD, MPH<br />
U. S. Department <strong>of</strong> Veterans Affairs<br />
Minneapolis, MN
The Case for Prostate Cancer Risk Reduction<br />
by 5-Alpha Reductase Inhibitors<br />
By Youssef S. Tanagho, MD, MPH, and Gerald L. Andriole, MD<br />
Overview: Prostate cancer remains a significant public<br />
health problem. The current approach with prostate-specific<br />
antigen (PSA)-based screening has questionable effects on<br />
prostate cancer–specific mortality but is clearly associated<br />
with overdiagnosis <strong>of</strong> prostate cancer, especially relatively<br />
low-risk and low-volume tumors. Methods to decrease overdiagnosis<br />
include alterations in screening practices and, potentially,<br />
the use <strong>of</strong> 5-alpha reductase inhibitors. This article<br />
reviews the major trials that have evaluated 5-alpha reductase<br />
inhibitors in this setting: the Prostate Cancer Prevention Trial<br />
(PCPT) and the Reduction by Dutasteride Prostate Cancer<br />
Events Trial (REDUCE). Although these trials enrolled different<br />
PROSTATE CANCER risk reduction is a potentially<br />
valuable strategy, as current approaches with PSAbased<br />
screening, which have a potentially small effect on<br />
prostate cancer mortality, are associated with a significant<br />
risk <strong>of</strong> overdiagnosis <strong>of</strong> prostate cancer. 1-3 Most men diagnosed<br />
with PSA-detected tumors in the United States receive<br />
aggressive treatments. 4,5 These treatments <strong>of</strong>ten have<br />
significant side effects and are costly in both human and<br />
economic terms. Therefore, one plausible strategy to improve<br />
the efficacy <strong>of</strong> PSA-based screening is to consider the<br />
use <strong>of</strong> 5-alpha reductase inhibitors, as they may reduce<br />
overdiagnosis <strong>of</strong> prostate cancer and improve the utility <strong>of</strong><br />
PSA as a marker for aggressive disease, 6-9 as well as reduce<br />
adverse sequelae <strong>of</strong> BPH.<br />
Overdiagnosis <strong>of</strong> Prostate Cancer<br />
Welch and Black 10 identified three factors that result in<br />
overdiagnosis <strong>of</strong> prostate cancer: 1) the existence <strong>of</strong> a silent<br />
disease reservoir; 2) the use <strong>of</strong> activities (such as screening)<br />
that lead to its detection; and 3) tumors with a long natural<br />
history and, hence, limited cancer-specific mortality. Prostate<br />
cancer may represent a “textbook” disease for overdiagnosis.<br />
Multiple studies have shown a high prevalence <strong>of</strong> this<br />
disease on autopsy, ranging from approximately 30% <strong>of</strong> men<br />
in their 30s to up to 80% <strong>of</strong> men in their 80s. Thus, there is<br />
a large silent reservoir <strong>of</strong> this disease that could be clinically<br />
detected. 11-12 Second, it has been estimated that close to<br />
70% <strong>of</strong> <strong>American</strong> men have undergone one or more PSA<br />
tests, and evidence suggests that older men, those in their<br />
70s and 80s, who stand to benefit the least from PSA-based<br />
screening represent the age stratum most likely to be<br />
screened. 13 Moreover, use <strong>of</strong> relatively low PSA thresholds<br />
as a guide to select patients for biopsy could result in the<br />
diagnosis <strong>of</strong> prostate cancer in a high proportion <strong>of</strong> men in<br />
their 60s to 80s. 14 It has been estimated that 6 to 10 times as<br />
many men are considered to have an abnormal PSA than are<br />
destined to die <strong>of</strong> prostate cancer. Finally, most cases <strong>of</strong><br />
prostate cancer have a long natural history and a limited<br />
cancer-specific mortality. Albertsen and colleagues 15 looked<br />
at survival among men with localized prostate cancer. Men<br />
with Gleason score 8 to 10 disease and no comorbidities had<br />
more than twice the chance <strong>of</strong> dying <strong>of</strong> other causes than <strong>of</strong><br />
prostate cancer within 10 years. Men who had one or more<br />
significant comorbidities and Gleason score 8 to 10 disease<br />
92<br />
patient populations, their findings are complementary and<br />
suggest a potential role for these agents in prostate cancer<br />
risk reduction. Use <strong>of</strong> 5-alpha reductase inhibitors results in<br />
an approximate 25% reduction in the detection <strong>of</strong> prostate<br />
cancer, reduces diagnosis <strong>of</strong> high-grade prostatic intraepithelial<br />
neoplasia (PIN), and improves benign prostatic hyperplasia<br />
(BPH)-related outcomes and the performance <strong>of</strong> PSA as a<br />
diagnostic test for aggressive prostate cancer. Side effects<br />
occur in a small percentage <strong>of</strong> men and consist <strong>of</strong> decreased<br />
sexual function and libido as well as gynecomastia. The risk <strong>of</strong><br />
high-grade tumor development while receiving these agents is<br />
uncertain.<br />
were about 5 times as likely to have a nonprostate cancer–<br />
related death. A contemporary clinical trial, the Göteborg<br />
screening study, demonstrated that in a screened population<br />
<strong>of</strong> about 20,000 Scandinavian men—a population demographically<br />
predisposed to a relatively high overall death<br />
rate from prostate cancer—only 122 died <strong>of</strong> prostate cancer<br />
and more than 3,800 died <strong>of</strong> other causes at 14 years <strong>of</strong><br />
follow-up. 3 Similarly, <strong>of</strong> 367 men with screen-detected, localized<br />
prostate cancers who were considered candidates for<br />
radical prostatectomy in the observation arm <strong>of</strong> the U.S.<br />
PIVOT trial, 31 died <strong>of</strong> prostate cancer, whereas 152 died <strong>of</strong><br />
other causes at a median 10-year follow-up. 16<br />
Estimates <strong>of</strong> the magnitude <strong>of</strong> overdiagnosis <strong>of</strong> prostate<br />
cancer can be made by comparing the screened to the “usual<br />
care” arms <strong>of</strong> randomized screening trials. The PLCO trial,<br />
which enrolled men across 10 participating U.S. centers,<br />
showed that the screened arm had a 17% (95% CI: 11–22)<br />
increased chance <strong>of</strong> diagnosis <strong>of</strong> prostate cancer compared<br />
with the usual care arm at 7 to 10 years <strong>of</strong> follow-up. 4<br />
However, it is known that the usual care arm for PLCO was<br />
heavily contaminated with PSA testing. If one were to<br />
compare the usual care arm <strong>of</strong> PLCO to a matched contemporary<br />
population in the United States, the usual care arm<br />
would have a 22% excess prostate cancer detection rate.<br />
Furthermore, men in the usual care arm <strong>of</strong> PLCO had more<br />
than twice as many prostate cancers discovered as would<br />
have been otherwise anticipated for a similar group <strong>of</strong> men<br />
in the pre-PSA era. 17 In ERSPC, a multicenter European<br />
trial, there was a 71% increase in the detection <strong>of</strong> prostate<br />
cancer in the screened arm compared with the usual care<br />
arm, and a 64% increase was noted in the Göteborg trial. In<br />
most European sites, screening was performed at a 4-year<br />
interval, whereas in Scandinavia, it generally occurred at a<br />
2-year interval. One illustration <strong>of</strong> the effect <strong>of</strong> overdiagnosis<br />
<strong>of</strong> prostate cancer was <strong>of</strong>fered by Caroll and colleagues, 18<br />
who used the Göteborg data to demonstrate that if 1,000<br />
From the Division <strong>of</strong> Urology, Washington University School <strong>of</strong> Medicine, St. Louis, MO.<br />
Authors’ disclosures <strong>of</strong> potential conflicts <strong>of</strong> interest are found at the end <strong>of</strong> this article.<br />
Address reprint requests to Gerald L. Andriole, MD, Division <strong>of</strong> Urologic Surgery,<br />
Washington University in St. Louis, 4960 Children’s Place, Campus Box 8242, St. Louis,<br />
MO 63110; email: andrioleg@wustl.edu.<br />
© <strong>2012</strong> by <strong>American</strong> <strong>Society</strong> <strong>of</strong> <strong>Clinical</strong> <strong>Oncology</strong>.<br />
1092-9118/10/1-10
PROSTATE CANCER RISK REDUCTION<br />
men were followed for 14 years, screening would avert the<br />
death <strong>of</strong> five <strong>of</strong> nine men destined to die <strong>of</strong> prostate cancer<br />
while leading to the diagnosis <strong>of</strong> prostate cancer in 120 men.<br />
In addition to the considerable laboratory expenses associated<br />
with PSA screening, the anxiety associated with<br />
abnormal PSA test results and the known significant side<br />
effects <strong>of</strong> prostate biopsy—including sepsis and hospitalization—represent<br />
significant “costs” <strong>of</strong> prostate cancer overdiagnosis.<br />
Indeed, the largest cost associated with prostate<br />
cancer overdiagnosis is overtreatment. In a Scandinavian<br />
study, Fall and colleagues showed that during the first week<br />
following the diagnosis <strong>of</strong> prostate cancer, patients experienced<br />
a 2.8-fold (95% CI: 2.5–3.2) increased relative risk <strong>of</strong> a<br />
major cardiovascular event and a 8.4-fold (95% CI: 1.9–22.7)<br />
increased risk <strong>of</strong> suicide. 19 In PLCO, more than 90%<br />
<strong>of</strong> screen-detected cancers have received aggressive therapy.<br />
4 Certainly, earlier detection and treatment <strong>of</strong> prostate<br />
cancer in the PSA era have contributed to the decreasing<br />
mortality <strong>of</strong> this disease. Nonetheless, the significant human<br />
and economic burden <strong>of</strong> prostate cancer overtreatment<br />
presents a considerable challenge to current PSA-based<br />
screening efforts.<br />
Agents to Reduce Overdiagnosis <strong>of</strong> Prostate Cancer<br />
Plausible approaches to prostate cancer risk reduction<br />
include a variety <strong>of</strong> agents. The 5-alpha reductase inhibitors<br />
and vitamin E and selenium have been the best studied.<br />
SELECT evaluated vitamin E and selenium and failed to<br />
show any reduction in the diagnosis <strong>of</strong> prostate cancer. 20 A<br />
more recent analysis <strong>of</strong> the data suggested that vitamin E<br />
may actually increase the detection <strong>of</strong> prostate cancer. 21<br />
Five-alpha reductase inhibitors are a group <strong>of</strong> drugs with<br />
antiandrogenic activity, commonly used in the treatment <strong>of</strong><br />
BPH. These agents exert their antiandrogenic effect by<br />
blocking the conversion <strong>of</strong> testosterone to the active form<br />
dihydrotestosterone (DHT); the latter acts as a potent cellular<br />
androgen that promotes prostate growth. Two isoenzymes<br />
<strong>of</strong> 5-alpha reductase are found in the human genome,<br />
Type 1 and Type 2. Type 2 appears to be the more important<br />
is<strong>of</strong>orm in the pathophysiology <strong>of</strong> BPH, whereas both Types<br />
1 and 2 are expressed in cancer and cancer-related (PIN)<br />
tissue. Finasteride is a selective inhibitor <strong>of</strong> 5-alpha reductase<br />
Type 2, while dutasteride is an inhibitor <strong>of</strong> both isoenzymes.<br />
Both drugs have been linked to decreased sexual<br />
KEY POINTS<br />
● Prostate-specific antigen (PSA)-based screening is<br />
associated with overdiagnosis <strong>of</strong> prostate cancer.<br />
● Most prostate cancers detected in the United States<br />
are treated aggressively.<br />
● Five-alpha reductase inhibitors decrease the detection<br />
<strong>of</strong> low-risk prostate cancers during PSA screening.<br />
● It is uncertain whether 5-alpha reductase inhibitors<br />
used in this setting promote aggressive prostate cancer.<br />
● Five-alpha reductase inhibitors improve the performance<br />
<strong>of</strong> PSA as a diagnostic test for aggressive<br />
prostate cancer.<br />
Table 1. Comparison <strong>of</strong> PCPT and REDUCE<br />
PCPT REDUCE<br />
Design<br />
Duration (yr) 7 4<br />
Number <strong>of</strong> patients 18,882 8,251<br />
Location United States International<br />
Age � 55 50–80<br />
Entry PSA � 3 2.5–10<br />
5-� reductase inhibition Type 2 only Types 1 and 2<br />
Negative baseline biopsy No Yes<br />
Follow-up biopsy 7 yr � FC 2yrand4yr� FC<br />
Cores per study biopsy 6 10<br />
Baseline Characteristics<br />
Median age 63 63<br />
Median PSA 1.1 5.7<br />
Prostate Cancer Results<br />
Placebo 24.4% 21.1%<br />
Treated 18.4% 16.3%<br />
Abbreviations: FC, for cause; PSA, prostate-specific antigen; yr, years.<br />
function and libido as well as gynecomastia in a small subset<br />
<strong>of</strong> men. In general, men apt to experience these side effects<br />
do so during the first year; thereafter, rates are similar<br />
between placebo and treated men. Most side effects are<br />
reversible if the drug is stopped.<br />
Five-alpha reductase inhibitors have been studied in two<br />
large prospective randomized trials, both <strong>of</strong> which have<br />
demonstrated a reduced incidence <strong>of</strong> prostate cancer diagnosis<br />
(see Table 1). PCPT evaluated men age 55 or older with<br />
serum PSA values <strong>of</strong> � 3 who were randomly selected to<br />
receive placebo or 5 mg per day <strong>of</strong> finasteride and underwent<br />
“for cause” and end-<strong>of</strong>-study biopsies at 7 years. 22 This trial<br />
showed a 25% (95% CI: 18.6–30.6, p � 0.001) statistically<br />
significant risk reduction in the diagnosis <strong>of</strong> prostate cancer<br />
in men treated with finasteride. However, this study has<br />
been criticized for three main reasons. First, there was a<br />
very high detection rate for prostate cancer over the 7-year<br />
course <strong>of</strong> the trial. This “low-risk” group <strong>of</strong> men in the<br />
placebo arm (who had a normal rectal exam and a PSA � 3)<br />
had a 24.4% chance <strong>of</strong> a prostate cancer diagnosis after 7<br />
years <strong>of</strong> follow-up. In contrast, average U.S. men have a 17%<br />
lifetime risk <strong>of</strong> prostate cancer. Second, there were fewer<br />
“for cause” biopsies in the finasteride arm. If the biopsy rates<br />
had been equal between the two arms, the difference in<br />
overall prostate cancer detection rates would have been<br />
narrowed. Finally, a major concern <strong>of</strong> PCPT was the finding<br />
on biopsy <strong>of</strong> more prostate cancers with a Gleason score <strong>of</strong> 7<br />
or higher in the finasteride arm. This excess <strong>of</strong> Gleason score<br />
7 to 10 cancer has been attributed to biases induced by<br />
finasteride—the major ones being the increased sensitivity<br />
<strong>of</strong> PSA and the “volume” bias. The latter reflects the effect <strong>of</strong><br />
prostate size on the ability <strong>of</strong> digital rectal exam and biopsy<br />
to detect prostate cancer. Four independent groups have<br />
adjusted the original data <strong>of</strong> PCPT for these biases against<br />
finasteride. The 27% excess Gleason score 7 to 10 disease in<br />
the finasteride arm is eliminated. 23-26 It was on the basis <strong>of</strong><br />
these findings that ASCO and the <strong>American</strong> Urological<br />
Association together recommended that men with a PSA <strong>of</strong><br />
3 or below who are undergoing screening with PSA should<br />
be encouraged to talk to their physicians about the risks<br />
and benefits <strong>of</strong> taking a 5-alpha reductase inhibitor. 27 More<br />
recently, the U.S. Food and Drug Administration (FDA)<br />
used other models to adjust for bias in PCPT. One <strong>of</strong> these<br />
models showed a persistent 1.51 (95% CI: 1.01 to 2.26)<br />
93
excess risk <strong>of</strong> Gleason score 8 to 10 prostate cancer in the<br />
finasteride arm. 28<br />
The REDUCE trial enrolled men between the age <strong>of</strong> 50<br />
and 75 who are considered to be at high risk for prostate<br />
cancer because <strong>of</strong> an elevated PSA (between 2.5 and 10). 29<br />
Patients were required to have undergone a negative<br />
prestudy biopsy that was centrally reviewed. Patients in the<br />
REDUCE trial received 0.5 mg <strong>of</strong> dutasteride per day or<br />
placebo and underwent protocol-specified biopsies at 2 years<br />
and 4 years as well as “for cause” biopsies whenever indicated.<br />
During the first 2 years <strong>of</strong> the REDUCE trial, there<br />
was statistically significant 22.4% (95% CI: 13.0–30.8, p �<br />
0.001) relative risk reduction in the diagnosis <strong>of</strong> prostate<br />
cancer.<br />
The design <strong>of</strong> the REDUCE trial with a planned second<br />
biopsy at 4-year follow-up for men without cancer on either<br />
the initial prestudy biopsy or the 2-year study-mandated<br />
biopsy is important to consider. One would anticipate that at<br />
randomization there would be an equal number <strong>of</strong> small<br />
volume cancers in each arm. After the discovery <strong>of</strong> 142 more<br />
cancers in the placebo arm and their removal from the study<br />
following the year-2 biopsies, there were more men and more<br />
cancers remaining in the dutasteride arm, thus giving rise to<br />
an imbalance between the two study groups for the remaining<br />
duration <strong>of</strong> the study. Despite these imbalances, a<br />
statistically significant 23.7% (95% CI: 9.9–35.3, p � 0.001)<br />
relative risk reduction in the detection <strong>of</strong> prostate cancer<br />
was again noted during the second round <strong>of</strong> biopsies in<br />
REDUCE. The finding <strong>of</strong> fewer cancer cases in the dutasteride<br />
arm despite three significant biases against dutasteride<br />
during the second round <strong>of</strong> biopsies (i.e., more cancers<br />
waiting to be discovered, more men undergoing biopsy, and<br />
a 35% smaller prostate) suggests that dutasteride is capable<br />
<strong>of</strong> continually suppressing tumor growth.<br />
There was no excess <strong>of</strong> Gleason score 7 to 10 cancers in the<br />
dutasteride arm <strong>of</strong> REDUCE. There was, however, a numerical<br />
increase in Gleason score 8 to 10 cancers in the dutasteride<br />
arm (29 vs. 19). This excess occurred exclusively<br />
during the second round <strong>of</strong> biopsies. During the first round,<br />
there were 17 Gleason score 8 to 10 cancers in the dutasteride<br />
arm and 18 in the placebo arm. During the second<br />
round <strong>of</strong> biopsies, there were 12 in the dutasteride arm and<br />
one in the placebo arm. An explanation for this excess biopsy<br />
Gleason score 8 to 10 during the second round <strong>of</strong> the<br />
REDUCE trial has been <strong>of</strong>fered by considering that the<br />
excess Gleason score 5 to 7 cancers diagnosed in the placebo<br />
group during the first round <strong>of</strong> biopsies almost certainly<br />
included the cancers <strong>of</strong> some men in whom biopsy-detectable<br />
Gleason score 8 to 10 cancers were actually present. Using<br />
data from the study <strong>of</strong> Choo and colleagues that considered<br />
men with Gleason score 4 to 7 prostate cancer who were on<br />
active surveillance and underwent follow-up biopsy, approximately<br />
8% <strong>of</strong> such patients were found to have Gleason<br />
score 8 to 10 disease on follow-up biopsy. Eight percent <strong>of</strong> the<br />
141 excess cancers detected in years 1 and 2 in the placebo<br />
arm <strong>of</strong> REDUCE would project 11 additional Gleason score<br />
8 to 10 tumors in the placebo arm. Thus, there was a<br />
virtually identical number <strong>of</strong> Gleason score 8 to 10 cancers<br />
in both arms during the entire study.<br />
Other studies have looked at the occurrence <strong>of</strong> Gleason<br />
score 8 to 10 cancers in men receiving dutasteride. These<br />
include the COMBAT and the REDEEM studies. Neither <strong>of</strong><br />
these trials showed any increase in Gleason score 8 to 10<br />
94<br />
TANAGHO AND ANDRIOLE<br />
cancer in the dutasteride-driven arm after 3 to 4 years,<br />
respectively. 30,31<br />
When one considers the raw data from the REDUCE trial<br />
and adjusts it for the bias induced by prostate shrinkage and<br />
increased PSA sensitivity in the dutasteride arm, the odds<br />
<strong>of</strong> Gleason 7 to 10 cancer in the dutasteride arm actually<br />
decreased by 38%.<br />
In preparation for the <strong>Oncology</strong> Drugs Advisory Committee<br />
meeting in December 2010, cancers in the REDUCE trial<br />
were reread using a modified Gleason score. The REDUCE<br />
protocol specified that tumors would be graded using the<br />
“classic” Gleason scoring system. When cancers in REDUCE<br />
were reread using a modified Gleason scoring system, similar<br />
to the one used in PCPT, the number <strong>of</strong> Gleason score 8<br />
to 10 cancers in the dutasteride arm rose from 29 to 32,<br />
whereas the number in the placebo arm fell from 19 to 16.<br />
Thus, there were twice as many Gleason score 8 to 10<br />
cancers in the dutasteride arm when the modified Gleason<br />
score was used. This doubling <strong>of</strong> Gleason score 8 to 10<br />
cancers suggested a safety signal and the potential induction<br />
<strong>of</strong> Gleason score 8 to 10 cancers by dutasteride, as might<br />
have been seen for finasteride in PCPT. Based on these<br />
considerations, the FDA concluded that if 5-alpha reductase<br />
inhibitors were used for prostate cancer risk reduction, one<br />
might potentially anticipate one additional modified-score<br />
Gleason score 8 to 10 cancer in the treated arm in order to<br />
avert three or four lower-grade cancers. The FDA has stated<br />
that health care pr<strong>of</strong>essionals should be aware that 5-alpha<br />
reductase inhibitors may increase the risk <strong>of</strong> prostate cancer<br />
and that 5-alpha reductase inhibitors are not approved for<br />
the prevention <strong>of</strong> prostate cancer. 28<br />
Effect <strong>of</strong> 5-Alpha Reductase Inhibitors on PSA<br />
Multiple studies have shown that 5-alpha reductase inhibitors<br />
result in a statistically significant improvement in<br />
the performance <strong>of</strong> PSA as a diagnostic test for prostate<br />
cancer. These include the PLESS BPH trial, 6 PCPT, 7 and<br />
the REDUCE trial. 8,9 Moreover, when one looks at the<br />
ability <strong>of</strong> PSA rise to detect clinically significant cancers, use<br />
<strong>of</strong> the National Comprehensive Cancer Network–recommended<br />
PSA velocity guidelines in REDUCE would have<br />
resulted in the detection <strong>of</strong> 64% <strong>of</strong> Gleason score 7 to 10<br />
cancers in the placebo arm, whereas 75% <strong>of</strong> Gleason score 7<br />
to 10 cancers in the dutasteride arm exhibited a PSA rise <strong>of</strong>f<br />
nadir. These data highlight the importance <strong>of</strong> monitoring<br />
PSA in men on 5-alpha reductase inhibitors. Most patients<br />
experience about a 50% reduction in PSA by 6 to 12 months;<br />
any rise from nadir should warrant a biopsy.<br />
Conclusion<br />
In summary, PSA-based screening is associated with<br />
considerable overdiagnosis <strong>of</strong> prostate cancer, which, in<br />
turn, results in a significant burden <strong>of</strong> overtreatment. Fivealpha<br />
reductase inhibitors decrease the detection <strong>of</strong> low-risk<br />
prostate cancers during PSA screening, while at the same<br />
time improving the performance <strong>of</strong> PSA as a diagnostic test<br />
for aggressive prostate cancer. Five-alpha reductase inhibitors<br />
may, therefore, play an important role in focusing<br />
prostate cancer screening efforts on the more lethal forms <strong>of</strong><br />
prostate cancer. Nevertheless, there remains an element <strong>of</strong><br />
uncertainty regarding a potential role for 5-alpha reductase<br />
inhibitors used in this setting in promoting aggressive<br />
prostate cancer.
PROSTATE CANCER RISK REDUCTION<br />
Authors’ Disclosures <strong>of</strong> Potential Conflicts <strong>of</strong> Interest<br />
Author<br />
Employment or<br />
Leadership<br />
Positions<br />
Consultant or<br />
Advisory Role<br />
Youssef S. Tanagho*<br />
Gerald L. Andriole Amarex; Amgen;<br />
Augmenix;<br />
Bayer; Bristol-<br />
Myers Squibb;<br />
Cambridge Endo;<br />
Caris MPI;<br />
GlaxoSmithKline;<br />
Janssen<br />
Biotech; Myriad<br />
Genetics; Ortho-<br />
<strong>Clinical</strong><br />
Diagnostics;<br />
STEBA Biotech;<br />
Viking Medical<br />
*No relevant relationships to disclose.<br />
1. Andriole GL, Crawford ED, Grubb RL 3rd, et al. Prostate cancer<br />
screening in the randomized Prostate, Lung, Colorectal, and Ovarian Cancer<br />
Screening Trial: mortality results after 13 years <strong>of</strong> follow-up. J Natl Cancer<br />
Inst. <strong>2012</strong>;104:125-132.<br />
2. Schroder FH, Hugosson J, Roobol MJ, et al. Screening and prostatecancer<br />
mortality in a randomized European study. N Engl J Med. 2009;360:<br />
1320-1328.<br />
3. Hugosson J, Carlsson S, Aus G, et al. Mortality results from the<br />
Göteborg randomised population-based prostate-cancer screening trial. Lancet<br />
Oncol. 2010;11:725-732.<br />
4. Andriole GL, Crawford ED, Grubb RL 3rd, et al. Mortality results from<br />
a randomized prostate-cancer screening trial. N Engl J Med. 2009;360:1310-<br />
1319.<br />
5. Cooperberg MR, Broering JM, and Carroll PR. Time trends and local<br />
variation in primary treatment <strong>of</strong> localized prostate cancer. J Clin Oncol.<br />
2010;28:1117-1123.<br />
6. Andriole GL, Guess HA, Epstein JI, et al. Treatment with finasteride<br />
preserves usefulness <strong>of</strong> prostate-specific antigen in the detection <strong>of</strong> prostate<br />
cancer: results <strong>of</strong> a randomized, double-blind, placebo-controlled clinical trial.<br />
PLESS Study Group. Proscar Long-term Efficacy and Safety Study. Urology.<br />
1998;52:195-202.<br />
7. Thompson IM, Chi C, Ankerst DP, et al. Effect <strong>of</strong> finasteride on the<br />
sensitivity <strong>of</strong> PSA for detecting prostate cancer. J Natl Cancer Inst. 2006;98:<br />
1128-1133.<br />
8. Andriole GL, Bostwick D, Brawley OW, et al. The effect <strong>of</strong> dutasteride on<br />
the usefulness <strong>of</strong> prostate specific antigen for the diagnosis <strong>of</strong> high grade and<br />
clinically relevant prostate cancer in men with a previous negative biopsy:<br />
results from the REDUCE study. J Urol. 2011;185:126-131.<br />
9. Marberger M, Freedland SJ, Andriole GL, et al. Usefulness <strong>of</strong> prostatespecific<br />
antigen (PSA) rise as a marker <strong>of</strong> prostate cancer in men treated with<br />
dutasteride: Lessons from the REDUCE study. BJU Int. Epub 2011 Jun 23.<br />
10. Welch HG, Black WC. Overdiagnosis in cancer. J Natl Cancer Inst.<br />
2010;102:605-613.<br />
11. Sakr WA, Grignon DJ, Haas BP, et al. Age and racial distribution <strong>of</strong><br />
prostatic intraepithelial neoplasia. Eur Urol. 1996;30:138-144.<br />
12. Powell IJ, Bock CH, Ruterbusch JJ, et al. Evidence supports a faster<br />
growth rate and/or earlier transformation to clinically significant prostate<br />
cancer in black than in white <strong>American</strong> men, and influences racial progression<br />
and mortality disparity. J Urol. 2010;183:1792-1796.<br />
13. Drazer MW, Huo D, Schonberg MA, et al. Population-based patterns<br />
and predictors <strong>of</strong> prostate-specific antigen screening among older men in the<br />
United States. J Clin Oncol. 2011;29:1736-1743.<br />
14. Welch HG, Schwartz LM, and Woloshin S. Prostate-specific antigen<br />
levels in the United States: implications <strong>of</strong> various definitions for abnormal.<br />
J Natl Cancer Inst. 2005;97:1132-1137.<br />
15. Albertsen PC, Moore DF, Shih W, et al. Impact <strong>of</strong> comorbidity on<br />
survival among men with localized prostate cancer. J Clin Oncol. 2011;29:<br />
1335-1341.<br />
16. Wilt TJ. The VA/NCI/AHRQ CSP#407: Prostate cancer Intervention<br />
Versus Observation Trial (PIVOT): main results from a randomized trial<br />
Stock<br />
Ownership Honoraria<br />
Envisioneering Amgen; Bayer;<br />
Bristol-Myers<br />
Squibb; Caris;<br />
GlaxoSmithKline;<br />
Janssen<br />
Biotech; Myriad<br />
Genetics; Ortho-<br />
<strong>Clinical</strong><br />
Diagnostics;<br />
STEBA Biotech<br />
REFERENCES<br />
Research<br />
Funding<br />
Caris;<br />
GlaxoSmithKline;<br />
STEBA Biotech<br />
Expert<br />
Testimony<br />
Other<br />
Remuneration<br />
comparing radical prostatectomy to watchful waiting in men with clinically<br />
localized prostate cancer. In: <strong>American</strong> Urological Association 2011 Annual<br />
Meeting; 2011 May 17.<br />
17. Pinsky PF, Blacka A, Kramer BS, et al. Assessing contamination and<br />
compliance in the prostate component <strong>of</strong> the Prostate, Lung, Colorectal and<br />
Ovarian (PLCO) Cancer Screening. Clin Trials. 2010;4:303-311.<br />
18. Carroll PR, Whitson JM, Cooperberg MR. Serum prostate-specific<br />
antigen for the early detection <strong>of</strong> prostate cancer: always, never, or only<br />
sometimes? J Clin Oncol. 2011;29:345-347.<br />
19. Fall K, Fang F, Mucci LA, et al. Immediate risk for cardiovascular<br />
events and suicide following a prostate cancer diagnosis: prospective cohort<br />
study. PLoS Med. 2009;6:1-8.<br />
20. Lippman SM, Klein EA, Goodman PJ, et al. Effect <strong>of</strong> selenium and<br />
vitamin E on risk <strong>of</strong> prostate cancer and other cancers: the Selenium and<br />
Vitamin E Cancer Prevention Trial (SELECT). JAMA. 2009;301:39-51. Epub<br />
2008 Dec 9.<br />
21. Klein EA, Thompson IM, Tangen CM, et al. Vitamin E and the risk <strong>of</strong><br />
prostate cancer: the Selenium and Vitamin E Cancer Prevention Trial<br />
(SELECT). JAMA. 2011;306:1549-1556.<br />
22. Thompson IM, Goodman PJ, Tangen CM, et al. The influence <strong>of</strong><br />
finasteride on the development <strong>of</strong> prostate cancer. N Engl J Med. 2003;349:<br />
215-224.<br />
23. Cohen YC, Liu KS, Heyden NL, et al. Detection bias due to the effect <strong>of</strong><br />
finasteride on prostate volume: a modeling approach for analysis <strong>of</strong> the<br />
Prostate Cancer Prevention Trial. J Natl Cancer Inst. 2007;99:1366-1374.<br />
24. Pinsky PF, Parnes H, and Ford L. Estimating rates <strong>of</strong> true high-grade<br />
disease in the Prostate Cancer Prevention Trial. Cancer Prev Res (Phila).<br />
2008;1:182-186.<br />
25. Redman MW, Tangen CM, Goodman PJ, et al. Finasteride does not<br />
increase the risk <strong>of</strong> high-grade prostate cancer: a bias-adjusted modeling<br />
approach. Cancer Prev Res (Phila). 2008;1:174-181.<br />
26. Kaplan SA, Roehrborn CG, Meehan AG, et al. PCPT: Evidence that<br />
finasteride reduces risk <strong>of</strong> most frequently detected intermediate- and highgrade<br />
(Gleason score 6 and 7) cancer. Urology. 2009;73:935-939.<br />
27. Kramer BS, Hagerty KL, Justman S, et al. Use <strong>of</strong> 5-� reductase<br />
inhibitors for prostate cancer chemoprevention: <strong>American</strong> <strong>Society</strong> <strong>of</strong> <strong>Clinical</strong><br />
<strong>Oncology</strong>/<strong>American</strong> Urological Association 2008 <strong>Clinical</strong> Practice Guideline.<br />
J Clin Oncol. 2009;27:1502-1516.<br />
28. Theoret MR, Ning YM, Zhang JJ, et al. The risks and benefits <strong>of</strong> 5-�<br />
reductase inhibitors for prostate cancer prevention. N Engl J Med. 2011;365:<br />
97-99.<br />
29. Andriole GL, Bostwick DG, Brawley OW, et al. Effect <strong>of</strong> dutasteride on<br />
the risk <strong>of</strong> prostate cancer. N Engl J Med. 2010;352:1192-1202.<br />
30. Roehrborn CG, Andriole GL, Wilson TH, et al. Effect <strong>of</strong> dutasteride on<br />
prostate biopsy rates and the diagnosis <strong>of</strong> prostate cancer in men with lower<br />
urinary tract symptoms and enlarged prostates in the combination <strong>of</strong> Avodart<br />
and Tamsulosin trial. Eur Urol. 2011;59:244-249.<br />
31. Fleshner NE, Lucia MS, Egerdie B, et al. Dutasteride in localised<br />
prostate cancer management: the REDEEM randomised, double-blind,<br />
placebo-controlled trial. Lancet. Epub <strong>2012</strong> Jan 23.<br />
95
Screening for Prostate Cancer with Prostate-<br />
Specific Antigen: What’s the Evidence?<br />
By Pamela M. Marcus, PhD, and Barnett S. Kramer, MD, MPH<br />
Overview: In October 2011, the U.S. Preventive Services Task<br />
Force (USPSTF, or “Task Force”) released draft recommendations<br />
on prostate cancer screening with prostate-specific<br />
antigen (PSA), concluding that “PSA-based screening results<br />
in small or no reduction in prostate cancer–specific mortality<br />
and is associated with harms related to subsequent evaluation<br />
and treatments, some <strong>of</strong> which may be unnecessary.” This<br />
statement was accompanied by a grade “D” recommendation,<br />
which indicates that in the Task Force’s judgment there “is<br />
moderate or high certainty that the service has no net benefit<br />
or that the harms outweigh the benefits.” The Task Force, an<br />
independent panel <strong>of</strong> nonfederal (U.S.) experts in prevention<br />
and evidence-based medicine, conducts systematic evidence<br />
reviews <strong>of</strong> preventive health care services and makes recommendations<br />
about preventive services in primary care. Task<br />
Force recommendations do not set U.S. federal policy but can<br />
THE USPSTF IS an independent panel <strong>of</strong> nonfederal<br />
(U.S.) primary care providers and scientists with expertise<br />
in prevention and evidence-based medicine. The Task<br />
Force conducts systematic evidence reviews <strong>of</strong> preventive<br />
health care services and makes recommendations about<br />
preventive services in primary care. Task Force recommendations<br />
do not set U.S. federal policy but can and do<br />
influence clinical practice as well as health care coverage<br />
and reimbursement. Therefore, they are important and<br />
<strong>of</strong>ten controversial, particularly when the recommendation<br />
questions a common medical intervention. Under the Task<br />
Force’s purview is cancer screening, a prevention strategy<br />
that has, over the years, proven to be contentious and<br />
emotionally charged.<br />
In October 2011, the Task Force released a draft <strong>of</strong><br />
recommendations on prostate cancer screening with PSA. 1<br />
They concluded that “PSA-based screening results in small<br />
or no reduction in prostate cancer–specific mortality and is<br />
associated with harms related to subsequent evaluation and<br />
treatments, some <strong>of</strong> which may be unnecessary.” This statement<br />
was accompanied by a grade “D” recommendation,<br />
which indicates that in the Task Force’s judgment there “is<br />
moderate or high certainty that the service has no net<br />
benefit or that the harms outweigh the benefits.” As <strong>of</strong> this<br />
writing, the recommendations are still in draft form, and the<br />
Task Force is assessing a large number <strong>of</strong> comments sent to<br />
them during the public comment period. Whatever the<br />
ultimate conclusions <strong>of</strong> the Task Force, prostate cancer<br />
screening with PSA will remain a Medicare benefit for men<br />
<strong>of</strong> all ages and risk factor pr<strong>of</strong>iles.<br />
Screening for prostate cancer with PSA is an example <strong>of</strong><br />
mass screening. A mass cancer screening program strives to<br />
screen asymptomatic persons with certain characteristics<br />
(usually based on age) and reduce the rate <strong>of</strong> disease-specific<br />
mortality in the population being screened. For a mass<br />
cancer screening program to be <strong>of</strong> value, mortality from the<br />
target cancer must be reduced, but harms associated with<br />
the screening process and downstream events triggered by<br />
screening cannot outweigh the benefit. Persons who are<br />
screened have no symptoms <strong>of</strong> the disease in question and<br />
therefore are healthy with regard to that illness. It is<br />
96<br />
and do influence reimbursement and clinical practice. In this<br />
article, we will present evidence the Task Force considered<br />
when making its decision, including two highly influential<br />
randomized controlled trials (RCTs) <strong>of</strong> prostate cancer<br />
screening, the European Randomized Study <strong>of</strong> Prostate Cancer<br />
(ERSPC) and the Prostate, Lung, Colorectal and Ovarian<br />
Cancer Screening Trial (PLCO). The two trials arrived at<br />
different conclusions about the efficacy <strong>of</strong> routine prostate<br />
cancer screening, but similar conclusions about the accompaniment<br />
<strong>of</strong> clinically relevant harms with prostate cancer<br />
screening, including overdiagnosis (screen detection <strong>of</strong> cancers<br />
that never would be diagnosed in the absence <strong>of</strong> screening).<br />
We also will present other available evidence on benefits<br />
and harms <strong>of</strong> PSA-based screening and consider that evidence<br />
and the findings <strong>of</strong> ERSPC and PLCO in conjunction with one<br />
another.<br />
difficult to make healthy individuals any healthier, especially<br />
with an intervention like cancer screening, which puts<br />
large numbers <strong>of</strong> people in harm’s way. Therefore, there<br />
should be strong evidence <strong>of</strong> benefit from cancer screening<br />
before mass screening programs are implemented.<br />
In this article, we discuss the evidence that led to the Task<br />
Force’s decision to recommend against routine prostate<br />
cancer screening with PSA. We first present prostate cancer<br />
statistics for the United States. Before presenting the evidence<br />
that influenced the Task Force’s recommendation, we<br />
provide an overview <strong>of</strong> cancer screening, as such knowledge<br />
is necessary for proper evaluation <strong>of</strong> evidence. For a thorough<br />
treatment <strong>of</strong> the topic, please see Prorok and colleagues.<br />
2<br />
Prostate Cancer in the United States<br />
Prostate cancer is the most frequently diagnosed nonskin<br />
cancer and the second-leading cause <strong>of</strong> cancer death in U.S.<br />
men. More than 240,000 cases are expected to be diagnosed<br />
in <strong>2012</strong> and more than 28,000 men are expected to die <strong>of</strong> the<br />
disease. 3 Early prostate cancer is treatable but usually has<br />
no symptoms; advanced prostate cancer, on the other hand,<br />
is symptomatic but not curable. More than 90% <strong>of</strong> prostate<br />
cancers in the United States are detected at a treatable<br />
stage, however, and the 5-year relative survival for those<br />
cancers approaches 100%. Because treatments <strong>of</strong>ten lead to<br />
urinary and erectile dysfunction, some clinicians have adopted<br />
a “watchful waiting” or “active surveillance” strategy<br />
rather than an initial surgical one for older patients and<br />
those who appear to have less aggressive tumors.<br />
Historically, prostate cancer incidence rates began to rise<br />
rapidly in the mid- to late 1980s, continued to do so until the<br />
From the Division <strong>of</strong> Cancer Control and Population Sciences, National Cancer Institute,<br />
Bethesda, MD.<br />
Authors’ disclosures <strong>of</strong> potential conflicts <strong>of</strong> interest are found at the end <strong>of</strong> this article.<br />
Address reprint requests to Pamela M. Marcus, PhD, Division <strong>of</strong> Cancer Control and<br />
Population Sciences, National Cancer Institute, 6130 Executive Blvd., Room 4106,<br />
Bethesda, MD 20892-7344; email: marcusp@mail.nih.gov.<br />
© <strong>2012</strong> by <strong>American</strong> <strong>Society</strong> <strong>of</strong> <strong>Clinical</strong> <strong>Oncology</strong>.<br />
1092-9118/10/1-10
PSA SCREENING: HARMS WITHOUT CLEAR BENEFIT<br />
mid-1990s, and then started to fall. This trend is thought by<br />
many to be primarily attributable to changes in use <strong>of</strong> PSA<br />
screening rather than radical changes in risk factors or<br />
classification schemes for invasiveness <strong>of</strong> disease. In the<br />
early 1990s, at the height <strong>of</strong> PSA screening, the age-adjusted<br />
prostate cancer incidence rate was in excess <strong>of</strong> 200 per<br />
100,000 man-years, but in 2008 (the most recent data<br />
available), it was about 150 per 100,000 man-years. 4 Prostate<br />
cancer mortality peaked in 1993 at about 39 per 100,000<br />
man-years but in 2008 had fallen to about 23 per 100,000<br />
man-years. 4<br />
Cancer Screening<br />
Cancer screening refers to the routine testing <strong>of</strong> persons<br />
who are asymptomatic for cancer <strong>of</strong> a particular organ. The<br />
goal <strong>of</strong> screening is to decrease the risk <strong>of</strong> death from a given<br />
cancer by detecting the cancer at a time when treatment can<br />
lead to cure. A reduction in disease-specific mortality with<br />
screening, relative to established care without that exam,<br />
KEY POINTS<br />
● A 2010 meta-analysis <strong>of</strong> prostate-specific antigen’s<br />
(PSA’s) effect on prostate cancer mortality that included<br />
five randomized controlled trials (including<br />
the two largest, the European Randomized Study <strong>of</strong><br />
Prostate Cancer and the Prostate, Lung, Colorectal<br />
and Ovarian Cancer Screening Trial) generated a<br />
summary risk ratio <strong>of</strong> 0.88 (95% CI: 0.71–1.09),<br />
indicating no statistically significant benefit <strong>of</strong> PSA<br />
screening.<br />
● If PSA screening does reduce prostate cancer mortality,<br />
the reduction is likely to be small.<br />
● The harms associated with prostate cancer screening<br />
(false-positive exams), diagnostic evaluation <strong>of</strong> positive<br />
exams (e.g., hematospermia, hematuria, sepsis),<br />
treatment <strong>of</strong> prostate cancer (e.g., urinary incontinence<br />
and impotence), and overdiagnosis (screen detection<br />
<strong>of</strong> cancers that never would be diagnosed in<br />
the absence <strong>of</strong> screening) are well-established and clinically<br />
relevant, and should be considered in concert with<br />
any benefit <strong>of</strong> PSA screening that might exist.<br />
● In October 2011, the U.S. Preventive Services Task<br />
Force (USPSTF) released a draft statement advising<br />
against prostate cancer screening with PSA and assigned<br />
a grade “D” recommendation, which states<br />
that in the Task Force’s judgment there “is moderate<br />
or high certainty that the service has no net benefit or<br />
that the harms outweigh the benefits.” Nevertheless,<br />
prostate cancer screening with PSA will remain a<br />
Medicare benefit for men <strong>of</strong> all ages and risk pr<strong>of</strong>iles.<br />
● The USPSTF recommendation reflects two core concepts<br />
in the field <strong>of</strong> cancer prevention and screening:<br />
1) it is difficult to make healthy persons even healthier;<br />
and 2) strong evidence <strong>of</strong> benefit should exist<br />
before routinely administering a clinical intervention<br />
to healthy persons if that measure is likely to incur<br />
harm.<br />
Table 1. The Potential Positive and Negative Consequences<br />
<strong>of</strong> Cancer Screening<br />
Positive<br />
Reduced risk <strong>of</strong> death from the target cancer<br />
Provides some reassurance to those who are concerned that they may have<br />
cancer<br />
Negative<br />
False reassurance if cancer is present, which may lead to disregard <strong>of</strong> symptoms<br />
Harms <strong>of</strong> the screening test (e.g., perforation from endoscopy)<br />
False-positive exams, including the need to undergo diagnostic evaluation and<br />
experience any harms that accompany it<br />
Earlier detection <strong>of</strong> cancer that does not lead to a change in outcome<br />
Identification <strong>of</strong> overdiagnosed* cancers, with accompanying unnecessary,<br />
potentially harmful diagnostic evaluation and treatment<br />
* Cancers that never would have been diagnosed in the absence <strong>of</strong> screening.<br />
indicates that screening works. RCTs are used to determine<br />
whether screening reduces cancer mortality, as that study<br />
design provides the most definitive evidence. Other study<br />
designs, such as case-control studies or comparisons <strong>of</strong><br />
temporal trends in screening and mortality, are subject to<br />
powerful confounding because <strong>of</strong> the inability to control for<br />
factors that are related to both screening and death due to<br />
the disease <strong>of</strong> interest. Metrics such as an increase in 5-year<br />
survival, number <strong>of</strong> cases detected, and number <strong>of</strong> earlystage<br />
cases are insufficient on their own to demonstrate a<br />
benefit <strong>of</strong> screening as a result <strong>of</strong> the presence <strong>of</strong> certain<br />
methodologic biases.<br />
There are both positive and negative consequences <strong>of</strong><br />
cancer screening (Table 1). Screening may benefit patients<br />
by reducing the risk <strong>of</strong> death from the target cancer, relative<br />
to what it would be in the absence <strong>of</strong> that screening exam. A<br />
negative screening exam can provide some reassurance to<br />
those who are concerned that they may have cancer, although<br />
occasionally the reassurance is false and may lead to<br />
disregard <strong>of</strong> cancer symptoms. Screening, however, can lead<br />
to substantial harm. Medical misadventures may occur as<br />
part <strong>of</strong> the screening exam itself. Many positive screening<br />
exams will be false positives, which lead to unnecessary<br />
anxiety as well as diagnostic evaluation that may be accompanied<br />
by harm. Screening can result in earlier detection <strong>of</strong><br />
a cancer that does not lead to a change in date <strong>of</strong> death; in<br />
this instance, the patient spends more <strong>of</strong> his or her life as a<br />
patient with cancer, but life expectancy and cause <strong>of</strong> death<br />
are unchanged. Screening also can result in the detection <strong>of</strong><br />
overdiagnosed cancers, cancers that never would have been<br />
diagnosed in the absence <strong>of</strong> screening. Patients with overdiagnosed<br />
cancers receive unnecessary treatment, experience<br />
unnecessary treatment-related morbidity, and can even die<br />
prematurely as a result <strong>of</strong> unnecessary treatment (e.g.,<br />
lethal cancers induced by radiotherapy, postoperative death,<br />
and heart disease or leukemia caused by chemotherapy).<br />
The U.K. National Screening Committee has compiled a<br />
list <strong>of</strong> criteria for “appraising the viability, effectiveness, and<br />
appropriateness” <strong>of</strong> a screening program 5 (www.screening.<br />
nhs.uk/criteria/fileid9287). We have identified and adapted<br />
the most relevant criteria to prostate cancer screening with<br />
PSA, and the 13 we consider most relevant are listed in<br />
Table 2. Among those criteria are at least five that either are<br />
known not to be true for PSA screening or whose veracity is<br />
uncertain: the natural history <strong>of</strong> the detected lesions is not<br />
fully understood, PSA screening appears not to preferentially<br />
detect lesions that are most likely to progress, there is<br />
97
Table 2. Criteria for Implementation <strong>of</strong> a Screening Program<br />
Applied to Prostate Cancer Screening*<br />
Criteria for Screening<br />
An important health problem<br />
Recognizable latent or early asymptomatic stage<br />
Available suitable test with agreed-upon cut<strong>of</strong>f values<br />
Test acceptable to the population<br />
Natural history <strong>of</strong> the detected lesions understood<br />
Preferential detection <strong>of</strong> lesions likely to progress<br />
Available, accepted treatment<br />
Available facilities for diagnosis and treatment<br />
Agreed-upon policy on whom to treat<br />
Strong evidence (RCTs) <strong>of</strong> mortality/morbidity reduction<br />
Benefits proven to outweigh harms<br />
Quality assurance standards in place (including monitoring)<br />
Affordable (cost-effectiveness)<br />
* Adapted from www.screening.nhs.uk/criteria/fileid9287.<br />
no standard clinical policy on whom to treat, strong evidence<br />
(RCTs) <strong>of</strong> mortality/morbidity reduction does not exist, and<br />
benefits have not been proven to outweigh harms.<br />
Evaluation <strong>of</strong> PSA Screening for Prostate Cancer:<br />
Prostate Cancer Mortality<br />
Five RCTs <strong>of</strong> PSA prostate cancer screening (either alone<br />
or in conjunction with another modality) have reported<br />
prostate cancer mortality results. 6 A 2010 meta-analysis <strong>of</strong><br />
those studies generated a summary risk ratio <strong>of</strong> 0.88 (95%<br />
CI: 0.71–1.09), indicating no statistically significant benefit<br />
<strong>of</strong> PSA screening. 6 The meta-analysis included ERSPC 7 and<br />
PLCO, 8 the two largest and most influential RCTs <strong>of</strong> prostate<br />
cancer screening. ERSPC, which randomly selected in<br />
excess <strong>of</strong> 180,000 men ages 50 to 74 from 1991 to 2003,<br />
reported a 20% prostate cancer relative mortality reduction<br />
(95% CI: 0.67–0.98) with screening using the core group <strong>of</strong><br />
men ages 55 to 69 (more than 160,000), although the<br />
prostate cancer mortality reduction was not statistically<br />
significant in the entire study population (rate ratio: 0.85,<br />
95% CI: 0.73–1.00). PLCO, which randomly selected almost<br />
77,000 men ages 55 to 74 from 1993 to 2001, reported a<br />
nonsignificant 9% (95% CI: 0.87 to 1.36) increase in prostate<br />
mortality with screening after 13 years <strong>of</strong> follow-up. Both<br />
trials, however, demonstrated clinically important levels <strong>of</strong><br />
harm associated with diagnostic evaluation <strong>of</strong> positive<br />
screens and treatment <strong>of</strong> screen-detected prostate cancers.<br />
Neither ERSPC nor PLCO was perfect. As to be expected,<br />
researchers with an a priori belief that prostate cancer<br />
screening is warranted believe that ERSPC’s shortcomings<br />
are not fatal, but those who take a more negative view <strong>of</strong><br />
prostate cancer screening feel that PLCO is the stronger <strong>of</strong><br />
the two trials. Those who make their decisions based on<br />
systematic evidence reviews, such as the Task Force, look at<br />
the evidence base as a whole and conclude that although a<br />
small benefit <strong>of</strong> PSA screening may be possible, the harms<br />
that can follow may outweigh that potentially small benefit.<br />
ERPSC was a multicenter trial: Finland, Switzerland,<br />
Italy, the Netherlands, Belgium, Switzerland, Spain, Portugal,<br />
and France participated, although data from the latter<br />
two countries were excluded from the initially published<br />
analyses. Given the unique characteristics <strong>of</strong> health care in<br />
each country, intervention arm participants had different<br />
experiences and access to different diagnostic evaluation<br />
procedures and treatments. The screening interval varied by<br />
98<br />
MARCUS AND KRAMER<br />
country (2 to 7 years), as did the use <strong>of</strong> additional modalities<br />
and the PSA value that defined a positive screen. In the<br />
Netherlands, Belgium, Switzerland, and Spain, randomization<br />
occurred after informed consent was received, but in<br />
Finland, Switzerland, and Italy, potential subjects were<br />
identified using population registries and randomly selected,<br />
and then those selected for screening were asked to<br />
provide written informed consent, thus raising the question<br />
<strong>of</strong> whether two equivalent arms were produced. A substantial<br />
number <strong>of</strong> men randomly selected for the control arm<br />
were unaware <strong>of</strong> their participation in the trial, so they were<br />
less likely to receive cutting-edge treatment, in particular,<br />
radical prostatectomy, because they had no relationship<br />
with the screening centers, which were academic centers. It<br />
has been demonstrated that participants in the screened<br />
arm were more likely to receive therapy with curative<br />
intent. Therefore, it is unclear whether screening led to a<br />
reduction in mortality or whether differences in the treatment<br />
were responsible.<br />
PLCO’s randomization was more likely than ERSPC’s to<br />
have produced equivalent screening arms, as all participants<br />
signed their informed consent forms before they were<br />
randomly selected. Furthermore, all 10 PLCO centers used<br />
the same criterion to define a positive PSA screen, and<br />
control participants knew they were in the trial and had a<br />
relationship with the specialty medical center where the<br />
intervention arm participants were screened. PLCO’s<br />
“Achilles’ heel” was a high rate <strong>of</strong> PSA screening in the<br />
control arm. The trial had been designed to have 90%<br />
statistical power to detect a 20% reduction in prostate<br />
cancer mortality in the presence <strong>of</strong> no more than 20%<br />
noncompliance (i.e., screening) in the control arm, but actual<br />
noncompliance, measured annually, was between 40% and<br />
54% at particular screening years. Given that rate <strong>of</strong> noncompliance,<br />
PLCO’s statistical power was 90% for a mortality<br />
reduction no smaller than 40%, which may be unrealistic<br />
for PSA screening.<br />
Evaluation <strong>of</strong> PSA Screening for Prostate Cancer:<br />
Harms<br />
Although ERSPC and PLCO came to different conclusions<br />
about whether PSA screening could reduce prostate cancer<br />
mortality, they both did come to the conclusion that screening<br />
could lead to nontrivial levels <strong>of</strong> adverse events due to<br />
diagnostic evaluation and treatment. They also both indicate<br />
that overdiagnosis occurs with prostate cancer screening.<br />
PSA screening itself, which requires a blood draw, has<br />
infrequent side effects, and the ones that do result—infection<br />
at the draw site, vasovagal reaction, discomfort—are<br />
considered to be minor. However, adverse events associated<br />
with prostatic biopsy are <strong>of</strong> more concern. In PLCO, almost<br />
13% <strong>of</strong> men had at least one false-positive PSA exam and <strong>of</strong><br />
those, more than 5% had a biopsy dictated by that exam 9 ;at<br />
the Netherlands ERSPC site, 50% <strong>of</strong> men who had a biopsy<br />
as a result <strong>of</strong> a positive exam had hematospermia, 23% had<br />
hematuria, and 4% had fever. 10 In addition, there are<br />
reports <strong>of</strong> a recent and consistent increase in urosepsis with<br />
resistant organisms for men undergoing PSA-motivated<br />
prostatic biopsy. Surveillance, Epidemiology, and End Results<br />
(SEER)-Medicare data indicate that the rate <strong>of</strong> hospitalization<br />
for biopsy-related infection within 30 days <strong>of</strong> that<br />
biopsy has quadrupled, from about 0.25% before 2001 to<br />
more than 1% in 2007. 11 Hospitalizations also occur for
PSA SCREENING: HARMS WITHOUT CLEAR BENEFIT<br />
Fig. 1. Prostate cancer in the United States, 1975 to<br />
2008: new cases and deaths.<br />
noninfectious biopsy-related complications, but not to the<br />
same degree. A similar pattern for biopsy-related hospitalizations<br />
was observed in Ontario, Canada, during the 2000s. 13<br />
The complications associated with prostate cancer treatment<br />
are unpleasant and life-altering. Urinary incontinence<br />
and erectile dysfunction occur as a result <strong>of</strong> treatment,<br />
although the percentage <strong>of</strong> men whose morbidities are<br />
attributable solely to treatment likely varies by patient<br />
characteristics, treatment received, the definition <strong>of</strong> incontinence<br />
and erectile dysfunction employed, whether the<br />
diagnosis is self-assigned, and the prevalence <strong>of</strong> these morbidities<br />
in the study population before screening, and thus is<br />
difficult to pinpoint. Nevertheless, those receiving prostatectomy<br />
have been shown in research settings to have a 2- to<br />
11-fold increase in risk <strong>of</strong> urinary incontinence and a 1.2- to<br />
2.1-fold increase in risk <strong>of</strong> erectile dysfunction, relative to<br />
men who are treated with “watchful waiting.” 1 In the Task<br />
Force’s 2011 review, 1 the smallest and largest absolute<br />
percentile differences for urinary incontinence following<br />
prostatectomy compared with watchful waiting were nine<br />
(19% vs. 10%) and 40 (44% vs. 4%). For erectile dysfunction,<br />
the figures were 21 (89% vs. 68%) and 36 (81% vs. 45%). In<br />
a community-based (U.S.) study <strong>of</strong> experiences <strong>of</strong> men diagnosed<br />
with prostate cancer in 1994 and 1995 who had<br />
radical prostatectomy, 87% <strong>of</strong> men reported that they had<br />
total urinary control prior to their surgery as compared with<br />
less than 40% at 6, 12, 24, and 60 months postsurgery. 14<br />
Eighty-one percent reported that they had erections firm<br />
enough for intercourse before their surgery, as compared<br />
with less than 30% at the same postsurgery time points. 14<br />
Radiation therapy may confer a lower relative risk <strong>of</strong> incontinence<br />
and erectile dysfunction, but androgen deprivation<br />
therapy may increase the risk <strong>of</strong> the latter. 15 Postoperative<br />
deaths or cardiovascular events occur about 0.5% <strong>of</strong> the<br />
time, although rates are age-dependent.<br />
Men with overdiagnosed prostate cancers are harmed the<br />
most by PSA screening. In the case <strong>of</strong> overdiagnosis, any<br />
expense, inconvenience, discomfort, sexual dysfunction,<br />
morbidity, hospitalization, or death that results from diagnostic<br />
evaluation for a positive screen and treatment <strong>of</strong> a<br />
diagnosed cancer is entirely unnecessary. Figure 1 provides<br />
convincing evidence <strong>of</strong> overdiagnosis: the number <strong>of</strong> prostate<br />
cancer cases in the United States was substantially higher<br />
in 2008 than it was in the early 1970s, but the prostate<br />
cancer mortality rates between the early 1970s and 2008<br />
were not very different. This increase in risk correlates with<br />
two changes in medical practice: an increase in transurethral<br />
resection <strong>of</strong> the prostate (which could serendipitously<br />
detect prostate cancers when performed for other reasons)<br />
beginning in the 1970s and the adoption <strong>of</strong> PSA in the late<br />
1980s. A true “epidemic” <strong>of</strong> prostate cancer cannot explain<br />
this rise in incidence because the rise was unaccompanied by<br />
extreme shifts in risk factor prevalence or the identification<br />
<strong>of</strong> a new, widespread exposure that strongly increased risk;<br />
even if it had, not all cancers would have been curable.<br />
Additional data supporting the existence <strong>of</strong> overdiagnosis<br />
come from a comparison <strong>of</strong> prostate cancer incidence and<br />
mortality in Connecticut and the Seattle-Puget Sound<br />
area, 16 two areas that differed with regard to the uptake <strong>of</strong><br />
PSA screening in the 1980s and 1990s. Connecticut, where<br />
PSA screening was less common, had consistently lower<br />
prostate cancer incidence rates from 1987 through 2001, but<br />
prostate cancer mortality rates were very similar, with<br />
Seattle data showing a possible, although not statistically<br />
significant, increase in prostate cancer mortality for men<br />
ages 75 to 79. A nearly identical reduction in the prostate<br />
cancer mortality rate in both regions over time has been<br />
observed; if PSA screening was reducing mortality in addition<br />
to reductions due to treatment, the reduction for Seattle<br />
would have been even more pronounced.<br />
Microsimulation methods have been used to estimate the<br />
percentage <strong>of</strong> prostate cancers that are considered to be<br />
overdiagnosed disease. Using ERSPC data, Draisma and<br />
colleagues estimated that among men screened annually<br />
from age 55 to 67, 50% <strong>of</strong> screen-detected cases would be<br />
overdiagnosed cases. 17 Using SEER data, overdiagnosis<br />
ranged from 23% to 42% <strong>of</strong> all screen-detected cases. 18<br />
Modeling using PLCO data is not yet available, but an<br />
excess <strong>of</strong> cases existed in the intervention arm in every<br />
study, including after 13 years <strong>of</strong> follow-up. 8 Two years after<br />
the last screen (i.e., the end <strong>of</strong> study year 7), the excess <strong>of</strong><br />
479 cases indicates that overdiagnosis could have accounted<br />
for about 17% <strong>of</strong> cases in the screened arm. That figure is an<br />
underestimate, as contamination occurred in the control<br />
arm. Although it is unclear whether the extent <strong>of</strong> overdiagnosis<br />
is moderate or extreme, it is clear that it exists with<br />
PSA screening.<br />
99
Conclusion<br />
One could argue that neither ERSPC nor PLCO findings<br />
can be considered definitive, as the shortcomings <strong>of</strong> those<br />
trials provide an unreliable measure <strong>of</strong> PSA’s actual benefit.<br />
In fact, the Physician Data Query Prevention and Screening<br />
Editorial Board concluded that the evidence for PSA screening<br />
is insufficient to determine whether that practice results<br />
in a mortality reduction. But board members concluded as<br />
well that “based on solid evidence, screening with PSA<br />
and/or digital rectal exam detects some prostate cancers<br />
that would never have caused important clinical problems<br />
and leads to some degree <strong>of</strong> overtreatment,” and that “based<br />
on solid evidence, current prostate cancer treatments, in-<br />
Authors’ Disclosures <strong>of</strong> Potential Conflicts <strong>of</strong> Interest<br />
Author<br />
Pamela M. Marcus*<br />
Barnett S. Kramer*<br />
*No relevant relationships to disclose.<br />
Employment or<br />
Leadership<br />
Positions<br />
Consultant or<br />
Advisory Role<br />
1. Chou R, Croswell JM, Dana T, et al. Screening for prostate cancer: a<br />
review <strong>of</strong> the evidence for the U.S. Preventive Services Task Force. Ann Intern<br />
Med. 2011;155:762-771.<br />
2. Prorok PC, Kramer BS, Gohagan JK. Screening theory and study design:<br />
the basics. In Kramer BS, Gohagan JK, Prorok PC (eds). Cancer Screening:<br />
Theory and Practice. New York: Marcel Dekker, 1999;29-53.<br />
3. Siegel R, Naishadham D, Jemal A. Cancer statistics, <strong>2012</strong>. CA Cancer<br />
J Clin. <strong>2012</strong>;62:10-29. Epub <strong>2012</strong> Jan 4.<br />
4. Howlader N, Noone AM, Krapcho M, et al. (eds). SEER Cancer Statistics<br />
Review 1975-2008. http://seer.cancer.gov/csr/1975_2008/. Accessed January<br />
30, <strong>2012</strong>.<br />
5. UK National Screening Committee. Programme Appraisal Criteria.<br />
http://www.screening.nhs.uk/criteria/fileid9287. Accessed January 30, <strong>2012</strong>.<br />
6. Djulbegovic M, Beyth RJ, Neuberger MM, et al. Screening for prostate<br />
cancer: systematic review and meta-analysis <strong>of</strong> randomised controlled trials.<br />
BMJ. 2010 Sep 14;341:c4543.<br />
7. Schröder FH, Hugosson J, Roobol MJ, et al. Screening and prostatecancer<br />
mortality in a randomized European study. N Engl J Med. 2009;360:<br />
1320-1328.<br />
8. Andriole GL, Crawford ED, Grubb RL 3rd, et al. Prostate cancer<br />
screening in the randomized Prostate, Lung, Colorectal, and Ovarian Cancer<br />
Screening Trial: Mortality results after 13 years <strong>of</strong> follow-up. J Natl Cancer<br />
Inst. <strong>2012</strong>;104:125-132.<br />
9. Croswell JM, Kramer BS, Kreimer AR, et al. Cumulative incidence <strong>of</strong><br />
false-positive results in repeated, multimodal cancer screening. Am Fam Med.<br />
2009;7:212-222.<br />
10. Ilic D, O’Connor D, Green S, et al. Screening for prostate cancer.<br />
Cochrane Database Syst Rev. 2006 Jul 19;3:CD004720.<br />
11. Loeb S, Carter HB, Berndt SI, et al. Complications after prostate biopsy:<br />
data from SEER-Medicare. J Urol. 2011;186:1830-1834. Epub 2011 Sep 23.<br />
100<br />
cluding radical prostatectomy and radiation therapy, result<br />
in permanent side effects in many men.” 19<br />
The harms associated with PSA screening for prostate<br />
cancer are well-established; the benefit <strong>of</strong> screening, however,<br />
is not and therefore remains theoretical. Given that a<br />
central principle <strong>of</strong> medicine is “first, do no harm,” the<br />
medical community must recognize that routine PSA screening<br />
may run counter to that principle. Rather than simply<br />
advocating the practice both in the <strong>of</strong>fice and in the public<br />
arena, health care providers must engage in informed decision<br />
making with men, as the U.S. Centers for Disease<br />
Control and Prevention currently supports, 20 to make sure<br />
they are aware <strong>of</strong> the potential benefits, harms, and remaining<br />
uncertainties.<br />
Stock<br />
Ownership Honoraria<br />
REFERENCES<br />
Research<br />
Funding<br />
Expert<br />
Testimony<br />
MARCUS AND KRAMER<br />
Other<br />
Remuneration<br />
12. Welch HG. Should I be tested for cancer? Maybe not and here’s why.<br />
Berkeley and Los Angeles, CA, University <strong>of</strong> California Press, 2004.<br />
13. Nam RK, Saskin R, Lee Y, et al. Increasing hospital admission rates for<br />
urological complications after transrectal ultrasound guided prostate biopsy.<br />
J Urol. 2010;183:963-969.<br />
14. Penson DF, McLerran D, Feng Z, et al. 5-year urinary and sexual<br />
outcomes after radical prostatectomy: results from the Prostate Cancer<br />
Outcomes Study. J Urol. 2005;173:1701-1705.<br />
15. Potosky AL, Davis WW, H<strong>of</strong>fman RM, et al. Five-year outcomes after<br />
prostatectomy or radiotherapy for prostate cancer: the Prostate Cancer<br />
Outcomes Study. J Natl Cancer Inst. 2004;96:1358-1367.<br />
16. Lu-Yao G, Albertsen PC, Stanford JL, Stukel TA, Walter-Corkery E,<br />
Barry MJ. Screening, treatment, and prostate cancer mortality in the Seattle<br />
area and Connecticut: fifteen-year follow-up. J Gen Intern Med. 2008;23:1809-<br />
1814.<br />
17. Draisma G, Boer R, Otto SJ, et al. Lead times and overdetection due to<br />
prostate-specific antigen screening: estimates from the European Randomized<br />
Study <strong>of</strong> Screening for Prostate Cancer. J Natl Cancer Inst. 2003;95:868-<br />
878.<br />
18. Draisma G, Etzioni R, Tsodikov A, et al. Lead time and overdiagnosis in<br />
Prostate-Specific Antigen screening: importance <strong>of</strong> methods and context.<br />
J Natl Cancer Inst. 2009;101:374-383.<br />
19. Department <strong>of</strong> Health and Human Services, National Institutes <strong>of</strong><br />
Health, National Cancer Institute. Prostate cancer screening. http://www.<br />
cancer.gov/cancertopics/pdq/screening/prostate/HealthPr<strong>of</strong>essional. Accessed<br />
March 5, <strong>2012</strong>.<br />
20. Department <strong>of</strong> Health and Human Services, Centers for Disease<br />
Control and Prevention. Prostate cancer screening: a decision guide. http://<br />
www.cdc.gov/cancer/prostate/pdf/prosguide.pdf. Accessed March 5, <strong>2012</strong>.
GLIOBLASTOMA: TAKING THE STANDARD OF CARE<br />
TO THE CUTTING EDGE<br />
CHAIR<br />
Mark R. Gilbert, MD<br />
University <strong>of</strong> Texas M. D. Anderson Cancer Center<br />
Houston, TX<br />
SPEAKERS<br />
Daniel J. Brat, MD, PhD<br />
Emory University<br />
Atlanta, GA<br />
Howard Colman, MD, PhD<br />
University <strong>of</strong> Utah<br />
Salt Lake City, UT
Glioblastoma: Biology, Genetics, and Behavior<br />
Overview: Glioblastoma (GBM) is a highly malignant, rapidly<br />
progressive astrocytoma that is distinguished pathologically<br />
from lower-grade tumors by necrosis and microvascular hyperplasia.<br />
The global pattern <strong>of</strong> growth changes dramatically<br />
with the development <strong>of</strong> GBM histology and is characterized<br />
by hypoxia-driven peripheral expansion from a growing necrotic<br />
core. Necrotic foci present centrally in GBM and are<br />
typically surrounded by “pseudopalisading” cells—a configuration<br />
that is relatively unique and long recognized as an<br />
ominous prognostic feature. Theses pseudopalisades are severely<br />
hypoxic, overexpress hypoxia inducible factor-1 (HIF-1),<br />
and secrete proangiogenic factors, such as vascular endothelial<br />
growth factor (VEGF) and interleukin 8 (IL-8). The microvascular<br />
hyperplasia that emerges in response promotes<br />
peripheral tumor expansion. Recent evidence suggests that<br />
pseudopalisades represent a wave <strong>of</strong> tumor cells actively<br />
migrating away from central hypoxia that arises following a<br />
GLIOBLASTOMA (GBM; World Health Organization<br />
[WHO] grade 4) is the highest grade astrocytoma and<br />
has a dismal prognosis. 1 Mean survival following the most<br />
advanced treatment, including neurosurgery, radiotherapy,<br />
and chemotherapy is only 60 to 70 weeks. 2 When patients<br />
receive only surgical resection, but are not treated with<br />
adjuvant therapy, mean survival is a mere 14 weeks, underscoring<br />
the tremendous natural growth properties <strong>of</strong> these<br />
tumors. Lower-grade infiltrative astrocytomas (i.e., WHO<br />
grade 2 and 3 astrocytomas) are also ultimately fatal, but<br />
have substantially slower growth rates and longer survivals<br />
(3 to 8 years). Only once lower-grade tumors have progressed<br />
to GBM do they demonstrate accelerated growth<br />
and rapid progression to death. Those GBMs that progress<br />
from lower-grade gliomas are referred to as “secondary”<br />
GBMs, whereas those GBMs that present without a lowergrade<br />
precursor are termed “primary” or “de novo” GBMs.<br />
De novo tumors account for approximately 90% <strong>of</strong> all GBMs<br />
and are both clinically and molecularly distinct from secondary<br />
GBMs, as will be discussed below.<br />
Growth Patterns in Grade 2–3 Astrocytomas<br />
Importantly, the biologic properties <strong>of</strong> GBM (grade 4)<br />
are quite distinct from those <strong>of</strong> lower-grade astrocytomas<br />
(grade 2 and 3) and suggest that it represents more than an<br />
incremental step in malignancy. The unique neuroimaging<br />
and pathologic features that emerge during the transition to<br />
GBM provide the best insight into potential mechanisms<br />
responsible for the enhanced growth. By magnetic resonance<br />
imaging (MRI), grade 2 and 3 astrocytomas show hyperintense<br />
T2-weighted (or fluid attenuated inversion recovery<br />
[FLAIR]) signal abnormalities, reflecting vasogenic edema<br />
generated in response to diffuse infiltration by individual<br />
tumor cells. Lower-grade tumors expand the involved brain,<br />
but show mild or no contrast enhancement, suggesting an<br />
intact blood-brain barrier and a lack <strong>of</strong> tumor necrosis.<br />
Radial growth rates are modest, with annual increases in<br />
diameter <strong>of</strong> 2 to 4 mm/yr. 3 Histologic sections <strong>of</strong> grade 2–3<br />
tumors reflect the imaging properties: neoplastic cells are<br />
seen diffusely infiltrating between neuronal and glial processes,<br />
leading to architectural distortion and edema. 4-5 As<br />
102<br />
By Daniel J. Brat, MD, PhD<br />
vascular insult. Vaso-occlusive and prothrombotic mechanisms<br />
in GBM could readily explain the presence <strong>of</strong> pseudopalisading<br />
necrosis in tissue sections, the rapid peripheral<br />
expansion on neuroimaging, and the dramatic shift to an<br />
accelerated rate <strong>of</strong> clinical progression as a result <strong>of</strong> hypoxiainduced<br />
angiogenesis. The genetic alterations that coincide<br />
with progression to GBM include amplification <strong>of</strong> epidermal<br />
growth factor receptor (EGFR), deletion <strong>of</strong> CDKN2A, and<br />
mutation or deletion <strong>of</strong> PTEN. Other diagnostic and prognostic<br />
tests used in neuro-oncology include assessment <strong>of</strong> 1p/19q,<br />
MGMT promoter methylation, IDH1, and p53. More recently,<br />
the Cancer Genome Atlas data have indicated that there are<br />
four robust transcriptional classes <strong>of</strong> GBM, referred to as<br />
proneural, neural, classical, and mesenchymal. These classes<br />
have genetic associations and may pave the road for future<br />
development <strong>of</strong> targeted therapies.<br />
astrocytomas progress from the lower end <strong>of</strong> grade 2 to the<br />
upper end <strong>of</strong> grade 3, the degree <strong>of</strong> nuclear anaplasia<br />
increases and the proliferative capacity creeps upward,<br />
resulting in a more densely cellular tumor with greater<br />
malignant potential. Thus, grade 2–3 astrocytomas represent<br />
a continuum <strong>of</strong> gradually increasing tumor grade and<br />
biologic potential, with clinical behavior generally correlating<br />
with the properties <strong>of</strong> individual tumor cells.<br />
Changes in Tumor Growth from Grade 2–3<br />
Astrocytomas to Glioblastoma<br />
Tumor dynamics are fundamentally altered in GBM in<br />
comparison to grade 2–3 astrocytomas. Radial growth rates<br />
for GBM can accelerate to values nearly 10 times greater<br />
than those in grade 2 astrocytomas. 3 MRI <strong>of</strong> GBM typically<br />
reveals a central, contrast-enhancing component (“rimenhancing<br />
mass”) emerging from within the infiltrative<br />
astrocytoma and rapidly expanding outward, causing a<br />
much larger T2-weighted signal abnormality in the tumor’s<br />
periphery.<br />
Studies <strong>of</strong> cellular proliferation strongly suggest a fundamental<br />
difference in driving biologic forces between the<br />
grade 2–3 astrocytomas and glioblastoma. 6 The most reliable<br />
marker <strong>of</strong> proliferation is the Ki-67/MIB-1 antibody,<br />
which identifies nuclei <strong>of</strong> cells in the G1, S, G2, and M<br />
phases <strong>of</strong> the cell cycle, but is not expressed in the resting<br />
phase, G0. In one study <strong>of</strong> proliferation in diffuse astrocytomas<br />
<strong>of</strong> grades 2, 3, and 4, Giannini and colleagues found that<br />
MIB-1 index was highly correlated with survival on multivariate<br />
analysis among grade 2 and 3 astrocytomas. However,<br />
when GBMs (grade 4) were included in the analysis,<br />
proliferation was no longer an independent marker <strong>of</strong> prog-<br />
From the Department <strong>of</strong> Pathology and Laboratory Medicine, Winship Cancer Institute,<br />
Emory University School <strong>of</strong> Medicine, Atlanta, GA.<br />
Author’s disclosure <strong>of</strong> potential conflicts <strong>of</strong> interest are found at the end <strong>of</strong> this article.<br />
Address reprint requests to Daniel J. Brat, MD, PhD, Department <strong>of</strong> Pathology and<br />
Laboratory Medicine, Emory University Hospital, H-176, 1364 Clifton Rd. NE, Atlanta, GA<br />
30322; email: dbrat@emory.edu<br />
© <strong>2012</strong> by <strong>American</strong> <strong>Society</strong> <strong>of</strong> <strong>Clinical</strong> <strong>Oncology</strong>.<br />
1092-9118/10/1-10
GLIOBLASTOMA: BIOLOGY, GENETICS AND BEHAVIOR<br />
nosis. 7 Similarly, the significance <strong>of</strong> MIB-1 prolifieration in<br />
prognosticating grade 2 and 3 astrocytomas was also supported<br />
by an investigation by Hsu and colleagues, who<br />
concluded that the MIB-1 index was useful in grade 2 and 3<br />
tumors, but not GBM. 8 When proliferation is studied in the<br />
setting <strong>of</strong> the GBM histology alone, the prognostic significance<br />
is not evident. For example, in one study <strong>of</strong> 116 newly<br />
diagnosed GBMs, the mean MIB-1 index was 12.5% and<br />
varied from 0 to 76.4%. In this series, MIB-1 proliferation was<br />
not associated with survival on either univariate or multivariate<br />
analysis. 9 Thus, taking the evidence from these and other<br />
studies together, proliferation is a biologically important,<br />
prognostically significant driving force in grade 2 and 3 astrocytomas.<br />
However, once the disease has progressed to GBM,<br />
biologic forces more significant than proliferation emerge.<br />
Forces That Drive Gliobastoma Progression<br />
What are these driving forces that emerge in GBM? The<br />
histopathologic features that distinguish GBM from lower<br />
grade astrocytomas are found near the contrast-enhancing<br />
rim on MRI and include the following: 1) foci <strong>of</strong> necrosis,<br />
usually with evidence <strong>of</strong> surrounding cellular pseudopalisades<br />
(“pseudopalisading necrosis”), and 2) microvascular<br />
hyperplasia, a form <strong>of</strong> angiogenesis morphologically recognized<br />
as endothelial proliferation within newly sprouted<br />
vessels. 1 In contrast to lower-grade astrocytomas, these two<br />
diagnostic findings <strong>of</strong> GBM are largely independent <strong>of</strong> tumor<br />
cell properties (i.e., degree <strong>of</strong> anaplasia, proliferation, etc.),<br />
yet carry an inordinate degree <strong>of</strong> prognostic power. Rather<br />
than mere markers, these structures are more likely to be<br />
mechanistically linked to the accelerated growth properties<br />
that characterize the grade 3 to 4 transition.<br />
An emerging model <strong>of</strong> tumor progression may explain the<br />
development <strong>of</strong> pseudopalisades, the relationship between<br />
pseudopalisades and angiogenesis, and the strong association<br />
between pseudopalisades and aggressive clinical behav-<br />
KEY POINTS<br />
● The growth pattern <strong>of</strong> glioblastoma is fundamentally<br />
distinct from lower-grade gliomas and represents<br />
more than a small incremental change in tumor<br />
grade.<br />
● Whereas tumor cell proliferation is prognosic in<br />
lower-grade gliomas, hypoxia-driven mechanisms become<br />
more relevant in glioblastoma.<br />
● Vasco-occlusion and thrombosis are typical in glioblastoma<br />
and likely initiate or propagate perfusionlimited<br />
hypoxia and necrosis associated with tumor<br />
progression.<br />
● Genetic tests that are performed for diagnostic and<br />
prognostic purposes for glioblastoma include assessment<br />
<strong>of</strong> EGFR, PTEN, 1p/19q, MGMT promoter<br />
methylation, IDH1, and p53.<br />
● Analysis <strong>of</strong> the Cancer Genome Atlas data has demonstrated<br />
four distinct transcriptional classes <strong>of</strong><br />
glioblastoma that have genomic correlates and prognostic<br />
significance: proneual, neural, classical, and<br />
mesenchymal.<br />
ior. 10,11 This model hypothesizes a sequence that begins<br />
with an infiltrating astocytoma <strong>of</strong> moderate to high cellularity<br />
(i.e., grade 3 astrocytoma) and continues with 1) vascular<br />
occlusion within the tumor that is <strong>of</strong>ten associated with<br />
intravascular thrombosis; 2) hypoxia in regions surrounding<br />
vascular pathology; 3) outward migration <strong>of</strong> tumor cells<br />
away from hypoxia, creating a peripherally moving wave<br />
(pseudopalisade) and central necrosis; 4) secretion <strong>of</strong><br />
hypoxia-inducible, proangiogenic factors (VEGF, IL-8) by<br />
pseudopalisading cells; 5) an exuberant angiogenic response<br />
creating microvascular proliferation in regions adjacent to<br />
central hypoxia; and 6) accelerated outward expansion <strong>of</strong><br />
tumor cells toward a new vasculature. The global growth<br />
properties <strong>of</strong> GBM within the brain reflect a coalescence <strong>of</strong><br />
these microscopic processes and result in a peripherally<br />
expanding tumor with a large degree <strong>of</strong> central necrosis.<br />
Pseudopalisades Are Actively Migrating Tumor Cells<br />
Pseudopalisading <strong>of</strong> cells around central degereneration<br />
has been recognized for nearly a century as both a defining<br />
feature <strong>of</strong> GBM and a morphologic finding that predicts<br />
aggressive behavior. A commonly held belief has been that<br />
pseudopalisades represent a rim <strong>of</strong> residual tumor cells<br />
around a centrally degenerating clone <strong>of</strong> highly proliferative<br />
cells. However, recent studies have refuted this and concluded<br />
that pseudopalisades represent a wave <strong>of</strong> actively<br />
migrating tumor cells that are moving away from an area <strong>of</strong><br />
central hypoxia. Pseudopalisading cells are known to be<br />
hypoxic, as demonstrated by their dramatic upregulation <strong>of</strong><br />
HIF-1, a nuclear transcription factor that orchestrates the<br />
cell’s adaptive response to low oxygen. 12 Gene expression<br />
studies performed on microdissected pseudopalisading cells<br />
have demonstrated upregulated gene transcripts in this<br />
population that suggest a response to a hypoxic microenvironment,<br />
including those related to glycolysis, angiogenesis,<br />
and cell cycle control. 13 Hypoxic GBM cells in culture that<br />
have similar upregulation <strong>of</strong> HIF-1� are more highly migratory<br />
than normoxic cells, and HIF-1 itself mediates many <strong>of</strong><br />
the critical promigratory mechanisms in gliomas and other<br />
neoplastic cells. 12 Moreover, hypoxic pseudopalisades express<br />
increased levels <strong>of</strong> extracellular matrix proteases<br />
associated with invasion, including MMP-2 and uPAR. 4,12<br />
Thus, the combined evidence suggests that the pseudopalisades<br />
in GBM are formed by a population <strong>of</strong> hypoxic,<br />
actively migrating neoplastic cells that have imposed themselves<br />
on a less mobile population, thereby creating a hypercellular<br />
zone around an evolving area <strong>of</strong> central necrosis.<br />
Pseudopalisades Are Hypoxic Tumor Cells Migrating<br />
Away from Vascular Pathology<br />
The hypoxia that leads to increased tumor cell migration<br />
to form pseudopalisades and to widespread invasiveness<br />
could result from limitations in vascular perfusion within<br />
the tumor (i.e., disruption <strong>of</strong> the blood supply) or from<br />
reduced oxygen diffusion within the neoplasm, in part due to<br />
increased metabolic demands <strong>of</strong> a growing tumor. Attenuated<br />
perfusion would lead to cell migration away from<br />
central blood vessels that no longer provides the necessary<br />
oxygen supply. Limitations in oxygen diffusion, on the other<br />
hand, would cause tumor cells at greatest distance from<br />
arterial supplies to become hypoxic and migrate toward<br />
viable vessels. These mechanisms are not mutually exclu-<br />
103
sive. However, a growing body <strong>of</strong> experimental and observational<br />
evidence favors the hypothesis that pseudopalisades<br />
represent tumor cells migrating away from a dysfunctional<br />
vasculature. 12,14,15 Perhaps less appreciated, abnormal vessels<br />
can <strong>of</strong>ten be noted within the lumina <strong>of</strong> at least a subset<br />
<strong>of</strong> pseudopalisades. A comprehensive survey <strong>of</strong> human GBM<br />
specimens found that over 50% <strong>of</strong> pseudopalisades had<br />
evidence <strong>of</strong> a central vascular lumen that was either degenerating<br />
or thrombosed. 12 Thus, pseudopalisades around necrosis<br />
appear to represent hypoxic tumor cells migrating<br />
away from vaso-occlusion and thrombosis.<br />
Intravascular Thrombosis Accentuates and Propagates<br />
Tumor Hypoxia<br />
While precise initiators <strong>of</strong> vascular pathology in GBM<br />
continue to be studied, it is becoming clear that intravascular<br />
thrombosis within these neoplasms can accentuate and<br />
propagate tumoral hypoxia and necrosis. Intravascular<br />
thrombosis within the tumoral tissue <strong>of</strong> GBM is a frequent<br />
intraoperative finding by the neurosurgeon. Even more<br />
impressively, thrombosed vessels within resected GBM<br />
specimens can almost always be identified under the microscope<br />
(noted histologically in over 90% <strong>of</strong> GBMs). 12,15 In<br />
contrast, the frequency <strong>of</strong> microscopic thrombosis is much<br />
lower in anaplastic astrocytoma (grade 3), a tumor that<br />
lacks necrosis and angiogenesis. In the uncommon instance<br />
when intravascular thrombosis is identified microscopically<br />
in anaplastic astrocytomas (AAs, WHO grade 3), it is predictive<br />
<strong>of</strong> aggressive growth, similar to that <strong>of</strong> GBM, suggesting<br />
it may precede the necrosis and angiogenesis that<br />
arise in GBM and are associated with a more rapid progression.<br />
15 Since the frequency <strong>of</strong> intravascular thrombosis<br />
within neoplastic tissue is much higher in GBMs (grade 4)<br />
than AAs (grade 3), critical pro-thrombotic events must<br />
occur in this transition. In many instances, the intravascular<br />
thrombosis can be seen within or adjacent to the regions<br />
<strong>of</strong> pseudopalisading necrosis, leading to the proposition that<br />
vaso-occlusion due to thrombosis could directly initiate or<br />
propagate hypoxia and necrosis in GBM.<br />
Multiple factors likely contribute to intravascular thrombosis<br />
in GBM, including abnormal blood flow within a<br />
distorted vasculature, increased interstitial edema, dysregulation<br />
<strong>of</strong> pro- and anticoagulant factors, and access <strong>of</strong><br />
plasma-clotting factors to tumoral tissue. Normal central<br />
nervous system (CNS) blood vessels allow only limited<br />
diffusion though their walls because <strong>of</strong> a highly restrictive<br />
blood-brain barrier, which is formed primarily by endothelial<br />
tight junctions, but also has contributions from astrocytic<br />
foot plates, extracellular matrix, and endothelialpericytic<br />
interactions. This barrier becomes breached in<br />
GBM and can be visualized radiologically by the presence <strong>of</strong><br />
contrast enhancement due to increased vascular permeability<br />
to contrast agents (e.g., gadolinium) and to proteins that<br />
bind them, such as albumin. Damaged vessels appear fenestrated,<br />
show detachment <strong>of</strong> pericytes, and exhibit extracellular<br />
matrix alterations. All the factors that contribute to<br />
increased permeability have not been defined, but VEGF<br />
secretion by neoplastic cells is known to cause vascular<br />
leakage. 16 One result is to bring plasma coagulation factors<br />
such as factor VII into the tissue spaces where they are<br />
activated by binding tissue factor and result in thrombosis.<br />
104<br />
DANIEL J. BRAT<br />
Angiogenesis Supports Peripheral Tumor Growth<br />
If emerging models <strong>of</strong> GBM progression are valid, then<br />
vascular pathology may underlie the development <strong>of</strong> hypoxia<br />
and necrosis in GBM. Although necrosis has long been<br />
recognized as a marker <strong>of</strong> aggressive behavior in diffuse<br />
gliomas, by itself it does not explain rapid tumor progression.<br />
Indeed, tumor cell death is the goal <strong>of</strong> most adjuvant<br />
therapies. Instead, pseudopalisades that surround necrosis<br />
in GBM are intimately related to microvascular hyperplasia,<br />
a defining morphologic feature <strong>of</strong> GBM that is most <strong>of</strong>ten<br />
noted in regions directly adjacent to pseudopalisades. 11 This<br />
exuberant angiogenic response attempts to lay down a new<br />
vasculature for rapid neoplastic expansion, yet the proper<br />
function <strong>of</strong> these distorted vessels has not been established.<br />
The formation <strong>of</strong> new blood vessels from pre-existing ones is<br />
tightly regulated process and follows a complex sequence in<br />
response to pro- and antiangiogenic factors. Initial phases<br />
require increased vascular permeability <strong>of</strong> parent vessels,<br />
extravasation <strong>of</strong> plasma, and deposition <strong>of</strong> proangiogenic<br />
matrix proteins. In response to the mitogenic effects <strong>of</strong><br />
proangiogenic cytokines, endothelial cells proliferate and<br />
migrate along a chemotactic gradient into the extracellular<br />
matrix. Once established, endothelial cells form tubes with a<br />
central lumen, elaborate a basement membrane, and eventually<br />
recruit pericytes and smooth muscle cells to surround<br />
the mature vessels.<br />
One <strong>of</strong> the most critical proangiogenic factors produced by<br />
pseudopalisades that is responsible for directing nearby<br />
angiogenesis in GBM is VEGF. As noted above, pseudopalisading<br />
cells are severely hypoxic and express high levels <strong>of</strong><br />
hypoxia-inducible transcription factors, including HIF-1.<br />
The VEGF gene contains a hypoxia-responsive element<br />
(HRE) within its promoter that binds HIF-1, thereby activating<br />
transcription. 17 VEGF concentrations in the cystic<br />
fluid <strong>of</strong> human GBMs can reach levels that are 200–300-fold<br />
higher than in serum. Inhibition <strong>of</strong> this HIF/VEGF pathway<br />
suppresses tumor growth experimentally. Once expressed<br />
and secreted, extracellular VEGF binds to its high affinity<br />
receptors, VEGFR-1 and VEGFR-2, which are upregulated<br />
on endothelial cells <strong>of</strong> high-grade gliomas, but not present in<br />
normal brain. Receptor activation then leads to angiogenesis<br />
in regions adjacent to pseudopalisades, eventually leading to<br />
a vascular density in GBMs that is among the highest <strong>of</strong> all<br />
human neoplasms.<br />
A second proangiogenic factor that is highly upregulated<br />
in GBMs is interleukin-8 (IL-8, CXCL8). 18 Much like VEGF,<br />
hypoxia/anoxia strongly stimulates IL-8 expression and its<br />
expression is also found at highest levels within the pseudopalisades<br />
<strong>of</strong> GBM. Unlike VEGF, IL-8 has a more punctate<br />
distribution within pseudopalisades, and it remains<br />
unclear if tumor cells or scattered infiltrating macrophages<br />
are most responsible for the majority <strong>of</strong> its expression.<br />
Hypoxic upregulation <strong>of</strong> IL-8 is not directly due to HIF<br />
activation, but is more likely due to activation <strong>of</strong> AP-1 by<br />
hypoxia/anoxia. The IL-8 receptors that could potentially<br />
contribute to IL-8 mediated tumorigenic and angiogenic<br />
responses in GBM include CXCR1 and CXCR2, both <strong>of</strong><br />
which are G-protein coupled.<br />
The precise type <strong>of</strong> angiogenesis that is most evident<br />
in GBM, microvascular hyperplasia, is characterized by<br />
numerous enlarged, rapidly dividing endothelial cells, pericytes,<br />
and smooth muscle cells that form tufted microaggre-
GLIOBLASTOMA: BIOLOGY, GENETICS AND BEHAVIOR<br />
gates at the leading edge <strong>of</strong> sprouting vessels. In its most<br />
florid form, angiogenesis takes the shape <strong>of</strong> “glomeruloid<br />
bodies”—a feature that is most characteristic <strong>of</strong> GBM, but is<br />
also an independent marker <strong>of</strong> poor prognosis in other forms<br />
<strong>of</strong> cancer. 19 Since necrosis and hypoxia are located in the<br />
GBM’s core and near the contrast-enhancing rim, hypoxiainduced<br />
angiogenesis occurs further peripherally, favoring neoplastic<br />
growth outward. The permissive nature <strong>of</strong> the CNS<br />
parenchymal matrix to diffuse infiltration by individual<br />
glioma cells allows for this burst <strong>of</strong> peripheral expansion. 4<br />
Genetic Testing<br />
A large and growing number <strong>of</strong> genetic alterations have<br />
been identified in the diffuse gliomas, some <strong>of</strong> which are<br />
used for diagnostic and prognostic purposes in neurooncology.<br />
The genetic alterations that coincide with progression<br />
to GBM include amplification <strong>of</strong> EGFR, deletion <strong>of</strong><br />
CDKN2A, and mutation or deletion <strong>of</strong> PTEN. Other diagnostic<br />
and prognostic tests used in neuro-oncology include<br />
assessment <strong>of</strong> 1p/19q, MGMT promoter methylation, IDH1,<br />
and p53. 1<br />
p53 Pathway Alteration<br />
Alterations in the p53 tumor suppressor pathway are<br />
frequent in infiltrative astrocytomas. The p53 pathway can<br />
be altered mechanisms including mutation <strong>of</strong> TP53 and<br />
deletion <strong>of</strong> the opposite allele, MDM2 gene amplification,<br />
and p14ARF gene deletion. Point mutations in the DNAbinding<br />
region <strong>of</strong> TP53 are the most frequent source <strong>of</strong><br />
inactivation. The majority (50% to 60%) <strong>of</strong> lower-grade<br />
(WHO grade 2) infiltrating astrocytomas show such TP53<br />
mutations, suggesting it occurs early. GBMs that arise from<br />
grade 2 and 3 astrocytomas, so-called secondary GBMs, have<br />
similar frequencies <strong>of</strong> TP53 mutations. TP53 mutations are<br />
less frequent (30%) in primary, or de novo, GBMs, yet the<br />
p53 pathway is altered by other mechanisms in these<br />
tumors. The prognostic significance <strong>of</strong> TP53 mutations and<br />
p53 protein expression in astrocytomas has been debated.<br />
Young age is a strong predictor <strong>of</strong> prolonged survival among<br />
patients with astrocytomas, especially GBM. Since TP53<br />
mutations occur more frequently in GBMs from young<br />
patients, their significance must be separated from the<br />
survival advantage <strong>of</strong> youth.<br />
For glioma classification, TP53 mutations currently have<br />
limited utility. Since p53 derived from mutant genes has a<br />
longer cellular half-life than wild type, the protein accumulates<br />
in the nucleus. Thus, positive nuclear immunohistochemical<br />
staining for the p53 protein correlates, albeit<br />
imperfectly, with the presence <strong>of</strong> TP53 mutations. Positive<br />
p53 immunostaining can be occasionally helpful for distinguishing<br />
astrocytic from oligodendroglial differentiation,<br />
since the oligodendrogliomas rarely harbor TP53 mutations.<br />
1p/19q<br />
There is a strong association <strong>of</strong> allelic losses on chromosomes<br />
1p and 19q and the oligodendroglioma phenotype, and<br />
60% to 80% <strong>of</strong> oligodendroglial neoplasms demonstrate combined<br />
1p and 19q losses. 1,6 Enthusiasm for defining genetic<br />
subsets <strong>of</strong> oligodendrogliomas increased substantially with<br />
the demonstration <strong>of</strong> prognostically distinct groups. 20<br />
1p and 19q losses are less frequent in other forms <strong>of</strong><br />
gliomas. For example, Smith and colleagues investigated the<br />
allelic losses <strong>of</strong> 1p and 19q in 115 diffuse gliomas. 21 Combined<br />
loss <strong>of</strong> 1p and 19q were seen in 11% <strong>of</strong> astrocytomas,<br />
31% <strong>of</strong> the mixed oligoastrocytomas, and 64% <strong>of</strong> oligodendrogliomas.<br />
Thus, while most oligodendrogliomas showed<br />
1p/19q loss, not all did. Moreover, a small percentage <strong>of</strong><br />
mixed gliomas and astrocytomas also showed similar deletions.<br />
More recent studies have demonstrated less frequent<br />
1p/19q losses in diffuse astrocytomas. 22<br />
Similar studies have interrogated the prognostic significance<br />
<strong>of</strong> combined loss <strong>of</strong> 1p and 19q in diverse types <strong>of</strong><br />
diffuse gliomas and found them to be predictive <strong>of</strong> prolonged<br />
overall survival only for patients with oligodendrogliomas.<br />
Combined 1p and 19q losses are not predictive <strong>of</strong> prolonged<br />
survival in astrocytomas or oligoastrocytomas <strong>of</strong> any<br />
grade. 22-23 However, diagnostic testing for 1p/19q is <strong>of</strong>ten<br />
used in cases <strong>of</strong> diffuse gliomas that have ambiguous morphology,<br />
since 1p/19q codeletion is highly associated with<br />
oligodedroglioma.<br />
EGFR<br />
Amplifications <strong>of</strong> the EGFR gene occur in approximately<br />
40% <strong>of</strong> GBMs and 10% <strong>of</strong> anaplastic astrocytomas. 1 Amplifications<br />
are much less frequent in low-grade astrocytomas<br />
and are considered a late genetic event in the progression <strong>of</strong><br />
tumors to GBM. Either wild-type or mutated forms <strong>of</strong> EGFR<br />
can be amplified. The most common EGFR amplification is a<br />
mutated form lacking exons 2–7, which results in a truncated<br />
cell surface protein with constitutive tyrosine kinase<br />
activity (EGFRvIII).<br />
The significance <strong>of</strong> EGFR gene amplification or EGFR<br />
protein overexpression as a prognostic marker in GBM has<br />
been debated. Most comprehensive studies have concluded<br />
that EGFR status is not prognostically significant in patients<br />
with GBM. However, therapies have been developed<br />
that are directed at the overexpressed EGFR in GBMs.<br />
Therefore, it may become critical to establish the EGFR<br />
status <strong>of</strong> GBM as a part <strong>of</strong> the pathologic diagnosis in order<br />
to predict pharmacologic responses to EGFR inhibitors. For<br />
example, Mellingh<strong>of</strong>f and colleagues demonstrated that<br />
those GBMs that have the best therapeutic response to<br />
EGFR inhibitors are characterized by the coexpression <strong>of</strong><br />
EGFRvIII and PTEN. 24<br />
Losses on Chromosome 10/PTEN Mutation<br />
Some <strong>of</strong> the most frequent genetic deletions in GBMs<br />
involve chromosome 10 and occur as losses <strong>of</strong> the entire<br />
chromosome or as losses <strong>of</strong> only the long or short arms. Most<br />
interest focuses on 10q, since it is commonly implicated in<br />
high-grade progression and is the location <strong>of</strong> PTEN. Located<br />
at 10q23.3, PTEN is mutated in 20% to 40% <strong>of</strong> GBMs and<br />
generally occurs in the setting <strong>of</strong> chromosome 10 allelic loss.<br />
Both chromosome 10 losses and PTEN mutations are much<br />
more frequent in diffuse forms <strong>of</strong> astrocytoma than oligodendrogliomas.<br />
They are also relatively late events in the<br />
progression to GBM, since both are much more frequent in<br />
GBMs than in AAs or grade 2 astrocytomas.<br />
The high frequency <strong>of</strong> chromosome 10 losses in GBMs<br />
(� 80%) and the tight correlation <strong>of</strong> its loss with the GBM<br />
phenotype might suggest that it would not be a useful<br />
prognostic marker across tumor grades. Indeed, in studies <strong>of</strong><br />
high-grade astrocytomas (AAs and GBMs), loss <strong>of</strong> heterozygosity<br />
(LOH) <strong>of</strong> chromosome 10 (either 10p or 10q) has been<br />
105
an independent predictor <strong>of</strong> poor prognosis in large part<br />
because LOH was highly associated with AAs that had short<br />
survivals. Some investigators have therefore suggested that<br />
AAs with LOH <strong>of</strong> 10q may behave clinically as GBMs.<br />
Overall, it appears that chromosome 10 losses could serve as<br />
a marker <strong>of</strong> astrocytic differentiation in diffuse high-grade<br />
gliomas and <strong>of</strong> poor prognosis among high-grade astrocytomas,<br />
with special relevance for predicting aggressive AAs.<br />
MGMT<br />
Current standard <strong>of</strong> care chemotherapy (temozolomide)<br />
used to treat GBM acts by crosslinking DNA by alkylating at<br />
the O6 <strong>of</strong> guanine. DNA crosslinking is reversed by the DNA<br />
repair enzyme MGMT (O6-methylguanine-DNA methyltransferase).<br />
Thus, low levels <strong>of</strong> MGMT expression would be<br />
expected to be associated with an enhanced response to<br />
therapy. The expression level <strong>of</strong> MGMT is determined in<br />
large part by the methylation status <strong>of</strong> the gene’s promoter.<br />
This “epigenetic silencing” <strong>of</strong> MGMT occurs in 40% to 50% <strong>of</strong><br />
GBMs and can be assessed by its promoter methylation<br />
status on polymerase chain reaction (PCR)-based tests <strong>of</strong><br />
genomic DNA. Epigenetic silencing <strong>of</strong> MGMT in tumoral<br />
tissue is associated with response to BCNU therapy and<br />
improved survival in patients with GBM.<br />
A recent investigation <strong>of</strong> temozolamide for the treatment<br />
<strong>of</strong> GBM found that epigenetic gene silencing <strong>of</strong> MGMT was<br />
associated with a longer survival, independent <strong>of</strong> treatment.<br />
2,25 The study also demonstrated a survival advantage<br />
among those patients treated with temozolomide and radiotherapy<br />
whose GBMs had a silenced MGMT gene. Since<br />
temozolamide is now standard <strong>of</strong> care for the treatment <strong>of</strong><br />
GBM, testing for MGMT status is becoming an important<br />
component <strong>of</strong> a complete diagnostic work-up.<br />
IDH1<br />
Mutations in isocitrate dehydrogenase 1 (IDH1) are frequent<br />
in grade 2 and 3 astrocytomas, oligodendrogliomas,<br />
and oligoastrocytomas, as well as the GBMs that progress<br />
from these lower-grade lesions (i.e., secondary GBMs). 26,27<br />
IDH2 mutations have also been noted in these same neoplasms,<br />
but at much lower frequency. Since IDH mutations<br />
occur in diffuse gliomas with astrocytic, oligodendroglial,<br />
and mixed histologies, it is believed that they are an early<br />
event, preceding molecular alterations such as TP53 mutations<br />
and 1p/19q codeletion, which are associated with<br />
astrocytic and oligodendroglial histologies, respectively.<br />
Within all histologic types and grade <strong>of</strong> diffuse glioma, the<br />
presence <strong>of</strong> IDH mutations is associated with prolonged<br />
survival. In GBMs, the finding <strong>of</strong> IDH1 mutations is<br />
strongly associated with patients with younger age and a<br />
substantially longer survival. 26,27<br />
Importantly, over 90% <strong>of</strong> IDH1 mutations in the diffuse<br />
gliomas occur at a specific site and are characterized by a<br />
base exchange <strong>of</strong> guanine to adenine within codon 132,<br />
resulting in an amino acid change from arginine to histidine<br />
Author’s Disclosure <strong>of</strong> Potential Conflicts <strong>of</strong> Interest<br />
Author<br />
Daniel J. Brat*<br />
*No relevant relationships to disclose.<br />
106<br />
Employment or<br />
Leadership<br />
Positions<br />
Consultant or<br />
Advisory Role<br />
(R132H). Because <strong>of</strong> this consistent protein alteration, a<br />
monoclonal antibody has been developed to the mutant<br />
protein, allowing its use in paraffin-embedded specimens<br />
(mIDH1R132H). 28 The initial characterization <strong>of</strong> this antibody<br />
was performed on 186 gliomas <strong>of</strong> various histologies<br />
and grades that had been sequenced for IDH1 mutations<br />
and showed excellent sensitivity and specificity <strong>of</strong> the antibody<br />
for mutations. Moreover, the ability <strong>of</strong> the antibody to<br />
detect only a minor component <strong>of</strong> the tissue as mutant may<br />
give this method greater sensitivity than sequencing for<br />
identifying R132H mutant gliomas.<br />
The Cancer Genome Atlas Project<br />
The Cancer Genome Atlas (TCGA) is a large-scale collaborative<br />
effort supported by the National Cancer Institute<br />
and the National Human Genome Research Institute that<br />
has provided an integrated platform for defining the molecular<br />
alterations <strong>of</strong> cancer that are associated with pathologic<br />
and radiologic features, clinical behaviors, and response to<br />
therapy. One <strong>of</strong> the first three malignancies targeted by the<br />
TCGA pilot project was GBM, resulting in the identification<br />
<strong>of</strong> specific genetic mutations, copy number variations, chromosomal<br />
translocations, gene and microRNA expression,<br />
and DNA methylation patterns in nearly 500 tumor samples.<br />
All data is publically available for intensive correlative<br />
analysis. The genomic component <strong>of</strong> this investigation confirmed<br />
frequent alterations in the p53 (TP53 mutations,<br />
p14ARF deletion, MDM2 amplification), RB (CDKN2A<br />
deletion, CDK4 amplification, RB1 deletion/mutation), and<br />
receptor tyrosine kinase signaling pathways (EGFR,<br />
PDGFRA, and ERBB2 amplification, PTEN and PI(3)K<br />
mutation, NF1 deletion). 29 The TCGA project also led to the<br />
identification <strong>of</strong> four distinct molecular subtypes based on<br />
transcriptional pr<strong>of</strong>iles: proneural, neural, classical, and<br />
mesenchymal. 30 These expression classes have variable associations<br />
with genomic alterations. For example, the proneural<br />
class has a high frequency <strong>of</strong> IDH1 mutations and<br />
PDGFR amplifications; NF1 mutations and deletions are<br />
most frequent in the mesenchymal class; and the classical<br />
class has a high frequency <strong>of</strong> EGFR amplifications. However,<br />
these genetic associations are not absolute, and GBMs<br />
with EGFR and PDGFR amplification, TP53 and PTEN<br />
mutations, and CDKN2A deletions are noted in each transcriptional<br />
class. The proneural gene expression signature is<br />
associated with improved clinical outcome. Much <strong>of</strong> this<br />
survival advantage is due to the inclusion <strong>of</strong> tumors with<br />
IDH mutations. This specific subset <strong>of</strong> GBMs, those within<br />
the proneural expression class and with IDH mutations, is<br />
tightly correlated with the CpG island methylator phenotype<br />
(G-CIMP). 31 The prognostic difference among the other<br />
three transcriptional classes in the TCGA data is not statistically<br />
significant. However, this robust transcriptional classification<br />
may lead to the identification <strong>of</strong> class-specific<br />
therapeutic targets that are directed at underlying molecular<br />
mechanisms.<br />
Stock<br />
Ownership Honoraria<br />
Research<br />
Funding<br />
Expert<br />
Testimony<br />
DANIEL J. BRAT<br />
Other<br />
Remuneration
GLIOBLASTOMA: BIOLOGY, GENETICS AND BEHAVIOR<br />
1. Louis DN, Ohgaki H, Wiestler OD, et al. WHO Classification <strong>of</strong> Tumours<br />
<strong>of</strong> the Central Nervous System. 4th ed. Lyon: Intl. Agency for Research; 2007.<br />
2. Stupp R, Mason WP, van den Bent MJ, et al. Radiotherapy plus<br />
concomitant and adjuvant temozolomide for glioblastoma. N Engl J Med.<br />
2005;352:987-996.<br />
3. Swanson KR, Bridge C, Murray JD, et al. Virtual and real brain tumors:<br />
using mathematical modeling to quantify glioma growth and invasion. J Neurol<br />
Sci. 2003;216:1-10.<br />
4. Bellail AC, Hunter SB, Brat DJ, et al. Microregional extracellular matrix<br />
heterogeneity in brain modulates glioma cell invasion. Int J Biochem Cell<br />
Biol. 2004;36:1046-1069.<br />
5. Gupta M, Djalilvand A, Brat DJ. Clarifying the diffuse gliomas: an<br />
update on the morphologic features and markers that discriminate oligodendroglioma<br />
from astrocytoma. Am J Clin Pathol. 2005;124:755-768.<br />
6. Brat DJ, Prayson RA, Ryken TC, et al. Diagnosis <strong>of</strong> malignant glioma:<br />
Role <strong>of</strong> neuropathology. J Neurooncol. 2008;89:287-311.<br />
7. Giannini C, Scheithauer BW, Burger PC, et al. Cellular proliferation in<br />
pilocytic and diffuse astrocytomas. J Neuropathol Exp Neurol. 1999;58:46-53.<br />
8. Hsu DW, Louis DN, Efird JT, et al. Use <strong>of</strong> MIB-1 (Ki-67) immunoreactivity<br />
in differentiating grade II and grade III gliomas. J Neuropathol Exp<br />
Neurol. 1997;56:857-65.<br />
9. Moskowitz SI, Jin T, Prayson RA. Role <strong>of</strong> MIB1 in predicting survival in<br />
patients with glioblastomas. J Neurooncol. 2006;76:193-200.<br />
10. Brat DJ, Van Meir EG. Vaso-occlusive and prothrombotic mechanisms<br />
associated with tumor hypoxia, necrosis, and accelerated growth in glioblastoma.<br />
Lab Invest. 2004;84:397-405.<br />
11. Rong Y, Durden DL, Van Meir EG, et al. “Pseudopalisading” necrosis in<br />
glioblastoma: a familiar morphologic feature that links vascular pathology,<br />
hypoxia, and angiogenesis. J Neuropathol Exp Neurol. 2006;65:529-539.<br />
12. Brat DJ, Castellano-Sanchez AA, Hunter SB, et al. Pseudopalisades in<br />
glioblastoma are hypoxic, express extracellular matrix proteases, and are<br />
formed by an actively migrating cell population. Cancer Res. 2004;64:920-927.<br />
13. Dong S, Nutt CL, Betensky RA, et al. Histology-based expression<br />
pr<strong>of</strong>iling yields novel prognostic markers in human glioblastoma. J Neuropathol<br />
Exp Neurol. 2005;64:948-955.<br />
14. Holash J, Maisonpierre PC, Compton D, et al. Vessel cooption, regression,<br />
and growth in tumors mediated by angiopoietins and VEGF. Science.<br />
1999;284:1994-1998.<br />
15. Tehrani M, Friedman T, Olson JJ, et al. Intravascular thrombosis in<br />
central nervous system malignancies: a potential role in astrocytoma progression<br />
to glioblastoma. Brain Pathol. 2008;18:164-171.<br />
16. Senger DR, Galli SJ, Dvorak AM, et al. Tumor cells secrete a vascular<br />
permeability factor that promotes accumulation <strong>of</strong> ascites fluid. Science.<br />
1983;219:983-985.<br />
REFERENCES<br />
17. Plate KH. Mechanisms <strong>of</strong> angiogenesis in the brain. J Neuropathol Exp<br />
Neurol. 1999;58:313-320.<br />
18. Brat DJ, Bellail AC, Van Meir EG. The role <strong>of</strong> interleukin-8 and its<br />
receptors in gliomagenesis and tumoral angiogenesis. Neuro-oncol. 2005;7:<br />
122-133.<br />
19. Straume O, Chappuis PO, Salvesen HB, et al. Prognostic importance <strong>of</strong><br />
glomeruloid microvascular proliferation indicates an aggressive angiogenic<br />
phenotype in human cancers. Cancer Res. 2002;62:6808-6811.<br />
20. Cairncross G, Berkey B, Shaw E, et al. Phase III trial <strong>of</strong> chemotherapy<br />
plus radiotherapy compared with radiotherapy alone for pure and mixed<br />
anaplastic oligodendroglioma: Intergroup Radiation Therapy <strong>Oncology</strong> Group<br />
Trial 9402. J Clin Oncol. 2006;24:2707-2714.<br />
21. Smith JS, Alderete B, Minn Y, et al. Localization <strong>of</strong> common deletion<br />
regions on 1p and 19q in human gliomas and their association with histological<br />
subtype. Oncogene. 1999;18:4144-4152.<br />
22. Perry A, Fuller CE, Banerjee R, et al. Ancillary FISH analysis for 1p<br />
and 19q status: preliminary observations in 287 gliomas and oligodendroglioma<br />
mimics. Front Biosci. 2003;8:a1-a9.<br />
23. Brat DJ, Seiferheld WF, Perry A, et al. Analysis <strong>of</strong> 1p, 19q, 9p, and 10q<br />
as prognostic markers for high-grade astrocytomas using fluorescence in situ<br />
hybridization on tissue microarrays from Radiation Therapy <strong>Oncology</strong> Group<br />
trials. Neuro-oncol. 2004;6:96-103.<br />
24. Mellingh<strong>of</strong>f IK, Wang MY, Vivanco I, et al. Molecular determinants <strong>of</strong><br />
the response <strong>of</strong> glioblastomas to EGFR kinase inhibitors. N Engl J Med.<br />
2005;353:<strong>2012</strong>-2024.<br />
25. Hegi ME, Diserens AC, Gorlia T, et al. MGMT gene silencing and<br />
benefit from temozolomide in glioblastoma. N Engl J Med. 2005;352:997-<br />
1003.<br />
26. Parsons DW, Jones S, Zhang X, et al. An integrated genomic analysis <strong>of</strong><br />
human glioblastoma multiforme. Science. 2008;321:1807-1812.<br />
27. Yan H, Parsons DW, Jin G, et al. IDH1 and IDH2 mutations in gliomas.<br />
N Engl J Med. 2009;360:765-773.<br />
28. Capper D, Weissert S, Balss J, et al. Characterization <strong>of</strong> R132H<br />
mutation-specific IDH1 antibody binding in brain tumors. Brain Pathol.<br />
2010;20:245-254.<br />
29. The Cancer Genome Atlas Network. Comprehensive genomic characterization<br />
defines human glioblastoma genes and core pathways. Nature.<br />
2008;455:1061-1068.<br />
30. Verhaak RG, Hoadley KA, Purdom E, et al. Integrated genomic<br />
analysis identifies clinically relevant subtypes <strong>of</strong> glioblastoma characterized<br />
by abnormalities in PDGFRA, IDH1, EGFR, and NF1. Cancer Cell. 2010;17:<br />
98-110.<br />
31. Noushmehr H, Weisenberger DJ, Diefes K, et al. Identification <strong>of</strong> a CpG<br />
island methylator phenotype that defines a distinct subgroup <strong>of</strong> glioma.<br />
Cancer Cell. 2010;17:510-522.<br />
107
Future Directions in Glioblastoma Therapy<br />
Overview: The standard <strong>of</strong> care for both newly diagnosed and<br />
recurrent glioblastoma (GBM) patients has changed significantly<br />
in the past 10 years. Surgery followed by radiation and<br />
concurrent and adjuvant temozolomide is now the wellestablished<br />
standard treatment for newly diagnosed GBM.<br />
More recently, bevacizumab has become a mainstay <strong>of</strong> treatment<br />
for recurrent GBM. However, despite these advances and<br />
significant improvements in patient outcomes, the management<br />
and treatment <strong>of</strong> GBM patients remains a challenging<br />
and frustrating endeavor. Difficulties in interpretation <strong>of</strong> imaging<br />
changes after initial treatment, as well as the effects <strong>of</strong><br />
antiangiogenic agents like bevacizumab on MRI characteristics,<br />
can make even the determination <strong>of</strong> disease progression<br />
complicated in multiple situations. Although a high percentage<br />
<strong>of</strong> patients benefit from antiangiogenic therapy in terms<br />
<strong>of</strong> radiographic response and progression-free survival, the<br />
GLIOBLASTOMA (GBM) IS the most common malignant<br />
brain tumor in adults, with approximately 15,000<br />
new GBM cases diagnosed each year in the United States.<br />
GBM is classified by the World Health Organization (WHO)<br />
as a grade 4 astrocytic tumor and is differentiated from<br />
lower grade astrocytomas by the pathologic features <strong>of</strong><br />
pseudopalisading necrosis and microvascular proliferation,<br />
among others.<br />
Standard Therapy for Newly Diagnosed GBM<br />
Radiographically, these tumors typically present as ring<br />
enhancing masses on contrast-enhance MRI imaging, <strong>of</strong>ten<br />
associated with marked infiltration and edema <strong>of</strong> surrounding<br />
brain. Standard treatment for newly diagnosed GBM<br />
includes maximal safe resection followed by chemoradiation,<br />
as defined by the phase III study by Stupp and colleagues. 1<br />
This standard treatment includes fractionated external<br />
beam radiation given over 6 weeks to a dose <strong>of</strong> 60 Gy<br />
combined with daily temozolomide at a dose <strong>of</strong> 75 mg/m 2 .<br />
This initial concurrent chemoradiation phase is followed by<br />
adjuvant temozolomide treatment at a dose <strong>of</strong> 150 mg/m 2<br />
day 1 through 5 out <strong>of</strong> 28 for the first cycle, with dose<br />
escalation to 200 mg/m 2 day 1 through 5 out <strong>of</strong> 28 for<br />
subsequent cycles. The duration <strong>of</strong> adjuvant therapy in the<br />
initial study was 6 months, but subsequent studies have<br />
continued adjuvant treatment out to 12 months or more.<br />
Although this regimen demonstrated significant improvement<br />
in median survival compared with radiation alone<br />
(14.6 months compared with 12.1 months, p � 0.001) and a<br />
benefit in 2-year survival rates (26.5% compared with<br />
10.4%), long-term survival for GBM patients remains disappointing.<br />
In a follow-up analysis <strong>of</strong> the initial phase III<br />
study, 2-, 3-, and 5-year survival rates were 27.2%, 16.0%,<br />
and 9.8%, respectively in the temozolomide chemoradiation<br />
group. 2 More recently, a phase III study comparing two dose<br />
schedules <strong>of</strong> adjuvant temozolomide after chemoradiation<br />
(RTOG 0525) was reported with no significant difference in<br />
overall survival between the standard and dose-dense temozolomide<br />
groups. 3 Together, these studies demonstrate that<br />
temozolomide chemoradiation has a significant benefit in<br />
newly diagnosed GBM compared to radiation alone (and<br />
prior chemotherapy/radiation combinations). However, de-<br />
108<br />
By Howard Colman, MD, PhD<br />
effects <strong>of</strong> bevacizumab on prolonging overall survival remain<br />
controversial. Furthermore, tumor progression after treatment<br />
with antiangiogenic agents carries a particularly poor prognosis<br />
and there is a general lack <strong>of</strong> effective therapies for this<br />
group <strong>of</strong> patients. These limitations in terms <strong>of</strong> standard<br />
treatments contrast with a relative wealth <strong>of</strong> new information<br />
regarding the molecular underpinnings <strong>of</strong> GBM. Data from<br />
several large-scale efforts to molecularly pr<strong>of</strong>ile GBM tumors<br />
including The Cancer Genome Atlas (TCGA) project have<br />
helped define specific molecular subtypes <strong>of</strong> GBM with distinct<br />
biology and clinical outcomes. These findings are helping<br />
to refine our understanding <strong>of</strong> the molecular heterogeneity<br />
and pathogenesis <strong>of</strong> these tumors and provide a basis for the<br />
future development <strong>of</strong> rational and targeted therapies for<br />
specific tumor subtypes.<br />
spite this improved standard <strong>of</strong> care, the majority <strong>of</strong> patients<br />
still die <strong>of</strong> their disease before 2 years.<br />
The Problem <strong>of</strong> Pseudoprogression<br />
One aspect <strong>of</strong> the management <strong>of</strong> GBM patients that<br />
causes significant uncertainty is the issue <strong>of</strong> interpretation<br />
<strong>of</strong> MRI imaging and disease progression. Gadoliniumenhanced<br />
MRI remains the mainstay <strong>of</strong> assessment <strong>of</strong> GBM<br />
status. However, changes in enhancement on MRI reflect<br />
changes in vascular permeability, and are thus an imaging<br />
surrogate for disease status rather than a reflection <strong>of</strong> true<br />
tumor burden. Other causes <strong>of</strong> increased vascular permeability<br />
can increase the extent and degree <strong>of</strong> enhancement on<br />
MRI, resulting in a potentially incorrect determination <strong>of</strong><br />
tumor progression. This need to differentiate “pseudoprogression”<br />
(a radiographic increase in MRI enhancement or<br />
edema resulting from tissue injury or inflammation without<br />
increased tumor activity) from true tumor progression represents<br />
a clinical and diagnostic challenge, particularly in<br />
patients who have recently completed radiation with concurrent<br />
chemotherapy. 4 Pseudoprogression typically stabilizes<br />
or improves without alteration in treatment within 3 to 6<br />
months. Pseudoprogression rates have been reported between<br />
9% and 31% in patients with malignant glioma after<br />
chemoradiation, with significantly higher incidence in tumors<br />
with MGMT promoter methylation. 5,6<br />
Since conventional MRI <strong>of</strong>ten cannot distinguish true<br />
progression from pseudoprogression, a lot <strong>of</strong> attention has<br />
been focused on other advanced imaging modalities. Although<br />
MRI perfusion, MRI spectroscopy, and PET can add<br />
diagnostic information and aid in clinical decision making,<br />
none <strong>of</strong> these are sufficiently sensitive or specific to definitively<br />
determine the underlying etiology. 7 An incorrect determination<br />
<strong>of</strong> progression in these patients can lead to<br />
From the Department <strong>of</strong> Neurosurgery and Huntsman Cancer Institute, University <strong>of</strong><br />
Utah, Salt Lake City, UT.<br />
Author’s disclosure <strong>of</strong> potential conflicts <strong>of</strong> interest are found at the end <strong>of</strong> this article.<br />
Address reprint requests to Howard Colman, MD, PhD, 175 North Medical Drive East,<br />
Salt Lake City, UT, 84132; email: howard.colman@hci.utah.edu.<br />
© <strong>2012</strong> by <strong>American</strong> <strong>Society</strong> <strong>of</strong> <strong>Clinical</strong> <strong>Oncology</strong>.<br />
1092-9118/10/1-10
FUTURE DIRECTIONS IN GBM THERAPY<br />
premature discontinuation <strong>of</strong> an effective therapy and/or<br />
initiation <strong>of</strong> less effective therapies. In addition, treatment<br />
<strong>of</strong> pseudoprogression as progression with an alteration in<br />
therapy can lead to a misinterpretation <strong>of</strong> subsequent imaging<br />
improvement as a “response” to the new therapy. For<br />
these reasons, a clear understanding regarding the issues<br />
and limitations <strong>of</strong> current imaging is crucial for clinicians<br />
managing these patients. To help guide these decisions, the<br />
recent recommendations by the Response Assessment in<br />
Neuro-<strong>Oncology</strong> (RANO) Working Group are that progression<br />
can only be determined with certainty in the 12 weeks<br />
after initial chemoradiation by pathologic confirmation or if<br />
the majority <strong>of</strong> the new enhancement is outside the original<br />
high-dose radiation field. 8 Patients in which the determination<br />
<strong>of</strong> progression cannot be made with certainty should not<br />
be enrolled in clinical trials during this time period, and<br />
patients who remain clinically stable or in whom pseudoprogression<br />
is suspected should continue with their current<br />
therapy.<br />
Standard Therapies for Recurrent GBM<br />
Following disease progression, historic outcomes for treatment<br />
<strong>of</strong> recurrent disease have been disappointing. In<br />
pooled analyses <strong>of</strong> multiple clinical trials for recurrent<br />
GBM, response rates have ranged from 4% to 7%; 6-month<br />
KEY POINTS<br />
● Radiation therapy with concurrent temozolomide and<br />
adjuvant temozolomide is the established standard<br />
treatment for newly diagnosed glioblastoma (GBM).<br />
● Pseudoprogression is an important complicating factor<br />
in the interpretation <strong>of</strong> MRI imaging <strong>of</strong> GBM<br />
patients and practitioners need to have a good awareness<br />
<strong>of</strong> this issue and familiarity with current recommendations<br />
by the Response Assessment in Neuro-<br />
<strong>Oncology</strong> (RANO) Working Group.<br />
● Although multiple therapeutic options exist for recurrent<br />
GBM, bevacizumab monotherapy remains the<br />
most common standard salvage therapy. However,<br />
bevacizumab treatment is associated with modest<br />
effects on overall survival, difficulties in imaging<br />
interpretation <strong>of</strong> progression, and the development <strong>of</strong><br />
resistance and poor prognosis at progression.<br />
● Understanding <strong>of</strong> the molecular alterations associated<br />
with distinct molecular subtypes and clinical<br />
outcome in GBM continues to increase. Key biomarkers<br />
with prognostic and/or tumor subtype associations<br />
include proneural and mesenchymal gene<br />
expression patterns, MGMT promoter methylation,<br />
IDH1 mutation, and CpG island methylation phenotype<br />
(CIMP).<br />
● The efficacy and evaluation <strong>of</strong> targeted therapy in<br />
GBM can be complicated by issues <strong>of</strong> drug permeability<br />
and difficulty in assessing biologic endpoints in<br />
the tumor, and these aspects have to be addressed in<br />
current and future clinical trials to increase the<br />
likelihood <strong>of</strong> further improvements in outcome.<br />
progression-free survival (PFS) rates ranged from 9% to<br />
15%; and median overall survival (OS) rates ranged from<br />
5–7 months. 9,10 Standard treatment options for recurrent<br />
disease are more varied than for newly diagnosed patients,<br />
with recent United States Food and Drug Administration<br />
(FDA) approvals adding to the number and types <strong>of</strong> standard<br />
treatment options. Reresection alone or with placement<br />
<strong>of</strong> polifeprosan 20 with carmustine is an option for patients<br />
with recurrence in noneloquent areas <strong>of</strong> the brain. 11,12<br />
Treatment with additional cytotoxic chemotherapy is frequently<br />
considered, with several potential choices. Retreatment<br />
with alternative schedules <strong>of</strong> temozolomide has some<br />
efficacy, with one study suggesting improved 6-month PFS<br />
and one-year survival in patients who had been stable for<br />
at least several months after completing standard adjuvant<br />
temozolomide, but with poorer outcomes in patients who<br />
were treated with an alternative temozolomide schedule at<br />
the time <strong>of</strong> failure on standard dose temozolomide. 13 Several<br />
other alkylating agents (CCNU, BCNU) and other cytotoxic<br />
chemotherapies (carboplatin, irinotecan, etoposide) are also<br />
options for initial or subsequent progressions. Evaluation <strong>of</strong><br />
data from the control arm using CCNU at 100–130 mg/m 2<br />
every 6 weeks from a recent phase III study demonstrated<br />
a modest radiographic response rate (4.3%) but a relatively<br />
good PFS-6 rate <strong>of</strong> 19%, 14 which was somewhat better than<br />
the enzastaurin (experimental) arm (11.1%, p � 0.13) and<br />
highlighted the potential activity <strong>of</strong> CCNU in this situation.<br />
Very recently, the FDA approved a device, the NovoTTF-<br />
100A system, for treatment <strong>of</strong> recurrent GBM based on a<br />
randomized study demonstrating similar survival outcomes<br />
in patients treated with this system compared with chemotherapy<br />
<strong>of</strong> the physicians’ choice.<br />
Antiangiogenic Therapies<br />
Glioblastoma has long been recognized as a promising<br />
tumor for antiangiogenic treatments because <strong>of</strong> its high<br />
vascularity. Perhaps the most significant development in<br />
the treatment <strong>of</strong> recurrent GBM in recent years has been<br />
the development <strong>of</strong> antiangiogenic therapies and the FDA<br />
granting <strong>of</strong> accelerated approval in March 2009 for bevacizumab.<br />
Although this agent has altered the standard options<br />
and approach for recurrent GBM, it has also become<br />
increasingly clear that this treatment has limitations in<br />
terms <strong>of</strong> disease control and requires clinicians to modify<br />
their standard interpretation <strong>of</strong> MRI imaging and determination<br />
<strong>of</strong> disease progression. This approval was based on<br />
two independent phase II studies demonstrating activity.<br />
15,16 Both studies showed a high radiographic response<br />
rate and an improvement in 6-month PFS with bevacizumab<br />
treatment compared to prior historic controls treated with<br />
cytotoxic chemotherapies and other targeted agents. In the<br />
multicenter BRAIN study, use <strong>of</strong> bevacizumab resulted in a<br />
radiographic response rate <strong>of</strong> 28% (95% CI 18.5%, 40.3%)<br />
and the median duration <strong>of</strong> response was 4.2 months (95%<br />
CI 3.0%, 5.7%). 15 Bevacizumab treatment also resulted in an<br />
increase in the percentage <strong>of</strong> patients treated with stable or<br />
decreasing corticosteroid dosages. A second study reported<br />
by Kreisl and colleagues demonstrated similar results with<br />
single agent bevacizumab in which 29% <strong>of</strong> patients achieved<br />
6 months <strong>of</strong> PFS. 16 Based on these results, two phase III<br />
studies testing the addition <strong>of</strong> bevacizumab to temozolomide<br />
chemoradiation in newly diagnosed GBM have been completed<br />
or are in progress, with results expected within the<br />
109
next 2 to 3 years. In addition to bevacizumab, several other<br />
agents targeting vascular endothelial growth factor (VEGF)<br />
or its receptor have been tested in clinical trials. In general,<br />
reported radiographic response rates and PFS rates have<br />
been similar or lower with these agents compared with<br />
bevacizumab. 17,18 In one recent phase III study testing the<br />
VEGF receptor inhibitor cediranib as monotherapy or with<br />
lomustine compared with lomustine alone in recurrent GBM<br />
found no significant difference in median PFS between these<br />
groups. 19<br />
The Problem <strong>of</strong> Bevacizumab Failure<br />
Although bevacizumab and other antiangiogenesis agents<br />
have demonstrated improved radiographic response rates<br />
and PFS rates in recurrent GBM compared to prior cytotoxic<br />
and targeted therapies, it is less clear that these agents are<br />
really improving OS. In addition, multiple studies to date<br />
have failed to demonstrate a benefit <strong>of</strong> combination therapy<br />
with bevacizumab. In the setting <strong>of</strong> bevacizumab failure,<br />
studies <strong>of</strong> various salvage therapies have demonstrated low<br />
radiographic response rates and 6-month PFS rates in the<br />
0% to 2% range. 20,21 This dismal prognosis indicates that<br />
progression in the setting <strong>of</strong> antiangiogenic therapy likely<br />
represents a change in biology <strong>of</strong> the tumor. However, the<br />
mechanisms <strong>of</strong> resistance and strategies for optimizing<br />
treatment at this stage <strong>of</strong> disease remain an active area <strong>of</strong><br />
investigation. Because <strong>of</strong> concern regarding rebound edema<br />
with discontinuation <strong>of</strong> bevacizumab, many practitioners<br />
tend to continue bevacizumab in this setting despite radiographic<br />
progression. Although data related to the question<br />
<strong>of</strong> continuation or discontinuation <strong>of</strong> bevacizumab at the<br />
time <strong>of</strong> bevacizumab progression continues to evolve, a<br />
recent single institution meta-analysis suggested an improved<br />
outcome with continuation <strong>of</strong> bevacizumab in this<br />
situation. 22<br />
Opportunities and Barriers for Improved<br />
Treatment <strong>of</strong> GBM<br />
Molecular subtypes and candidate therapeutic targets in<br />
GBM. One <strong>of</strong> the recent therapeutic themes in oncology is<br />
that the identification <strong>of</strong> signature molecular alterations in<br />
particular tumors in combination with drugs that potently<br />
and specifically target that alteration can have dramatic<br />
effects on patient outcome. The term “oncogene addiction”<br />
describes the hypothesis that tumor growth and survival is<br />
dependent on activity <strong>of</strong> one key gene or pathway 23 and can<br />
be seen in successes <strong>of</strong> targeted therapies in specific tumor<br />
subtypes such as imatinib for BCR-ABL-positive CML or<br />
BRAF inhibitors for metastatic melanoma with BRAF mutations.<br />
Although similar dramatic advances in GBM have<br />
not been observed yet, notable increases in our knowledge <strong>of</strong><br />
the molecular underpinnings <strong>of</strong> GBM are raising hopes for<br />
targeted breakthroughs in therapy.<br />
Strong data indicate that the pathologic entity defined<br />
under the single WHO diagnosis <strong>of</strong> GBM consists <strong>of</strong> multiple<br />
molecular subtypes <strong>of</strong> tumors. These tumors all share the<br />
histopathologic features <strong>of</strong> GBM, but demonstrate very<br />
distinct biology and molecular ontology. These data may<br />
point the way to more effective or “personalized” treatment<br />
based on the molecular features <strong>of</strong> an individual patient’s<br />
tumor. 24 In particular, several studies have identified gene<br />
expression differences between GBM subtypes. One prominent<br />
subtype, referred to as “mesenchymal,” is characterized<br />
110<br />
by increased levels <strong>of</strong> expression <strong>of</strong> genes associated with<br />
mesenchymal differentiation, extracellular matrix, invasion,<br />
and angiogenesis. These “mesenchymal” tumors are associated<br />
with worse prognosis and may be associated with<br />
“primary” GBMs. Another robust gene expression GBM<br />
subtype identified and validated in multiple data sets is<br />
characterized by increased expression <strong>of</strong> genes associated<br />
with normal neural tissues or neural development, and has<br />
been called “proneural.” These proneural tumors are associated<br />
with better prognosis and “secondary” GBMs. 25-27<br />
The recent integrated molecular analysis from The Cancer<br />
Genome Atlas Network (TCGA) effort in GBM further found<br />
that specific somatic mutation and DNA copy number alterations<br />
were associated with specific gene expression subtypes.<br />
27 Proneural tumors were associated with significantly<br />
higher rates <strong>of</strong> point mutations in IDH1 (p � 0.01) or p53<br />
genes (p � 0.1), and amplification <strong>of</strong> PDGRA (p � 0.01). In<br />
contrast, mesenchymal tumors were characterized by higher<br />
rates <strong>of</strong> chromosomal deletions involving NF1 or mutations<br />
in NF1. Another gene expression subtype called Classical<br />
was associated with higher rates <strong>of</strong> amplification <strong>of</strong><br />
chromosome 7 (including amplification <strong>of</strong> EGFR) and loss<br />
<strong>of</strong> chromosome 10. Since these molecular alterations are<br />
associated with differences in clinical behavior and prognosis,<br />
these and other less-dominant molecular alterations<br />
are potential candidates for therapies targeting particular<br />
tumor subtypes.<br />
Predictive and Prognostic Biomarkers in GBM<br />
HOWARD COLMAN<br />
In addition to a better understanding <strong>of</strong> the molecular<br />
pathogenesis <strong>of</strong> gliomas, the growing molecular pr<strong>of</strong>iling<br />
data have led to the identification <strong>of</strong> biomarkers that are<br />
beginning to be used to help guide patient therapy. <strong>Clinical</strong>ly<br />
relevant biomarkers can be roughly divided into those<br />
that are prognostic and predictive. Prognostic markers are<br />
associated with patient outcome or natural history <strong>of</strong> the<br />
disease in patients without treatment or in patients receiving<br />
nontargeted therapies. Predictive biomarkers are associated<br />
with differential outcome to treatment with a specific<br />
targeted therapy. Recent studies have identified several<br />
molecular markers prognostic <strong>of</strong> patient outcome to standard<br />
therapy (radiation and temozolomide) in GBM. These<br />
include MGMT promoter methylation 28 and a multigene<br />
predictor based on gene expression levels <strong>of</strong> multiple<br />
genes. 25 Mutation in the genes IDH1 and IDH2 was recently<br />
described in low- and high-grade gliomas. 29 Mutation <strong>of</strong><br />
IDH1 or IDH2 is observed in a small subset <strong>of</strong> GBM tumors<br />
and is strongly associated with the proneural gene expression<br />
phenotype and better prognosis. In addition, evidence<br />
from TCGA has also identified an epigenetically defined<br />
tumor subset that overlaps with IDH1 mutation. This<br />
CIMP, is analogous to prior descriptions <strong>of</strong> CIMP in colon<br />
cancer and other tumors, although the identity <strong>of</strong> the genes<br />
methylated in GBM is distinct from these other tumors. 30<br />
CIMP-positive tumors are positively associated with IDH1<br />
mutation and show improved prognosis, compared to CIMPnegative<br />
tumors. Future work is required to define the<br />
causative role <strong>of</strong> CIMP, IDH mutation, and their relation to<br />
gene expression phenotype and gliomagenesis in GBM and<br />
lower grade gliomas.
FUTURE DIRECTIONS IN GBM THERAPY<br />
Challenges for the Evaluation and Efficacy <strong>of</strong><br />
Targeted Therapies in GBM<br />
As in other solid tumors, there are currently a large<br />
number <strong>of</strong> ongoing trials in GBM testing various targeted<br />
therapies and other approaches to overcoming resistance<br />
including those targeting DNA repair, receptor tyrosine<br />
kinases and other signaling pathways, and pathways involved<br />
in tumor stem-cell maintenance and regulation.<br />
Unfortunately, the use <strong>of</strong> targeted therapies in brain tumors<br />
and their evaluation in clinical trials face some additional<br />
challenges that are different than with solid tumors<br />
elsewhere in the body. First, many small molecule inhibitors<br />
have relatively poor penetration into the brain and<br />
cerebrospinal fluid, resulting in lower effective dose. Second,<br />
evaluating the biologic effects <strong>of</strong> a drug on its target in<br />
brain tumors is challenging because <strong>of</strong> its anatomic location<br />
and increased difficulty in doing repeat biopsies during or<br />
Author’s Disclosure <strong>of</strong> Potential Conflicts <strong>of</strong> Interest<br />
Author<br />
Employment or<br />
Leadership<br />
Positions<br />
Consultant or<br />
Advisory Role<br />
Howard Colman Castle<br />
Biosciences<br />
1. Stupp R, Mason WP, van den Bent MJ, et al. Radiotherapy plus<br />
concomitant and adjuvant temozolomide for glioblastoma. N Engl J Med.<br />
2005;352:987-996.<br />
2. Stupp R, Hegi ME, Mason WP, et al. Effects <strong>of</strong> radiotherapy with<br />
concomitant and adjuvant temozolomide versus radiotherapy alone on survival<br />
in glioblastoma in a randomised phase III study: 5-year analysis <strong>of</strong> the<br />
EORTC-NCIC trial. Lancet. 2009;10:459-466.<br />
3. Gilbert MR, et al. M.P. RTOG 0525: A randomized phase III trial<br />
comparing standard adjuvant temozolomide (TMZ) with a dose-dense (dd)<br />
schedule in newly diagnosed glioblastoma (GBM). Neuro-<strong>Oncology</strong>. 2011;<br />
13:3s (suppl; abstr 46).<br />
4. de Wit MC, de Bruin HG, Eijkenboom W, et al. Immediate postradiotherapy<br />
changes in malignant glioma can mimic tumor progression.<br />
Neurology. 2004;63:535-537.<br />
5. Brandsma D, van den Bent MJ. Pseudoprogression and pseudoresponse<br />
in the treatment <strong>of</strong> gliomas. Curr Opin Neurol Neurosurg. 2009;22:633-638.<br />
6. Brandes AA, Franceschi E, Tosoni A, et al. MGMT promoter methylation<br />
status can predict the incidence and outcome <strong>of</strong> pseudoprogression after<br />
concomitant radiochemotherapy in newly diagnosed glioblastoma patients.<br />
J Clin Oncol. 2008;26:2192-2197.<br />
7. Brandsma D, Stalpers L, Taal W, et al. <strong>Clinical</strong> features, mechanisms, and<br />
management <strong>of</strong> pseudoprogression in malignant gliomas. Lancet. 2008;9:453-461.<br />
8. Wen PY, Macdonald DR, Reardon DA, et al. Updated response assessment<br />
criteria for high-grade gliomas: response assessment in neuro-oncology<br />
working group. J Clin Oncol. 2010;28:1963-1972.<br />
9. Wong ET, Hess KR, Gleason MJ, et al. Outcomes and prognostic factors<br />
in recurrent glioma patients enrolled onto phase II clinical trials. J Clin<br />
Oncol. 1999;17:2572-2578.<br />
10. Lamborn KR, Yung WK, Chang SM, et al. Progression-free survival: an<br />
important end point in evaluating therapy for recurrent high-grade gliomas.<br />
Neuro-<strong>Oncology</strong>. 2008;10:162-170.<br />
11. Barker FG 2nd, Chang SM, Gutin PH, et al. Survival and functional<br />
status after resection <strong>of</strong> recurrent glioblastoma multiforme. Neurosurgery.<br />
1998;42:709-720.<br />
12. Brem H, Piantadosi S, Burger PC, et al. Placebo-controlled trial <strong>of</strong><br />
safety and efficacy <strong>of</strong> intraoperative controlled delivery by biodegradable<br />
polymers <strong>of</strong> chemotherapy for recurrent gliomas. The Polymer-brain Tumor<br />
Treatment Group. Lancet. 1995;345:1008-1012.<br />
13. Perry JR, Bélanger K, Mason WP, et al. Phase II trial <strong>of</strong> continuous<br />
dose-intense temozolomide in recurrent malignant glioma: RESCUE study.<br />
J Clin Oncol. 2010;28:2051-2057.<br />
14. Wick W, Puduvalli VK, Chamberlain MC, et al. Phase III study <strong>of</strong><br />
after treatment compared to some other systemic cancers.<br />
Nonetheless, if experience from other solid tumors is any<br />
indication, future successful clinical trials in recurrent<br />
GBM will likely need to have a larger emphasis on pharmacodynamics<br />
and biologic tissue effects as important endpoints.<br />
In summary, the dramatic increase in our understanding<br />
<strong>of</strong> the molecular underpinnings and subtypes <strong>of</strong> glioblastoma,<br />
along with increases in the number and type <strong>of</strong><br />
available small molecule inhibitors, indicates that improvement<br />
in patient outcome in GBM with appropriate targeted<br />
therapy is indeed feasible. However, the success and rapidity<br />
<strong>of</strong> these developments will hinge on successful integration<br />
<strong>of</strong> molecular and preclinical data with the more efficient<br />
and innovative clinical trial designs in order to identify<br />
those key pathways and drugs that will have the highest<br />
benefit for particular GBM subtypes.<br />
Stock<br />
Ownership Honoraria<br />
REFERENCES<br />
Research<br />
Funding<br />
Expert<br />
Testimony<br />
Other<br />
Remuneration<br />
enzastaurin compared with lomustine in the treatment <strong>of</strong> recurrent intracranial<br />
glioblastoma. J Clin Oncol. 2010;28:1168-1174.<br />
15. Friedman HS, Prados MD, Wen PY, et al. Bevacizumab alone and in<br />
combination with irinotecan in recurrent glioblastoma. J Clin Oncol. 2009;<br />
27:4733-4740.<br />
16. Kreisl TN, Kim L, Moore K, et al. Phase II trial <strong>of</strong> single-agent<br />
bevacizumab followed by bevacizumab plus irinotecan at tumor progression<br />
in recurrent glioblastoma. J Clin Oncol. 2009;27:740-745.<br />
17. Norden AD, Drappatz J, Wen PY. Antiangiogenic therapies for highgrade<br />
glioma. Nat Rev Neurol. 2009;5:610-620.<br />
18. Reardon DA, Turner S, Peters KB, et al. A review <strong>of</strong> VEGF/VEGFRtargeted<br />
therapeutics for recurrent glioblastoma. J Natl Compr Canc Netw.<br />
2011;9:414-427.<br />
19. Batchelor T, Mulholland P, Nyns B, et al. The efficacy <strong>of</strong> cediranib as<br />
monotherapy and in combination with lomustine compared to lomustine alone<br />
in patients with recurrent glioblastoma: A phase III randomized study. Neuro<br />
Oncol. 2010;12(suppl 4).<br />
20. Iwamoto FM, Abrey LE, Beal K, et al. Patterns <strong>of</strong> relapse and prognosis<br />
after bevacizumab failure in recurrent glioblastoma. Neurology. 2009;73:1200-1206.<br />
21. Quant EC, Norden AD, Drappatz J, et al. Role <strong>of</strong> a second chemotherapy<br />
in recurrent malignant glioma patients who progress on bevacizumab. Neuro<br />
Oncol. 2009;11:550-555.<br />
22. Reardon DA, Vredenburgh JJ, Desjardins A, et al. Bevacizumab (BV)<br />
continuation following BV progression: Meta-analysis <strong>of</strong> five consecutive recurrent<br />
glioblastoma (GBM) trials. J Clin Oncol. 2011;29:45s(suppl; abstr 2030).<br />
23. Weinstein IB, Joe A. Oncogene addiction. Cancer Res. 2008;68:3077-3080.<br />
24. Colman H, Aldape K. Molecular predictors in glioblastoma: toward<br />
personalized therapy. Arch Neurol. 2008;65:877-883.<br />
25. Colman H, Zhang L, Sulman EP, et al. A multigene predictor <strong>of</strong><br />
outcome in glioblastoma. Neuro-<strong>Oncology</strong>. 2010;12:49-57.<br />
26. Phillips HS, Kharbanda S, Chen R, et al. Molecular subclasses <strong>of</strong><br />
high-grade glioma predict prognosis, delineate a pattern <strong>of</strong> disease progression,<br />
and resemble stages in neurogenesis. Cancer Cell. 2006;9:157-173.<br />
27. Verhaak RG, Hoadley KA, Purdom E, et al. Integrated genomic analysis<br />
identifies clinically relevant subtypes <strong>of</strong> glioblastoma characterized by abnormalities<br />
in PDGFRA, IDH1, EGFR, and NF1. Cancer Cell. 2010;17:98-110.<br />
28. Hegi ME, Diserens AC, Gorlia T, et al. MGMT gene silencing and benefit<br />
from temozolomide in glioblastoma. N Engl J Med. 2005;352:997-1003.<br />
29. Yan H, Parsons DW, Jin G, et al. IDH1 and IDH2 mutations in gliomas.<br />
N Engl J Med. 2009;360:765-773.<br />
30. Noushmehr H, Weisenberger DJ, Diefes K, et al. Identification <strong>of</strong> a CpG<br />
island methylator phenotype that defines a distinct subgroup <strong>of</strong> glioma.<br />
Cancer Cell. 2010;17:510-522.<br />
111
Establishing the Standard <strong>of</strong> Care for<br />
Patients with Newly Diagnosed and<br />
Recurrent Glioblastoma<br />
Overview: The current standard <strong>of</strong> care for patients with<br />
newly diagnosed glioblastoma includes maximal safe tumor<br />
resection followed by concurrent external-beam radiation with<br />
daily low-dose temozolomide followed by 6 to 12 months <strong>of</strong><br />
adjuvant temozolomide, typically by using a cycle <strong>of</strong> 5 consecutive<br />
days out <strong>of</strong> 28. Efforts to improve on these results<br />
from the European Organisation for Research and Treatment<br />
<strong>of</strong> Cancer (EORTC)/National Cancer Institute <strong>of</strong> Canada (NCIC)<br />
trial using either dose-dense chemotherapy strategies or<br />
combinations with signal transduction modulators have, to<br />
date, been unsuccessful. Two large international randomized<br />
trials examining the efficacy <strong>of</strong> adding bevacizumab, an antiangiogenic<br />
agent, to the standard treatment have been completed,<br />
with expectations <strong>of</strong> results within in the next 2 years.<br />
For recurrent glioblastoma, there are no firmly established<br />
GLIOBLASTOMA IS classified as a grade 4 glioma by<br />
using the now universally accepted WHO grading<br />
criteria. It is the most common malignant primary brain<br />
neoplasm, accounting for 54% <strong>of</strong> all gliomas. 1 The annual<br />
incidence in the United States is estimated to be between<br />
3 and 4 per 100,000 <strong>of</strong> the general population. Although<br />
glioblastoma is found in all age groups, the highest incidence<br />
is in middle age to elderly. Although distinct pathologic<br />
characteristics such as microvascular proliferation and necrosis<br />
are used to define the disease, recent advances in<br />
molecular characterization demonstrate that the uniform<br />
histologic criteria belie tremendous genetic heterogeneity.<br />
Primary glioblastoma, developing as a glioblastoma, has<br />
distinct molecular characteristics from the secondary glioblastoma<br />
which evolves from the progression transformation<br />
from lower grade gliomas. 2 These differences may influence<br />
treatment decisions and in the case <strong>of</strong> mutations in the<br />
IDH1 gene which is found almost exclusively in secondary<br />
glioblastoma, clear implications on prognosis. 3<br />
The prognosis for patients with glioblastoma remains<br />
poor. Overall, the 1-year survival rate in population-based<br />
studies is less than 40%, although patients enrolled on<br />
clinical trials do have better reported outcomes. 1 Several<br />
clinical prognostic factors have been identified, including<br />
age, performance status, and extent <strong>of</strong> initial tumor resection.<br />
4 Several <strong>of</strong> the important clinical prognostic factors<br />
have been combined by using recursive partitioning analysis<br />
into a six-class system, allowing easier stratification accounting<br />
in clinical trials. 5 Median survival for glioblastoma<br />
ranges from 6 months to 18 month according to clinical<br />
prognostic factors in the recursive partitioning analysis.<br />
This underscores the importance <strong>of</strong> clinical factors in outcome<br />
and the potential for misinterpretation <strong>of</strong> clinical trial<br />
outcomes if there is an imbalance <strong>of</strong> these factors. Additionally,<br />
molecular prognostic factors have been identified that<br />
are independent <strong>of</strong> the clinical factors. These include IDH1<br />
mutation, presence <strong>of</strong> a glioma-specific CpG Island hypermethylation<br />
phenotype (G-CIMP), a gene expression pr<strong>of</strong>ile,<br />
and the methylation <strong>of</strong> the promoter region <strong>of</strong> the methylguanine<br />
methyltransferase (MGMT) gene. 3,6-8<br />
112<br />
By Mark R. Gilbert, MD<br />
standards <strong>of</strong> care. Although intracavitary insertion <strong>of</strong> carmustineimpregnated<br />
polymers has been approved by the U.S. Food<br />
and Drug Administration (FDA), this strategy is not widely<br />
used. Bevacizumab has been FDA approved for recurrent<br />
glioblastoma, but no randomized trial has clearly demonstrated<br />
a survival benefit. Alternative dosing schedules <strong>of</strong><br />
temozolomide (i.e., metronomic) has modest activity even in<br />
patients with prior temozolomide exposure. <strong>Clinical</strong> trials<br />
testing small-molecule signal transduction modulators have<br />
been disappointing, although most report a small response<br />
rate, suggesting that molecularly definable tumor subpopulations<br />
may help guide treatment decisions. Successful<br />
implementation <strong>of</strong> marker-based treatment would lead to<br />
personalized care and the creation <strong>of</strong> individualized standards<br />
<strong>of</strong> care.<br />
Treatment <strong>of</strong> Newly Diagnosed Glioblastoma<br />
The seminal work <strong>of</strong> the Brain Tumor Study Group in<br />
the 1970s established the efficacy <strong>of</strong> external-beam radiation<br />
for glioblastoma. 9 Their studies demonstrated a more than<br />
doubling <strong>of</strong> survival from 3 to 4 months with surgery and<br />
corticosteroids to more than 9 months with radiation treatment.<br />
The addition <strong>of</strong> chemotherapy, typically a nitrosourea<br />
such as lomustine or carmustine, did not substantially<br />
improve survival compared with radiation alone. In fact, a<br />
large meta-analysis compiling 12 randomized trials and<br />
including more than 3,000 patients found only a 6% improvement<br />
in 1-year survival when chemotherapy was<br />
added to surgery and conventional radiation. 10<br />
However, a new standard <strong>of</strong> care for patients with glioblastoma<br />
was established with the 2005 publication <strong>of</strong> the<br />
results <strong>of</strong> a clinical trial performed by the European Organisation<br />
for Research and Treatment <strong>of</strong> Cancer (EORTC) and<br />
National Cancer Institute <strong>of</strong> Canada (NCIC). 11 This was a<br />
large, randomized phase III trial that compared regional<br />
external-beam radiation (60 Gy in 2-Gy fractions) as the<br />
standard arm with the experimental treatment comprising<br />
concurrent regional radiation (60 Gy in 2-Gy fractions) with<br />
daily temozolomide (75 mg/m 2 for 42 consecutive days). The<br />
chemoradiation was then followed by six cycles <strong>of</strong> singleagent<br />
temozolomide (150 to 200 mg/m 2 /day for 5 consecutive<br />
days <strong>of</strong> every 28 day cycle). The treatment schema is<br />
provided in Fig. 1. The clinical trial demonstrated statistically<br />
significant improvements in overall survival (14.6 vs.<br />
12.1 months), progression-free survival, and the 2-year survival<br />
rate (26.5% vs. 10%). Long-term follow-up continued to<br />
demonstrate improved outcomes with a 5-year survival rate<br />
From the Department <strong>of</strong> Neuro-oncology, University <strong>of</strong> Texas M. D. Anderson Cancer<br />
Center, Houston, TX.<br />
Author’s disclosure <strong>of</strong> potential conflicts <strong>of</strong> interest are found at the end <strong>of</strong> this article.<br />
Address reprint requests to Mark R. Gilbert, MD, Department <strong>of</strong> Neuro-oncology,<br />
University <strong>of</strong> Texas M. D. Anderson Cancer Center, 1515 Holcombe Blvd., Houston, TX<br />
77030; email: mrgilbert@mdanderson.org.<br />
© <strong>2012</strong> by <strong>American</strong> <strong>Society</strong> <strong>of</strong> <strong>Clinical</strong> <strong>Oncology</strong>.<br />
1092-9118/10/1-10
ESTABLISHING TREATMENTS FOR GLIOBLASTOMA<br />
Fig. 1. Treatment schema for the chemoradiation regimen<br />
from the EORTC/NCI-Canada <strong>Clinical</strong> Trial.<br />
<strong>of</strong> 9.8% in the combined treatment patients compared with<br />
1.9% in the radiation-only group. 12<br />
A companion laboratory correlative study to the EORTC/<br />
NCIC trial evaluating patient outcomes on the basis <strong>of</strong> the<br />
methylation status <strong>of</strong> the methylguanine methyltransferase<br />
(MGMT) gene was performed by Hegi and colleagues. 8<br />
Hypermethylation <strong>of</strong> the promoter region was postulated to<br />
decrease or eliminate gene expression and that with absent<br />
or low intracellular MGMT, a major mechanism <strong>of</strong> repair <strong>of</strong><br />
alkylating agent–mediated DNA damage. Therefore, in the<br />
presence <strong>of</strong> the methylated MGMT promoter region, the<br />
addition <strong>of</strong> temozolomide would result in a better response<br />
rate. They were able to analyze 206 and found that 45% were<br />
had MGMT promoter methylation. These studies did demonstrate<br />
an improved outcome for patients with tumors<br />
harboring methylated MGMT gene promoter regions. However,<br />
although not as marked, the outcomes for patients<br />
with unmethylated tumors experienced improvement with<br />
the addition <strong>of</strong> temozolomide, suggesting that MGMT methylation<br />
status is clearly prognostic but not fully predictive <strong>of</strong><br />
treatment benefit.<br />
These findings, correlating MGMT promoter methylation<br />
with outcomes, suggested that the MGMT enzyme may be<br />
an important therapeutic target and depletion <strong>of</strong> intracellular<br />
MGMT in tumor cells may enhance temozolomide<br />
efficacy. Prolonged exposure to temozolomide has been demonstrated<br />
to decrease MGMT activity in peripheral-blood<br />
mononuclear cells in patients receiving 21 consecutive days<br />
<strong>of</strong> treatment. 13 This approach, using a dose-dense treatment<br />
schedule, was tested in Radiation Therapy <strong>Oncology</strong> Group<br />
KEY POINTS<br />
● A standard <strong>of</strong> care consisting <strong>of</strong> concurrent radiation<br />
and temozolomide followed by adjuvant temozolomide<br />
has been established for patients with newly<br />
diagnosed glioblastoma.<br />
● Recently completed and ongoing randomized clinical<br />
trials in newly diagnosed glioblastoma seek to enhance<br />
the current therapy by adding synergistic<br />
therapies.<br />
● There are very few treatments established for recurrent<br />
glioblastoma despite extensive testing <strong>of</strong> many<br />
signal transduction modulators.<br />
● Antiangiogenic therapies look promising for recurrent<br />
glioblastoma. Bevacizumab is administered frequently<br />
in this setting.<br />
● Given the molecular heterogeneity <strong>of</strong> glioblastoma,<br />
further advances will likely require a comprehensive<br />
effort merging tumor pr<strong>of</strong>iling with targeted<br />
therapies.<br />
(RTOG) 0525, an international collaborative phase III<br />
trial. 14 Enrollment criteria included adults older than 18<br />
years with available tumor tissue blocks with more than<br />
1cm 2 <strong>of</strong> tumor and Karn<strong>of</strong>sky performance score <strong>of</strong> 70 or<br />
greater. All patients received concurrent radiation with<br />
daily temozolomide and then were randomly assigned to<br />
either standard adjuvant temozolomide (150 to 200 mg/m 2 /<br />
day for 5 consecutive days <strong>of</strong> a 28-day cycle) or dose-dense<br />
temozolomide (75 to 100 mg/m 2 /day for 21 consecutive days<br />
<strong>of</strong> a 28-day cycle). Patients could receive treatment for up<br />
to 12 cycles <strong>of</strong> adjuvant therapy. The study accrued 1,173<br />
patients with 833 undergoing random assignment. The two<br />
treatment arms were well balanced in all parameters including<br />
age, gender, performance status, extent <strong>of</strong> tumor resection<br />
and MGMT methylation status. The results, presented<br />
at the 47 th ASCO Annual Meeting (June 4–8, 2011, Chicago,<br />
IL), demonstrated that there was no improvement in either<br />
overall or progression-free survival with the use <strong>of</strong> the<br />
dose-dense temozolomide schedule. Additionally, subset<br />
analysis by MGMT methylation status did not show a<br />
selective benefit for dose-dense treatment for tumors with<br />
either methylated or unmethylated MGMT gene promoter<br />
phenotype. However, this prospective study did demonstrate<br />
that MGMT methylation status is clearly prognostic.<br />
Other strategies have been evaluated as potential enhancers<br />
<strong>of</strong> the established efficacy <strong>of</strong> the chemoradiation regimen.<br />
A series <strong>of</strong> signal transduction modulators and other<br />
similar agents have been evaluated, typically in single-arm<br />
phase II trials. Agents such as erlotinib, talampanel, and<br />
Poly ICLC have been tested with promising results compared<br />
with the historic controls provided by the EORTC<br />
study. 15,16 However, the validity <strong>of</strong> this comparison has been<br />
questioned, and concerns regarding patient population differences<br />
and the more recent availability <strong>of</strong> effective salvage<br />
regimens have limited interest in pursuing large-scale clinical<br />
trials.<br />
Interest in the use <strong>of</strong> antiangiogenic strategies has generated<br />
several large-scale randomized clinical trials. Bevacizumab,<br />
a humanized monoclonal antibody targeting<br />
vascular endothelial growth factor (VEGF) A, has demonstrated<br />
activity in patients with recurrent glioblastoma<br />
(discussed in more detail later herein). This has led to the<br />
development and completion <strong>of</strong> two randomized, doubleblind<br />
placebo-controlled trials for patients with newly diagnosed<br />
glioblastoma. The results <strong>of</strong> the two studies (RTOG<br />
0825 and AvaGLIA) are expected within the next 2 years.<br />
Additionally, data from a phase II trial adding the integrin<br />
inhibitor cilengitide to the conventional chemoradiation regimen<br />
showed an improved outcome in newly diagnosed,<br />
MGMT-methylated glioblastoma compared with historic<br />
controls and led to a phase III placebo-controlled randomized<br />
trial. 17 This study recently completed accrual and<br />
outcome results are expected soon. Other studies including<br />
randomized phase II trials evaluating mammalian target<br />
113
<strong>of</strong> rapamycin (mTOR) inhibition, VEGFR inhibition (cediranib),<br />
or poly (ADP-ribose) polymerase (PARP) inhibition<br />
are either underway or in planning.<br />
In summary, the pivotal clinical trial performed by the<br />
EORTC and NCIC using radiation and temozolomide established<br />
the current standard <strong>of</strong> care for patients with newly<br />
diagnosed glioblastoma. Efforts to improve on these results<br />
with intensification <strong>of</strong> the chemotherapy were unsuccessful.<br />
<strong>Clinical</strong> studies have been performed or are underway to<br />
determine whether there are synergistic therapies that can<br />
further enhance outcomes for this patient population and<br />
thereby establish a new standard <strong>of</strong> care.<br />
Treatment <strong>of</strong> Recurrent Glioblastoma<br />
The treatment <strong>of</strong> recurrent glioblastoma remains unsatisfactory.<br />
Tumor resection for recurrent disease does not<br />
appear to substantially affect either the 6-month progressionfree<br />
survival rate nor overall survival from the time <strong>of</strong><br />
progression. 18 However, relief <strong>of</strong> tumor-induced mass effect<br />
may be clinically beneficial and the confirmation <strong>of</strong> true<br />
recurrence (rather than treatment-related necrosis or pseudoprogression)<br />
may be extremely helpful in treatment decisions.<br />
A wide variety <strong>of</strong> approaches have been studied for recurrent<br />
glioblastoma. These include local strategies <strong>of</strong>ten requiring<br />
a surgical procedure and systemic chemotherapy<br />
treatments either as single agent or in combination.<br />
Locoregional Treatment Strategies<br />
A variety <strong>of</strong> treatment approaches have been tested targeting<br />
the tumor directly, recognizing that spread <strong>of</strong> glioblastoma<br />
outside <strong>of</strong> the nervous system is rare. Use <strong>of</strong> a<br />
carmustine-impregnated bioerodable polymer was tested for<br />
recurrent disease and in a placebo-controlled randomized<br />
trial showed a modest, but statistically significant improvement<br />
in the 6-month survival rate. 19 These findings led to<br />
FDA approval, but this treatment approach is not used<br />
widely today.<br />
Attempts to improve delivery <strong>of</strong> agents directly to tumor<br />
led to the institution <strong>of</strong> convectional enhanced delivery.<br />
Intratumoral catheters delivering small volumes <strong>of</strong> fluid<br />
under pressure have been shown to spread widely from the<br />
point <strong>of</strong> infusion. This method was used to deliver cintredekin<br />
besudotox (interleukin [IL]-13-PE38QQR, IL-13<br />
pseudomona extoxin) in a series <strong>of</strong> clinical trials. Despite<br />
strong preclinical data, there was no improvement in outcome<br />
compared with the carmustine wafer in a phase III<br />
clinical trial. 20 Alternative strategies are looking at using<br />
convection-enhanced delivery for chemotherapies, particularly<br />
those like topotecan that do not cross the blood-brain<br />
barrier with systemic delivery. 21<br />
Direct injection <strong>of</strong> vectors containing gene therapies has<br />
also been evaluated. Preclinical studies using transfection <strong>of</strong><br />
the herpes thymidine kinase gene into tumor cells by using<br />
a retroviral vector demonstrated high efficacy after administration<br />
<strong>of</strong> acyclovir or ganciclovir. However, clinical trials<br />
in both recurrent and newly diagnosed glioblastoma failed<br />
to demonstrate efficacy, potentially the consequence <strong>of</strong><br />
poor vector delivery by using direct injection into the walls<br />
<strong>of</strong> the tumor cavity. 22 More recently, replication-competent<br />
viruses are being evaluated in clinical trials, potentially<br />
reducing the problem <strong>of</strong> delivery with continued local pro-<br />
114<br />
MARK R. GILBERT<br />
duction and spread <strong>of</strong> the virus. For example, the delta-24<br />
virus, a replication-competent adenovirus that requires amplification<br />
<strong>of</strong> the Rb pathway to replicate, is currently being<br />
studied in a clinical trial. 23<br />
Systemic Chemotherapy<br />
Metronomic and dose-dense temozolomide. Dose-dense temozolomide<br />
has also been investigated in patients with<br />
recurrent glioblastoma. 24 Two alternative dosing schedules,<br />
alternating weekly schedule or 21 consecutive days <strong>of</strong> a<br />
28-day cycle, have been used most commonly. Several phase<br />
II trials have evaluated the “21 <strong>of</strong> 28-day” regimen with<br />
6-month progression-free survival rates ranging from 18% to<br />
30% even in patients with prior temozolomide exposure.<br />
Similarly, retrospective analyses <strong>of</strong> patients treated with<br />
the “week on, week <strong>of</strong>f ” dose-dense schedule with temozolomide<br />
report a 6-month progression-free survival rate as high<br />
as 36%. Both schedules are associated with a high rate <strong>of</strong><br />
lymphopenia, although this toxicity may be more pronounced<br />
with the 21 <strong>of</strong> 28-day schedule.<br />
Metronomic schedules <strong>of</strong> temozolomide have been evaluated<br />
in a prospective clinical trial. 25 Temozolomide was<br />
administered at low dose (50 mg/m 2 /day) continuously to<br />
patients who had previously been treated with standard<br />
dosing <strong>of</strong> temozolomide as a component <strong>of</strong> first-line treatment<br />
(as described previously herein for newly diagnosed<br />
glioblastoma). Interestingly, patients rechallenged after<br />
early failure (� 6 months <strong>of</strong> adjuvant temozolomide) or after<br />
a treatment-free interval experienced 6-month progressionfree<br />
survival rates <strong>of</strong> 27% and 36%, respectively. Patients<br />
with treatment failure during adjuvant treatment but more<br />
than six cycles fared poorly. These results suggest that in<br />
select subpopulations, re-treatment with temozolomide may<br />
be effective.<br />
Resistance modulation with PARP inhibitors. There is<br />
increasing interest in exploring inhibitors <strong>of</strong> PARP as a<br />
novel strategy to enhance the efficacy <strong>of</strong> temozolomide and<br />
address the recent finding that recurrent glioblastomas<br />
acquire either MSH6 mismatch gene–inactivating mutations<br />
or hypermethylation <strong>of</strong> the promoter region leading to<br />
reduced expression. 26 Several PARP inhibitors are in early<br />
clinical trials, typically in combination with a variety <strong>of</strong><br />
temozolomide dosing schedules. The dose-limiting toxicity<br />
is likely to be myelosuppression, supporting the need for<br />
pharmcodynamic studies confirming maximal synergy <strong>of</strong><br />
PARP inhibition with temozolomide-induced tumor DNA<br />
damage.<br />
Nitrosoureas and other chemotherapy agents. Nitrosoureas<br />
such as carmustine and lomustine were the standard<br />
treatment for recurrent malignant gliomas but their<br />
use steadily declined with the introduction <strong>of</strong> temozolomide,<br />
which is better tolerated and rarely causes cumulative<br />
myelotoxicity. There has been a recent resurgence <strong>of</strong> nitrosourea<br />
use after the report <strong>of</strong> a 6-month progression-free<br />
survival rate <strong>of</strong> 19% in patients with prior temozolomide<br />
exposure. 27 Other chemotherapy agents such as irinotecan,<br />
carboplatin, cisplatin and procarbazine are used but the<br />
response rates in recurrent disease have been modest.<br />
Targeted agents. As discussed previously, glioblastoma<br />
has been classified into two types primarily on the basis <strong>of</strong><br />
genetic features. Although there is some similarity in genetic<br />
alterations, such as loss <strong>of</strong> phosphotase and tensin<br />
homolog on chromosome 10 (PTEN), loss <strong>of</strong> cyclin-dependent
ESTABLISHING TREATMENTS FOR GLIOBLASTOMA<br />
Table 1. Single-Agent Targeted Therapies for Recurrent Glioblastoma<br />
Agent Study Year Target<br />
kinase inhibitor p16 INK4A and amplification <strong>of</strong> CDK 4,<br />
primary glioblastoma typically has higher incidence <strong>of</strong> epidermal<br />
growth factor receptor (EGFR) amplification with<br />
inactivation <strong>of</strong> PTEN and p16 tumor suppression genes. In<br />
contrast, secondary glioblastoma multiforme is characterized<br />
by mutation <strong>of</strong> TP53 gene and more recently, almost<br />
exclusively demonstrate mutations in the isocitrate dehydrogenase<br />
(IDH) 1 and 2 genes.<br />
Epidermal growth factor (EGF) and its receptor EGFR,<br />
platelet-derived growth factor (PDGF) A and B and their<br />
receptors PDGFR � and �, VEGF and its receptor (VEGFR),<br />
insulin-like growth factor (IGF)-1 and IGF receptor (IGFR),<br />
transforming growth factor (TGF)-�, fibroblast growth factor<br />
(FGF), and hepatocyte growth factor (HGF) are thought to<br />
be critical to the pathogenesis and survival <strong>of</strong> glioblastoma.<br />
Activation <strong>of</strong> these tyrosine kinase receptors triggers three<br />
major downstream pathways: mitogen-activated protein<br />
kinase (MAPK); phosphoinositide 3 kinase (PI3K)/Akt; and<br />
phospholipase C� (PLC�) and protein kinase C (PKC). These<br />
signaling pathways regulate cell proliferation and differentiation<br />
and prevent apoptosis, and are therefore logical<br />
targets <strong>of</strong> treatment for glioblastoma.<br />
Unfortunately, despite these molecular findings in a high<br />
percentage <strong>of</strong> glioblastoma, as indicated in Table 1, studies<br />
with single-agent signal transduction modulators demonstrated<br />
only modest results at best. In particular, even in the<br />
presence <strong>of</strong> amplification or mutation <strong>of</strong> EGFR, treatment<br />
with a potent EGFR inhibitor such as erlotinib did not result<br />
in a high response rate. Subsequent analyses suggested that<br />
response to erlotinib occurred only when the downstream<br />
component <strong>of</strong> the pathway was not already constitutively<br />
activated as indicated by the presence <strong>of</strong> a functional PTEN<br />
gene regulating Akt. 28 These results underscore the complex<br />
nature <strong>of</strong> signal transduction modulation strategies with<br />
pathway overlap and downstream effectors. Attempts and<br />
combination regimens <strong>of</strong> signal transduction modulators has<br />
been complicated by overlapping toxicities.<br />
Antiangiogenic agents. One <strong>of</strong> the hallmarks <strong>of</strong> glioblastoma<br />
is prominent angiogenesis, making this component <strong>of</strong><br />
tumor biology a logical target. The newly formed blood<br />
vessels are <strong>of</strong>ten poorly developed, with incomplete tight<br />
junction between endothelial cells and tortuous paths with<br />
blind loops. These features lead to peritumoral edema and<br />
account for much <strong>of</strong> the imaging enhancement by leakage <strong>of</strong><br />
systemic administration <strong>of</strong> contrast material before either<br />
computed tomography or magnetic resonance imaging. A<br />
Trial<br />
Phase<br />
No. <strong>of</strong><br />
Patients<br />
6-Month<br />
Progression-Free<br />
Survival (%)<br />
Erlotinib van den Bent29 2009 EGFR II 110 11<br />
Gefitinib Rich30 2004 EGFR II 53 13; 14<br />
Imatinib Raymond31 2008 C-ABL, C-KIT, PDGFR II 51 16<br />
Pazopanib Iwamoto32 2010 VEGFR, PDGFR II 35 3<br />
Vorinostat Galanis33 2009 HDAC II 66 15<br />
Tipifarnib Cloughesy34 2006 Farnesyltransferase II 67 12<br />
Galanis<br />
Temsirolimus<br />
35 2005 II 65 8<br />
Chang36 2005 mTOR 41 3<br />
Wick<br />
Enzastaurin<br />
27 2010 III 174 11<br />
Kreisl37 2010 PKC I/II 72 7<br />
Abbreviations: EGFR, epidermal growth factor receptor; C-ABL, a non-receptor protein tyrosine kinase; C-KIT, a cell surface protein that binds stem cell factor;<br />
C-MET, met proto-oncogene; PDGFR, platelet-derived growth factor receptor; VEGFR, vascular endothelial growth factor receptor; HDAC, histone deacetylase; mTOR,<br />
mammalian target <strong>of</strong> rapamycin; PKC, protein kinase C.<br />
variety <strong>of</strong> antiangiogenic agents have been tested, as outlined<br />
in Table 2.<br />
Bevacizumab, a humanized monoclonal antibody, works<br />
by sequestering the circulating ligand VEGF-A, resulting in<br />
angiogenesis inhibition. Several phase II studies have been<br />
performed in patients with recurrent glioblastoma. Most<br />
studies demonstrate a response rate ranging from 25% to<br />
40% and a 6-month progression-free survival in the same<br />
range. Importantly, nearly all patients receiving bevacizumab<br />
are able to either reduce or stop corticosteroid use.<br />
These findings, most notably in a multicenter randomized<br />
noncomparative phase II trial, led to the accelerated approval<br />
<strong>of</strong> bevacizumab for patients with recurrent glioblastoma.<br />
39<br />
A variety <strong>of</strong> other antiangiogenic agents have been evaluated<br />
including aflibercept (VEGF-Trap), a decoy VEGFR<br />
fused to the Fc portion <strong>of</strong> an immunoglobulin molecule.<br />
However, results <strong>of</strong> a phase II trial in recurrent glioblastoma<br />
revealed only modest activity. Cediranib is a small-molecule<br />
tyrosine kinase inhibitor <strong>of</strong> the spectrum <strong>of</strong> VEGFR that<br />
showed early efficacy, but a subsequent phase III trial with<br />
lomustine failed to demonstrate added benefit from the<br />
cediranib. Carbozantinib (XL184), a small-molecule agent<br />
that targets both VEGFR and c-Met also showed efficacy in<br />
early studies, but systemic toxicities <strong>of</strong>ten precluded extended<br />
use. Cilengitide is a novel antiangiogenic agent that<br />
targets the � v� 3 integrin that is required for endothelial cell<br />
migration for neovascularization. There may be additional<br />
tumor signal pathway effects that further enhance efficacy.<br />
Single-agent activity has been modest, but as described<br />
herein, synergistic efficacy with chemoradiation is being<br />
tested in newly diagnosed glioblastoma.<br />
Table 2. Antiangiogenic Therapies Investigated for<br />
Recurrent Glioblastoma<br />
Agent Study Year Target<br />
Trial<br />
Phase<br />
No. <strong>of</strong><br />
Patients<br />
6-Month<br />
Progression-Free<br />
Survival (%)<br />
Cediranib Batchelor 38 2010 pan-VEGFR II 31 26<br />
Bevacizumab Friedman 39 2009 VEGF-A II 85 36<br />
Cilengitide Reardon 40 2008 � v� 3 integrin 81 15<br />
Gilbert 41 2011 � v� 5 integrin II 26 12<br />
XL-184 Wen 42 2010 VEGFR, C-MET II 124 21<br />
Abbreviations: VEGFR, vascular endothelial growth factor receptor; C-MET,<br />
met proto-oncogene; EGFR, epidermal growth factor receptor; VEGF, vascular<br />
endothelial growth factor.<br />
115
Conclusion<br />
A standard <strong>of</strong> care has been established for patients with<br />
newly diagnosed glioblastoma on the basis <strong>of</strong> level 1 evidence<br />
for a randomized clinical trial. Current investigations<br />
are exploring agents to add to the established chemoradiation<br />
regimen that would augment activity without a marked<br />
increase in toxicity. Many <strong>of</strong> these trials are also including<br />
molecular characterization <strong>of</strong> the tumors in an effort to<br />
define patient subpopulations likely to benefit (or not benefit)<br />
from the new regimen, thereby enhancing the risk to<br />
benefit for individual patients.<br />
Author’s Disclosure <strong>of</strong> Potential Conflicts <strong>of</strong> Interest<br />
Author<br />
Employment or<br />
Leadership<br />
Positions<br />
Consultant or<br />
Advisory Role<br />
Mark R. Gilbert Abbott<br />
Laboratories;<br />
Genentech;<br />
GlaxoSmithKline;<br />
Merck<br />
1. 2005-2006 Statistical Report: Primary Brain Tumors in the United<br />
States Statistical Report, 1998-2002 (Years Data Collected). Hinsdale, IL:<br />
Central Brain Tumor Registry <strong>of</strong> the United States; 2006.<br />
2. Wen PY, Kesari S. Malignant gliomas in adults. N Engl J Med.<br />
2008;359:492-507.<br />
3. Yan H, Parsons DW, Jin G, et al. IDH1 and IDH2 mutations in gliomas.<br />
N Engl J Med. 2009;360:765-773.<br />
4. Weller M, Felsberg J, Hartmann C, et al. Molecular predictors <strong>of</strong><br />
progression-free and overall survival in patients with newly diagnosed<br />
glioblastoma: a prospective translational study <strong>of</strong> the German Glioma Network.<br />
J Clin Oncol. 2009;27:5743-5750.<br />
5. Curran WJ Jr, Scott CB, Horton J, et al. Recursive partitioning analysis<br />
<strong>of</strong> prognostic factors in three Radiation Therapy <strong>Oncology</strong> Group malignant<br />
glioma trials. J Natl Cancer Inst. 1993;85:704-710.<br />
6. Noushmehr H, Weisenberger DJ, Diefes K, et al: Identification <strong>of</strong> a CpG<br />
island methylator phenotype that defines a distinct subgroup <strong>of</strong> glioma.<br />
Cancer Cell 17:510-522.<br />
7. Colman H, Zhang L, Sulman EP, et al. A multigene predictor <strong>of</strong> outcome<br />
in glioblastoma. Neuro Oncol. 2010;12:49-57.<br />
8. Hegi ME, Diserens AC, Gorlia T, et al. MGMT gene silencing and benefit<br />
from temozolomide in glioblastoma. N Engl J Med. 2005;352:997-1003.<br />
9. Walker MD, Green SB, Byar DP, et al. Randomized comparisons <strong>of</strong><br />
radiotherapy and nitrosoureas for the treatment <strong>of</strong> malignant glioma after<br />
surgery. N Engl J Med. 1980;303:1323-1329.<br />
10. Stewart LA. Chemotherapy in adult high-grade glioma: a systematic<br />
review and meta-analysis <strong>of</strong> individual patient data from 12 randomised<br />
trials. Lancet. 2002;359:1011-1018.<br />
11. Stupp R, Mason WP, van den Bent MJ, et al. Radiotherapy plus<br />
concomitant and adjuvant temozolomide for glioblastoma. N Engl J Med.<br />
2005;352:987-996.<br />
12. Stupp R, Hegi ME, Mason WP, et al. Effects <strong>of</strong> radiotherapy with<br />
concomitant and adjuvant temozolomide versus radiotherapy alone on survival<br />
in glioblastoma in a randomised phase III study: 5-year analysis <strong>of</strong> the<br />
EORTC-NCIC trial. Lancet Oncol. 2009;10:459-466<br />
13. Tolcher AW, Gerson SL, Denis L, et al. Marked inactivation <strong>of</strong> O6alkylguanine-DNA<br />
alkyltransferase activity with protracted temozolomide<br />
schedules. Br J Cancer. 2003;88:1004-1011.<br />
14. Gilbert MR, Wang M, Aldape K, et al. RTOG 0525: a randomized phase<br />
III trial comparing standard adjuvant temozolomide (TMZ) with a dose-dense<br />
(dd) schedule in newly diagnosed glioblastoma (GBM). J Clin Oncol 2011;29<br />
(suppl; abstr 141s).<br />
15. Prados MD, Chang SM, Butowski N, et al. Phase II study <strong>of</strong> erlotinib<br />
plus temozolomide during and after radiation therapy in patients with newly<br />
diagnosed glioblastoma multiforme or gliosarcoma. J Clin Oncol. 2009;27:<br />
579-584.<br />
16. Grossman SA, Ye X, Piantadosi S, et al. Survival <strong>of</strong> patients with newly<br />
116<br />
A similar standard does not currently exist for recurrent<br />
glioblastoma, although antiangiogenic agents particularly<br />
bevacizumab, demonstrate tumor response and tumor control<br />
although this is <strong>of</strong>ten short-lived and to date, there are<br />
no salvage regimens after bevacizumab failure. Given the<br />
marked molecular heterogeneity <strong>of</strong> recurrent glioblastoma,<br />
future advances will likely require a major initiative to<br />
perform clinical trials encompassing correlations or patient<br />
selection on the basis <strong>of</strong> tumor characteristics. This initiative<br />
will require both an increase in resource allocation and<br />
a widespread collaborative effort to be successful.<br />
Stock<br />
Ownership Honoraria<br />
REFERENCES<br />
Genentech;<br />
Merck<br />
Research<br />
Funding<br />
Genentech;<br />
Merck<br />
Expert<br />
Testimony<br />
MARK R. GILBERT<br />
Other<br />
Remuneration<br />
diagnosed glioblastoma treated with radiation and temozolomide in research<br />
studies in the United States. Clin Cancer Res. 2010;16:2443-2449.<br />
17. Stupp R, Goldbrunner B, Neyns B, et al. Phase I/IIa trial <strong>of</strong> cilengitide<br />
(EMD121974) and temozolomide with concomitant radiotherapy, followed<br />
by temozolomide and cilengitide maintenance therapy in patients (pts) with<br />
newly diagnosed glioblastoma (GBM). J Clin Oncol. 2007;25 (suppl; abstr<br />
2000).<br />
18. Clarke JL, Ennis MM, Yung WK, et al. Is surgery at progression a<br />
prognostic marker for improved 6-month progression-free survival or overall<br />
survival for patients with recurrent glioblastoma? Neuro Oncol. 2011;13:<br />
1118-1124.<br />
19. Brem H, Piantadosi S, Burger PC, et al. Placebo-controlled trial <strong>of</strong><br />
safety and efficacy <strong>of</strong> intraoperative controlled delivery by biodegradable<br />
polymers <strong>of</strong> chemotherapy for recurrent gliomas. The Polymer-brain Tumor<br />
Treatment Group. Lancet. 1995;345:1008-1012.<br />
20. Kunwar S, Prados MD, Chang SM, et al. Direct intracerebral delivery<br />
<strong>of</strong> cintredekin besudotox (IL13-PE38QQR) in recurrent malignant glioma: a<br />
report by the Cintredekin Besudotox Intraparenchymal Study Group. J Clin<br />
Oncol. 2007;25:837-844.<br />
21. Anderson RC, Elder JB, Brown MD, et al. Changes in the immunologic<br />
phenotype <strong>of</strong> human malignant glioma cells after passaging in vitro. Clin<br />
Immunol. 2002;102:84-95.<br />
22. Rainov NG. A phase III clinical evaluation <strong>of</strong> herpes simplex virus type<br />
1 thymidine kinase and ganciclovir gene therapy as an adjuvant to surgical<br />
resection and radiation in adults with previously untreated glioblastoma<br />
multiforme. Hum Gene Ther. 2000;11:2389-2401.<br />
23. Jiang H, Gomez-Manzano C, Lang FF, et al. Oncolytic adenovirus:<br />
preclinical and clinical studies in patients with human malignant gliomas.<br />
Curr Gene Ther. 2009;9:422-427.<br />
24. Wick W, Platten M, Weller M. New (alternative) temozolomide regimens<br />
for the treatment <strong>of</strong> glioma. Neuro Oncol. 2009;11:69-79.<br />
25. Perry JR, Rizek P, Cashman R, et al. Temozolomide rechallenge in<br />
recurrent malignant glioma by using a continuous temozolomide schedule:<br />
the “rescue” approach. Cancer. 2008;113:2152-2157.<br />
26. Yip S, Miao J, Cahill DP, et al. MSH6 mutations arise in glioblastomas<br />
during temozolomide therapy and mediate temozolomide resistance. Clin<br />
Cancer Res. 2009;15:4622-4629.<br />
27. Wick W, Puduvalli VK, Chamberlain MC, et al. Phase III study <strong>of</strong><br />
enzastaurin compared with lomustine in the treatment <strong>of</strong> recurrent intracranial<br />
glioblastoma. J Clin Oncol. 2010;28:1168-1174.<br />
28. Haas-Kogan DA, Prados MD, Tihan T, et al. Epidermal growth factor<br />
receptor, protein kinase B/Akt, and glioma response to erlotinib. J Natl<br />
Cancer Inst. 2005;97:880-887.<br />
29. van den Bent MJ, Brandes AA, Rampling R, et al. Randomized phase II<br />
trial <strong>of</strong> erlotinib versus temozolomide or carmustine in recurrent glioblastoma:<br />
EORTC brain tumor group study 26034. J Clin Oncol. 2009;27:1268-1274.
ESTABLISHING TREATMENTS FOR GLIOBLASTOMA<br />
30. Rich JN, Reardon DA, Peery T, et al. Phase II trial <strong>of</strong> gefitinib in<br />
recurrent glioblastoma. J Clin Oncol. 2004;22:133-142.<br />
31. Raymond E, Brandes A, Van Oosterom A, et al. Multicentre phase II<br />
study <strong>of</strong> imatinib mesylate in patients with recurrent glioblastoma: an<br />
EORTC/NDDG/BTG Intergroup Study. J Clin Oncol. 2004;22 (suppl; abstr<br />
1501).<br />
32. Iwamoto FM, Lamborn KR, Kuhn JG, et al. A phase I/II trial <strong>of</strong> the<br />
histone deacetylase inhibitor romidepsin for adults with recurrent malignant<br />
glioma: North <strong>American</strong> Brain Tumor Consortium Study 03-03. Neuro Oncol.<br />
2010;13:509-516.<br />
33. Galanis E, Jaeckle KA, Maurer MJ, et al. Phase II trial <strong>of</strong> vorinostat in<br />
recurrent glioblastoma multiforme: a North Central Cancer Treatment Group<br />
study. J Clin Oncol. 2009;27:2052-2058.<br />
34. Cloughesy TF, Wen PY, Robins HI, et al. Phase II trial <strong>of</strong> tipifarnib in<br />
patients with recurrent malignant glioma either receiving or not receiving<br />
enzyme-inducing antiepileptic drugs: A North <strong>American</strong> Brain Tumor Consortium<br />
Study. J Clin Oncol. 2006;24:3651-3656.<br />
35. Galanis E, Buckner JC, Maurer MJ, et al. Phase II trial <strong>of</strong> temsirolimus<br />
(CCI-779) in recurrent glioblastoma multiforme: a North Central Cancer<br />
Treatment Group Study. J Clin Oncol. 2005;23:5294-5304.<br />
36. Chang SM, Wen P, Cloughesy T, et al. Phase II study <strong>of</strong> CCI-779 in<br />
patients with recurrent glioblastoma multiforme. Invest New Drugs. 2005;23:<br />
357-361.<br />
37. Kreisl TN, Kotliarova S, Butman JA, et al. A phase I/II trial <strong>of</strong><br />
enzastaurin in patients with recurrent high-grade gliomas. Neuro Oncol.<br />
2010;12:181-189.<br />
38. Batchelor TT, Duda DG, di Tomaso E, et al. Phase II study <strong>of</strong> cediranib,<br />
an oral pan-vascular endothelial growth factor receptor tyrosine kinase<br />
inhibitor, in patients with recurrent glioblastoma. J Clin Oncol. 2010;28:<br />
2817-2823.<br />
39. Friedman HS, Prados MD, Wen PY, et al. Bevacizumab alone and in<br />
combination with irinotecan in recurrent glioblastoma. J Clin Oncol. 2009;<br />
27:4733-4740.<br />
40. Reardon DA, Fink KL, Mikkelsen T, et al. Randomized phase II study<br />
<strong>of</strong> cilengitide, an integrin-targeting arginine-glycine-aspartic acid peptide, in<br />
recurrent glioblastoma multiforme. J Clin Oncol. 2008;26:5610-5617.<br />
41. Gilbert MR, Kuhn J, Lamborn KR, et al. Cilengitide in patients with<br />
recurrent glioblastoma: the results <strong>of</strong> NABTC 03-02, a phase II trial with<br />
measures <strong>of</strong> treatment delivery. J Neurooncol. 2011;106:147-153.<br />
42. Wen PY, Prados M, Schiff D, et al: Phase II study <strong>of</strong> XL814 (BMS<br />
907351), an inhibitor <strong>of</strong> MET, VEGFR2 and RET in patients (pts) with<br />
progressive glioblastoma (GB). J Clin Oncol. 2010;28 (suppl; abstr 2006).<br />
117
IMAGING IN NEURO-ONCOLOGY: PRACTICAL<br />
PRIMER FOR THE PRACTICING PHYSICIAN<br />
CHAIR<br />
Susan M. Chang, MD<br />
University <strong>of</strong> California, San Francisco<br />
San Francisco, CA<br />
SPEAKERS<br />
Whitney B. Pope, MD, PhD<br />
University <strong>of</strong> California, Los Angeles<br />
Los Angeles, CA<br />
Andrew D. Norden, MD, MPH<br />
Dana-Farber Cancer Institute<br />
Boston, MA
Current Concepts in Brain Tumor Imaging<br />
By Andrew D. Norden, MD, MPH, Whitney B. Pope, MD, PhD, and Susan M. Chang, MD<br />
Overview: Magnetic resonance imaging (MRI) is the most<br />
useful imaging tool in the evaluation <strong>of</strong> patients with brain<br />
tumors. Most information is supplied by standard anatomic<br />
images that were developed in the 1980s and 1990s. More<br />
recently, functional imaging including diffusion and perfusion<br />
MRI has been investigated as a way to generate predictive and<br />
prognostic biomarkers for high-grade glioma evaluation, but<br />
additional research is needed to establish the added benefits<br />
<strong>of</strong> these indices to standard MRI. Response critieria for<br />
high-grade gliomas have recently been updated by the Response<br />
Assessment in Neuro-<strong>Oncology</strong> (RANO) working<br />
THE FIRST published magnetic resonance image (MRI)<br />
was from a paper in the journal Nature by Nobel Prize<br />
Laureate Paul Lauterbur in 1973. 1 By 1981, magnetic resonance<br />
had been used for imaging the brain, demonstrated<br />
the pathologic appearance <strong>of</strong> glioblastoma (GBM), and<br />
compared favorably to computed tomography (CT) as the<br />
“posterior fossa was visualized with substantially less artifact.<br />
...” 2 MRI was noted to detect tumors not seen on CT as<br />
early as 1982, and this led quickly to an explosion <strong>of</strong> articles<br />
evaluating MRI <strong>of</strong> brain tumors and other intracranial<br />
pathology. T1- and T2-weighted images were quickly adopted<br />
as standard imaging along with multiple planar scanning.<br />
Gadolinium-based contrast agents were introduced<br />
around 1984, 3 as were high-field strength superconducting<br />
(1.5 Tesla) scanners. 4 Another advance in brain tumor<br />
imaging occurred in the late 1990s with the introduction <strong>of</strong><br />
fluid-attenuated inversion recovery (FLAIR) sequences that<br />
generated strongly T2-weighted images although signal associated<br />
with cerebrospinal fluid (CSF) was suppressed. 5<br />
Today, MRI remains the imaging modality <strong>of</strong> choice for<br />
tumor diagnosis, characterization, and assessment <strong>of</strong> treatment<br />
response.<br />
Conventional MRI in Neuro-<strong>Oncology</strong><br />
MRI has traditionally been used to evaluate tumor location,<br />
size and extent, mass effect, involvement <strong>of</strong> critical<br />
structures such as adjacent blood vessels, and compromise <strong>of</strong><br />
the blood-brain barrier (which results in contrast enhancement).<br />
The typical MR scan for a patient with glioma<br />
includes sagittal T1, axial T1, T2, FLAIR and postcontrast<br />
axial and coronal T1-weighted images. Recently pulse sequences<br />
sensitive to physiology, rather than just anatomy,<br />
are being more commonly used (see following). Changes in<br />
enhancing tumor size based on bidimensional measurements<br />
<strong>of</strong> postcontrast T1-weighted images are the basis for<br />
both the Macdonald and later RANO criteria for evaluating<br />
tumor response. 6 Conversely, nonenhancing tumor is assessed<br />
qualitatively in RANO, but not at all in the Macdonald<br />
criteria. Nonenhancing tumor is typified by areas <strong>of</strong><br />
increased T2 signal intensity associated with mass effect<br />
and architectural distortion such as blurring <strong>of</strong> the graywhite<br />
interface. 7 Edema and treatment effect including<br />
gliosis also result in increased T2 signal, which can make<br />
nonenhancing tumor difficult to quantify. FLAIR is more<br />
sensitive to T2 signal abnormalities, as a result <strong>of</strong> the<br />
nulling <strong>of</strong> CSF, thereby overcoming the limitation <strong>of</strong> partial<br />
volume averaging in the cortical and periventricular regions<br />
group. The new criteria account for nonenhancing tumor in<br />
addition to the contrast-enhancing abnormalities on which<br />
older criteria relied. This issue has recently come to the fore<br />
with the introduction <strong>of</strong> the antiangiogenic agent bevacizumab<br />
into standard treatment for recurrent glioblastoma. Because<br />
<strong>of</strong> its potent antipermeability effect, contrast enhancement is<br />
markedly reduced in patients who receive bevacizumab. The<br />
RANO criteria also address the phenomenon <strong>of</strong> pseudoprogression,<br />
in which there may be transient MRI worsening <strong>of</strong> a<br />
glioblastoma following concurrent radiotherapy and temozolomide.<br />
as can be seen in standard T2 images. However, FLAIR also<br />
reduces gray-white differentiation in comparison to typical<br />
T2-weighted images, which can diminish the image reader’s<br />
ability to distinguish the T2 changes that are a result <strong>of</strong><br />
tumor compared with T2 changes that are the result <strong>of</strong><br />
edema and/or gliosis. Thus, T2 and FLAIR images can<br />
provide complementary information, and both should be<br />
acquired for evaluation <strong>of</strong> patients with brain tumors.<br />
Although relying on changes in enhancing tumor previously<br />
worked well for evaluating treatment response, the<br />
widespread adoption <strong>of</strong> bevacizumab therapy for recurrent<br />
GBM highlights the limitations <strong>of</strong> this approach. This limitation<br />
stems largely from bevacizumab’s antipermeability<br />
effect. GBMs are characterized by extensive abnormal vasculature<br />
with a leaky blood-brain barrier. 8 As a result,<br />
contrast material extravasates out <strong>of</strong> tumor vessels, leading<br />
to increased signal on gadolinium-enhanced T1-weighted<br />
images. Bevacizumab sequesters vascular endothelial<br />
growth factor (VEGF), a potent permeability factor and<br />
promoter <strong>of</strong> angiogenesis, and thereby acts to diminish<br />
contrast enhancement. Therefore a reduction in contrast<br />
enhancement following bevacizumab infusion may not necessarily<br />
reflect a cytotoxic tumor effect. Relying on the<br />
change in contrast enhancement alone can thus misrepresent<br />
treatment response, a phenomenon known as “pseudoresponse.”<br />
9,10<br />
Although the RANO criteria include FLAIR or T2 hyperintensity<br />
changes as potentially indicative <strong>of</strong> nonenhancing<br />
tumor progression, no quantification <strong>of</strong> nonenhancing tumor<br />
is performed. This is because <strong>of</strong> difficulties in determining<br />
the borders <strong>of</strong> T2 abnormal regions as well as differentiating<br />
gliosis and other treatment effects from tumor. This has<br />
spurred interest in physiologic imaging as a way <strong>of</strong> obtaining<br />
quantitative data on tumor burden, although to date, this<br />
goal has not been fully realized.<br />
From Dana-Farber Cancer Institute and Brigham and Women’s Hospital, Boston, MA;<br />
David Geffen School <strong>of</strong> Medicine, University <strong>of</strong> California, Los Angeles, Los Angeles, CA;<br />
University <strong>of</strong> California, San Francisco, Department <strong>of</strong> Neurological Surgery, San Francisco,<br />
CA.<br />
Authors’ disclosures <strong>of</strong> potential conflicts <strong>of</strong> interest are found at the end <strong>of</strong> this article.<br />
Address reprint requests to Susan M. Chang, MD, University <strong>of</strong> California, San<br />
Francisco, Department <strong>of</strong> Neurological Surgery, 400 Parnassus Ave., A808, San Francisco,<br />
CA 94143-0372; email: changs@neurosurg.ucsf.edu.<br />
© <strong>2012</strong> by <strong>American</strong> <strong>Society</strong> <strong>of</strong> <strong>Clinical</strong> <strong>Oncology</strong>.<br />
1092-9118/10/1-10<br />
119
Functional Imaging in Neuro-<strong>Oncology</strong><br />
Magnetic Resonance Perfusion<br />
Given the antiangiogenic effects <strong>of</strong> bevacizumab, perfusion<br />
imaging is intuitively an appealing technique for assessing<br />
the effect <strong>of</strong> drug treatment. Several methods <strong>of</strong><br />
obtaining perfusion data have been developed, the two most<br />
common <strong>of</strong> which are dynamic susceptibility contrast<br />
(DSC) imaging and dynamic contrast enhanced (DCE)<br />
imaging. DSC is used to generate maps <strong>of</strong> relative cerebral<br />
(or tumor) blood volume (rCBV), and relative cerebral blood<br />
flow (rCBF), among other metrics. DCE is generally used to<br />
measure the permeability constant Ktrans, which is a metric<br />
<strong>of</strong> capillary leakiness. Many groups have investigated the<br />
role <strong>of</strong> perfusion imaging in the evaluation <strong>of</strong> gliomas. For<br />
instance, maximal rCBV has prognostic value in astrocytoma,<br />
even when controlling for tumor grade. 11 A number <strong>of</strong><br />
studies have shown that high rCBV or increasing rCBV is<br />
associated with a worse prognosis across tumor grades. 12<br />
Perfusion imaging has been used to assess response to<br />
standard (radiation and cytotoxic chemotherapy) as well as<br />
antiangiogenic treatment. For patients with GBM who receive<br />
standard therapy, the percentage change in rCBV from<br />
pre- to post-treatment measurements is predictive <strong>of</strong> 1-year<br />
survival. 13 In patients with recurrent GBM who are treated<br />
with antiangiogenic therapy, a “vascular normalization index”<br />
that is generated by combining Ktrans, microvessel<br />
KEY POINTS<br />
● Magnetic resonance imaging (MRI) is the imaging<br />
modality <strong>of</strong> choice to evaluate brain tumor location,<br />
size and extent, mass effect, involvement <strong>of</strong> critical<br />
structures such as adjacent blood vessels, and compromise<br />
<strong>of</strong> the blood-brain barrier.<br />
● Perfusion and diffusion magnetic resonance imaging<br />
tools are examples <strong>of</strong> physiologic imaging modalities<br />
that may provide quantitative data on tumor burden,<br />
although further investigation is required to define<br />
their roles in routine care and clinical trials.<br />
● In approximately 20% to 30% <strong>of</strong> glioblastoma cases,<br />
the first MRI obtained after radiation therapy and<br />
concurrent temozolomide meets the criteria for progressive<br />
disease, but subsequent follow-up scans<br />
show lesion shrinkage or stability. This has been<br />
termed pseudoprogression and is among the most<br />
common causes <strong>of</strong> misdiagnosed tumor recurrence.<br />
● Because <strong>of</strong> its antipermeability effect, antiangiogenic<br />
therapy with bevacizumab results in marked reduction<br />
<strong>of</strong> tumor enhancement, which reduces the sensitivity<br />
<strong>of</strong> magnetic resonance imaging for diagnosing<br />
tumor recurrence.<br />
● Response critieria for high-grade gliomas have recently<br />
been updated by the Response Assessment in<br />
Neuro-<strong>Oncology</strong> working group. The new criteria account<br />
for nonenhancing tumor in addition to the<br />
contrast-enhancing abnormalities on which older criteria<br />
relied.<br />
120<br />
volume, and circulating collagen IV is predictive <strong>of</strong> survival,<br />
even though it is acquired only 1 day after treatment<br />
initiation. 14 Yet many studies that assess treatment response<br />
to bevacizumab therapy have not been particularly<br />
successful in relating changes in CBV or CBF to outcomes.<br />
15,16 Thus challenges continue in the application <strong>of</strong><br />
perfusion indices as a biomarker <strong>of</strong> treatment response in<br />
this setting. The use <strong>of</strong> perfusion as an early response (1-day<br />
posttreatment initiation marker 14 ) could have a clinical<br />
effect, however.<br />
Perfusion imaging has also been applied to the issue <strong>of</strong><br />
differentiating true from pseudoprogression. In general,<br />
true progression has higher perfusion than radiationinduced<br />
changes, although there can be overlap in<br />
values. 17-20 Standardized protocols may help improve the<br />
use <strong>of</strong> perfusion imaging in multicenter trials, but overall<br />
accuracy also must be increased to have the desired clinical<br />
affect.<br />
Magnetic Resonance Diffusion<br />
Diffusion-weighting imaging (DWI) is another potential<br />
physiology-based magnetic resonance biomarker for the<br />
evaluation <strong>of</strong> gliomas. Based on the movement <strong>of</strong> water<br />
molecules, diffusion data are typically reported as the<br />
apparent diffusion coefficient (ADC); decreased water motion<br />
corresponds to lower ADC values. Since water molecules<br />
are generally more motion-restricted intracellularly<br />
compared to extracellularly, necrosis and cell death or<br />
lysis can result in increased ADC. Similarly, edema increases<br />
ADC by expanding the interstitial, extracellular<br />
fluid volume, which also allows for freer movement <strong>of</strong><br />
water molecules. Conversely, lower ADC is associated<br />
with increased cell density. 23 Because <strong>of</strong> these properties,<br />
DWI has the potential to indicate treatment response in<br />
gliomas.<br />
Functional diffusion maps (fDMs) are used to assess<br />
the voxel-wise changes in ADC measured in the same<br />
patient over time. 21-24 This can increase the sensitivity in<br />
detecting subtle changes in tumor cell density. Initially<br />
fDMs were used to predict response to cytotoxic chemotherapy<br />
and radiotherapy within the contrast-enhancing tumor<br />
bed, 22-24 but recent studies have demonstrated their effective<br />
use outside regions <strong>of</strong> contrast enhancement 21,25,26 and<br />
as tools for studying the effects <strong>of</strong> antiangiogenic treatment.<br />
27,28<br />
Another approach to using diffusion imaging data is to<br />
construct histograms <strong>of</strong> ADC values. ADC histogram analysis<br />
has been investigated as a predictive biomarker <strong>of</strong><br />
bevacizumab treatment response in the setting <strong>of</strong> recurrent<br />
GBM. 29 For this application, tumors with low ADC values<br />
before initiation <strong>of</strong> bevacizumab were more likely to progress<br />
by 6 months compared with tumors with high ADC<br />
values. These results have since been confirmed in a retrospective<br />
analysis <strong>of</strong> a large multicenter trial. 30 However,<br />
prospectively validated biomarkers for response to antiangiogenesis<br />
treatment are not currently available for clinical<br />
use.<br />
Imaging <strong>of</strong> Recurrent Gliomas<br />
NORDEN, POPE, AND CHANG<br />
The Macdonald criteria defined progressive disease, or<br />
recurrence, as “. . . � 25% increase in size <strong>of</strong> enhancing
CONCEPTS IN BRAIN TUMOR IMAGING<br />
tumor or any new tumor on CT or MRI scans, or neurologically<br />
worse, and steroids stable or increased.” 31 Although<br />
these were the best available criteria for nearly 20 years,<br />
their focus on enhancing tumor limited their usefulness<br />
because low-grade diffuse gliomas typically lack an enhancing<br />
component and high-grade gliomas <strong>of</strong>ten contain nonenhancing<br />
elements. Furthermore, processes other than tumor<br />
frequently influence the extent <strong>of</strong> contrast enhancement.<br />
Examples include treatment-induced inflammation or<br />
necrosis, postoperative changes, seizures, and infarction.<br />
Medications may also affect contrast enhancement. Corticosteroids,<br />
for example, perhaps the most commonly used class<br />
<strong>of</strong> medication in neuro-oncology, are known to reduce contrast<br />
enhancement in a large portion <strong>of</strong> patients with gliomas.<br />
32<br />
False Positives<br />
A number <strong>of</strong> phenomena may lead to an inaccurate diagnosis<br />
<strong>of</strong> tumor recurrence on posttreatment MRI scans.<br />
Since publication <strong>of</strong> a large, randomized trial in 2005,<br />
standard-<strong>of</strong>-care therapy for GBM includes involved-field<br />
radiation therapy with concurrent and adjuvant temozolomide.<br />
33 Most neuro-oncologists obtain an initial posttreatment<br />
MRI scan 4 weeks following the radiation therapy<br />
phase. In approximately 20% to 30% <strong>of</strong> cases, this MRI<br />
meets the criteria for progressive disease, but without<br />
further treatment other than adjuvant temozolomide, subsequent<br />
follow-up scans show lesion shrinkage or stability.<br />
34,35 This has been termed pseudoprogression and is<br />
among the most common causes <strong>of</strong> misdiagnosed tumor<br />
recurrence.<br />
Knowledge <strong>of</strong> this phenomenon has changed practice for<br />
many neuro-oncologists, who now accept the first postradiation<br />
MRI scan as a new baseline; so long as a patient is<br />
not highly symptomatic from apparent progression, a<br />
reasonable approach is to proceed with one-to-three cycles<br />
<strong>of</strong> adjuvant temozolomide before deciding whether the<br />
changes reflect pseudoprogression or true progression. In<br />
general, pseudoprogression that is related to radiation<br />
therapy occurs within 3 months following treatment, but<br />
cases have been described as late as 6 months following<br />
treatment. For patients who develop severe or symptomatic<br />
worsening in this window <strong>of</strong> time, surgery may be<br />
required. Although the pathologic substrate <strong>of</strong> pseudoprogression<br />
remains to be determined with certainty, in one<br />
article seven patients underwent surgical resection for pseudoprogression<br />
and subsequent histopathology showed only<br />
necrosis. 36<br />
Given an increasing recognition <strong>of</strong> pseudoprogression in<br />
the temozolomide treatment era and the knowledge that<br />
temozolomide has radiosensitizing properties, many have<br />
suggested that combining temozolomide with radiation therapy<br />
increases the risk <strong>of</strong> pseudoprogression compared with<br />
radiation therapy alone. 36 However, the rates <strong>of</strong> pseudoprogression<br />
reported in recent studies 34,35 are similar to the<br />
rates reported before the widespread use <strong>of</strong> temozolomide. 37<br />
Another potentially important risk factor for pseudoprogression<br />
is the DNA repair gene O 6 -methylguanine-DNA methyltransferase<br />
(MGMT) promoter methylation. Recent data<br />
demonstrate that MGMT promoter methylation status is<br />
predictive <strong>of</strong> response to temozolomide and a favorable<br />
prognosis in patients with GBM. 38 This finding is intuitive<br />
because MGMT is an endogenous DNA repair enzyme that<br />
mediates resistance to alkylating chemotherapeutics like<br />
temozolomide. In a study <strong>of</strong> 103 patients newly diagnosed<br />
with GBM, pseudoprogression was diagnosed in 32 (31%)<br />
patients. Ninety-one percent <strong>of</strong> subjects whose tumors had<br />
methylated MGMT promoters developed pseudoprogression<br />
as compared with 41% <strong>of</strong> patients whose tumors had unmethylated<br />
MGMT promoters, and the difference was statistically<br />
significant. 35 Also <strong>of</strong> interest is that overall<br />
survival was higher in patients whose tumors developed<br />
pseudoprogression. The findings suggest that pseudoprogression<br />
may occur when concurrent radiation therapy and<br />
temozolomide are particularly active against residual tumor,<br />
although this remains to be confirmed in prospective<br />
studies.<br />
Pseudoprogression is problematic from the perspective <strong>of</strong><br />
clinical trial end points in addition to the clinical challenges<br />
it creates. Accrual <strong>of</strong> patients whose tumors spontaneously<br />
“respond” falsely elevates response rates and prolongs<br />
progression-free survival in clinical trials for recurrent disease.<br />
For this reason, the recently adopted RANO criteria<br />
(Table 1) stipulate that progression cannot be diagnosed<br />
within the first 12 weeks after radiation therapy unless the<br />
majority <strong>of</strong> new enhancement is outside the radiation field<br />
or there is histopathologic confirmation <strong>of</strong> tumor progression.<br />
6<br />
Treatments other than radiation therapy are less common<br />
causes <strong>of</strong> misdiagnosed progression. It is well known, for<br />
example, that enhancement at the margin <strong>of</strong> an operative<br />
cavity <strong>of</strong>ten develops 48 to 72 hours after surgery and may<br />
persist for weeks or months. In some cases, perioperative<br />
infarction may result in new enhancement that is visually<br />
indistinguishable from tumor recurrence. In a study <strong>of</strong> 50<br />
GBM resections, diffusion-weighted MRI sequences showed<br />
restricted diffusion consistent with infarction in 35 (70%)<br />
cases. Of these, 15 (43%) showed enhancement that might<br />
have been confused with recurrent disease. 39 Finally, local<br />
therapies may provoke MRI changes that are suspicious for<br />
recurrence. Examples include carmustine wafers and experimental<br />
therapies administered by direct intracerebral injection<br />
or convection-enhanced delivery.<br />
False Negatives<br />
As noted above, the Macdonald criteria were not designed<br />
to assess nonenhancing tumors. This is problematic because<br />
the majority <strong>of</strong> low-grade gliomas and an important minority<br />
<strong>of</strong> high-grade gliomas do not enhance, particularly in<br />
older adults. 40 Furthermore, nonenhancing elements are<br />
present in almost all enhancing high-grade gliomas. Another<br />
challenge in assessing nonenhancing tumor is that<br />
these abnormalities typically appear hyperintense on T2- or<br />
FLAIR sequences and cannot be readily distinguished from<br />
edema, gliosis, toxic leukoencephalopathy, microangiopathy,<br />
or radiation-induced white matter disease.<br />
Unfortunately, the widespread use <strong>of</strong> antiangiogenic<br />
therapy has only compounded the problem. The prototypic<br />
antiangiogenic agent is bevacizumab, a humanized monoclonal<br />
antibody against VEGF that received accelerated<br />
United States Food and Drug Administration approval<br />
for recurrent high-grade glioma in 2009. Although bevacizumab<br />
is responsible for high radiographic response rates<br />
and marked prolongation <strong>of</strong> progression-free survival, 41,42<br />
121
Table 1. RANO Criteria for Response Assessment Incorporating MRI and <strong>Clinical</strong> Factors†<br />
Complete response Requires all <strong>of</strong> the following: complete disappearance <strong>of</strong> all enhancing measurable and nonmeasurable disease sustained for at least 4<br />
wk; no new lesions; stable or improved nonenhancing (T2/FLAIR) lesions; patients must be <strong>of</strong>f corticosteroids (or on physiologic<br />
replacement doses only); and stable or improved clinically. Note: Patients with nonmeasurable disease only cannot have a complete<br />
response; the best response possible is stable disease<br />
Partial response Requires all <strong>of</strong> the following: � 50% decrease compared with baseline in the sum <strong>of</strong> products <strong>of</strong> perpendicular diameters <strong>of</strong> all<br />
measurable enhancing lesions sustained for at least 4 wk; no progression <strong>of</strong> nonmeasurable disease; no new lesions; stable or<br />
improved nonenhancing (T2/FLAIR) lesions on same or lower dose <strong>of</strong> corticosteroids compared with baseline scan; the corticosteroid<br />
dose at the time <strong>of</strong> the scan evaluation should be no greater than the dose at time <strong>of</strong> baseline scan; and stable or improved clinically.<br />
Note: Patients with nonmeasurable disease only cannot have a partial response; the best response possible is stable disease<br />
Stable disease Requires all <strong>of</strong> the following: does not qualify for complete response, partial response, or progression; stable nonenhancing (T2/FLAIR)<br />
lesions on same or lower dose <strong>of</strong> corticosteroids compared with baseline scan. In the event that the corticosteroid dose was increased<br />
for new symptoms and signs without confirmation <strong>of</strong> disease progression on neuroimaging, and subsequent follow-up imaging shows<br />
that this increase in corticosteroids was required because <strong>of</strong> disease progression, the last scan considered to show stable disease will<br />
be the scan obtained when the corticosteroid dose was equivalent to the baseline dose<br />
Progressive disease Defined by any <strong>of</strong> the following: � 25% increase in sum <strong>of</strong> the products <strong>of</strong> perpendicular diameters <strong>of</strong> enhancing lesions compared with<br />
the smallest tumor measurement obtained either at baseline (if no decrease) or best response, on stable or increasing doses <strong>of</strong><br />
corticosteroids*; significant increase in T2/FLAIR nonenhancing lesion on stable or increasing doses <strong>of</strong> corticosteroids compared with<br />
baseline scan or best response after initiation <strong>of</strong> therapy* not caused by comorbid events (eg, radiation therapy, demyelination,<br />
ischemic injury, infection, seizures, postoperative changes, or other treatment effects); any new lesion; clear clinical deterioration not<br />
attributable to other causes apart from the tumor (eg, seizures, medication adverse effects, complications <strong>of</strong> therapy, cerebrovascular<br />
events, infection, and so on) or changes in corticosteroid dose; failure to return for evaluation as a result <strong>of</strong> death or deteriorating<br />
condition; or clear progression <strong>of</strong> nonmeasurable disease<br />
Abbreviations: MRI, magnetic resonance imaging; FLAIR, fluid-attenuated inversion recovery.<br />
† All measurable and nonmeasurable lesions must be assessed using the same techniques as at baseline.<br />
* Stable doses <strong>of</strong> corticosteroids include patients not on corticosteroids.<br />
Reprinted with permission. 6 © 2010 by <strong>American</strong> <strong>Society</strong> <strong>of</strong> <strong>Clinical</strong> <strong>Oncology</strong>.<br />
the results <strong>of</strong> ongoing studies are needed to determine<br />
whether it improves overall survival. Some data suggest<br />
that anti-VEGF therapies may prolong survival by virtue <strong>of</strong><br />
potently reducing vascular permeability. 43 This is a wellknown<br />
effect <strong>of</strong> antiangiogenic treatment that likely accounts<br />
for the reduction in contrast enhancement typically<br />
seen within days <strong>of</strong> the first dose <strong>of</strong> bevacizumab or similar<br />
agents. 44<br />
Although antiangiogenic drugs are generally welltolerated<br />
and lack the toxicities that are typical <strong>of</strong> cytotoxic<br />
agents, substantial preclinical data obtained since the late<br />
1990s suggest that blocking VEGF signaling promotes an<br />
infiltrative tumor growth pattern based on co-option <strong>of</strong><br />
existing cerebral vasculature. 45-48 Several uncontrolled, retrospective<br />
clinical studies reported that high-grade gliomas<br />
treated with bevacizumab showed increased infiltrative tumor<br />
growth on T2/FLAIR sequences. 49-52 Indeed, patients<br />
for whom there is MRI evidence <strong>of</strong> prolonged disease control<br />
on bevacizumab sometimes experience clinical progression<br />
that seems consistent with this phenomenon. In a small<br />
study that pathologically characterized the areas <strong>of</strong> nonenhancing,<br />
infiltrative tumor, there were thin-walled, relatively<br />
normal-appearing blood vessels and elevated levels <strong>of</strong><br />
insulin-like growth factor binding protein-2 and matrix<br />
metalloprotease-2. 53<br />
Some recent data challenge the assumption that antiangiogenic<br />
therapy promotes an infiltrative growth pattern. In<br />
one study, patients treated with bevacizumab-treated were<br />
compared with matched pairs <strong>of</strong> patients treated with conventional<br />
therapies. 54 The rates <strong>of</strong> distant or diffuse recurrence<br />
were approximately 20% in both groups with no<br />
significant difference detected. Another study examined preand<br />
posttreatment recurrence patterns in patients with<br />
recurrent GBM treated with bevacizumab or bevacizumab/<br />
irinotecan on the BRAIN trial. 55 Among patients treated<br />
122<br />
NORDEN, POPE, AND CHANG<br />
with bevacizumab alone, 16% experienced a change in recurrence<br />
pattern from local to diffuse. For reasons that are<br />
unclear, a higher proportion (39%) <strong>of</strong> patients treated with<br />
bevacizumab and irinotecan experienced this change. In a<br />
retrospective report that examined recurrence patterns in<br />
80 patients with GBM who were treated with bevacizumab,<br />
there was no significant difference in recurrence pattern<br />
after bevacizumab treatmentas compared with before bevacizumab<br />
treatment. 56 Approximately 70% to 80% <strong>of</strong> recurrences<br />
were described as local, which is consistent with older<br />
data. 57<br />
Whether antiangiogenic therapy actually promotes infiltrative<br />
tumor growth or merely unmasks it is an unanswered<br />
question. By controlling peritumoral edema and<br />
durably reducing contrast enhancement, bevacizumab may<br />
lead to the false impression <strong>of</strong> increased nonenhancing<br />
tumor progression compared with the prebevacizumab era.<br />
Further data are required to address this. Regardless, there<br />
is no question that bevacizumab therapy renders the radiological<br />
diagnosis <strong>of</strong> progressive disease more challenging<br />
than before. There does appear to be a subset <strong>of</strong> patients<br />
with recurrent high-grade glioma whose tumors progress<br />
primarily in a nonenhancing fashion who would be missed<br />
when the conventional Macdonald criteria is applied. The<br />
RANO criteria for progressive disease therefore include<br />
“. . . significant increase in T2/FLAIR nonenhancing lesion<br />
on stable or increasing doses <strong>of</strong> corticosteroids not caused by<br />
comorbid events (eg, radiation therapy, demyelination, ischemic<br />
injury, infection, seizures, postoperative changes, or<br />
other treatment effects).” 6 Although the ability to determine<br />
the etiology <strong>of</strong> T2/FLAIR changes is challenging, these<br />
criteria represent an important step toward optimally assessing<br />
progressive disease in the current era <strong>of</strong> antiangiogenic<br />
therapy.
CONCEPTS IN BRAIN TUMOR IMAGING<br />
Authors’ Disclosures <strong>of</strong> Potential Conflicts <strong>of</strong> Interest<br />
Author<br />
Employment or<br />
Leadership<br />
Positions<br />
Consultant or<br />
Advisory Role<br />
Stock<br />
Ownership Honoraria<br />
Research<br />
Funding<br />
Andrew D. Norden*<br />
Whitney B. Pope Genentech<br />
Susan M. Chang Novartis;<br />
Schering-Plough<br />
*No relevant relationships to disclose.<br />
1. Lauterbur PC. Image formation by induced local interactions. Examples<br />
employing nuclear magnetic resonance. 1973. Clin Orthop Relat Res. 1989;<br />
3-6.<br />
2. Doyle FH, Gore JC, Pennock JM, et al. Imaging <strong>of</strong> the brain by nuclear<br />
magnetic resonance. Lancet. 1981;2:53-57.<br />
3. Carr DH, Gadian DG. Contrast agents in magnetic resonance imaging.<br />
Clin Radiol. 1985;36:561-568.<br />
4. Bilaniuk LT, Zimmerman RA, Wehrli FW, et al. Cerebral magnetic<br />
resonance: comparison <strong>of</strong> high and low field strength imaging. Radiology.<br />
1984;153:409-414.<br />
5. Kates R, Atkinson D, Brant-Zawadzki M. Fluid-attenuated inversion<br />
recovery (FLAIR): clinical prospectus <strong>of</strong> current and future applications. Top<br />
Magn Reson Imaging 1996;8:389-396.<br />
6. Wen PY, Macdonald DR, Reardon DA, et al. Updated response assessment<br />
criteria for high-grade gliomas: response assessment in neuro-oncology<br />
working group. J Clin Oncol. 2010;28:1963-1972.<br />
7. Pope WB, Sayre J, Perlina A, et al. MR imaging correlates <strong>of</strong> survival in<br />
patients with high-grade gliomas. AJNR Am J Neuroradiol. 2005;26:2466-<br />
2474.<br />
8. Rees JH, Smirniotopoulos JG, Jones RV, et al. Glioblastoma multiforme:<br />
radiologic-pathologic correlation. Radiographics. 1996;16:1413-1438; quiz<br />
1462-1463.<br />
9. Clarke JL, Chang S. Pseudoprogression and pseudoresponse: challenges<br />
in brain tumor imaging. Curr Neurol Neurosci Rep. 2009;9:241-246.<br />
10. Brandsma D, van den Bent MJ. Pseudoprogression and pseudoresponse<br />
in the treatment <strong>of</strong> gliomas. Curr Opin Neurol. 2009;22:633-638.<br />
11. Hirai T, Murakami R, Nakamura H, et al. Prognostic value <strong>of</strong> perfusion<br />
MR imaging <strong>of</strong> high-grade astrocytomas: long-term follow-up study. AJNR<br />
Am J Neuroradiol. 2008;29:1505-1510.<br />
12. Law M, Young RJ, Babb JS, et al. Gliomas: predicting time to progression<br />
or survival with cerebral blood volume measurements at dynamic<br />
susceptibility-weighted contrast-enhanced perfusion MR imaging. Radiology.<br />
2008;247:490-498.<br />
13. Mangla R, Singh G, Ziegelitz D, et al. Changes in relative cerebral blood<br />
volume 1 month after radiation-temozolomide therapy can help predict<br />
overall survival in patients with glioblastoma. Radiology. 2010;256:575-584.<br />
14. Sorensen AG, Batchelor TT, Zhang WT, et al. A “vascular normalization<br />
index” as potential mechanistic biomarker to predict survival after a<br />
single dose <strong>of</strong> cediranib in recurrent glioblastoma patients. Cancer Res.<br />
2009;69:5296-5300.<br />
15. Bidault F, Sahnoun, M., Rousseau, V., et al.: High-Grade Gliomas<br />
Treated with Bevacizumab: Assessment <strong>of</strong> Tumor Response with Dynamic<br />
Susceptibility Contrast Perfusion MRI (DSC-MRI) presented at the Radiological<br />
<strong>Society</strong> <strong>of</strong> North America Annual Meeting, Chicago, IL., 2011.<br />
16. Sawlani RN, Raizer J, Horowitz SW, et al. Glioblastoma: A method for<br />
predicting response to antiangiogenic chemotherapy by using MR perfusion<br />
imaging-pilot study. Radiology. 2010;255:622-8.<br />
17. Barajas RF Jr, Chang JS, Segal MR, et al. Differentiation <strong>of</strong> recurrent<br />
glioblastoma multiforme from radiation necrosis after external beam radiation<br />
therapy with dynamic susceptibility-weighted contrast-enhanced perfusion<br />
MR imaging. Radiology. 2009;253:486-496.<br />
18. Hu LS, Baxter LC, Smith KA, et al. Relative cerebral blood volume<br />
values to differentiate high-grade glioma recurrence from posttreatment<br />
radiation effect: direct correlation between image-guided tissue histopathology<br />
and localized dynamic susceptibility-weighted contrast-enhanced perfusion<br />
MR imaging measurements. AJNR Am J Neuroradiol. 2009;30:552-558.<br />
19. Sugahara T, Korogi Y, Tomiguchi S, et al. Posttherapeutic intraaxial<br />
brain tumor: the value <strong>of</strong> perfusion-sensitive contrast-enhanced MR imaging<br />
for differentiating tumor recurrence from nonneoplastic contrast-enhancing<br />
tissue. AJNR Am J Neuroradiol. 2000;21:901-909.<br />
20. Lev MH, Ozsunar Y, Henson JW, et al. Glial tumor grading and<br />
outcome prediction using dynamic spin-echo MR susceptibility mapping<br />
compared with conventional contrast-enhanced MR: confounding effect <strong>of</strong><br />
REFERENCES<br />
Expert<br />
Testimony<br />
Other<br />
Remuneration<br />
elevated rCBV <strong>of</strong> oligodendrogliomas [corrected]. AJNR Am J Neuroradiol.<br />
2004;25:214-221.<br />
21. Ellingson BM, Malkin MG, Rand SD, et al. Validation <strong>of</strong> functional<br />
diffusion maps (fDMs) as a biomarker for human glioma cellularity. J Magn<br />
Reson Imaging 2010;31:538-548.<br />
22. M<strong>of</strong>fat BA, Chenevert TL, Meyer CR, et al. The functional diffusion<br />
map: an imaging biomarker for the early prediction <strong>of</strong> cancer treatment<br />
outcome. Neoplasia. 2006;8:259-267.<br />
23. Hamstra DA, Chenevert TL, M<strong>of</strong>fat BA, et al. Evaluation <strong>of</strong> the<br />
functional diffusion map as an early biomarker <strong>of</strong> time-to-progression and<br />
overall survival in high-grade glioma. Proc Natl Acad Sci USA.2005;102:<br />
16759-16764.<br />
24. Hamstra DA, Galbán CJ, Meyer CR, et al. Functional diffusion map as<br />
an early imaging biomarker for high-grade glioma: correlation with conventional<br />
radiologic response and overall survival. J Clin Oncol. 2008;26:3387-<br />
3394.<br />
25. Ellingson B, Malkin M, Rand S, et al. Functional diffusion maps<br />
applied to FLAIR abnormal areas are valuable for the clinical monitoring <strong>of</strong><br />
recurrent brain tumors. Proc Intl Soc Mag Reson Med. 2009;102:17:285.<br />
26. Ellingson BM, Rand SD, Malkin MG, et al. Utility <strong>of</strong> functional<br />
diffusion maps to monitor a patient diagnosed with gliomatosis cerebri.<br />
J Neurooncol. 2010;97:419-423.<br />
27. Ellingson BM, Malkin MG, Rand SD, et al. Comparison <strong>of</strong> cytotoxic and<br />
anti-angiogenic treatment responses using functional diffusion maps in<br />
FLAIR abnormal regions. Proc Intl Soc Mag Reson Med. 2009;102:17:1010.<br />
28. Ellingson BM, Malkin MG, Rand SD, et al. Volumetric analysis <strong>of</strong><br />
functional diffusion maps is a predictive imaging biomarker for cytotoxic and<br />
anti-angiogenic treatments in malignant gliomas. J Neurooncol. 2011:102;95-<br />
103.<br />
29. Pope WB, Kim HJ, Huo J, et al. Recurrent glioblastoma multiforme:<br />
ADC histogram analysis predicts response to bevacizumab treatment. Radiology.<br />
2009;252:182-189.<br />
30. Pope WB, et al. Title forthcoming. J Neurooncol. In press.<br />
31. Macdonald DR, Cascino TL, Schold SC Jr, et al. Response criteria for<br />
phase II studies <strong>of</strong> supratentorial malignant glioma. J Clin Oncol. 1990;8:<br />
1277-1280.<br />
32. Watling CJ, Lee DH, Macdonald DR, et al. Corticosteroid-induced<br />
magnetic resonance imaging changes in patients with recurrent malignant<br />
glioma. J Clin Oncol. 1994;12:1886-1889.<br />
33. Stupp R, Mason WP, van den Bent MJ, et al. Radiotherapy plus<br />
concomitant and adjuvant temozolomide for glioblastoma. N Engl J Med.<br />
2005;352:987-996.<br />
34. Brandes AA, Franceschi E, Tosoni A, et al. MGMT promoter methylation<br />
status can predict the incidence and outcome <strong>of</strong> pseudoprogression after<br />
concomitant radiochemotherapy in newly diagnosed glioblastoma patients.<br />
J Clin Oncol. 2008;26:2192-2197.<br />
35. Taal W, Brandsma D, de Bruin HG, et al. Incidence <strong>of</strong> early pseudoprogression<br />
in a cohort <strong>of</strong> malignant glioma patients treated with chemoirradiation<br />
with temozolomide. Cancer. 2008;113:405-410.<br />
36. Chamberlain MC, Glantz MJ, Chalmers L, et al. Early necrosis following<br />
concurrent Temodar and radiotherapy in patients with glioblastoma.<br />
J Neurooncol. 2007;82:81-83.<br />
37. de Wit MC, de Bruin HG, Eijkenboom W, et al. Immediate postradiotherapy<br />
changes in malignant glioma can mimic tumor progression.<br />
Neurology. 2004;63:535-537.<br />
38. Hegi ME, Diserens AC, Gorlia T, et al. MGMT gene silencing and<br />
benefit from temozolomide in glioblastoma. N Engl J Med. 2005;352:997-1003.<br />
39. Ulmer S, Braga TA, Barker FG 2nd, et al. <strong>Clinical</strong> and radiographic<br />
features <strong>of</strong> peritumoral infarction following resection <strong>of</strong> glioblastoma. Neurology.<br />
2006;67:1668-1670.<br />
40. Scott JN, Brasher PM, Sevick RJ, et al. How <strong>of</strong>ten are nonenhancing<br />
supratentorial gliomas malignant? A population study. Neurology. 2002;59:<br />
947-949.<br />
123
41. Friedman HS, Prados MD, Wen PY, et al. Bevacizumab alone and in<br />
combination with irinotecan in recurrent glioblastoma. J Clin Oncol. 2009;<br />
27:4733-4740.<br />
42. Kreisl TN, Kim L, Moore K, et al. Phase II trial <strong>of</strong> single-agent<br />
bevacizumab followed by bevacizumab plus irinotecan at tumor progression<br />
in recurrent glioblastoma. J Clin Oncol. 2009;27:740-745.<br />
43. Kamoun WS, Ley CD, Farrar CT, et al. Edema control by cediranib, a<br />
vascular endothelial growth factor receptor-targeted kinase inhibitor, prolongs<br />
survival despite persistent brain tumor growth in mice. J Clin Oncol.<br />
2009;27:2542-2552.<br />
44. Pope WB, Lai A, Nghiemphu P, et al. MRI in patients with high-grade<br />
gliomas treated with bevacizumab and chemotherapy. Neurology. 2006;66:<br />
1258-1260.<br />
45. Ebos JM, Lee CR, Cruz-Munoz W, et al. Accelerated metastasis after<br />
short-term treatment with a potent inhibitor <strong>of</strong> tumor angiogenesis. Cancer<br />
Cell. 2009;15:232-239.<br />
46. Keunen O, Johansson M, Oudin A, et al. Anti-VEGF treatment reduces<br />
blood supply and increases tumor cell invasion in glioblastoma. Proc Natl<br />
Acad Sci USA.2011;108:3749-3754.<br />
47. Lamszus K, Kunkel P, Westphal M. Invasion as limitation to antiangiogenic<br />
glioma therapy. Acta Neurochir Suppl. 2003;88:169-177.<br />
48. Pàez-Ribes M, Allen E, Hudock J, et al. Antiangiogenic therapy elicits<br />
malignant progression <strong>of</strong> tumors to increased local invasion and distant<br />
metastasis. Cancer Cell. 2009;15:220-231.<br />
124<br />
NORDEN, POPE, AND CHANG<br />
49. Fischer I, Cunliffe CH, Bollo RJ, et al. High-grade glioma before and<br />
after treatment with radiation and Avastin: initial observations. Neuro Oncol.<br />
2008;10:700-708.<br />
50. Iwamoto FM, Abrey LE, Beal K, et al. Patterns <strong>of</strong> relapse and prognosis<br />
after bevacizumab failure in recurrent glioblastoma. Neurology. 2009;73:<br />
1200-1206.<br />
51. Norden AD, Young GS, Setayesh K, et al. Bevacizumab for recurrent<br />
malignant gliomas: efficacy, toxicity, and patterns <strong>of</strong> recurrence. Neurology.<br />
2008;70:779-787.<br />
52. Narayana A, Kunnakkat SD, Medabalmi P, et al. Change in pattern <strong>of</strong><br />
relapse after antiangiogenic therapy in high-grade glioma. Int J Radiat Oncol<br />
Biol Phys. <strong>2012</strong>;82:77-82.<br />
53. de Groot JF, Fuller G, Kumar AJ, et al. Tumor invasion after treatment<br />
<strong>of</strong> glioblastoma with bevacizumab: radiographic and pathologic correlation in<br />
humans and mice. Neuro Oncol. 2010;12:233-242.<br />
54. Wick A, Dörner N, Schäfer N, et al. Bevacizumab does not increase the<br />
risk <strong>of</strong> remote relapse in malignant glioma. Ann Neurol. 2011;69:586-592.<br />
55. Pope WB, Xia Q, Paton VE, et al. Patterns <strong>of</strong> progression in patients<br />
with recurrent glioblastoma treated with bevacizumab. Neurology. 2011;76:<br />
432-437.<br />
56. Chamberlain MC. Radiographic patterns <strong>of</strong> relapse in glioblastoma.<br />
J Neurooncol. 2011;101:319-323.<br />
57. Hochberg FH, Pruitt A. Assumptions in the radiotherapy <strong>of</strong> glioblastoma.<br />
Neurology. 1980;30:907-911.
PERSPECTIVES ON HEADLINE-MAKING NEWS IN<br />
NEURO-ONCOLOGY<br />
CHAIR<br />
Timothy F. Cloughesy, MD<br />
University <strong>of</strong> California, Los Angeles<br />
Los Angeles, CA<br />
SPEAKERS<br />
E. Antonia Chiocca, MD, PhD<br />
The Ohio State University Medical Center<br />
Columbus, OH<br />
Philip H. Gutin, MD<br />
Memorial Sloan-Kettering Cancer Center<br />
New York, NY
Noninvasive Application <strong>of</strong> Alternating<br />
Electric Fields in Glioblastoma: A Fourth<br />
Cancer Treatment Modality<br />
Overview: Tumor treating fields (TTF) therapy is a novel<br />
antimitotic, electric field–based treatment for cancer. This<br />
nonchemical, nonablative treatment is unlike any <strong>of</strong> the established<br />
cancer treatment modalities, such as surgery, radiation,<br />
and chemotherapy. Recently, it has entered clinical use<br />
after a decade <strong>of</strong> intensive translational research. TTF therapy<br />
is delivered to patients by a portable, battery-operated, medical<br />
device using noninvasive transducer arrays placed on the<br />
skin surface surrounding the treated tumor. TTF therapy is<br />
THE DEFINITION <strong>of</strong> the electric field is attributed to<br />
Michael Faraday in the 1820s and was later formulated<br />
by James Clerk Maxwell in his electromagnetic theory in<br />
1865. 1 It is a field <strong>of</strong> electric forces that surround a source<br />
charge. When a test charge is placed within an electric field,<br />
a force acts on it. Negative charges attract positive charges,<br />
while similar signed charges repel each other. As seen in<br />
Fig. 1A, an electric field surrounding a source charge can be<br />
described using diverging lines <strong>of</strong> force. The closer the test<br />
charge is to the source charge, the closer the lines <strong>of</strong> force are<br />
to each other, which represents higher field intensity.<br />
To understand the effects <strong>of</strong> electric fields within cells, it is<br />
important to introduce three definitions. First, electric fields<br />
can be uniform or nonuniform. A uniform electric field is<br />
represented by parallel lines <strong>of</strong> force (Fig. 1B). A nonuniform<br />
electric field is represented by converging or diverging lines<br />
<strong>of</strong> force (Fig. 1A and 1D). Second, an electric field can be a<br />
constant field or a time-varying field, resulting in electrostatic<br />
or electrodynamic phenomena, respectively. In a constant<br />
field, the source charges remain the same over time. A<br />
test charge will move in one direction within a constant<br />
electric field toward the oppositely charged source (Fig. 1B).<br />
In a time-varying or alternating electric field, the charge <strong>of</strong><br />
the sources alternates over time (Fig. 1C). Third, the test<br />
charge can be an electric charge or an electric dipole (an<br />
element with a positive charge on one end and a negative<br />
charge on the opposite end). An electric charge will move<br />
back and forth, while a dipole will rotate within an alternating<br />
uniform electric field and align with the direction <strong>of</strong> the<br />
field. In a nonuniform converging electric field, both dipoles<br />
and charges move in the direction <strong>of</strong> the higher field intensity<br />
through a process known as dielectrophoresis (Fig. 1D).<br />
Mechanism <strong>of</strong> Action <strong>of</strong> TTF Therapy<br />
Over 100 years after Maxwell’s original publication,<br />
Yoram Palti, MD, PhD, hypothesized that properly tuned<br />
alternating electric fields at physiological intensities (i.e.,<br />
1–3 V/cm) would disrupt the mitotic process <strong>of</strong> dividing<br />
cancer cells. 2,3 Dr. Palti hypothesized and subsequently<br />
demonstrated in vitro that at frequencies between 100 and<br />
300 kHz, alternating electric fields disrupt the formation <strong>of</strong><br />
the mitotic spindle during metaphase and lead to dielectrophoretic<br />
movement <strong>of</strong> charged and/or polar molecules and<br />
organelles during anaphase and telophase, disrupting normal<br />
cytokinesis and leading to apoptosis. 2,3 According to this<br />
model, the first mechanism <strong>of</strong> action is explained by the fact<br />
126<br />
By Philip H. Gutin, MD, and Eric T. Wong, MD<br />
now a U.S. Food and Drug Administration (FDA)–approved<br />
treatment for patients with recurrent glioblastoma (GBM) who<br />
have exhausted surgical and radiation treatments. This article<br />
will introduce the basic science behind TTF therapy, its<br />
mechanism <strong>of</strong> action, the preclinical findings that led to its<br />
clinical testing, and the clinical safety and efficacy data<br />
available to date, as well as <strong>of</strong>fer future research directions on<br />
this novel treatment modality for cancer.<br />
that the tubulin subunits are one <strong>of</strong> the most polar molecules<br />
in the cell. These tubulin subunits align in the direction<br />
<strong>of</strong> the applied electric field (Fig. 2A), interfering with<br />
the normal polymerization <strong>of</strong> the mitotic spindle, which<br />
results in formation <strong>of</strong> abnormal mitotic figures in vitro. 3<br />
The second mechanism <strong>of</strong> action is explained by examining<br />
the change in shape <strong>of</strong> the electric field within a dividing cell<br />
from anaphase to telophase. When the cell division axis is<br />
aligned with the direction <strong>of</strong> the electric field, the field lines<br />
that enter the cell at one end converge at the cytokinetic<br />
furrow between the developing daughter cells and then<br />
diverge on the opposite side (Fig. 2B). This nonuniform<br />
electric field within the cell generates dielectrophoretic<br />
forces that act on polar and charged elements in the cell,<br />
pushing them toward the cytokinetic furrow leading to<br />
violent blebbing <strong>of</strong> the plasma membrane. 3 This finding was<br />
also validated by researchers from Beth Israel Deaconess<br />
Medical Center and may be mediated by improper placement<br />
<strong>of</strong> the contractile elements that form the cytokinetic<br />
ring on anaphase entry. 4<br />
Preclinical Studies <strong>of</strong> the Antitumor Effects <strong>of</strong><br />
TTF Therapy<br />
Between 2004 and 2010, a series <strong>of</strong> publications and<br />
conference presentations addressed the issue <strong>of</strong> the applicability<br />
range <strong>of</strong> TTF therapy to different in vitro and in vivo<br />
cancer models either alone or in combination with standard<br />
chemotherapy. 3,5-8 Tables 1 and 2 summarize the state-<strong>of</strong>the-art<br />
preclinical research with TTF therapy. TTF therapy<br />
has been shown to effectively inhibit cancer cell growth in<br />
various cell lines in vitro (Table 1). This effect was clearly<br />
dose (field intensity) dependent in the range <strong>of</strong> 1 to 3 V/cm. 5<br />
The optimal frequency for the inhibitory effect <strong>of</strong> TTF<br />
therapy differed between cell types and was inversely related<br />
to cell size (Table 1; e.g., glioma cell cultures at 200<br />
kHz 3,5 ). In addition, based on the directional nature <strong>of</strong> TTF<br />
From the Department <strong>of</strong> Neurosurgery, Memorial Sloan-Kettering Cancer Center Brain<br />
Tumor Center, New York, NY; and Brain Tumor Center and Neuro-<strong>Oncology</strong> Unit, Beth<br />
Israel Deaconess Medical Center, Boston, MA.<br />
Authors’ disclosures <strong>of</strong> potential conflicts <strong>of</strong> interest are found at the end <strong>of</strong> this article.<br />
Address reprint requests to Philip H. Gutin, MD, Department <strong>of</strong> Neurosurgery, C-703,<br />
Memorial Sloan-Kettering Cancer Center, 1275 York Ave., New York, NY 10065; email:<br />
gutinp@mskcc.org.<br />
© <strong>2012</strong> by <strong>American</strong> <strong>Society</strong> <strong>of</strong> <strong>Clinical</strong> <strong>Oncology</strong>.<br />
1092-9118/10/1-10
TTF THERAPY IN GLIOBLASTOMA<br />
Fig. 1. Electric field theory. (A) Opposite charges attract.<br />
(B) A constant, uniform, electric field. (C) Charges<br />
and dipoles in a time-varying, uniform electric field. (D) A<br />
dipole in a time-varying, nonuniform electric field (dielectrophoresis).<br />
therapy, its antimitotic effect in cultures was enhanced by<br />
sequentially applying more than one field direction to the<br />
treated cells. 5 The combination <strong>of</strong> TTF therapy with different<br />
chemotherapeutic agents has been shown to have at<br />
least additive if not synergistic effects. 7,9 Specifically, the<br />
combination <strong>of</strong> TTF therapy with temozolomide in glioma<br />
cell lines was shown to be additive. Interestingly, in breast<br />
cancer cells, TTF therapy showed overt synergism with<br />
taxanes (e.g., paclitaxel), probably a result <strong>of</strong> the temporal<br />
KEY POINTS<br />
● Tumor treating fields (TTF) therapy is an emerging,<br />
low-toxicity treatment modality for solid tumors<br />
based on the delivery <strong>of</strong> antimitotic alternating electric<br />
fields to the tumor, which interfere with cytokinesis<br />
and microtubule assembly that eventually lead<br />
to cell death.<br />
● As a monotherapy, TTF therapy is at least as effective<br />
as currently available active chemotherapy and<br />
biologic therapies for the treatment <strong>of</strong> recurrent glioblastoma<br />
(GBM).<br />
● The efficacy <strong>of</strong> this noninvasive treatment modality is<br />
achieved with significantly less toxicity and a better<br />
quality <strong>of</strong> life compared with chemotherapy.<br />
● Preliminary data suggest TTF therapy acts synergistically<br />
with temozolomide and other chemotherapy in<br />
both preclinical and clinical trials.<br />
● Future research should focus on integrating TTF<br />
therapy into the treatment <strong>of</strong> GBM in the adjuvant<br />
and maintenance settings, as well as in the treatment<br />
<strong>of</strong> other solid tumor malignancies.<br />
proximity <strong>of</strong> taxanes’ effect in metaphase and TTF therapy’s<br />
mitotic interference on cell entry into anaphase. 5<br />
TTF therapy has been tested in numerous in vivo cancer<br />
models (Table 2). 3,5,8,10 Noninvasive application <strong>of</strong> TTF<br />
therapy to animals was performed using electrically insulated<br />
transducer arrays placed on the head or torso surrounding<br />
the region <strong>of</strong> the tumor. Inhibition <strong>of</strong> tumor growth<br />
was seen in each <strong>of</strong> these models when the correct frequency<br />
<strong>of</strong> TTF therapy was applied. Specifically, 200 kHz TTF<br />
therapy applied in two sequential and perpendicular field<br />
directions lead to significant (p � 0.01) inhibition <strong>of</strong> a<br />
syngeneic, orthotopic F-98 glioma in rats after 7 days <strong>of</strong><br />
treatment. 5 An additional syngeneic, orthotopic model <strong>of</strong><br />
non-small cell lung cancer in mice showed that 150 kHz TTF<br />
therapy significantly (p � 0.01) inhibited tumor growth<br />
within 7 days <strong>of</strong> treatment. 8,11 Furthermore, the additive<br />
effect <strong>of</strong> TTF therapy with chemotherapy seen in vitro was<br />
recapitulated in different in vivo models. 5,8 Finally, in a<br />
metastatic tumor model using a squamous carcinoma tumor<br />
implanted in the kidney capsule <strong>of</strong> rabbits, TTF therapy<br />
applied to the abdomen blocked metastatic spread <strong>of</strong> tumor<br />
from the kidney to the lungs. 10,27<br />
Translating TTF Therapy into <strong>Clinical</strong> Use<br />
Since TTF therapy is a physical antimitotic modality with<br />
no half-life, its application should be continuous. Kinetic<br />
modeling was used to predict the minimal treatment duration<br />
needed with TTF therapy. 12 Based on these data, a<br />
minimal treatment course <strong>of</strong> 4 weeks was defined and<br />
implemented in clinical studies. In vivo animal experiments<br />
and pilot clinical data subsequently verified the 4-week<br />
minimal treatment duration. 12 Such continuous delivery<br />
was made possible by the development <strong>of</strong> a portable, batteryoperated,<br />
medical device that patients can use at home<br />
(NovoTTF-100A, Novocure, Haifa, Israel). Finally, extensive<br />
toxicity studies <strong>of</strong> TTF therapy were performed in healthy<br />
127
mice, rats, and rabbits. 5,9 <strong>Clinical</strong>, laboratory, and pathologic<br />
analyses showed that TTF therapy is well tolerated and<br />
does not lead to systemic toxicity in animals. As expected by<br />
the frequency range <strong>of</strong> TTF therapy (100–300 kHz), these<br />
electric fields do not have any effect on excitable tissues<br />
(neural, muscular, or cardiac), nor do they cause significant<br />
heating. 13-15<br />
Histology Cell Line<br />
Table 1. In Vitro Evidence Overview<br />
Fig. 2. Effects <strong>of</strong> tumor treating fields therapy on<br />
intracellular structures during mitosis. (A) During metaphase,<br />
tubulin dimers align with the external electric<br />
field, interfering with the formation <strong>of</strong> the mitotic spindle.<br />
(B) During cytokinesis, the nonuniform electric field<br />
formed within the dividing cell drives charged and polar<br />
macro-molecules and organelles toward the cleavage<br />
furrow.<br />
<strong>Clinical</strong> Testing <strong>of</strong> TTF Therapy as a Monotherapy<br />
The NovoTTF device was first applied to patients in a<br />
small feasibility trial in Switzerland in 2003. 16 In 2004, TTF<br />
therapy was tested in a pilot clinical trial in patients with<br />
recurrent GBM (Table 3). 5 This single-center, single-arm<br />
trial included patients with favorable prognostic character-<br />
Optimal/Effective<br />
TTF Frequency (kHz)<br />
Additive/Synergistic<br />
with Chemotherapy Reference<br />
High-grade glioma F-98; C-6; RG-2 200 Temozolomide (dacarbazine) Can Res, 20043 U-118; U-87 Proc Natl Acad Sci USA, 20075 Breast adenocarcinoma Normal: 120 Cyclophosphamide Can Res, 20043 MDA-MB-231<br />
MCF7 Doxorubicin Neuro Oncol, 20114 Multiple drug resistant: 120 Paclitaxel BMC Cancer, 20107 MDA-MB-231Dox<br />
AA8/EmtR1 Doxorubicin<br />
MCF7/Mx Paclitaxel<br />
Non-small cell lung cancer (adenocarcinoma) H1299 150 Paclitaxel ERS, 20108 LLC Pemetrexed AACR, 20076 Can Res, 20043 Colorectal adenocarcinoma CT-26 100* NA Can Res, 20043 Malignant melanoma B16F1 Patricia 100 NA Can Res, 20043 Prostate PC-3 100* NA Can Res, 20043 Cervical cancer HeLa 200* NA Neuro Oncol, 20114 Abbreviations: TTF, tumor treating fields; NA, not available (was not reported by the authors).<br />
* Effect seen at this frequency; additional frequencies were not tested.<br />
128<br />
GUTIN AND WONG
TTF THERAPY IN GLIOBLASTOMA<br />
istics. Treatment with the device was well tolerated, and no<br />
treatment-related serious adverse events were reported.<br />
Most patients developed grade 1 to 2 contact dermatitis<br />
beneath the transducer arrays on the scalp. Efficacy endpoints<br />
were very encouraging with a 20% objective response<br />
rate, progression-free survival (PFS) at 6 months <strong>of</strong> 50%,<br />
median time to progression (TTP) <strong>of</strong> 26 weeks, and median<br />
overall survival (OS) <strong>of</strong> 62.2 weeks (14.4 months). Compared<br />
to the historic results <strong>of</strong> salvage chemotherapy, these results<br />
showed clear activity <strong>of</strong> TTF therapy when used as a<br />
monotherapy in recurrent GBM. 17<br />
Based on the results <strong>of</strong> this pilot trial, a pivotal phase III,<br />
multicenter, randomized (1:1) clinical study was initiated in<br />
patients with recurrent GBM (Table 3). The randomized<br />
study, which recruited 237 patients between 2006 and 2009,<br />
compared the efficacy and safety <strong>of</strong> monotherapy with the<br />
NovoTTF device to that <strong>of</strong> the best available active chemotherapy<br />
according to physician’s choice. Thirty-six patients<br />
received bevacizumab, 36 received nitrosureas, 12 received<br />
temozolomide, and 33 received other agents. This was the<br />
largest randomized study in recurrent GBM to be completed<br />
to date. The results <strong>of</strong> the study were presented at the 2010<br />
Table 2. In Vivo Evidence Overview<br />
Tumor Type Anatomic Location Animal Model Frequency (kHz) Effect <strong>of</strong> TTF References<br />
GBM Right hemisphere Rat 200 Tumor growth inhibition with 2 and 3 field directions Proc Natl Acad Sci USA,20075 Non-small cell lung cancer Lung parenchyma Mouse 150 1. Tumor growth inhibition with 2 field directions ERS, 20108 2. Additive tumor inhibition with pemetrexed<br />
Malignant melanoma Intradermal Mouse 100 Tumor growth inhibition with 1 and 2 field directions Can Res, 20043 Proc Natl Acad Sci USA,20075 Malignant melanoma Intravenous Mouse 100 Inhibition <strong>of</strong> metastatic seeding in the lungs Clin Exp Metastasis, 200910 VX-2 (anaplastic) Kidney capsule Rabbit 150–200 1. Tumor growth inhibition seen with 2 field directions Clin Exp Metastasis, 200910 Abbreviation: GBM, glioblastoma.<br />
2. Increase in median survival AACR, 2009 27<br />
3. Inhibition <strong>of</strong> metastatic seeding in the lungs Neuro Oncol, 2010 12<br />
4. Additive tumor inhibition with paclitaxel<br />
Table 3. <strong>Clinical</strong> Evidence Overview<br />
ASCO Annual Meeting and were updated at the 2011<br />
<strong>Society</strong> for Neuro-<strong>Oncology</strong> (SNO) Annual Meeting. 18,19<br />
Baseline characteristics <strong>of</strong> patients were balanced between<br />
the two treatment groups. In both groups, patients had poor<br />
prognostic predictors compared with previous clinical trials<br />
<strong>of</strong> recurrent GBM (90% <strong>of</strong> patients were at their second or<br />
subsequent recurrence; 20% had failed bevacizumab before<br />
entering the trial; and the average tumor diameter was<br />
above 5 cm). In the conservative intent-to-treat (ITT) analysis,<br />
the study showed that patients with recurrent GBM<br />
treated with NovoTTF alone had comparable OS to that <strong>of</strong><br />
patients who received chemotherapy and/or bevacizumab<br />
(6.6 months vs. 6.0 months, respectively; p � 0.26; hazard<br />
ratio [HR] � 0.86; Table 3). Although NovoTTF did not show<br />
superiority over active chemotherapies, it was clear that it<br />
was at least as effective as these treatments. Secondary<br />
endpoints in the trial were supportive: blinded radiology<br />
review showed that PFS at 6 months was 21.4% in the<br />
NovoTTF group compared with 15.2% in the chemotherapy<br />
group (p � 0.24). There were more radiological responses<br />
seen in the NovoTTF group compared with the chemotherapy<br />
group (12% vs. 6%, respectively; p � 0.07), including<br />
Trial Phase<br />
Overall Survival<br />
Progression-Free<br />
Survival (PFS)<br />
at 6 Months or<br />
(# <strong>of</strong> Subjects)<br />
(Months)<br />
Hazard<br />
Median PFS (Weeks)<br />
Indication (Analysis Group)<br />
Analysis TTF Chemo Ratio (p) TTF Chemo P value References<br />
Recurrent GBM (at first relapse) Phase I-II (n � 10) 14.5 m 6.0 m* Non-randomized 50% 15%* NA Proc Natl Acad Sci USA,<br />
20075 ITT Analysis<br />
Recurrent GBM (at second and<br />
Phase III (n � 237) 6.6 m 6.0 m HR � 0.86 21.4% 15.2% p � 0.24 J Clin Oncol, 2010<br />
fourth relapse)<br />
(p � 0.26)<br />
18<br />
ITT analysis Neuro Oncol, 201119 Recurrent GBM (treated patients only) Phase III (n � 210) 7.8 m 6.0 m HR � 0.67 26.2% 15.2% p � 0.03 J Clin Oncol, 2010<br />
(p � 0.012)<br />
18<br />
PP Analysis Neuro Oncol, 201119 Recurrent GBM (KPS � 80, age � 61) Phase III (n � 110) 8.8 m 6.6 m HR � NA 25.6% 7.7% NA Neuro Oncol, 2010<br />
(p � 0.01)<br />
19<br />
Subgroup analysis<br />
Recurrent GBM (after bevacizumab failure) Phase III (n � 43) 4.4 m 3.1 m (p � 0.02) NA NA NA Neuro Oncol, 201020 Subgroup analysis<br />
Recurrent GBM (TTF versus bevacizumab) Phase III (n � 156) 6.6 m 5.0 m HR � 0.65 21% 21% p � 0.05 Neuro Oncol, 2011<br />
(p � 0.048)<br />
21<br />
Subgroup analysis<br />
Newly diagnosed GBM<br />
I-II (n � 10) 39� m 14.7 m* (p � 0.002) 90% 50%* NA BMC Med Phys, 2009<br />
(together with temozolomide)<br />
9<br />
ITT Analysis 155 w 26 w<br />
Relapsed advanced NSCLC<br />
I-II (n � 42) 13.8 m 8.2 m* NA 28 w 12 w* ESMO, 2010<br />
(together with pemetrexed)<br />
25<br />
ITT Analysis ERS, 20108 Expert Opin Investig Drugs,<br />
201011 Abbreviations: GBM, glioblastoma; ITT, intention to treat; NA, not available (was not reported by the authors); HR, hazard ratio; PP, per protocol; KPS, Karn<strong>of</strong>sky<br />
performance status; TTF, tumor treating fields; NSCLC, non-small cell lung cancer.<br />
* Single-arm trials with literature control.<br />
129
three sustained complete responses in the NovoTTF group<br />
compared with none in the chemotherapy group. These<br />
results were accompanied by significantly (p � 0.05) less<br />
treatment-related adverse events with NovoTTF compared<br />
with chemotherapy. Patients in the NovoTTF group reported<br />
a higher quality <strong>of</strong> life compared with patients<br />
treated with chemotherapy. This analysis was based on the<br />
European Organisation for Research and Treatment <strong>of</strong> Cancer<br />
QLQ-C30 and mirrored the lack <strong>of</strong> chemotherapy-related<br />
toxicities in the NovoTTF group. Interestingly, patients in<br />
the NovoTTF group reported better cognitive and emotional<br />
functioning and much less pain than patients in the chemotherapy<br />
group, although these domains <strong>of</strong> the questionnaires<br />
are not related to known side effects <strong>of</strong> chemotherapy.<br />
To date, several exploratory analyses <strong>of</strong> the study data<br />
have been performed. The first analysis compared patients<br />
who received the same “amount” <strong>of</strong> therapy in both groups.<br />
This prospectively defined per-protocol analysis excluded<br />
patients from both groups who received less than one predefined<br />
treatment course. The analysis demonstrated superior<br />
survival in the NovoTTF group compared with the<br />
chemotherapy group (7.8 months vs. 6.0 months; p � 0.012,<br />
HR � 0.67). 18,19 The rationale behind this analysis is that<br />
TTF is a physical modality with no half-life, so that the<br />
moment the therapy is stopped, its antimitotic effect stops<br />
as well. In contrast, chemotherapies have measurable<br />
plasma and tissue half-life, which results in continued<br />
efficacy and toxicity long after a dose has been given.<br />
Therefore, to achieve pharmacokinetic balance in the<br />
“amount” <strong>of</strong> treatment in both groups, this analysis used a<br />
simplified criterion that one course <strong>of</strong> chemotherapy (e.g., 1<br />
day <strong>of</strong> carmustine or 5 days <strong>of</strong> temozolomide) is equivalent to<br />
four weeks <strong>of</strong> continuous TTF therapy.<br />
Two more analyses <strong>of</strong> the study data were presented at<br />
the 2010 and 2011 SNO Annual Meetings. 20,21 The first<br />
study analyzed known clinical prognostic factors <strong>of</strong> age and<br />
Karn<strong>of</strong>sky performance status (KPS). This analysis demonstrated<br />
that in patients age 60 and younger with a KPS<br />
greater than 70, treatment with NovoTTF resulted in superior<br />
OS compared with chemotherapy (8.8 months vs. 6.6<br />
months; p � 0.01). This survival advantage could be attributed<br />
to better compliance with TTF therapy in this group <strong>of</strong><br />
patients. In support <strong>of</strong> this finding, a statistically significant<br />
correlation was seen in the NovoTTF group between treatment<br />
compliance (as measured by the device computerized<br />
log file) and OS (p � 0.0475).<br />
The second analysis is a post hoc, exploratory analysis <strong>of</strong><br />
the treatment <strong>of</strong> 120 patients with NovoTTF compared with<br />
36 patients with bevacizumab. Although without a prespecified<br />
analysis in the trial, patients in the study treated with<br />
NovoTTF lived significantly longer than those treated with<br />
bevacizumab (6.6 months vs. 5.0 months, respectively; p �<br />
0.048, HR � 0.65). 21 This analysis included all ITT patients<br />
who received either bevacizumab or NovoTTF. Patient characteristics<br />
were almost identical and, in fact, favored the<br />
bevacizumab group prognostically. Clearly, this analysis<br />
cannot be taken as final evidence <strong>of</strong> superiority <strong>of</strong> NovoTTF<br />
over bevacizumab; however, it should be treated as<br />
hypothesis-generating data for future clinical studies. Finally,<br />
in the 43 patients who entered the study after bevacizumab<br />
therapy failure (approximately 20% <strong>of</strong> patients in<br />
both groups), OS was significantly longer with TTF therapy<br />
130<br />
than with chemotherapy (4.4 months vs. 3.1 months, respectively;<br />
p � 0.02). The data for the chemotherapy-treated<br />
group is in line with previous publications, which showed<br />
that following bevacizumab failure, the survival <strong>of</strong> patients<br />
with recurrent GBM is limited. 22<br />
Based on the results <strong>of</strong> this pivotal phase III study, the<br />
FDA approved the NovoTTF-100A device on April 8, 2011,<br />
through the premarket approval (PMA) regulatory pathway.<br />
The PMA pathway is reserved for class III (high-risk)<br />
medical devices and requires preclinical, clinical, and manufacturing<br />
evidence, including review <strong>of</strong> both efficacy and<br />
safety data by a panel <strong>of</strong> independent experts. The FDA<br />
concluded that the study results showed NovoTTF to be<br />
comparable in efficacy to active chemotherapy, without<br />
many <strong>of</strong> the side effects associated with chemotherapies and<br />
with a better quality <strong>of</strong> life. 23<br />
<strong>Clinical</strong> Trials Evaluating TTF Therapy in Combination<br />
with Chemotherapy<br />
Two studies <strong>of</strong> combined TTF therapy and chemotherapy<br />
have been published to date. The first was a single-arm,<br />
single-center trial performed in 2006 in patients with newly<br />
diagnosed GBM. 9 Patients received the Stupp protocol with<br />
TTF therapy added to maintenance temozolomide. 24 This<br />
trial showed promising PFS and OS data (PFS � 14 months;<br />
OS � 39 months; Table 3) and served as the basis for an<br />
ongoing, multicenter, pivotal phase III, randomized clinical<br />
study comparing TTF therapy and temozolomide with temozolomide<br />
alone in the maintenance stage <strong>of</strong> the Stupp<br />
protocol.<br />
The second study tested TTF therapy together with pemetrexed<br />
in 42 patients with pretreated, advanced non-small<br />
cell lung cancer. 8,11,25 Efficacy and safety with this combined<br />
treatment paradigm were promising. Time to local<br />
disease progression in the lungs and liver (where TTF was<br />
applied) was 28 weeks, and OS was 13.8 months. In contrast,<br />
TTP and OS for pemetrexed alone were previously<br />
reported to be 12 weeks and 8.3 months, respectively. 26<br />
TTF therapy is still in its early days. However, it has an<br />
established mechanism <strong>of</strong> action, and a growing body <strong>of</strong><br />
preclinical evidence has shown its wide applicability in solid<br />
tumor malignancies either alone or in combination with<br />
standard chemotherapies. Objective antitumor activity and<br />
an unprecedented safety pr<strong>of</strong>ile <strong>of</strong> this treatment modality<br />
have been seen in patients with recurrent GBM. Although<br />
TTF monotherapy has been shown to be at least as effective<br />
as the best available chemotherapies today for recurrent<br />
GBM, in-depth analysis <strong>of</strong> the phase III study data identified<br />
at least two subgroups where TTF therapy was superior<br />
to chemotherapy and could be <strong>of</strong>fered to patients as an<br />
alternative to chemotherapy: younger patients with a better<br />
functional status and patients in whom bevacizumab treatment<br />
has failed in the past.<br />
Conclusion<br />
GUTIN AND WONG<br />
The approval <strong>of</strong> TTF therapy for recurrent GBM ushers in<br />
a fourth modality <strong>of</strong> cancer treatment. More importantly,<br />
TTF treatment has a superior safety pr<strong>of</strong>ile, and its minor<br />
side effects do not appear to overlap with those <strong>of</strong> cytotoxic<br />
chemotherapies, targeted agents, or antiangiogenesis drugs.<br />
Therefore, the rational combination <strong>of</strong> TTF therapy with<br />
specific pharmacologic agents may enhance tumor cell death
TTF THERAPY IN GLIOBLASTOMA<br />
because <strong>of</strong> potential additive or synergistic effects. First, as<br />
demonstrated in preclinical and clinical models, chemotherapy<br />
administered together with TTF therapy may result in<br />
additive or synergistic tumor control without increasing<br />
systemic toxicities. Second, TTF treatment could be combined<br />
with targeted agents that block survival signaling<br />
within the tumor cell. This block may be sufficiently strong<br />
to enhance the cytotoxic effect <strong>of</strong> TTF therapy or vice versa.<br />
Authors’ Disclosures <strong>of</strong> Potential Conflicts <strong>of</strong> Interest<br />
Third, the combination <strong>of</strong> TTF and antiangiogenesis agents<br />
may be another promising path that combines different<br />
antitumor treatments to improve tumor control. Lastly, the<br />
proper scheduling <strong>of</strong> TTF therapy with other agents is<br />
unknown. Additional research may shed light on the optimal<br />
scheduling that may achieve a synergistic effect on tumor<br />
growth leading to long-term tumor control and enhanced<br />
patient survival.<br />
Employment or<br />
Leadership Consultant or Stock<br />
Research<br />
Expert<br />
Other<br />
Author<br />
Positions Advisory Role Ownership Honoraria Funding Testimony Remuneration<br />
Philip H. Gutin Novocure Novocure<br />
Eric T. Wong Novocure<br />
1. Maxwell JC. A Dynamical Theory <strong>of</strong> the Electromagnetic Field. Royal<br />
<strong>Society</strong> Transactions. CLV:1865.<br />
2. Palti Y, Schneiderman R, Gurvich Z, et al. Cell proliferation arrest and<br />
tumor cell destruction by low intensity, frequency tuned electric fields. 2004<br />
AACR Meeting Abstracts. (abstr 1000).<br />
3. Kirson ED, Gurvich Z, Schneiderman R, et al. Disruption <strong>of</strong> cancer cell<br />
replication by alternating electric fields. Cancer Res. 2004;64:3288-3295.<br />
4. Lee SX, Wong ET, Swanson KD. Mitosis Interference <strong>of</strong> Cancer Cells<br />
During Anaphase By Electric Field from NovoTTF-100A. Neuro Oncol.<br />
2011;13:iii10-iii25 (suppl 3; abstr CB-17).<br />
5. Kirson ED, Dbaly V, Tovarys F, et al. Alternating electric fields arrest<br />
cell proliferation in animal tumor models and human brain tumors. Proc Natl<br />
Acad Sci USA.2007;104:10152-10157.<br />
6. Schneiderman R, Shmueli E, Kirson E, et al. Synergism between<br />
chemotherapy and alternating electric fields in the inhibition <strong>of</strong> cancer cell<br />
proliferation in-vitro. 2007 AACR Meeting Abstracts. (abstr 2276).<br />
7. Schneiderman RS, Shmueli E, Kirson ED, et al. TTFields alone and in<br />
combination with chemotherapeutic agents effectively reduce the viability <strong>of</strong><br />
MDR cell sub-lines that over-express ABC transporters. BMC Cancer. 2010;<br />
10:229.<br />
8. Weinberg U, Fresard I, Kueng M, et al. An Open Label Pilot Study <strong>of</strong><br />
Tumor Treating Fields (TTFields) in Combination with Pemetrexed for<br />
Advanced Non-small Cell Lung Cancer (NSCLC). 2010 ERS Annual Congress.<br />
(abstr 363).<br />
9. Kirson ED, Schneiderman RS, Dbaly V, et al. Chemotherapeutic treatment<br />
efficacy and sensitivity are increased by adjuvant alternating electric<br />
fields (TTFields). BMC Med Phys. 2009;9:1.<br />
10. Kirson ED, Giladi M, Gurvich Z, et al. Alternating electric fields<br />
(TTFields) inhibit metastatic spread <strong>of</strong> solid tumors to the lungs. Clin Exp<br />
Metastasis. 2009;26:633-640.<br />
11. Pless M, Weinberg U. Tumor treating fields: Concept, evidence and<br />
future. Expert Opin Investig Drugs. 2010;20:1099-1106.<br />
12. Kirson ED, Wasserman Y, Izhaki A, et al. Modeling tumor growth<br />
kinetics and its implications for TTFields treatment planning. Neuro Oncol.<br />
2010;12:iv36-iv57 (suppl 4; abstr NO-54).<br />
13. Palti Y. Stimulation <strong>of</strong> muscles and nerves by means <strong>of</strong> externally<br />
applied electrodes. Bull Res Counc Isr Sect E Exp Med. 1962;10:54-56.<br />
14. Shizgal P, Mathews G. Electrical stimulation <strong>of</strong> the rat diencephalon:<br />
Differential effects <strong>of</strong> interrupted stimulation on on- and <strong>of</strong>f-responding.<br />
Brain Res. 1977;129:319-333.<br />
REFERENCES<br />
15. Yearwood TL, Hershey B, Bradley K, et al. Pulse width programming in<br />
spinal cord stimulation: A clinical study. Pain Physician. 2010;13:321-335.<br />
16. Salzberg M, Kirson E, Palti Y, et al. A pilot study with very lowintensity,<br />
intermediate-frequency electric fields in patients with locally advanced<br />
and/or metastatic solid tumors. Onkologie. 2008;31:362-365.<br />
17. Wong ET, Hess KR, Gleason MJ, et al. Outcomes and prognostic factors<br />
in recurrent glioma patients enrolled onto phase II clinical trials. J Clin<br />
Oncol. 1999;17:2572-2578.<br />
18. Stupp R, Kanner A, Engelhard H, et al. A prospective, randomized,<br />
open-label, phase III clinical trial <strong>of</strong> NovoTTF-100A versus best standard <strong>of</strong><br />
care chemotherapy in patients with recurrent glioblastoma. J Clin Oncol.<br />
2010;28:18s (suppl; abstr LBA2007).<br />
19. Wong ET, Ram Z, Gutin PH, et al. Updated survival data <strong>of</strong> the phase<br />
III clinical trial <strong>of</strong> NovoTTF-100A versus best standard chemotherapy for<br />
recurrent glioblastoma. Neuro Oncol. 2011;13:iii85-iii91 (suppl 3; abstr OT-<br />
09).<br />
20. Ram Z, Gutin PH, Stupp R. Subgroup and quality <strong>of</strong> life analyses <strong>of</strong> the<br />
phase III clinical trial <strong>of</strong> NovoTTF-100A versus best standard chemotherapy<br />
for recurrent glioblastoma. Neuro Oncol. 2010;12:iv36-iv57 (suppl 4; abstr<br />
NO-55).<br />
21. Ram Z, Gutin PH, Wong ET. Comparing the effect <strong>of</strong> NovoTTF to<br />
Bevacizumab in Recurrent GBM: A Post-Hoc Sub-Analysis <strong>of</strong> the Phase III<br />
Trial Data. Neuro Oncol. 2011;13:iii41-iii68 (suppl 3; abstr NO-50).<br />
22. Iwamoto FM, Abrey LE, Beal K, et al. Patterns <strong>of</strong> relapse and prognosis<br />
after bevacizumab failure in recurrent glioblastoma. Neurology. 2009;73:<br />
1200-1206.<br />
23. FDA: NovoTTF-100A Information for Use, 2011. http://www.access<br />
data.fda.gov/cdrh_docs/pdf10/P100034c.pdf. Accessed February 23, <strong>2012</strong>.<br />
24. Stupp R, Mason WP, van den Bent MJ, et al. Radiotherapy plus<br />
concomitant and adjuvant temozolomide for glioblastoma. N Engl J Med.<br />
2005;352:987-996.<br />
25. Pless M, Betticher DC, Buess M, et al. A phase II study <strong>of</strong> tumor<br />
treating fields (TTFields) in combination with pemetrexed for advanced non<br />
small cell lung cancer (NSCLC). Ann Oncol. 2010:viii122-viii161.<br />
26. Hanna N, Shepherd FA, Fossella FV, et al. Randomized phase III trial<br />
<strong>of</strong> pemetrexed versus docetaxel in patients with non-small-cell lung cancer<br />
previously treated with chemotherapy. J Clin Oncol. 2004;22:1589-1597.<br />
27. Kirson E, Gurvich Z, Izhaki A, et al. Alternating electric fields<br />
(TTFields) inhibit metastatic spread <strong>of</strong> solid tumors to the lungs in-vivo. 2009<br />
AACR Meeting Abstracts. (abstr 151).<br />
131
ADAPTIVE CLINICAL TRIAL DESIGNS:<br />
WHAT ARE THEY AND SHOULD THEY BE USED?<br />
CHAIR<br />
Mary Weber Redman, PhD<br />
Fred Hutchinson Cancer Research Center<br />
Seattle, WA<br />
SPEAKERS<br />
J. Jack Lee, PhD, DDS<br />
University <strong>of</strong> Texas M. D. Anderson Cancer Center<br />
Houston, TX<br />
Marc Buyse, ScD<br />
International Drug Development Institute<br />
Louvain-la-Neuve, Belgium
Limitations <strong>of</strong> Adaptive <strong>Clinical</strong> Trials<br />
Overview: Adaptive designs are aimed at introducing flexibility<br />
in clinical research by allowing important characteristics<br />
<strong>of</strong> a trial to be adapted during the course <strong>of</strong> the trial based<br />
on data coming from the trial itself. Adaptive designs can be<br />
used in all phases <strong>of</strong> clinical research, from phase I to phase<br />
III. They tend to be especially useful in early development,<br />
WITH THE large number <strong>of</strong> promising new molecules<br />
that are currently available for clinical testing, clinical<br />
trials must detect a drug’s benefit (or harm) as quickly<br />
as possible. In parallel to the explosion in the number <strong>of</strong><br />
drugs awaiting clinical testing, the costs <strong>of</strong> clinical trials<br />
have sky-rocketed, which adds to the pressure <strong>of</strong> optimizing<br />
trials, to the extent possible, in terms <strong>of</strong> sample sizes,<br />
timelines, and risk <strong>of</strong> failure. A new class <strong>of</strong> designs has<br />
emerged to address these challenges, collectively known as<br />
adaptive designs. In this chapter, we review different types<br />
<strong>of</strong> adaptive designs and briefly mention some situations in<br />
which such designs can be useful. Much <strong>of</strong> this chapter,<br />
however, is devoted to a discussion <strong>of</strong> the limitations and<br />
drawbacks <strong>of</strong> adaptive designs, which might partly explain<br />
why these designs have not been commonly used and might<br />
in the future have less <strong>of</strong> an effect on clinical research than<br />
claimed by their advocates.<br />
Types <strong>of</strong> Adaptive Designs<br />
One <strong>of</strong> the difficulties surrounding adaptive designs is<br />
that the term is used to encompass different situations. For<br />
clarity, we divide group adaptive designs into three broad<br />
categories.<br />
The first category is treatment effect–independent adaptive<br />
designs. In these designs, some <strong>of</strong> the design features<br />
can be adapted on observation <strong>of</strong> predefined patient characteristics<br />
(such as baseline prognostic factors) or outcomes<br />
(such as response rate or hazard rate overall or in the control<br />
group) but in ignorance <strong>of</strong> the treatment effect.<br />
The second category is treatment effect–dependent adaptive<br />
designs. In these designs, one or more <strong>of</strong> the design<br />
features (such as the sample size, the patient inclusion<br />
criteria, the treatment groups being compared, the treatment<br />
allocation ratio, or even the primary endpoint) can be<br />
adapted, depending on the observed treatment effect.<br />
The third category is other types <strong>of</strong> adaptive designs,<br />
which include the continual reassessment method (CRM) for<br />
phase I trials and seamless phase II/III designs.<br />
Treatment Effect–Independent Adaptive Designs<br />
In these designs, adaptations do not depend on the treatment<br />
effect. As such, these adaptations raise few issues and<br />
have little effect on the statistical inference; in particular,<br />
they do not inflate the type I error. In fact, such adaptations<br />
are so mild that trials using them are not referred to<br />
as “adaptive”. We provide two examples but do not discuss<br />
these designs in detail.<br />
Covariate-Adaptive Randomization<br />
One instance <strong>of</strong> treatment effect–independent adaptation<br />
is covariate-adaptive randomization, for which the probabil-<br />
By Marc Buyse, ScD<br />
when the paucity <strong>of</strong> prior data makes their flexibility a key<br />
benefit. The need for adaptive designs lessened as new<br />
treatments progress to later phases <strong>of</strong> development, when<br />
emphasis shifts to confirmation <strong>of</strong> hypotheses using fully<br />
prespecified, well-controlled designs.<br />
ity <strong>of</strong> allocating the next patient to one <strong>of</strong> the trial’s treatment<br />
groups is computed dynamically to ensure good<br />
balance among the treatment groups with respect to important<br />
prognostic factors (center or country, clinicopathologic<br />
features, and, increasingly, biomarkers measured at baseline).<br />
A common implementation <strong>of</strong> this approach is minimization,<br />
for which a predefined algorithm is used to minimize<br />
the imbalance between the distributions <strong>of</strong> important prognostic<br />
factors at baseline among treatment groups. When<br />
minimization is used, the treatment group the next patient<br />
is allocated to can depend on the baseline characteristics <strong>of</strong><br />
previously accrued patients but not on their outcome. 1<br />
Sample Size Increases<br />
Another type <strong>of</strong> treatment effect–independent adaptation<br />
consists <strong>of</strong> a sample size increase if the incidence <strong>of</strong> the event<br />
<strong>of</strong> interest is much lower than expected in the control group<br />
(to preserve the power <strong>of</strong> the trial) or if the event rate is<br />
much lower than expected overall (to preserve the timelines<br />
<strong>of</strong> event-driven analyses when the outcome <strong>of</strong> interest is a<br />
time-to-event, such as disease-free survival or overall survival).<br />
Again, these sample size increases are implemented<br />
in ignorance <strong>of</strong> the treatment effect; hence, they generally<br />
have no effect on type I error and, if implemented appropriately,<br />
raise no special statistical concerns. 2<br />
Treatment Effect–Dependent Adaptive Designs<br />
These designs are truly adaptive ins<strong>of</strong>ar as adaptations<br />
depend on the observed treatment effect, which requires<br />
caution to be exercised and a proper statistical approach to<br />
be used. If, for instance, the sample size <strong>of</strong> a trial was<br />
increased (or decreased) simply because the observed treatment<br />
effect was smaller (or larger) than anticipated, the<br />
final results <strong>of</strong> the trial could be biased. 3 For instance, a<br />
randomly large treatment effect could lead to a reduction<br />
in sample size even though the true effect were as expected.<br />
Note that a group sequential design is not subject to this<br />
problem because its sample size is fixed and can only be<br />
decreased if the trial is stopped for efficacy or futility at an<br />
interim analysis. 4<br />
From the International Drug Development Institute, Houston, TX; Interuniversity Institute<br />
for Biostatistics and Statistical Bioinformatics, Hasselt University, Diepenbeek,<br />
Belgium.<br />
Author’s disclosures <strong>of</strong> potential conflicts <strong>of</strong> interest are found at the end <strong>of</strong> this article.<br />
Address reprint requests to Marc Buyse, ScD, IDDI Inc, 363 N. Sam Houston Pkwy. E.,<br />
Suite 1100, Houston, TX 77060; email: marc.buyse@iddi.com.<br />
© <strong>2012</strong> by <strong>American</strong> <strong>Society</strong> <strong>of</strong> <strong>Clinical</strong> <strong>Oncology</strong>.<br />
1092-9118/10/1-10<br />
133
Sample Size Recalculation<br />
The most obvious adaptation to consider for an ongoing<br />
comparative (phase III) trial is a sample size recalculation if<br />
the treatment effect turns out to be vastly different from<br />
that assumed to design the trial. It might make sense, for<br />
example, to consider an increase in sample size if the<br />
assumed treatment effect was grossly overestimated (e.g., as<br />
a result <strong>of</strong> highly promising phase II results), but there are<br />
serious theoretical and practical objections to doing so (Table<br />
1).<br />
Theoretically, the treatment effect assumed in the protocol<br />
was the smallest effect considered to be clinically worthwhile;<br />
hence, there is no reason to lower this effect merely to<br />
reach statistical significance. In practice, however, the sample<br />
size is <strong>of</strong>ten based on a larger treatment effect than the<br />
minimum considered worthwhile, especially when the findings<br />
<strong>of</strong> phase II trials suggest that a larger treatment effect<br />
can potentially be achieved. Commonly, other considerations<br />
come into play, including budgetary constraints that tend to<br />
drive sample sizes down. It is tempting, especially for small<br />
companies with limited financial resources, to shoot for large<br />
treatment effects, hope for the best, and increase the sample<br />
size only if needed. Great caution must, however, be exercised<br />
when a sample size increase is triggered by the<br />
observation <strong>of</strong> a smaller than expected treatment benefit,<br />
ins<strong>of</strong>ar as such an adaptation can lead to changes in the<br />
types <strong>of</strong> patients accrued into the trial so that the trial after<br />
the adaptation is no longer the same as before the adaptation.<br />
It has also been proven that for any adaptive design one<br />
might consider, there exists a group sequential design that<br />
uniformly outperforms it (i.e., has better power for any given<br />
KEY POINTS<br />
● Adaptive designs can be useful in specific situations<br />
(e.g., in phase I trials aimed at determining a maximum<br />
tolerated dose or in dose-finding trials when a<br />
wide range <strong>of</strong> doses needs to be investigated).<br />
● Despite the hype surrounding adaptive designs, they<br />
have been used infrequently so far; the flexibility<br />
afforded by adaptive designs might not be compensated<br />
for by the potential loss in credibility associated<br />
with their use.<br />
● Although adaptive designs can be useful to rescue<br />
trials that are based on incorrect assumptions, they<br />
should not generally replace well-designed trials that<br />
use conventional approaches (e.g., group sequential<br />
designs that are more statistically efficient).<br />
● Generally speaking, treatment effect–independent<br />
adaptations raise few difficulties, whereas treatment<br />
effect–dependent adaptations are to be implemented<br />
with caution and use <strong>of</strong> appropriate statistical methods.<br />
● Covariate-adaptive randomization is useful to minimize<br />
imbalances in prognostic factors among treatment<br />
groups; in contrast, outcome-adaptive<br />
randomization is unhelpful statistically, difficult logistically,<br />
and unnecessary ethically.<br />
134<br />
Table 1. Pros and Cons <strong>of</strong> Adaptive Sample Size Increases<br />
Pros Cons<br />
Potential to rescue a trial that would<br />
miss statistical significance<br />
Flexibility if design considerations<br />
were inadequate<br />
treatment effect). 5 In other words, adaptive designs are<br />
inefficient compared with group sequential designs. This<br />
mathematical result has pr<strong>of</strong>ound consequences because it<br />
implies that even the most astute adaptive design can, in<br />
theory, be replaced by a group sequential counterpart that<br />
has better statistical power and none <strong>of</strong> the dangers created<br />
by adaptations made to an ongoing trial. In fairness to<br />
adaptive designs, the superior efficiency <strong>of</strong> group sequential<br />
designs assumes that the appropriate spending function can<br />
be prespecified and that interim analyses come at no cost,<br />
which is not the case in practice. 6<br />
Figure 1 compares a group sequential design with an<br />
adaptive design in the same situation. This figure illustrates<br />
the putative advantage <strong>of</strong> the adaptive design that starts<br />
with a small sample size based on a large treatment effect<br />
and increases the sample size if the observed effect is<br />
smaller than anticipated, whereas the group sequential<br />
design starts with a large sample size based on a small<br />
treatment effect and stops early if the observed effect is<br />
larger than anticipated. Stated differently, adaptive designs<br />
take an optimistic view and adapt if required, whereas<br />
group sequential designs take a pessimistic view and stop<br />
early if indicated. In the example <strong>of</strong> Fig. 1, the group<br />
sequential design is based on a difference in proportions <strong>of</strong><br />
0.05 and requires 1,400 patients but can stop after 350, 700,<br />
or 1,050 patients if the difference is equal to approximately<br />
0.15, 0.075, or 0.05, respectively. In contrast, the adaptive<br />
design is based on a difference in proportions <strong>of</strong> 0.10 and<br />
requires only 400 patients, but the sample size can be<br />
increased if the conditional power at 200 patients or 400<br />
patients is too low.<br />
The conditional power is the power that the trial is<br />
expected to have at its final analysis, given the data from the<br />
already accrued patients and assuming that the protocolspecified<br />
difference will apply to all patients still to be<br />
accrued. Note that at the beginning <strong>of</strong> the trial, the conditional<br />
power is simply equal to the power because no data<br />
are available yet. As stated above, it is essential, when<br />
recalculating the sample size in an adaptive manner, to use<br />
appropriate statistical methods. Several methods for this<br />
exist; Tsiatis and Mehta provide a review and examples <strong>of</strong><br />
these methods. 5<br />
Outcome-Adaptive Randomization<br />
MARC BUYSE<br />
Statistically inefficient<br />
Emphasis on statistical significance<br />
rather than clinical relevance<br />
Changes in patient population or other<br />
temporal trends<br />
Can <strong>of</strong>ten be substituted by nonadaptive<br />
sample size increases that do not affect<br />
the type I error<br />
In outcome-adaptive randomization, the probability <strong>of</strong><br />
allocating the next patient to one <strong>of</strong> the trial’s treatment<br />
groups is computed dynamically to favor the treatment<br />
group with the best outcome so far. 7 For instance, if response<br />
(tumor shrinkage) was the outcome <strong>of</strong> interest,<br />
patients would be allocated preferentially to the treatment<br />
group with the highest response rate. This approach has
LIMITATIONS OF ADAPTIVE CLINICAL TRIALS<br />
been called a “play-the-winner” strategy. 8 A crucial distinction<br />
needs to be made between covariate-adaptive randomization<br />
and outcome-adaptive randomization. Indeed,<br />
although the former raises no particular issue, the latter is<br />
fraught with problems.<br />
First, the outcome that is used to adapt the randomization<br />
has to be observed early and reliably, and it must be<br />
reasonably predictive <strong>of</strong> important clinical endpoints for the<br />
adaptation to succeed at placing more patients in the better<br />
treatment group.<br />
Second, adaptive randomization can result in major imbalances<br />
among treatment arms, which in turn negatively<br />
affects the statistical power <strong>of</strong> the trial.<br />
Third, the statistical inference is complicated because the<br />
treatment assignments and the responses are correlated; as<br />
a consequence, rerandomization tests must be used instead<br />
<strong>of</strong> traditional likelihood-based tests.<br />
Fourth, adaptive randomization can cause accrual bias (if<br />
patients wait for the probability <strong>of</strong> receiving the better<br />
treatment to increase) and/or selection bias (if patients are<br />
aware <strong>of</strong> the emerging difference among the treatment<br />
groups).<br />
Last but not least, it is incorrect to claim that adaptive,<br />
randomization is ethically superior to fixed randomization<br />
because equipoise mandates that allocation to any <strong>of</strong> the<br />
treatment groups be considered equally desirable. It might<br />
make sense to allocate more patients to the experimental<br />
group than to the control group, but the justification for<br />
doing so is that more information is needed about a new<br />
treatment than about a well-established standard treatment.<br />
When such is the case, a fixed unequal allocation ratio<br />
(such as a 2:1 ratio in favor <strong>of</strong> the experimental group) will<br />
do just as well as adaptive randomization, without being<br />
subject to the problems listed above. 9<br />
Other Types <strong>of</strong> Adaptive Designs<br />
Fig. 1. Comparison between a group sequential design and an adaptive design.<br />
There are many other types <strong>of</strong> adaptive designs, some <strong>of</strong><br />
which do not fall into the two clear-cut categories discussed.<br />
We briefly mention two types <strong>of</strong> designs that have attractive<br />
properties but have also been rarely used in practice.<br />
CRM in Phase I Trials<br />
Classic phase I cancer trials are aimed at determining the<br />
maximum tolerated dose (MTD) <strong>of</strong> a new drug or combination<br />
<strong>of</strong> drugs. 10 They are usually performed according to a<br />
fixed design called the “3 � 3” design. The design proceeds in<br />
cohorts <strong>of</strong> three patients, with the first cohort being treated<br />
at the minimum dose <strong>of</strong> interest and the next cohorts being<br />
treated at increasing dose levels according to a predetermined<br />
dose escalation scheme. The dose escalation proceeds<br />
until at least one dose-limiting toxic effect is observed in a<br />
cohort <strong>of</strong> three patients, in which case a second cohort <strong>of</strong><br />
three patients is treated at the same dose level. The dose<br />
escalation stops as soon as at least two patients experience<br />
a dose-limiting toxic effect, either in the first cohort <strong>of</strong> three<br />
patients treated at that dose level or in the two cohorts <strong>of</strong><br />
three patients treated at that dose level. Although this<br />
design is used in almost every phase I cancer trial today, it<br />
has several limitations. 11 First, too many patients may be<br />
treated at low doses, with virtually no chance <strong>of</strong> efficacy.<br />
Second, dose escalation may be too slow because <strong>of</strong> an<br />
excessive number <strong>of</strong> escalation steps, resulting in trials that<br />
take longer than needed to get to the MTD. Third, too few<br />
patients may be treated near the MTD, resulting in substantial<br />
residual uncertainty about the dose recommended for<br />
further trials, which raises ethical concerns. Indeed, if the<br />
recommended dose is chosen too low, it can fail to have<br />
efficacy in phase II trials, whereas if it is chosen too high, it<br />
can put patients at unacceptable risk in phase II trials. Last<br />
but not least, the “3 � 3” design makes no allowance for<br />
patient variability. An adaptive design known as the CRM<br />
was originally proposed by O’Quigley et al 12 in the early<br />
1990s. This design involves a statistical approach based on<br />
an assumed dose-response relationship, which is described<br />
through a mathematical function that links the probability<br />
<strong>of</strong> a dose-limiting toxic effect and the dose level. The CRM<br />
design is adaptive ins<strong>of</strong>ar as the dose to use for the next<br />
patient is determined from the toxic effects experienced by<br />
all the patients already treated so far. Many modifications to<br />
the CRM have been proposed, and simulation studies have<br />
135
shown that it generally outperforms the “3 � 3” design. 13,14<br />
Ironically, however, the CRM is not as popular as it should<br />
be given its attractive properties, probably because it requires<br />
calculations to be performed for the next dose to be<br />
determined, whereas the “3 � 3” design does not.<br />
Seamless Phase II/III Designs<br />
Table 2. Comparisons <strong>of</strong> Three Approaches for Late <strong>Clinical</strong> Development<br />
Approach Benefits Costs and Risks<br />
Phase II trial followed by phase III trial Standard approach Longest development time<br />
No regulatory risk<br />
Phase III trial provides independent confirmation <strong>of</strong><br />
phase II results<br />
Largest total sample size<br />
Seamless phase II/III trial Can adapt the phase III based on the phase II results Negotiation with regulatory agencies essential<br />
Statistical method well established<br />
Upfront commitment for phase II only<br />
Less experience with adaptive designs<br />
Phase III trial with interim analyses Much experience with group sequential designs Requires large upfront commitment<br />
No regulatory risk Difficult to design or start phase III based on scanty<br />
Optimal statistical approach if several interim looks<br />
are planned<br />
early data<br />
It is possible to embed a phase II trial into a phase III trial<br />
so that the transition between the two phases is operationally<br />
seamless—as opposed to performing a randomized<br />
phase II trial followed by a separate phase III trial. The<br />
simplest version <strong>of</strong> this approach consists <strong>of</strong> using a classic<br />
phase II design to screen for activity based on response and<br />
to calculate the sample size required <strong>of</strong> the phase III trial<br />
based on the final outcome <strong>of</strong> interest, such as time to<br />
progression or survival. Because the purpose <strong>of</strong> the phase II<br />
trial is only to stop for futility on the basis <strong>of</strong> lack <strong>of</strong> activity,<br />
there is no inflation in type I error. One-stage or two-stage<br />
phase II designs can be used, as well as a selection design if<br />
several experimental arms are simultaneously screened. In<br />
all cases, the phase II and phase III portions <strong>of</strong> the trial are<br />
designed independently <strong>of</strong> each other. Table 2 contrasts a<br />
seamless transition from phase II to phase III with two other<br />
approaches. Some authors have proposed to use a Bayesian<br />
approach to expand the phase II trial to a phase III trial. 15<br />
A different approach that is particularly useful in selecting<br />
one or more doses <strong>of</strong> a new investigational agent is to use<br />
Author’s Disclosures <strong>of</strong> Potential Conflicts <strong>of</strong> Interest<br />
an inferentially seamless design in which several doses are<br />
tested in the phase II portion <strong>of</strong> the trial and to select only<br />
the most promising one to continue in the phase III portion.<br />
16 Various designs have been proposed that control the<br />
overall significance level <strong>of</strong> the trial, whether or not there<br />
are adaptations <strong>of</strong> some design aspects at the end <strong>of</strong> the<br />
phase II trial. Jennison and Turnbull <strong>of</strong>fer a review and<br />
useful guidance. 17,18<br />
Conclusion<br />
Employment or<br />
Leadership Consultant or Stock<br />
Author<br />
Positions Advisory Role Ownership Honoraria<br />
Marc Buyse IDDI IDDI<br />
1. Rosenberger WF, Lachin JM. Randomization in <strong>Clinical</strong> Trials: Theory<br />
and Practice. New York, NY: Wiley; 2002.<br />
2. Gould AL. Planning and revising the sample size for a trial. Stat Med.<br />
1995;14:1039-1051.<br />
3. Mehta CR. Sample size reestimation for confirmatory clinical trials. In:<br />
Harrington D (ed). Designs for <strong>Clinical</strong> Trials: Perspectives on Current Issues.<br />
New York, NY: Springer; 2011.<br />
4. Kim KM. Sequential designs for clinical trials. In: Harrington D (ed).<br />
Designs for <strong>Clinical</strong> Trials: Perspectives on Current Issues. New York, NY:<br />
Springer; 2011.<br />
5. Tsiatis AA, Mehta C. On the inefficiency <strong>of</strong> the adaptive design for<br />
monitoring clinical trials. Biometrika. 2003;90:367-e78.<br />
6. Brannath W, Bauer P, Posch M. On the efficiency <strong>of</strong> adaptive designs for<br />
flexible interim decisions in clinical trials. J Stat Planning Infer. 2006;136:<br />
1956-1961.<br />
136<br />
REFERENCES<br />
There is little doubt that clinical research is in need <strong>of</strong><br />
reengineering to provide efficient readouts <strong>of</strong> efficacy and<br />
safety on the large number <strong>of</strong> treatments currently developed<br />
by pharmaceutical, biotechnology, and medical device<br />
companies. Running large-scale trials that have a high<br />
probability <strong>of</strong> failure is clearly undesirable. Many innovative<br />
methods have been proposed, including adaptive designs,<br />
Bayesian designs, and biomarker-based designs. Some recent<br />
trial designs combine all <strong>of</strong> these ideas and constitute,<br />
as such, exciting models for further developments. 19,20 The<br />
future will tell which <strong>of</strong> these innovations are useful and<br />
when they constitute a definite improvement over classic<br />
approaches. In the meantime, oncologists involved in clinical<br />
research must be aware <strong>of</strong> the limitations <strong>of</strong> each approach<br />
and adjust their expectations accordingly. <strong>Clinical</strong> trial<br />
sponsors might wish to consult regulatory guidances already<br />
available on adaptive designs. 21,22<br />
Research<br />
Funding<br />
Expert<br />
Testimony<br />
MARC BUYSE<br />
Other<br />
Remuneration<br />
7. Hu F, Rosenberger WF. The Theory <strong>of</strong> Response-Adaptive Randomization<br />
in <strong>Clinical</strong> Trials. New York, NY: Wiley; 2006.<br />
8. Wei LJ, Durham S. The randomized play-the-winner rule in medical<br />
trials. J Am Stat Assoc. 1978;73:840-843.<br />
9. Korn EL, Freidlin B. Outcome-adaptive randomization: is it useful?<br />
J Clin Oncol. 2003;90:367-378.<br />
10. Eisenhauer E, Twelves C, Buyse M. Phase I <strong>Clinical</strong> Trials in Cancer.<br />
Oxford, England: Oxford University Press; 2006.<br />
11. Cheung YK. Designs for phase I trials. In: Harrington D (ed). Designs<br />
for <strong>Clinical</strong> Trials: Perspectives on Current Issues. New York, NY: Springer;<br />
2011.<br />
12. O’Quigley J, Pepe M, Fisher L. Continual reassessment method: a<br />
practical design for phase 1 clinical trials in cancer. Biometrics. 1990;46:33-48.<br />
13. Chevret S. The continual reassessment method in cancer phase I<br />
clinical trials: a simulation study. Stat Med. 1993;12:1093-1108.
LIMITATIONS OF ADAPTIVE CLINICAL TRIALS<br />
14. Goodman SN, Zahurak ML, Piantadosi S. Some practical improvements<br />
in the continual reassessment method for phase I studies. Stat Med.<br />
1995;14:1149-1161.<br />
15. Inoue LY, Thall PF, Berry DA. Seamlessly expanding a randomized<br />
phase II trial to phase III. Biometrics. 2002;58:823-831.<br />
16. Bretz F, Koenig F, Brannath W, et al. Adaptive designs for confirmatory<br />
clinical trials. Stat Med. 2009;28:1181-1217.<br />
17. Jennison C, Turnbull BW. Confirmatory seamless Phase II/III clinical<br />
trials with hypotheses selection at interim: opportunities and limitations.<br />
Biom J. 2006;48:650-655.<br />
18. Jennison C, Turnbull BW. Adaptive seamless designs: selection and<br />
prospective testing <strong>of</strong> hypotheses. J Biopharm Statist. 2007;17:1135-1161.<br />
19. Zhou X, Liu S, Kim ES, Herbst RS, Lee JJ. Bayesian adaptive design<br />
for targeted therapy development in lung cancer—a step toward personalized<br />
medicine. Clin Trials. 2008;5:181-193.<br />
20. Barker AD, Sigman CC, Kell<strong>of</strong>f GJ, Hylton NM, Berry DA, Esserman<br />
LJ. I-SPY 2: an adaptive breast cancer trial design in the setting <strong>of</strong> neoadjuvant<br />
chemotherapy. Clin Pharmacol Ther. 2009;86:97-100.<br />
21. European Medicines Agency (EMA), Committee for Medicinal Products<br />
for Human Use (CHMP). Reflection Paper on Methodological Issues in<br />
Confirmatory <strong>Clinical</strong> Trials Planned With an Adaptive Design. http://home.<br />
att.ne.jp/red/akihiro/emea/245902enadopted.pdf. Accessed February 10,<br />
<strong>2012</strong>.<br />
22. U.S. Department <strong>of</strong> Health and Human Services, Food and Drug<br />
Administration (FDA). Draft Guidance for Industry: Adaptive Design <strong>Clinical</strong><br />
Trials for Drugs and Biologics. http://www.fda.gov/downloads/Drugs/<br />
GuidanceComplianceRegulatoryInformation/Guidances/UCM201790.pdf. Accessed<br />
February 10, <strong>2012</strong>.<br />
137
ENDPOINTS FOR CANCER TRIALS IN <strong>2012</strong>:<br />
STATISTICAL, REGULATORY, AND CLINICAL<br />
PERSPECTIVES (eQ&A)<br />
CHAIR<br />
Daniel J. Sargent, PhD<br />
Mayo Clinic<br />
Rochester, MN<br />
SPEAKERS<br />
Ann T. Farrell, MD<br />
U. S. Food and Drug Administration<br />
Silver Spring, MD<br />
Deborah Watkins Bruner, PhD, RN<br />
Abramson Cancer Center, University <strong>of</strong> Pennsylvania<br />
Philadelphia, PA
Capturing the Patient Perspective:<br />
Patient-Reported Outcomes as <strong>Clinical</strong><br />
Trial Endpoints<br />
Deborah Watkins Bruner, RN, PhD, Benjamin Movsas, MD, and Ethan Basch, MD<br />
Overview: Just as clinical trial design and rigor have evolved<br />
with improvements in methods and processes, so too have<br />
methods for capturing patient data in clinical trials. Substantial<br />
evidence suggests that standard physician reporting <strong>of</strong><br />
symptoms for which we lack objective diagnostics (e.g., pain)<br />
is <strong>of</strong>ten discordant with patient self-report. Current reporting<br />
using the National Cancer Institute Common Terminology<br />
Criteria for Adverse Events (CTC[AE]) for symptom capture<br />
relies on a filtering system, from patient to physician to<br />
medical record to medical record abstraction to data entry,<br />
with each step requiring interpretation and the possibility <strong>of</strong><br />
error. In contrast, patient-reported outcomes (PROs) eliminate<br />
the filter and rely on direct report. Furthermore, the lack <strong>of</strong><br />
validation and training in use <strong>of</strong> the CTC(AE) creates an<br />
inadequate data capture system. Inadequacies might be observed<br />
as underreporting or overreporting symptom preva-<br />
THE TITLE <strong>of</strong> this chapter, “Capturing the Patient<br />
Perspective: Patient-Reported Outcomes as <strong>Clinical</strong><br />
Trial Endpoints,” implies that the patient perspective is<br />
separate from other clinical trial endpoints. All endpoints<br />
involve patient data. Why then the emphasis on the patient<br />
perspective? Currently, all cancer clinical trials use the<br />
National Cancer Institute (NCI) Common Terminology Criteria<br />
for Adverse Events (CTC[AE]), a descriptive system<br />
that is used for adverse event reporting. The CTC(AE) uses<br />
a grading (severity) scale for each adverse event term. 1<br />
Grading is performed by either the physician or research<br />
assistants, who abstract information from the patient’s<br />
medical records. Either method <strong>of</strong> grading requires interpreting<br />
or filtering the patient experience. In contrast, “A<br />
patient-reported outcome (PRO) is a measurement <strong>of</strong> any<br />
aspect <strong>of</strong> a patient’s health status that comes directly from<br />
the patient (i.e., without interpretation <strong>of</strong> the patient’s<br />
responses by a physician or anyone else).” 2<br />
The Pros and Cons for Use or Nonuse <strong>of</strong> PROs in<br />
<strong>Clinical</strong> Trials<br />
What do we gain over standard CTC(AE) reporting by<br />
including the patient perspective in clinical trials? Table 1<br />
lists the complementary benefits <strong>of</strong> the CTC(AE) and PROs.<br />
However, given limited resources, what is the downside <strong>of</strong><br />
experienced physicians and research associates grading a<br />
patient’s symptoms? The downside is they might not be<br />
correctly grading symptoms and/or they might be missing<br />
key information; either can occur for a host <strong>of</strong> reasons.<br />
Lack <strong>of</strong> Reliability and Validity <strong>of</strong> the Current<br />
Grading System<br />
The CTC(AE) was not designed as a physician checklist or<br />
patient interview, and there are a variety <strong>of</strong> ways in which<br />
the grading is performed and by whom it is performed,<br />
which are described elsewhere. 3 This variation in grading<br />
leads to questions about training to use the grading system,<br />
and the answer is there is no training. This lack <strong>of</strong> training<br />
lence and severity compared with patient self-report.<br />
Inaccuracies in symptom reporting can lead to missing important<br />
prognostic information, lack <strong>of</strong> understanding <strong>of</strong> patient<br />
adherence with therapies, and lack <strong>of</strong> information for patient<br />
decision making. They can also lead to opportunities lost in<br />
terms <strong>of</strong> labeling claims and comparative effectiveness analyses.<br />
New developments in patient-reported outcome (PRO)<br />
reporting, including the PRO-CTC(AE) and models for incorporation<br />
<strong>of</strong> PROs in clinical trials, might facilitate routine<br />
PRO reporting complementary to CTC(AE) in clinical trials. In<br />
addition, the cadre <strong>of</strong> validated PRO instruments already in<br />
existence allows for more in-depth, hypothesis-driven evaluations.<br />
For standard toxicity reporting, the time has come for<br />
mandatory routine PRO symptom reporting complementary to<br />
the CTC(AE).<br />
is evident in the rare instances when the CTC(AE) has been<br />
tested for interrater reliability, which is modest at best.<br />
Kabaet al 4 evaluated the reliability <strong>of</strong> CTC version 2.0 with<br />
five experienced research associates independently reviewing<br />
17 different medical records and grading toxicities.<br />
Agreement among raters was lowest for symptoms that are<br />
difficult to directly observe and highest for symptoms that<br />
have an associated laboratory test, for example, agreement<br />
for nausea was 0.47 (0.23–0.71, 95% CI) and agreement for<br />
febrile neutropenia was 0.88 (0.73–1, 95% CI). 4 Atkinson<br />
and colleagues 5 assessed interrater reliability <strong>of</strong> multiple<br />
CTC(AE) symptoms. In a study at an NCI-designated comprehensive<br />
cancer center, patients receiving active chemotherapy<br />
routinely had toxicity assessments performed by<br />
two clinicians. The ratings were performed independently,<br />
without access to the other physicians’ reports, on a medical<br />
record form that included the name <strong>of</strong> each symptom, a<br />
checkbox to note the grade <strong>of</strong> the toxicity, and a key with<br />
definitions <strong>of</strong> each CTC(AE) grade for each symptom. Agreement<br />
among raters was low for most symptoms (e.g. diarrhea<br />
[0.58], constipation [0.5], nausea [0.52], and vomiting<br />
[0.46]). In addition to poor interrater reliability, no assessments<br />
<strong>of</strong> validity <strong>of</strong> the CTC(AE) have been published,<br />
meaning it is unknown whether grade 1 is clinically different<br />
from grade 2, etc. By contrast, the use <strong>of</strong> PROs in clinical<br />
trials is held to stringent and rigorous development and<br />
psychometric validation as described in detail in the U.S.<br />
Food and Drug Administration (FDA) guidance for use <strong>of</strong><br />
PROs. 2<br />
From the Nell Hodgson Woodruff School <strong>of</strong> Nursing, Winship Cancer Institute, Emory<br />
University, Atlanta, GA; Radiation <strong>Oncology</strong> Department, Henry Ford Health System,<br />
Detroit, MI; Memorial Sloan-Kettering Cancer Center, New York, NY.<br />
Authors’ disclosures <strong>of</strong> potential conflicts <strong>of</strong> interest are found at the end <strong>of</strong> this article.<br />
Address reprint requests to Deborah Watkins Bruner, RN, PhD, Nell Hodgson Woodruff<br />
School <strong>of</strong> Nursing, Winship Cancer Institute, Emory University, 1520 Clifton Rd., Room<br />
232, Atlanta, GA 30322-4201; email: dwbrune@emory.edu.<br />
© <strong>2012</strong> by <strong>American</strong> <strong>Society</strong> <strong>of</strong> <strong>Clinical</strong> <strong>Oncology</strong>.<br />
1092-9118/10/1-10<br />
139
Table 1. Complementary Use <strong>of</strong> Common Terminology Criteria for Adverse Events (CTC[AE]) and Patient-Reported Outcomes (PROs)<br />
in <strong>Clinical</strong> Trials<br />
Underreporting Prevalence <strong>of</strong> Symptoms<br />
Data have demonstrated that PROs <strong>of</strong>ten identify more<br />
symptoms than standard CTC(AE) reporting. In a phase III<br />
trial <strong>of</strong> a cream compared with a placebo to prevent breast<br />
cancer radiation therapy–related dermatitis, PROs delineated<br />
a wider spectrum <strong>of</strong> toxicity than CTC(AE) and provided<br />
more information on rash, redness, pruritus, and<br />
annoyance measures compared with CTC(AE) findings <strong>of</strong><br />
rash and pruritus. 6<br />
Underreporting and Overreporting Levels <strong>of</strong> Severity<br />
A burgeoning body <strong>of</strong> evidence indicates that physicians<br />
<strong>of</strong>ten underreport or overreport the level <strong>of</strong> severity <strong>of</strong><br />
symptoms compared with patient self-report. Watkins-<br />
Bruner and colleagues, 7 in the NCI-sponsored Radiation<br />
Therapy <strong>Oncology</strong> Group (RTOG), reported discordance between<br />
physician and patient reporting in RTOG’s first published<br />
PRO report (RTOG 90-20), “A Phase II Trial <strong>of</strong><br />
Radiotherapy Plus a Radiosensitizer for Locally Advanced<br />
Prostate Cancer.” In a comparison <strong>of</strong> patient self-report <strong>of</strong><br />
symptoms on a validated measure with physician ratings<br />
KEY POINTS<br />
● The current cancer clinical trial reporting system<br />
using the National Cancer Institute Common Terminology<br />
Criteria for Adverse Events (CTC[AE]) is<br />
inadequate, by itself, to capture subjective symptoms.<br />
● Patient-reported outcome (PRO) measures, rigorously<br />
developed and validated, provide important and<br />
accurate information required for robust clinical trials<br />
reporting.<br />
● The risk <strong>of</strong> not including PROs in routine reporting <strong>of</strong><br />
clinical trials is substantial inaccuracies in reporting<br />
<strong>of</strong> prevalence and severity <strong>of</strong> symptoms related to<br />
treatments, missing important prognostic information,<br />
lack <strong>of</strong> understanding <strong>of</strong> patient adherence with<br />
therapies, and lack <strong>of</strong> information for patient decision<br />
making.<br />
● Opportunities lost in not including PROs in routine<br />
clinical reporting include lack <strong>of</strong> information for labeling<br />
claims and comparative effectiveness.<br />
● Validated and reliable PROs exist that are easily<br />
available for both routine reporting <strong>of</strong> symptoms and<br />
more complex evaluations <strong>of</strong> hypotheses-driven endpoints.<br />
CTC(AE) PROs<br />
Primary use Toxic effects reporting Health status reporting<br />
Most useful for Objective assessment (e.g., diagnostic test, imaging,<br />
overt sign, such as bleeding)<br />
Subjective assessment (e.g., cannot be seen, felt, heard, observed,<br />
or clinically tested by physician)<br />
Best captures Severity, need for physician intervention Severity, function, effect on quality <strong>of</strong> life and treatment adherence<br />
Valid Not tested Yes a<br />
Reliable No Yes a<br />
Data capture method Through layers <strong>of</strong> interpretation Directly from the patient<br />
Time <strong>of</strong> data capture As it occurs/as physician picks it up At designated time points<br />
a Legacy instruments psychometrically tested to varying degrees; for current U.S. Food and Drug Administration use, must conform to stringent guidelines<br />
140<br />
<strong>of</strong> the same symptoms on a toxicity scale, disagreement<br />
between patient and physician reports <strong>of</strong> severity <strong>of</strong> dysuria,<br />
diarrhea, and erectile dysfunction ranged from 13% to 45%.<br />
The RTOG determined that the findings warranted routine<br />
use <strong>of</strong> PROs in clinical trials. A larger initiative to assess<br />
how well PROs correspond to reports <strong>of</strong> the same symptoms<br />
as rated by the CTC(AE) across NCI cancer clinical trial<br />
cooperative groups included a pooled analysis <strong>of</strong> 12 lung<br />
cancer clinical trials conducted by three cooperative groups. 8<br />
Data were assessed in 1,013 patients who had both adverse<br />
events and PROs recorded. Agreement between incidence <strong>of</strong><br />
any grade adverse event and a substantial decrease in PROs<br />
ranged from 27% to 61% for grade 2 and higher and 36% to<br />
61% for grade 3 and higher toxicities. Agreement between<br />
the incidence <strong>of</strong> a specific grade adverse event and a substantial<br />
decrease in a corresponding PRO ranged from 0% to<br />
38% for grade 2 and higher and 0% to 14% for grade 3 and<br />
higher toxicities. 8<br />
Missing Prognostic Information<br />
BRUNER, MOVSAS, AND BASCH<br />
Data from clinical trials have shown validated PROs to be<br />
prognostic for survival in almost every cancer disease site,<br />
especially in more advanced stage disease. For example, in<br />
an RTOG clinical trial <strong>of</strong> a radioprotectant for radiationinduced<br />
esophagitis in the treatment <strong>of</strong> non–small cell lung<br />
cancer, baseline patient-reported scores on the European<br />
Organisation for Research and Treatment <strong>of</strong> Cancer<br />
Quality-<strong>of</strong>-Life Questionnaire replaced known prognostic<br />
factors (performance status, stage, sex, age, race, marital<br />
status, histologic features, and tumor location) as the sole<br />
predictor <strong>of</strong> long-term overall survival. A 10-point higher<br />
baseline quality-<strong>of</strong>-life score corresponded to a decrease in<br />
the hazard <strong>of</strong> death by approximately 10% (p � 0.004). 9 A<br />
Gynecologic <strong>Oncology</strong> Group analysis <strong>of</strong> a phase III trial<br />
<strong>of</strong> cisplatin with or without topotecan in advanced cervical<br />
cancer found the patient-reported Functional Assessment <strong>of</strong><br />
Cancer Therapy–General (FACT-G) plus the FACT-Cervix<br />
subscale had a similar association with survival as the<br />
RTOG study even though different PRO measures were used<br />
and different disease sites were involved. A 10-point higher<br />
baseline score on FACT-Cervix corresponded to a decrease<br />
in hazard <strong>of</strong> death by approximately 10% (p � 0.002). 10 A<br />
pharmaceutical company–sponsored, phase III trial demonstrated<br />
efficacy <strong>of</strong> sunitinib, which prolonged progressionfree<br />
survival for patients with metastatic renal cell cancer.<br />
Three baseline PRO measures were individually predictive<br />
<strong>of</strong> progression-free survival. Higher scores at baseline on<br />
FACT-G, the FACT–Kidney Symptom Index disease-related<br />
symptoms subscale, and EuroQol Group’s visual analog<br />
scale were significantly associated with longer progression-
PATIENT-REPORTED OUTCOMES AS TRIAL ENDPOINTS<br />
free survival (hazard ratios, 0.93, 0.89, and 0.91; p � 0.001,<br />
p � 0.001, and p � 0.008, respectively). These data indicate<br />
that the risk <strong>of</strong> tumor progression or death was approximately<br />
7%, 11%, and 9% lower, respectively, for every higher<br />
change score <strong>of</strong> 5 points on FACT-G, 2 points on the<br />
FACT–Kidney Symptom Index disease-related symptoms<br />
subscale, and 10 points on EuroQol Group’s visual analog<br />
scale. 11<br />
Gotay et al 12 confirmed this trend <strong>of</strong> the prognostic significance<br />
<strong>of</strong> PROs, Gotay et al in a systematic literature review<br />
<strong>of</strong> cancer clinical trials in which they found that 36 (92%) <strong>of</strong><br />
39 studies had at least one PRO that was prognostic <strong>of</strong><br />
survival (p � 0.05) in multivariate analysis. Of the 39<br />
clinical trials that met inclusion criteria, most were lung (12<br />
trials) and breast (eight trials) cancer trials, with a total <strong>of</strong><br />
13,874 patients. The European Organisation for Research<br />
and Treatment <strong>of</strong> Cancer Quality-<strong>of</strong>-Life Questionnaire was<br />
the most common PRO used, found in 56% <strong>of</strong> the trials.<br />
Effect sizes varied, with 24 <strong>of</strong> 36 (66%) having a small effect<br />
size, eight (23%) having a moderate effect size, and 4 (11%)<br />
having a large effect size. Adjusting for performance status,<br />
treatment arm, stage, weight loss, and serum markers,<br />
PROs still provided distinct prognostic information for survival<br />
beyond standard clinical measures.<br />
Lack <strong>of</strong> Understanding <strong>of</strong> Patient Adherence<br />
Aromatase inhibitors provide an excellent example <strong>of</strong> how<br />
not including PROs or not including them at the right time<br />
points can mislead our understanding <strong>of</strong> patient adherence.<br />
A large pivotal clinical trial <strong>of</strong> 5,187 patients demonstrated<br />
the efficacy <strong>of</strong> an aromatase inhibitor to decrease local or<br />
metastatic recurrence or new contralateral breast cancer.<br />
The toxicities associated with aromatase inhibitors found<br />
using CTC version 2.0 were reported as primarily grade 1 or<br />
2, and those that were higher in the aromatase inhibitor<br />
group compared with the placebo group were hot flashes,<br />
arthritis, arthralgia, and myalgia. At a median <strong>of</strong> 2.4 years<br />
<strong>of</strong> follow-up, no significance was found between discontinued<br />
use <strong>of</strong> the aromatase inhibitor compared with placebo (4.5%<br />
vs. 3.6%, p � 0.11). 13 In the quality-<strong>of</strong>-life analysis <strong>of</strong> the<br />
same trial, Whelan and colleagues 14 found small but meaningful<br />
worsening in mean change scores from baseline in the<br />
PRO domains <strong>of</strong> vasomotor effects, bodily pain, vitality,<br />
physical functioning, and sexual function, with the last two<br />
domains having no corresponding detection using CTC reporting.<br />
However, the findings <strong>of</strong> studies in which adherence<br />
was evaluated over shorter durations suggest better adherence<br />
than studies with longer follow-up. One-year aromatase<br />
inhibitor adherence rates range from 77% to 88%,<br />
whereas 4- to 5-year prescription-based adherence rates<br />
were 27% to 49% in other studies. 15 PROs can help provide<br />
the answer to decreasing adherence over time. PRO data<br />
from two pivotal phase IV aromatase inhibitor trials found<br />
high prevalence rates for joint pain (59.6%), hot flushes<br />
(52%), lost interest in sexual intercourse (51.4%), and lack<br />
<strong>of</strong> energy (40.3%). PROs resulted in higher prevalence rates<br />
compared with physician ratings for most symptoms. 16 A<br />
recent study found that 36% <strong>of</strong> women who were prescribed<br />
aromatase inhibitors for 5 years reported ending treatment<br />
because <strong>of</strong> symptom bother. 16<br />
Lack <strong>of</strong> Information for Patient and<br />
<strong>Clinical</strong> Decision Making<br />
Patients <strong>of</strong>ten seek information that would help them<br />
make decisions regarding treatments and symptom management.<br />
Without systematic PRO assessments in clinical trials,<br />
patients might seek out less rigorously gathered data.<br />
For example, www.patientslikeme.com/ is a popular Webbased<br />
tool that seeks to provide users with information to<br />
help make treatment decisions and manage symptoms. An<br />
analysis <strong>of</strong> approximately 2,000 unique PRO reports on<br />
modafinil use across five disease conditions indicated how<br />
patients are easily able to gather information on the use <strong>of</strong><br />
the drug for management <strong>of</strong> symptoms. However, less than<br />
1% <strong>of</strong> members who reported taking modafinil reported<br />
taking it for an approved purpose (narcolepsy and excessive<br />
daytime sleepiness resulting from sleep apnea). Patients are<br />
able to access information on the other self-reported uses <strong>of</strong><br />
the drug, for example, general fatigue (68%), excessive<br />
daytime sleepiness (16%), and difficulty concentrating (3%),<br />
as well as perceived effectiveness <strong>of</strong> ameliorating each<br />
symptom. 18 In addition to the <strong>of</strong>f-label use <strong>of</strong> the drug,<br />
patients report adverse effects they attribute to the drug.<br />
Routine reporting <strong>of</strong> PROs in clinical trials and a commitment<br />
to lay broadcasting beyond academic publication<br />
would provide more accurate information for decision making.<br />
Lack <strong>of</strong> Information for Labeling Claims<br />
Few cancer drugs have received FDA approval for a<br />
symptomatic benefit. A recent exception is the November<br />
2011 approval <strong>of</strong> ruxolitinib for the treatment <strong>of</strong> myel<strong>of</strong>ibrosis.<br />
FDA approval for symptomatic benefit before 2011 goes<br />
back to 1998. However, between 1998 and 2011 almost 70<br />
anticancer drugs have been approved. Incyte Corp. obtained<br />
this rare symptom benefit labeling claim through close<br />
interaction with the FDA. The primary endpoint was reduction<br />
in spleen volume, and the secondary PRO endpoint was<br />
the total symptom score using a specifically developed<br />
measure following the FDA PRO guidelines. The company<br />
worked with the FDA for 2 years to finalize instrument<br />
approval. The PRO measured six symptoms: night sweats,<br />
itching, abdominal discomfort, pain under the ribs on the<br />
left side, early satiety, and bone or muscle pain. Along with<br />
rigorous instrument development, the clinical trial hypothesized<br />
a large effect <strong>of</strong> 50% or greater reduction in total<br />
symptom score from baseline to week 24. Forty-six percent<br />
<strong>of</strong> patients in the ruxolitinib arm met the PRO threshold<br />
compared with 5.3% in the placebo group (p � 0.0001).<br />
Richard Pazdur, the director <strong>of</strong> FDA’s Office <strong>of</strong> Hematology<br />
<strong>Oncology</strong> Products, was quoted as saying, “[The PRO] was a<br />
secondary endpoint, but in our mind this is why we gave the<br />
application full approval. One could quibble about the importance<br />
<strong>of</strong> reduction in spleen size, but with reduction in all<br />
the symptoms,” full approval was warranted. 19<br />
Lack <strong>of</strong> Information for Comparative Effectiveness<br />
Without PRO data in clinical trials, we would have to rely<br />
solely on toxicity data for comparative effectiveness <strong>of</strong> treatment<br />
modalities, but we have a preponderance <strong>of</strong> literature<br />
showing that physician-reported toxicities do not tell the full<br />
story. For example, in RTOG 0126, a phase III clinical trial,<br />
patients were randomly assigned to high-dose (79.2-Gy) or<br />
141
Table 2. Barriers and Solutions to Standard Use <strong>of</strong> PROs in <strong>Clinical</strong> Trials<br />
Barrier Solution<br />
Lack <strong>of</strong> “<strong>of</strong>f-the-shelf ” validated instruments Need to make PROs publically available<br />
Inadvertent unmasking Require a large effect size<br />
Substantiate symptomatic benefits via objective measures (such as radiographic<br />
or serum biomarker responses)<br />
Missing data e-products for data capture, real-time reminders, real-time monitoring, telephone<br />
interviews when patient is too ill to complete PRO<br />
There may be concerns that systematic assessment will document more low<br />
grade toxicities<br />
Additional resources for routine inclusion <strong>of</strong> PROs in clinical trials for symptom<br />
assessment may be needed since the NCI generally only funds PROs that<br />
answer a specific hypothesis<br />
standard-dose (70.2-Gy) radiation therapy for localized prostate<br />
cancer. A preliminary analysis <strong>of</strong> the high-dose arm,<br />
comparing 3-dimensional conformal radiotherapy (3DCRT)<br />
and intensity-modulated radiotherapy (IMRT), where each<br />
institution declared its choice <strong>of</strong> 3DCRT or IMRT before trial<br />
participation, was conducted. A total <strong>of</strong> 763 patients were<br />
in the high-dose arm and included in the toxicity analysis, 20<br />
and 499 patients (65%) were included in the PRO analysis. 21<br />
Although the CTC version 2.0 toxicity report indicated a<br />
benefit in favor IMRT for grade 2 or higher gastrointestinal<br />
toxicities (22% 3DCRT vs. 15% IMRT, p � 0.039) and in<br />
favor <strong>of</strong> IMRT for combined gastrointestinal and genitourinary<br />
grade 2 or higher toxicities (15.1% 3DCRT vs. 9.7%<br />
IMRT, p � 0.042), PROs did not corroborate a benefit<br />
patients could note. Corresponding PRO gastrointestinal<br />
and genitourinary variables indicated no substantial differences<br />
between 3DCRT and IMRT. An important difference<br />
that PROs demonstrated in favor <strong>of</strong> IMRT not found on<br />
PROs better identify baseline symptoms related to disease or prior treatment than<br />
current standard physician toxic effect reporting, making it clearer during a<br />
trial which symptoms are attributable to study drug compared with preexisting<br />
causes<br />
Standard PRO reporting should be viewed no differently than standard CTC(AE)<br />
reporting and routinely incorporated as such<br />
Abbreviations: CTC(AE), Common Terminology Criteria for Adverse Events; NCI, National Cancer Institute; PRO, patient-reported outcome.<br />
142<br />
BRUNER, MOVSAS, AND BASCH<br />
physician reporting was erectile dysfunction; <strong>of</strong> note, the<br />
IMRT benefit was primarily in men younger than 70 years<br />
(p � 0.04). Given the higher cost <strong>of</strong> IMRT over 3DCRT for<br />
the treatment <strong>of</strong> prostate cancer, the PRO data add an<br />
important dimension to the comparative effectiveness equation.<br />
Although the benefits <strong>of</strong> inclusion <strong>of</strong> PROs in clinical trials<br />
and the risks <strong>of</strong> noninclusion have been delineated, there<br />
are some barriers to routine inclusion <strong>of</strong> PROs in clinical<br />
trials. However, solutions to these barriers exist (Table 2).<br />
Numerous validated PRO instruments with the breadth and<br />
depth to answer hypothesis-specific clinical trial questions<br />
are available, but a more parsimonious method would help<br />
routine inclusion in clinical trials.<br />
A major step forward has been the development <strong>of</strong> the<br />
PRO version <strong>of</strong> the CTC-AE (PRO-CTC[AE]). 22 The PRO-<br />
CTC(AE) has been designed to capture a comprehensive<br />
range <strong>of</strong> symptoms and functioning that enhances monitor-<br />
Fig. 1. Radiation Therapy <strong>Oncology</strong><br />
Group (RTOG) outcomes model for guiding<br />
inclusion <strong>of</strong> patient-reported outcomes<br />
in clinical trials.<br />
Abbreviations: RT, radiation therapy;<br />
PSA, prostate specific antigen; QALYs,<br />
quality-adjusted life years; PROs, patient<br />
reported outcomes.
PATIENT-REPORTED OUTCOMES AS TRIAL ENDPOINTS<br />
ing <strong>of</strong> adverse events in cancer clinical trials. Approximately<br />
78 CTC(AE) items have been converted to PROs, including,<br />
for example, items purely or with a large subjective component,<br />
such as pain, fatigue, nausea or vomiting, elimination,<br />
sexual function, swallowing, and dyspnea. This NCI investment<br />
lays the groundwork for standard inclusion <strong>of</strong> PROs in<br />
clinical trials as a companion to the CTC(AE).<br />
In addition to having publicly accessible validated measures<br />
for PRO assessments, a framework for incorporation<br />
<strong>of</strong> PROs into clinical trials would help guide consistent<br />
assessments. To that end, the RTOG developed an outcomes<br />
model for guidance <strong>of</strong> clinical trials concept development.<br />
This model was first developed as a triad <strong>of</strong> outcomes,<br />
including clinical (e.g., standard survival and toxicity reporting),<br />
humanistic (patient-reported symptoms, quality<br />
<strong>of</strong> life, functional status, and preferences), and economic<br />
(quality-adjusted life-years and comparisons <strong>of</strong> effectiveness).<br />
23 The outcomes model has evolved with the increasing<br />
focus on the complexity and interrelationships <strong>of</strong> additional<br />
variables, including biomarkers and physical parameters<br />
(particular to radiation treatment planning) that can influence<br />
and be influenced by patient-reported outcomes. Figure<br />
1 depicts an example <strong>of</strong> the most recent iteration <strong>of</strong> this<br />
model used to guide inclusion <strong>of</strong> endpoints on an active trial<br />
(RTOG 0543), “A Phase III Trial <strong>of</strong> Short Term Androgen<br />
Deprivation With Pelvic Lymph Node or Prostate Bed Only<br />
Authors’ Disclosures <strong>of</strong> Potential Conflicts <strong>of</strong> Interest<br />
Author<br />
Employment or<br />
Leadership<br />
Positions<br />
Consultant or<br />
Advisory Role<br />
Deborah Watkins Bruner Endo<br />
Pharmaceuticals<br />
Benjamin Movsas*<br />
Ethan Basch*<br />
*No relevant relationships to disclose.<br />
Radiotherapy in Prostate Cancer Patients With a Rising<br />
PSA After Radical Prostatectomy.”<br />
Further advances in the area <strong>of</strong> PROs have recently<br />
emerged from the September 2011 NCI-sponsored clinical<br />
trials planning meeting, “Identification <strong>of</strong> Core Symptoms<br />
and Health-Related Quality <strong>of</strong> Life Domains for Use in<br />
Cancer Research.” An extensive process <strong>of</strong> literature and<br />
source data review identified a lengthy list <strong>of</strong> symptoms<br />
ranked by prevalence. A panel <strong>of</strong> experts narrowed the list to<br />
a core set <strong>of</strong> symptoms that can serve as a guide to common<br />
reporting across all cancer clinical trials. If adopted, this<br />
core set <strong>of</strong> PROs could facilitate comparison and combination<br />
<strong>of</strong> data across cancer clinical trials around the world.<br />
The publications from this meeting are in development.<br />
Conclusion<br />
This chapter provides a review <strong>of</strong> what we stand to lose<br />
without the routine incorporation <strong>of</strong> PROs in clinical trials.<br />
It is important to also consider what we stand to gain,<br />
including more comprehensive reporting <strong>of</strong> prevalence <strong>of</strong><br />
symptoms, improved accuracy in reporting <strong>of</strong> levels <strong>of</strong> severity,<br />
increased prognostic specificity, greater understanding<br />
<strong>of</strong> patient adherence, better information for patient and<br />
clinical decision making, additional targets for labeling<br />
claims, and information for comparative effectiveness. Most<br />
importantly, we give voice to the one who counts most, the<br />
patient.<br />
Stock<br />
Ownership Honoraria<br />
Bristol-Myers<br />
Squibb<br />
1. National Cancer Institute. Common Terminology Criteria for Adverse<br />
Events (CTCAE) and Common Toxicity Criteria (CTC). 2009. http://evs.nci.<br />
nih.gov/ftp1/CTCAE/About.htm. Accessed March 20, <strong>2012</strong>.<br />
2. U.S. Food and Drug Administration. Guidance for Industry: Patient-<br />
Reported Outcome Measures: Use in Medical Product Development to Support<br />
Labeling Claims. 2009. http://www.fda.gov/downloads/Drugs/Guidance<br />
ComplianceRegulatoryInformation/Guidances/UCM193282.pdf. Accessed March<br />
20, <strong>2012</strong>.<br />
3. Bruner DW. Should patient-reported outcomes be mandatory for toxicity<br />
reporting in cancer clinical trials? J Clin Oncol. 2007;25:5345-5347.<br />
4. Kaba H, Fukuda H, Yamamoto S, Ohashi Y. Reliability at the National<br />
Cancer Institute-Common Toxicity Criteria version 2.0 [in Japanese]. Gan To<br />
Kagaku Ryoho. 2004;31:1187-1192.<br />
5. Atkinson TM, Li Y, C<strong>of</strong>fey CW, et al. Reliability <strong>of</strong> adverse symptom<br />
event reporting by clinicians. Qual Life Res. Epub 2011 Oct 8.<br />
6. Neben-Wittich MA, Atherton PJ, Schwartz DJ, et al. Comparison <strong>of</strong><br />
provider-assessed and patient-reported outcome measures <strong>of</strong> acute skin<br />
toxicity during a Phase III trial <strong>of</strong> mometasone cream versus placebo during<br />
breast radiotherapy: the North Central Cancer Treatment Group (N06C4).<br />
Int J Radiat Oncol Biol Phys. 2011;81:397-402.<br />
7. Watkins-Bruner D, Scott C, Lawton C, et al. RTOG’s first quality <strong>of</strong> life<br />
study-RTOG 90-20: a phase II trial <strong>of</strong> external beam radiation with etanidazole<br />
for locally advanced prostate cancer. Int J Radiat Oncol Biol Phys.<br />
1995;33:901-906.<br />
8. Bruner DW, Gotay C, Moinpour C, et al. Value-added <strong>of</strong> patient-reported<br />
outcomes (pro) in cooperative group oncology clinical trials: a pooled analysis.<br />
In: Proceedings <strong>of</strong> the 15th Annual Meeting <strong>of</strong> the International <strong>Society</strong> for<br />
REFERENCES<br />
Research<br />
Funding<br />
Expert<br />
Testimony<br />
Other<br />
Remuneration<br />
Quality <strong>of</strong> Life Research; October 22-25. 2008; Montevideo, Uruguay. Abstract<br />
28.<br />
9. Movsas B, Moughan J, Sarna L, et al. Quality <strong>of</strong> life supersedes the<br />
classic prognosticators for long-term survival in locally advanced non-smallcell<br />
lung cancer: an analysis <strong>of</strong> RTOG 9801. J Clin Oncol. 2009;27:5816-5822.<br />
10. Monk BJ, Huang HQ, Cella D, Long HJ III. Quality <strong>of</strong> life outcomes<br />
from a randomized phase III trial <strong>of</strong> cisplatin with or without topotecan in<br />
advanced carcinoma <strong>of</strong> the cervix: a Gynecologic <strong>Oncology</strong> Group Study.<br />
J Clin Oncol 2005;23:4617-4625.<br />
11. Cella D, Cappelleri JC, Bushmakin A, et al. Quality <strong>of</strong> life predicts<br />
progression-free survival in patients with metastatic renal cell carcinoma<br />
treated with sunitinib versus interferon alfa. J Oncol Pract. 2009;5:66-70.<br />
12. Gotay CC, Kawamoto CT, Bottomley A, Efficace F. The prognostic<br />
significance <strong>of</strong> patient-reported outcomes in cancer clinical trials. J Clin<br />
Oncol. 2008;26:1355-1363.<br />
13. Goss PE, Ingle JN, Martino S, et al. A randomized trial <strong>of</strong> letrozole in<br />
postmenopausal women after five years <strong>of</strong> tamoxifen therapy for early-stage<br />
breast cancer. N Engl J Med. 2003;349:1793-1802.<br />
14. Whelan TJ, Goss PE, Ingle JN, et al. Assessment <strong>of</strong> quality <strong>of</strong> life in<br />
MA.17: a randomized, placebo-controlled trial <strong>of</strong> letrozole after 5 years <strong>of</strong><br />
tamoxifen in postmenopausal women. J Clin Oncol. 2005;23:6931-6940.<br />
15. Gotay C, Dunn J. Adherence to long-term adjuvant hormonal therapy<br />
for breast cancer. Exp Rev Pharmacoecon Outcomes Res. 2011;11:709-715.<br />
16. Oberguggenberger A, Hubalek M, Sztankay M, et al. Is the toxicity <strong>of</strong><br />
adjuvant aromatase inhibitor therapy underestimated? Complementary information<br />
from patient-reported outcomes (PROs). Breast Cancer Res Treat.<br />
2011;128:553-561.<br />
17. Wagner L, Zhao F, Chapman J, et al. Patient-reported predictors <strong>of</strong><br />
143
treatment discontinuation: quality <strong>of</strong> life among postmenopausal women with<br />
primary breast cancer randomized to exemestane or anastrozole. Cancer Res.<br />
2011;71(suppl):S6.<br />
18. Frost J, Okun S, Vaughan T, et al. Patient-reported outcomes as a<br />
source <strong>of</strong> evidence in <strong>of</strong>f-label prescribing: analysis <strong>of</strong> data from Patients<br />
LikeMe. J Med Internet Res. 2011;13:e6.<br />
19. McCallister E, Usdin S. A pr<strong>of</strong>essional trial. BioCentury. 2011;19:A1-<br />
A4.<br />
20. Michalski JM, Yan Y, Watkins-Bruner D, et al. Preliminary analysis <strong>of</strong><br />
3D-CRT vs. IMRT on the high dose arm <strong>of</strong> the RTOG 0126 prostate cancer<br />
trial: toxicity report. Int J Radiat Oncol Biol Phys. 2011;81(suppl):S1-S2.<br />
144<br />
BRUNER, MOVSAS, AND BASCH<br />
21. Watkins-Bruner D, Hunt D, Michalski J, et al. Preliminary analysis<br />
<strong>of</strong> 3DCRT vs IMRT on the high dose arm <strong>of</strong> the RTOG 0126 prostate cancer<br />
trial: patient reported outcomes. Int J Radiat Oncol Biol Phys. 2011;<br />
81(suppl):S44.<br />
22. National Cancer Institute. Patient-Reported Outcomes (PRO) version<br />
<strong>of</strong> the Common Terminology Criteria for Adverse Events (PRO-CTCAE).<br />
2010. http://outcomes.cancer.gov/tools/pro-ctcae.html. Accessed March 20,<br />
<strong>2012</strong>.<br />
23. Bruner DW. Outcomes research in cancer symptom management trials:<br />
the Radiation Therapy <strong>Oncology</strong> Group (RTOG) conceptual model. J Natl<br />
Cancer Inst Monogr. 2007;37:12-15.
NEW LOOKS AND CHALLENGES FOR<br />
COOPERATIVE GROUPS<br />
CHAIR<br />
Stephen S. Grubbs, MD<br />
Helen F. Graham Cancer Center at Christiana Care<br />
Newark, DE<br />
SPEAKERS<br />
Robert Leo Comis, MD<br />
Coalition <strong>of</strong> Cancer Cooperative Groups<br />
Philadelphia, PA<br />
Gregory H. Reaman, MD<br />
Children’s National Medical Center<br />
Washington, DC
The New National Cancer Institute National<br />
<strong>Clinical</strong> Trials Network<br />
Overview: The National Cancer Institute (NCI) Cooperative<br />
Group Program has been reviewed by three published studies<br />
in the last 7 years evaluating the efficiency and effectiveness<br />
<strong>of</strong> this national oncology clinical trials system. The recommendations<br />
for improvement from these reports have<br />
prompted NCI to transform the Cooperative Group Program<br />
THE NCI Cooperative Group Program has been an<br />
essential contributor to the discovery <strong>of</strong> new cancer<br />
therapies over the last 50 years. The program comprises<br />
networks <strong>of</strong> cancer centers and community oncology practices<br />
that develop and conduct large-scale multicenter<br />
clinical trials and promote discovery <strong>of</strong> correlative cancer<br />
science. The cooperative groups have established the therapies<br />
now routinely used to diagnose, prevent, and treat<br />
cancers.<br />
In 2007, the NCI director requested the Institute <strong>of</strong><br />
Medicine (IOM) conduct a consensus study <strong>of</strong> clinical trials<br />
and the NCI Cooperative Group Program and provide recommendations<br />
to improve the system. This evaluation was<br />
preceded by an NCI review <strong>of</strong> the system by the <strong>Clinical</strong><br />
Trials Working Group (CTWG) in June 2005. The IOM<br />
convened a 17-member committee representing a broad<br />
range <strong>of</strong> the oncology research community. The activity<br />
was supported by funding from the NCI and other cancer<br />
research organizations, including ASCO. A report was<br />
issued in April 2010 titled, “A National Cancer <strong>Clinical</strong><br />
Trials System for the 21 st Century.” Key features <strong>of</strong> the<br />
report are listed in Table 1.<br />
The NCI has responded to the IOM, CTWG, and the<br />
March 2010 NCI Operational Efficiency Work Group<br />
(OEWG) reports with a proposed transformation and reorganization<br />
<strong>of</strong> the cooperative groups into a new NCI <strong>Clinical</strong><br />
Trials Network (NCTN). Network goals are listed in Table 2.<br />
To achieve these goals, the NCI has embarked on cooperative<br />
group consolidation and reconfiguration <strong>of</strong> the network<br />
evaluation process and funding.<br />
The NCI has recommended the 10 cooperative groups<br />
consolidate to four adult-oriented groups and one pediatric<br />
group. The existing nine adult-oriented groups have elected<br />
to merge or affiliate into the four entities listed in Table 3.<br />
The NCI is in the process <strong>of</strong> developing a new funding<br />
opportunity announcement (FOA) to be published in July<br />
<strong>2012</strong> that will establish the new evaluation and funding <strong>of</strong><br />
the network. The NCI has articulated the goals <strong>of</strong> the new<br />
system evaluation (Table 4). NCI oversight and prioritization<br />
<strong>of</strong> the network will utilize the existing Disease Steering<br />
Committees and a newly proposed Cross-Disease/Trials<br />
Oversight Panel. More efficient clinical trials development<br />
and activation have already been implemented by strict<br />
timelines established in the OEWG report.<br />
The new FOA is expected in July <strong>2012</strong>, and competing<br />
applications for the new FOA are to be submitted by November<br />
<strong>2012</strong>, with NCI review in February 2013. Network<br />
awards are expected in late 2013 with funding to begin in<br />
2014.<br />
146<br />
By Stephen S. Grubbs, MD<br />
into a new NCI National <strong>Clinical</strong> Trials Network (NCTN) to<br />
improve the efficiency <strong>of</strong> large clinical trials and increase the<br />
speed <strong>of</strong> cancer translational research. The new NCTN <strong>of</strong>fers<br />
community-based clinical investigators new opportunities to<br />
advance cancer research in their community setting but also<br />
presents challenges in promoting community-based research.<br />
KEY POINTS<br />
● The Institute <strong>of</strong> Medicine, the <strong>Clinical</strong> Trials Working<br />
Group, and the Operational Efficiency Working<br />
Group have published evaluations recommending improvements<br />
<strong>of</strong> the National Cancer Institute (NCI)<br />
Cooperative Group Program.<br />
● The NCI is transforming the Cooperative Group Program<br />
into a National <strong>Clinical</strong> Trials Network by<br />
group consolidation and a new evaluation and funding<br />
model.<br />
● Nine adult disease–oriented cooperative groups have<br />
responded by consolidating into four groups, joining<br />
the existing single pediatric-oriented cooperative<br />
group.<br />
● Community clinical trial sites and investigators will<br />
potentially gain access to a greater variety <strong>of</strong> trials,<br />
will benefit from operational standardization between<br />
groups, and may achieve increase per case<br />
reimbursement.<br />
● Community sites, however, will be challenged by<br />
diminished volunteer investigator time and pressure<br />
to produce high accrual volume and remain at risk for<br />
per case funding below actual cost.<br />
Community Needs from the NCI NCTN<br />
Community clinical investigators are essential participants<br />
in a successful national oncology clinical trials system.<br />
They have access to large numbers <strong>of</strong> newly diagnosed<br />
patients and contribute significant accruals to phase III and<br />
cancer control and prevention trials. The Community <strong>Clinical</strong><br />
<strong>Oncology</strong> Program and cancer center community affiliate<br />
sites contribute over 50% <strong>of</strong> the patients annually to<br />
the cooperative group total accrual. The new NCTN <strong>of</strong>fers<br />
opportunities and challenges for the community clinical<br />
investigators.<br />
A successful community clinical trials site requires five<br />
From the Helen F. Graham Cancer Center, Delaware Christiana Care CCOP Newark,<br />
DE.<br />
Author’s disclosure <strong>of</strong> potential conflicts <strong>of</strong> interest are found at the end <strong>of</strong> this article.<br />
Address reprint requests to Stephen S. Grubbs, MD, Helen F. Graham Cancer Center,<br />
Delaware Christiana Care CCOP, 4701 Ogletown-Stanton Rd., Newark, DE, 19713-2055;<br />
email: ssgrubbs@cbg.org.<br />
© <strong>2012</strong> by <strong>American</strong> <strong>Society</strong> <strong>of</strong> <strong>Clinical</strong> <strong>Oncology</strong>.<br />
1092-9118/10/1-10
NCI NATIONAL CLINICAL TRIALS NETWORK<br />
main areas <strong>of</strong> support from NCTN and NCI: (1) clinical trial<br />
access, (2) group participation, (3) operation and regulatory<br />
standardization, (4) patient recruitment enhancement, and<br />
(5) adequate funding and support (Table 5). However, the<br />
new proposed network poses both old and new challenges,<br />
including NCTN membership issues, site trial portfolio<br />
management, volunteer investigator time, and funding in a<br />
time <strong>of</strong> stagnant and diminishing government resources<br />
(Table 6).<br />
The new NCI NCTN <strong>of</strong>fers the promise <strong>of</strong> a more efficient<br />
clinical trials system that will translate the rapidly expand-<br />
Author’s Disclosure <strong>of</strong> Potential Conflicts <strong>of</strong> Interest<br />
Author<br />
Stephen S. Grubbs*<br />
Table 1. IOM Committee Recommendations<br />
Goal I. Improve the speed and efficiency <strong>of</strong> the design, launch, and conduct <strong>of</strong><br />
clinical trials.<br />
§ Review and consolidate some front-<strong>of</strong>fice operations <strong>of</strong> the cooperative groups<br />
on the basis <strong>of</strong> peer review.<br />
§ Consolidate back-<strong>of</strong>fice operations and improve processes.<br />
§ Streamline and harmonize government oversight.<br />
§ Improve collaboration among stakeholders.<br />
Goal II. Incorporate innovative science and trial design into cancer clinical trials.<br />
§ Support and use biorepositories.<br />
§ Develop and evaluate novel trial designs.<br />
§ Develop standards for new technologies.<br />
Goal III. Improve the means <strong>of</strong> prioritization, selection, support, and completion <strong>of</strong><br />
cancer clinical trials.<br />
§ Reevaluate the National Cancer Institute’s role in the clinical trials system.<br />
§ Increase the accrual volume, diversity, and speed <strong>of</strong> clinical trials.<br />
§ Increase funding for the cooperative group program.<br />
Goal IV. Incentivize the participation <strong>of</strong> patients and physicians in clinical trials.<br />
§ Support clinical investigators.<br />
§ Cover the cost <strong>of</strong> patient care in clinical trials.<br />
Table 2. National Cancer Institute <strong>Clinical</strong> Trials Network<br />
Transformation Goals<br />
§ Prioritize molecular characterization resources and develop molecularly driven<br />
trial design.<br />
§ Improve prioritization <strong>of</strong> phase III portfolio across disease entities.<br />
§ Remove disincentives to study less common diseases.<br />
§ Create a shared IT structure.<br />
§ Harmonize procedures.<br />
§ Integrate imaging technology.<br />
§ Integrate national tissue banking resources.<br />
§ Grant open access to the network for clinical and transitional investigations.<br />
Table 3. Adult Cooperative Group Consolidation<br />
● <strong>American</strong> College <strong>of</strong> Surgeons <strong>Oncology</strong> Group, Cancer and Leukemia Group B,<br />
North Central Cancer Treatment Group (Alliance for <strong>Clinical</strong> Trials in <strong>Oncology</strong>)<br />
● Gynecologic <strong>Oncology</strong> Group, National Surgical Adjuvant Breast and Bowel<br />
Project, Radiation Therapy <strong>Oncology</strong> Group (Neuroinformatics Research Group<br />
[NRG] <strong>Oncology</strong>)<br />
● <strong>American</strong> College <strong>of</strong> Radiology’s Imaging Network, Eastern Cooperative<br />
<strong>Oncology</strong> Group<br />
● Southwest <strong>Oncology</strong> Group<br />
*No relevant relationships to disclose.<br />
Employment or<br />
Leadership<br />
Positions<br />
Consultant or<br />
Advisory Role<br />
ing scientific knowledge <strong>of</strong> cancer to practice and provide<br />
our patients and their families with innovative cancer management<br />
strategies in a more timely fashion. Surely our<br />
community clinical investigators will embrace NCTN and<br />
contribute to its success.<br />
Stock<br />
Ownership Honoraria<br />
Table 4. National <strong>Clinical</strong> Trials Network Group Review<br />
and Funding<br />
§ Modified U10 program<br />
§ Increased per case reimbursement for high accruing sites<br />
§ National Cancer Institute (NCI) external peer review <strong>of</strong> groups in the same<br />
review cycle<br />
§ New review criteria to include evaluation <strong>of</strong> collaboration within the network and<br />
other NCI-funded programs (e.g., Specialized Programs <strong>of</strong> Research Excellence),<br />
operational efficiency, and overall scientific quality.<br />
Table 5. Community <strong>Clinical</strong> Trial Site Desirables<br />
I. <strong>Clinical</strong> Trial Access<br />
§ Phase II and III treatment trials<br />
§ Cancer control and prevention trials<br />
§ Comparative effectiveness trials<br />
§ Innovative scientific questions<br />
§ Common diseases and expanded performance status eligibility<br />
II. Group Participation<br />
§ Group identity<br />
§ Interaction with academic colleagues<br />
§ Leadership opportunities in both scientific and operation activities<br />
§ Publication authorships<br />
§ Pr<strong>of</strong>essional recognition<br />
III. Operational and Regulatory Standardization<br />
§ Common protocol templates and data submission<br />
§ Consolidated audit system<br />
§ Regulatory burden easement<br />
IV. Patient Recruitment Enhancement<br />
§ Electronic health record cues<br />
§ <strong>Clinical</strong> trial marketing to the public<br />
§ Underserved community accrual<br />
V. Adequate Funding<br />
§ Reimbursement for actual cost<br />
Table 6. Community <strong>Clinical</strong> Trial Site Challenges<br />
I. Network Membership<br />
§ High volume sites versus broad membership inclusion<br />
§ Single versus multiple group membership<br />
§ Trial accrual credit assignment<br />
II. Site Trial Portfolio Management<br />
§ Access to all four groups’ trials<br />
§ Orphan or less common disease trial resource allocation<br />
II. Volunteer Investigator Time<br />
§ Engagement<br />
§ Promotion<br />
§ Value<br />
III. Reimbursement<br />
§ Tiered per case reimbursement<br />
§ Anticipated static or decreasing National Cancer Institute budget<br />
Research<br />
Funding<br />
Expert<br />
Testimony<br />
Other<br />
Remuneration<br />
147
1. Nass J, Moses H, Mendelsohn J (eds). A National Cancer <strong>Clinical</strong> Trials<br />
System for the 21 st Century: Reinvigorating the NCI Cooperative Group<br />
Program. Washington, DC: National Academies Press, 2010.<br />
2. Abrams J. Transforming the NCI Trials System. 14 th Meeting <strong>of</strong> the<br />
148<br />
REFERENCES<br />
STEPHEN S. GRUBBS<br />
<strong>Clinical</strong> Trials and Translational Research Advisory Committee. July 13, 2011.<br />
3. Mack A, Nass S. Implementing a National Cancer <strong>Clinical</strong> Trials System<br />
for the 21 st Century: Workshop Summary. Washington, DC: National Academies<br />
Press, 2011.
Successful Integration <strong>of</strong> Cooperative<br />
Groups: The Origin <strong>of</strong> the Children’s<br />
<strong>Oncology</strong> Group<br />
Overview: In March 2000, the four legacy pediatric cooperative<br />
groups <strong>of</strong>ficially merged to become the Children’s <strong>Oncology</strong><br />
Group (COG). This was accomplished by the ratification <strong>of</strong><br />
a new constitution by the respective executive committees<br />
and voting membership <strong>of</strong> the four legacy groups. The actual<br />
merger was preceded by a 12 to 18 month period <strong>of</strong> planning,<br />
negotiation, and transition, overseen by a Transition Committee<br />
<strong>of</strong> select executive leadership under the direction <strong>of</strong> the<br />
four current chairs <strong>of</strong> the existing pediatric groups. Despite<br />
the constant threat <strong>of</strong> budget reductions and questions related<br />
to the judicious use <strong>of</strong> National Cancer Institute (NCI)<br />
funds to support four pediatric groups when “children constitute<br />
only 3% <strong>of</strong> the US cancer problem,” the decision to unify<br />
was initiated and driven internally. The merger was envisioned<br />
as an opportunity to create efficiency by reducing duplicative<br />
systems and processes, which was becoming increasingly<br />
apparent as more planned clinical trials required intergroup<br />
collaboration. It was also recognized that such intergroup<br />
efforts would become more <strong>of</strong> a reality as clinical trial para-<br />
RECENT DECISIONS to restructure the NCI’s clinical<br />
trial enterprise in accordance with recommendations<br />
from the Institute <strong>of</strong> Medicine and external advisory groups<br />
have intended to reduce the number <strong>of</strong> cooperative groups<br />
through a process <strong>of</strong> integration. This integration <strong>of</strong> legacy<br />
groups and reorganization <strong>of</strong> the nation’s cancer clinical<br />
trials program has resulted in considerable angst for both<br />
well-established clinical investigators and those beginning<br />
their careers. All are now transitioning to a new reality <strong>of</strong><br />
team science and collaboration in clinical cancer research.<br />
The well-intentioned motivation to move a somewhat fragmented,<br />
competitive, duplicative system with parallel and<br />
under-resourced infrastructures to a system where scientific<br />
discovery can be rapidly translated to design, implementation,<br />
and conduct <strong>of</strong> high-impact studies and dissemination<br />
<strong>of</strong> results to improve care standards cannot be questioned.<br />
Nonetheless, reasonable concerns remain: the proposed<br />
model <strong>of</strong> centralized infrastructure support—including critical<br />
information systems, uncertainty about future budget<br />
allocations and funding instruments, and the lack <strong>of</strong> concrete<br />
metrics by which to judge the performance and success<br />
<strong>of</strong> this reorganization—continue during a period <strong>of</strong> intense<br />
transition. The anxiety <strong>of</strong> the integrated group leadership<br />
and their membership is understandable and familiar.<br />
Although the current changes are decidedly more far<br />
reaching in scope, this is not the first integration or merger<br />
within NCI’s Cooperative Group Program. In 2000, the<br />
Children’s Cancer Group, Pediatric <strong>Oncology</strong> Group (POG),<br />
National Wilms Tumor Study Group, and Intergroup Rhabdomyosarcoma<br />
Study Group merged to become COG. More<br />
than two decades before this, the pediatric divisions <strong>of</strong> the<br />
Cancer and Leukemia Group B and the Southwest <strong>Oncology</strong><br />
Group came together to form POG. 1,2 With the consolidation<br />
<strong>of</strong> the pediatric cooperative groups, well-recognized<br />
names and acronyms disappeared, but sense <strong>of</strong> mission and<br />
a strong foundation arising from legacy group accomplishments<br />
provided a framework for success. Naysayers may<br />
By Gregory H. Reaman, MD<br />
digms were built on risk-adjusted approaches. <strong>Clinical</strong>ly,<br />
biologically, and molecularly defined homogeneous subgroups<br />
<strong>of</strong> patients were <strong>of</strong> insufficient sample size within each group<br />
to design and conduct studies within a reasonable time frame.<br />
In essence, this merger was motivated by an overwhelming<br />
sense <strong>of</strong> necessity to preserve our mission <strong>of</strong> defining and<br />
delivering compassionate and state-<strong>of</strong>-the-art care through<br />
scientific discovery. The merger process itself was challenging,<br />
time consuming, not supported by any supplemental<br />
funding, and at times painful. What has emerged as a result<br />
is the largest pediatric cancer research organization in the<br />
world. Accomplishments in epidemiology, biology, translational<br />
science, and improved clinical outcomes for some<br />
pediatric cancers would have never been achieved without the<br />
merger. The very fact that outcome improvements were not<br />
realized in every type <strong>of</strong> pediatric cancer is testimony to the<br />
commitment <strong>of</strong> the COG membership to continue to look and<br />
move forward.<br />
comment that pediatric oncology is now different. In fact,<br />
pediatric oncology is different; the overall 5-year event-free<br />
survival rates for children with cancer exceed 80%. The<br />
clinical management <strong>of</strong> childhood cancer is integrally linked<br />
to clinical and translational research. Improvement in survival<br />
outcomes and cure rates are, in large part, a direct<br />
result <strong>of</strong> the fact that the overwhelming majority <strong>of</strong> children<br />
with cancer are enrolled on clinical trials. 3<br />
The decision to unify the pediatric groups was conceived<br />
and driven internally by a perceived affront to their missions<br />
and a very real threat to their scientific agenda. 2 This<br />
resulted from the increasing focus on risk-adjusted therapy<br />
approaches for childhood cancer and the increasing number<br />
<strong>of</strong> subgroups defined by clinical and biologic prognostic<br />
factors. Designing phase III clinical trials in homogenous<br />
patient groups within each pediatric cancer diagnosis that<br />
could be completed within reasonable timelines was severely<br />
hampered by sample size constraints initially mandating<br />
intergroup collaborations. Now such trials increasingly require<br />
consideration <strong>of</strong> international group collaborations.<br />
Also, any future plans to explore the feasibility <strong>of</strong> “personalized<br />
therapy” approaches with molecularly targeted<br />
agents in pediatric cancers would mandate a coordinated<br />
and consolidated effort. Infrastructure duplication and lack<br />
<strong>of</strong> process and procedure harmonization—even within the<br />
pediatric groups—were exaggerated by continual budget<br />
shortfalls. Despite the expectation that consolidation <strong>of</strong> the<br />
pediatric groups would result in improved efficiencies and<br />
From the Center for Drug Evaluation and Research, US Food and Drug Administration,<br />
Silver Spring, MD, and the Division <strong>of</strong> <strong>Oncology</strong>, Children’s National Medical Center,<br />
Washington, DC 20010.<br />
Author’s disclosure <strong>of</strong> potential conflicts <strong>of</strong> interest are found at the end <strong>of</strong> this article.<br />
Address reprint requests to Gregory H. Reaman, MD, Children’s National Medical<br />
Center, Washington, DC 20010; email: greaman@childrensnational.org.<br />
© <strong>2012</strong> by <strong>American</strong> <strong>Society</strong> <strong>of</strong> <strong>Clinical</strong> <strong>Oncology</strong>.<br />
1092-9118/10/1-10<br />
149
economies <strong>of</strong> scale, no predetermined metrics were developed<br />
and no resources were made available to collect and<br />
analyze data to evaluate success. In fact, annual budget<br />
reductions after the merger precluded any comprehensive,<br />
detailed self-assessment.<br />
There were considerable immediate challenges in the<br />
transition process <strong>of</strong> unifying culturally and organizationally<br />
distinct, procedurally and operationally disparate, competing<br />
clinical trial networks. COG leadership focused<br />
initially on individual, discipline-specific, and institutional<br />
membership requirements; financial planning and new models<br />
for funds distribution (given the planned discontinuation<br />
<strong>of</strong> institutional U-10 awards); consolidation <strong>of</strong> administrative<br />
functions; and coordination <strong>of</strong> data management and<br />
statistical operations. The last <strong>of</strong> these was particularly<br />
problematic since three separate sites served as statistics<br />
and data management centers using different and noncompatible,<br />
internally generated, electronic, Web-based remote<br />
data entry and registration systems. The importance <strong>of</strong><br />
providing adequate and centralized IT support and an<br />
enterprise-wide <strong>Clinical</strong> Data Management System cannot<br />
be overemphasized.<br />
It was a priority to maintain active legacy study conduct<br />
(80 phase III, 22 phase II, and 7 phase I) open studies using<br />
different data management platforms while planning for the<br />
importation <strong>of</strong> all legacy as well as new data to a single<br />
common platform. Of equal and—given resource constraints—competing<br />
priority was the development and implementation<br />
<strong>of</strong> new biologically based and hypothesisdriven<br />
clinical trials.<br />
Despite the lack <strong>of</strong> prespecified metrics, many successes<br />
were realized as a direct result <strong>of</strong> the consolidation. Three<br />
years after the merger, annual enrollment on therapeutic<br />
studies increased by more than 25% and enrollment on<br />
KEY POINTS<br />
● The merger <strong>of</strong> the legacy pediatric cooperative groups<br />
was self-motivated and self-directed.<br />
● No preestablished metrics <strong>of</strong> success were developed<br />
other than to maintain focus and achieve mission<br />
success.<br />
● Supplemental resources during a challenging transition<br />
period were not made available.<br />
● The lack <strong>of</strong> supplemental resources precluded the<br />
Children’s <strong>Oncology</strong> Group’s ability to perform a<br />
detailed assessment <strong>of</strong> evaluating efficiency, and progressive<br />
budget reductions made evaluation <strong>of</strong> any<br />
cost-effectiveness impossible.<br />
● Significant accomplishments in pediatric cancer were<br />
made possible only as a direct result <strong>of</strong> its planned<br />
consolidation and successful integration.<br />
Author’s Disclosure <strong>of</strong> Potential Conflicts <strong>of</strong> Interest<br />
Author<br />
Gregory H. Reaman*<br />
*No relevant relationships to disclose.<br />
150<br />
Employment or<br />
Leadership<br />
Positions<br />
Consultant or<br />
Advisory Role<br />
nontherapeutic (epidemiology and biology) studies more<br />
than doubled compared with combined enrollment statistics<br />
<strong>of</strong> the legacy groups. 4 COG became the largest childhood<br />
cancer research organization in the world and expanded its<br />
international collaborations. New clinical and biologically<br />
based classification systems were developed for both acute<br />
lymphoblastic and myeloid leukemia and neuroblastoma.<br />
5,6,7 Cytogenetic and molecular genetic predictors <strong>of</strong><br />
outcome were defined for these same diseases as well as for<br />
Wilms tumor and medulloblastoma. The prognostic significance<br />
<strong>of</strong> minimal residual disease in a number <strong>of</strong> tumor<br />
systems (acute leukemia, lymphoma, neuroblastoma) was<br />
established and incorporated into the expansion <strong>of</strong> riskadjusted<br />
treatment approaches to these diseases. 6,7 Expanding<br />
active programs in biomarker development for patient<br />
enrichment and risk classification was made possible by<br />
the retrospective analysis <strong>of</strong> independent data sets from<br />
legacy groups as test and validation cohorts. Early in its<br />
history, COG opened a frontline, randomized clinical trial<br />
for osteosarcoma, designed and conducted in collaboration<br />
with multiple European cooperative groups, which led to the<br />
development <strong>of</strong> the Cancer Therapy Evaluation Program<br />
guidelines for performance <strong>of</strong> international clinical trials.<br />
By consolidating legacy bio-specimen banks, COG’s Biopathology<br />
Center has become the national and international<br />
resource for clinically well-annotated specimens <strong>of</strong> virtually<br />
every type <strong>of</strong> pediatric cancer. Important genomic investigations<br />
in collaboration with the NCI’s Cancer Genome<br />
Atlas and the TARGET initiative have been made possible<br />
through this resource. These investigations have identified<br />
novel genetic lesions in acute leukemia and neuroblastoma,<br />
which are being evaluated as predictive biomarkers and<br />
for their potential therapeutic exploitation. 7-11 Since the<br />
consolidation, statistically significant improvements in<br />
5-year event-free survival rates have been achieved in all<br />
prognostic subgroups <strong>of</strong> acute lymphoblastic leukemia,<br />
acute myeloid leukemia, high-risk neuroblastoma, and medulloblastoma.<br />
7,8,12,13 COG has developed Long-Term<br />
Follow-Up Guidelines for childhood cancer survivors, which<br />
are regularly updated in accordance with increasing evidence<br />
and utilized worldwide. 7 Relevant portions <strong>of</strong> these<br />
exposure-dependent risk assessments and recommended<br />
surveillance parameters have been extrapolated for adult<br />
cancer survivors.<br />
Despite the initial challenges and anxiety experienced<br />
during the early days <strong>of</strong> the consolidation, and despite the<br />
criticism in response to the reduction in healthy competition,<br />
COG has emerged as more successful and impactful<br />
than the sum <strong>of</strong> its legacy groups. With thoughtful transition,<br />
focus on mission, assurance <strong>of</strong> adequate centralized<br />
resources and utilization <strong>of</strong> well-constructed, predetermined<br />
metrics, the current cooperative group reorganization<br />
should be viewed as a real opportunity to further advance<br />
clinical and translational research.<br />
Stock<br />
Ownership Honoraria<br />
Research<br />
Funding<br />
Expert<br />
Testimony<br />
GREGORY H. REAMAN<br />
Other<br />
Remuneration
COG: GIVING NEW MEANING TO COOPERATIVE<br />
1. Reaman GH. Pediatric oncology; current views and outcomes. Pediatr<br />
Clin Nort Am. 2002;49:1305-1318.<br />
2. Reaman GH. <strong>Clinical</strong> advances in pediatric hematology and oncology:<br />
Cooperative group research. Clin Adv Hematol Oncol. 2005;3:133-135.<br />
3. Anderson BD, Smith MA, Reaman GH, et al. Views <strong>of</strong> <strong>American</strong><br />
oncologists about the purposes <strong>of</strong> clinical trials. J Natl Cancer Inst. 2003;95:<br />
630-631.<br />
4. Steele JR, Wellemeyer AS, Hansen MJ, et al. Childhood cancer research<br />
network; A North <strong>American</strong> pediatric cancer registry. Cancer Epidemiol<br />
Biomarkers Prev. 2006;15:1241-1242.<br />
5. Schultz KR, Pullen DJ, Sather HN, et al. Risk and response based<br />
classification <strong>of</strong> childhood B precursor acute lymphoblastic leukemia: A<br />
combined analysis <strong>of</strong> prognostic markers from the Pediatric <strong>Oncology</strong> Group<br />
(POG) and the Children’s Cancer Group (CCG). Blood. 2007;109:926-935.<br />
6. Borowitz MJ, Devidas M, Hunger S, et al. <strong>Clinical</strong> significance <strong>of</strong><br />
minimal residual disease in childhood acute lymphoblastic leukemia and its<br />
relationship to other prognostic factors: A Children’s <strong>Oncology</strong> Group study.<br />
Blood. 2008;111:5477-5485.<br />
7. O’Leary M, Krailo M, Anderson JR, et al. Progress in childhood cancer:<br />
REFERENCES<br />
50 years <strong>of</strong> research collaboration: A report from the Children’s <strong>Oncology</strong><br />
Group. Semin Oncol. 2008;35:484-493.<br />
8. Hunger SP, Devidas M, Camitta B, et al. Improved survival for children<br />
with acute lymphoblastic leukemia (ALL) from 1990–2005: A report from the<br />
Children’s <strong>Oncology</strong> Group. J Clin Oncol. In press.<br />
9. Hunger SP, Raetz EA, Loh ML, et al. Improving outcomes for high-risk<br />
ALL: Translating new discoveries into clinical care. Pediatr Blood Cancer.<br />
2011;56:984-993.<br />
10. Mullighan CC, Su X, Zhang J, et al. Deletion <strong>of</strong> IZKF1 and prognosis in<br />
acute lymphoblastic leukemia. N Engl J Med. 2009;360:470-480.<br />
11. Mullighan CC, Zhang J, Harvey RC, et al. JAK mutations in high risk<br />
childhood acute lymphoblastic leukemia. Proc Natl Acad Sci USA.2009;106:<br />
914-918.<br />
12. O’Leary MC, Reaman GH. “Principles <strong>of</strong> pediatric oncology.” Hong WK,<br />
Bast RC, Hait W, et al (eds). Cancer Medicine 8 th Edition. Shelton, CT: BC<br />
Decker, 2010, pp. 1723-1741.<br />
13. Smith MA, Seibel NL, Altekruse SF, et al. Outcomes for children and<br />
adolescents with cancer: Challenges for the 21 st century. J Clin Oncol.<br />
2010;28:2625-2634.<br />
151
OVERCOMING DISPARITIES IN CLINICAL TRIAL<br />
ACCRUAL AMONG AFRICAN AMERICANS<br />
CHAIR<br />
Sandra M. Swain, MD<br />
Washington Hospital Center<br />
Washington, DC<br />
SPEAKERS<br />
Jean G. Ford, MD<br />
Johns Hopkins School <strong>of</strong> Medicine<br />
Baltimore, MD<br />
Deliya R. Banda, PhD<br />
Washington Hospital Center<br />
Washington, DC<br />
Worta McCaskill-Stevens, MD<br />
National Cancer Institute<br />
Bethesda, MD
A Critical Review <strong>of</strong> the Enrollment <strong>of</strong> Black<br />
Patients in Cancer <strong>Clinical</strong> Trials<br />
By Deliya R. Banda, PhD, Diane St. Germain, RN, MS, Worta McCaskill-Stevens, MD,<br />
Jean G. Ford, MD, and Sandra M. Swain, MD<br />
Overview: Although clinical trials represent a vital opportunity<br />
for improvements in cancer treatment, data show that a<br />
small proportion <strong>of</strong> patients with newly diagnosed cancer<br />
participate in clinical research. Black patients continue to<br />
have a worse prognosis for most cancers compared with other<br />
patients <strong>of</strong> other races/ethnicities. Racial/ethnic- and agerelated<br />
disparities in clinical trial accrual are also well documented.<br />
The recruitment and retention <strong>of</strong> minorities in these<br />
trials present an even greater challenge despite regulatory<br />
efforts and initiatives to increase representation. Treatment<br />
data from homogenous populations prevent us from under-<br />
ACCRUAL TO clinical trials remains a longstanding<br />
challenge, with 5% to 10% <strong>of</strong> patients with cancer from<br />
the general population participating and even lower rates <strong>of</strong><br />
black patients participating. 1 For the general population,<br />
the reasons are numerous, including lack <strong>of</strong> trial availability,<br />
lack <strong>of</strong> physician engagement and awareness <strong>of</strong> available<br />
trials, ineligibility, lack <strong>of</strong> insurance coverage, time<br />
commitment, and lack <strong>of</strong> knowledge and misperceptions<br />
regarding the conduct <strong>of</strong> clinical trials. 2-5 In addition to<br />
these factors, a multitude <strong>of</strong> additional subgroup-specific<br />
reasons have been reported, including a mistrust <strong>of</strong> the<br />
research and medical system, awareness <strong>of</strong> historical clinical<br />
research abuse, a lack <strong>of</strong> access to state-<strong>of</strong>-the-art care, a<br />
higher incidence <strong>of</strong> comorbidities (resulting in higher rates<br />
<strong>of</strong> ineligibility), religious and cultural beliefs that influence<br />
attitudes toward clinical trials, economic issues, and logistical<br />
concerns, such as child care and transportation. 6-11<br />
Cancer rates and incidence are disproportionately higher<br />
for some cancers in black patients compared with white<br />
patients. 12 Inclusion <strong>of</strong> black patients in clinical trials is<br />
therefore vital to gain an understanding <strong>of</strong> cancer biology<br />
and response to treatment in this population and to provide<br />
access to state-<strong>of</strong>-the-art care. Lack <strong>of</strong> access to state-<strong>of</strong>-theart<br />
care contributes to advanced disease at diagnosis, treatment<br />
morbidity, and decreased survival. 13,14 It may also<br />
increase the rate <strong>of</strong> comorbidities that deem many <strong>of</strong> this<br />
population ineligible for clinical trials. Among 235 black<br />
patients screened for protocol eligibility, only 8.5% were<br />
eligible. Patients were deemed ineligible because <strong>of</strong> comorbidities<br />
(with respiratory failure, HIV positivity, and anemia<br />
accounting for most), advanced disease stage, poor performance<br />
status, premature death, and short life expectancy. 9<br />
It is imperative that the rate <strong>of</strong> accrual <strong>of</strong> underserved<br />
populations increase to maximize the generalizability <strong>of</strong><br />
results from clinical trials. Several efforts have been made<br />
to increase accrual <strong>of</strong> underrepresented populations to<br />
clinical trials. The National Institutes <strong>of</strong> Health Revitalization<br />
Act <strong>of</strong> 1993 mandates the inclusion <strong>of</strong> women and<br />
minorities in federally sponsored cancer clinical trials. The<br />
aim is to conduct subset analyses to determine whether<br />
there are differences in effect based on race and sex. The<br />
National Cancer Institute (NCI) established the Minority-<br />
Based Community <strong>Clinical</strong> <strong>Oncology</strong> Program in 1990,<br />
standing therapeutic response and the true safety pr<strong>of</strong>ile <strong>of</strong><br />
novel therapies. Patient-, physician-, and system-level factors<br />
that affect trial participation have been extensively studied.<br />
However, years <strong>of</strong> accrual data remain largely unchanged,<br />
suggesting the challenge lies in effectively addressing these<br />
factors. Furthermore, data showing that black patients tend to<br />
have more advanced stage cancers at the time <strong>of</strong> diagnosis in<br />
fact beg their overrepresentation on clinical trials. An inability<br />
to successfully enroll diverse populations in clinical trials only<br />
exacerbates racial/ethnic differences in cancer treatment and<br />
survivorship.<br />
which supports institutions with 40% minority population in<br />
their catchment area to accrue racial minorities to treatment,<br />
cancer control, and prevention trials. This program<br />
has had successful accrual rates <strong>of</strong> 60% (Table 1). The<br />
National Medical Association (NMA) developed I.M.P.A.C.T.<br />
(Increase Minority Participation and Awareness <strong>of</strong> <strong>Clinical</strong><br />
Trials), an initiative to increase awareness, knowledge, and<br />
participation <strong>of</strong> black physicians and patients in biomedical<br />
research and clinical trials. There has been an increase in<br />
minority accrual to clinical trials sponsored by the NCI<br />
(Fig. 1). This may be attributable to enhanced efforts or an<br />
increase in minority populations.<br />
The University <strong>of</strong> California–Davis Cancer Center has<br />
recently embarked on a national effort to boost clinical trial<br />
recruitment <strong>of</strong> minorities as part <strong>of</strong> a major grant from the<br />
National Center on Minority Health and Health Disparities.<br />
15 This project, EMPaCT (Enabling Minority Participation<br />
in <strong>Clinical</strong> Trials), will engage five other centers and<br />
assess existing efforts to accrue minorities into trials. The<br />
goal is to develop consensus on evidence-based accrual<br />
models that can be adopted by cancer centers across the<br />
country.<br />
There has been a heightened understanding <strong>of</strong> the issues<br />
as evidenced by the plethora <strong>of</strong> published studies documenting<br />
the barriers to clinical trial enrollment in underserved<br />
populations. 3,4,13,16-18 However, methods to increase minority<br />
accrual elude clinicians largely because <strong>of</strong> the lack <strong>of</strong><br />
evidence-based strategies and practices in the literature.<br />
One <strong>of</strong> the most extensive literature reviews identified only<br />
14 articles that examined strategies to accrue underrepresented<br />
populations to cancer clinical trials. Notably, the<br />
efficacy or effectiveness <strong>of</strong> the strategies was evaluated in<br />
five <strong>of</strong> the 14 studies. 19 The first was a randomized two-arm<br />
study comparing the use <strong>of</strong> a media campaign involving<br />
newspapers and fliers with a clinic registry strategy where<br />
From the Washington Cancer Institute at Medstar Washington Hospital Center, Washington,<br />
DC; Medstar Health Research Institute, Hyattsville, MD; Division <strong>of</strong> Cancer<br />
Prevention, National Cancer Institute, Bethesda, MD; Johns Hopkins Center to Reduce<br />
Cancer Disparities, Johns Hopkins Bloomberg School <strong>of</strong> Public Health, Baltimore, MD.<br />
Authors’ disclosures <strong>of</strong> potential conflicts <strong>of</strong> interest are found at the end <strong>of</strong> this article.<br />
Address reprint requests to Deliya R. Banda, PhD, 110 Irving St. NW, Suite CG-111,<br />
Washington, DC 20010; email: Deliya.R.Banda@MedStar.net.<br />
© <strong>2012</strong> by <strong>American</strong> <strong>Society</strong> <strong>of</strong> <strong>Clinical</strong> <strong>Oncology</strong>.<br />
1092-9118/10/1-10<br />
153
Fiscal Year<br />
women were invited to participate in person. The media<br />
campaign proved more effective as evidenced by more<br />
women consenting to participate and fewer no-shows in this<br />
group compared with the clinic registry arm. 20 In the second<br />
study, a randomized study <strong>of</strong> older black men in a screening<br />
trial, three progressively intensive recruitment methods<br />
were compared. Intervention arms received different combinations<br />
<strong>of</strong> standard compared with enhanced mailings; baseline<br />
information gathered via telephone, via mail, or during<br />
a church session; reminders by telephone or mail; and<br />
consent forms mailed or administered during a church<br />
session. The most intensive <strong>of</strong> the intervention arms yielded<br />
the higher enrollment compared with the control and other<br />
intervention arms. 21 In the third study, enrollment in worksite<br />
cancer prevention programs was assessed comparing<br />
passive (telephone contact using telephone numbers provided<br />
by the company) with active recruitment strategies (a<br />
signup sheet). Enrollment and retention were lower among<br />
employees recruited by passive methods than by active<br />
methods. Passive methods, however, resulted in a representative<br />
employee sample. 22 In the fourth study, the use <strong>of</strong><br />
minority outreach recruiters, a specialized recruitment<br />
manual, and engagement <strong>of</strong> consultants for minority recruit-<br />
KEY POINTS<br />
Table 1. Minority-Based Community <strong>Clinical</strong> <strong>Oncology</strong> Program (MB-CCOP) Accrual, 2000–2009<br />
No. <strong>of</strong><br />
Funded MB-CCOPs<br />
No. <strong>of</strong><br />
Treatment Accruals<br />
● Accrual to cancer clinical trials remains low and<br />
largely unchanged despite initiatives and interventions<br />
to increase rates beyond the estimated 5% to<br />
10%.<br />
● In the face <strong>of</strong> poorer prognosis, even lower rates <strong>of</strong><br />
accrual <strong>of</strong> minorities in cancer trials is compounded<br />
by cultural-specific attitudes and sociodemographic<br />
factors that must be addressed in addition to the<br />
factors identified for other patients with cancer.<br />
● An intervention approach that directly addresses<br />
attitudes through use <strong>of</strong> a culturally appropriate<br />
video has been shown to affect the intention <strong>of</strong> black<br />
patients with cancer to enroll in therapeutic clinical<br />
trials.<br />
● Trial design factors, including eligibility criteria,<br />
data collection, and quality, are among the systemwide<br />
factors that also must be addressed to improve<br />
accrual.<br />
No. <strong>of</strong> Prevention<br />
and Control Accruals<br />
No. <strong>of</strong><br />
Overall Accruals<br />
No. <strong>of</strong><br />
Minority Patients<br />
ment failed to yield increases in recruitment <strong>of</strong> minority<br />
participants in a prostate cancer prevention trial. 23 In the<br />
final study, from the Southeast Cancer Control Consortium,<br />
no improvement was found in enrollment <strong>of</strong> patients with<br />
cancer into clinical trials after their intervention. This<br />
intervention involved use <strong>of</strong> nurse facilitators, quarterly<br />
newspapers, and health educators. 24<br />
The problem is clearly complex, with multiple contributing<br />
factors. Developing strategies that target factors that<br />
yield the greatest effect is thus a challenge. Among black<br />
patients, this is further complicated by the need to ensure<br />
such strategies are culturally sensitive. This review describes<br />
barriers to minority clinical trial accrual, a novel<br />
approach to change black patients’ attitudes toward clinical<br />
trial participation and strategies to enhance accrual.<br />
Factors That Influence Black Patients’ Decisions to<br />
Participate in Cancer <strong>Clinical</strong> Trials<br />
A comprehensive review identified 150 distinct barriers to<br />
accrual <strong>of</strong> underrepresented populations to cancer-related<br />
clinical trials. The most frequently reported <strong>of</strong> these are<br />
presented in Table 2. 13 Often, barriers are categorized<br />
according to their source (e.g., patient, physician, or institution<br />
or environment). A conceptual framework was developed<br />
that proposes that these barriers may influence accrual<br />
through their effects on awareness <strong>of</strong> trials, the opportunity<br />
to participate, and the decision to accept or refuse participation.<br />
The most frequently reported barrier related to<br />
awareness <strong>of</strong> trials was lack <strong>of</strong> clinical trials education.<br />
Education is crucial and can address other barriers, such as<br />
mistrust <strong>of</strong> research and the medical system, perceived<br />
harm, and fear. The most frequently reported barrier related<br />
to opportunity to participate was physician attitudes and<br />
patient eligibility. The most frequently reported barrier to<br />
acceptance <strong>of</strong> enrollment was mistrust <strong>of</strong> research and the<br />
medical system and costs <strong>of</strong> participation. 13<br />
Black Patients’ Attitudes Toward <strong>Clinical</strong><br />
Trial Participation<br />
Overall Minority<br />
Participation, %<br />
2000 8 425 358 783 427 55<br />
2001 10 642 541 1183 672 57<br />
2002 11 567 682 1249 949 76<br />
2003 11 521 930 1451 1249 86<br />
2004 13 673 467 1140 718 63<br />
2005 13 709 428 1137 569 50<br />
2006 13 684 393 1077 612 57<br />
2007 14 805 776 1581 962 61<br />
2008 13 895 733 1628 1051 65<br />
2009 14 851 461 1312 830 63<br />
Total (2000–2009) 14 851 461 1312 830 63<br />
Courtesy <strong>of</strong> Worta McCaskill-Stevens, MD, MS, program director, MB-CCOP, National Institutes <strong>of</strong> Health.<br />
154<br />
BANDA ET AL<br />
Many <strong>of</strong> the attitudes black patients have toward clinical<br />
trial participation stem from historical abuses, such as the<br />
U.S. Public Health Services Syphillis Study at Tuskegee,<br />
and related concerns <strong>of</strong> ethical misconduct and mistrust. 6,25-27<br />
In focus group interviews with 33 black patients and a black<br />
moderator, most participants viewed clinical research negatively<br />
and expressed concern they would be treated as a<br />
guinea pig. They also believed black patients would not
BLACK PATIENTS AND CANCER CLINICAL TRIALS<br />
Fig. 1. Minority enrollment in National Cancer Institute clinical<br />
trials, 2000–2010. Courtesy <strong>of</strong> Worta McCaskill-Stevens, MD,<br />
MS, program director, Minority Based Community <strong>Clinical</strong> <strong>Oncology</strong><br />
Program, National Institutes <strong>of</strong> Health.<br />
Abbreviations: CTEP, Cancer Therapy Evaluation Program; DCP,<br />
Division <strong>of</strong> Cancer Prevention; RRP, Radiation Research Program.<br />
benefit from clinical research because <strong>of</strong> racism and the<br />
inability to pay for medical care. 7<br />
Participants were asked specifically about the U.S. Public<br />
Health Services study and were found to have misinformation<br />
and misperceptions about the study. Many believed it<br />
was a conspiracy and, as a result, expected that investigators<br />
would be dishonest and not provide full disclosure <strong>of</strong> the<br />
risks related to research. They also believed in other conspiracies,<br />
such as the “creation” <strong>of</strong> HIV, military experiments<br />
with Agent Orange, and Central Intelligence Agency<br />
distribution <strong>of</strong> crack cocaine in black communities. 7<br />
In another set <strong>of</strong> focus groups with black women, the<br />
theme <strong>of</strong> trust also emerged. 28 Women were suspicious <strong>of</strong><br />
how clinical research was funded and expressed concerns<br />
regarding the termination <strong>of</strong> programs that specifically help<br />
the black community. 28 Focus group participants were presented<br />
with a hypothetical phase III randomized clinical<br />
trial in which adjuvant oral and intravenous chemotherapy<br />
were compared. Some participants did not understand the<br />
concept <strong>of</strong> randomization and lost interest in participating<br />
because <strong>of</strong> a lack <strong>of</strong> choice <strong>of</strong> treatment arm. Other participants,<br />
especially older women, were in favor <strong>of</strong> “natural”<br />
treatments and stressed the importance <strong>of</strong> prayer or God as<br />
healer. Most women were willing or at least open to considering<br />
a clinical trial using an oral chemotherapy agent.<br />
Physician Attitudes Toward Enrolling Black Patients<br />
in <strong>Clinical</strong> Trials<br />
Physicians play an important role in clinical trial accrual.<br />
Their perceptions and support <strong>of</strong> clinical research influence<br />
the discussions they have with patients and consequently<br />
the patients’ decision to participate in clinical trials. A<br />
survey <strong>of</strong> 166 black physicians revealed several challenges <strong>of</strong><br />
Table 2. Frequently Reported Patient Barriers to Minority<br />
Accrual to <strong>Clinical</strong> Trials 13<br />
Mistrust <strong>of</strong> research and medical system<br />
Perceived harm<br />
Cost<br />
Patient demographics<br />
Transportation<br />
Lack <strong>of</strong> education regarding clinical trials<br />
Time commitment<br />
Fear<br />
Family issues<br />
participating in clinical research, 29 including additional<br />
paperwork or telephone calls, masked drug assignment,<br />
excess patient care costs, losing the patient from their<br />
medical practice, and the potential effect <strong>of</strong> clinical research<br />
on the managed care status <strong>of</strong> their medical practice. Although<br />
most physicians surveyed (97%) believed clinical<br />
trials were important, only 63% referred patients and 64%<br />
encouraged patients to participate in clinical trials. This<br />
validated the focus group reports <strong>of</strong> mistrust <strong>of</strong> the medical<br />
system, lack <strong>of</strong> awareness <strong>of</strong> clinical trials, communication<br />
with study staff, and economic issues as factors that play a<br />
role in minority participation in clinical trials. 7,28,29<br />
The Eastern <strong>Oncology</strong> Cooperative Group (ECOG) conducted<br />
a pilot project with the NMA to determine whether<br />
an outreach intervention program would increase minority<br />
accrual. 30 The program consisted <strong>of</strong> five physician workshops<br />
focused on the identification <strong>of</strong> physician and patient<br />
barriers to minority accrual, potential solutions to address<br />
these barriers, and the willingness <strong>of</strong> the participating<br />
physicians to refer their patients to cancer clinical trials.<br />
The two most frequent physician barriers for the NMA<br />
physicians were lack <strong>of</strong> awareness and information about<br />
clinical trials and lack <strong>of</strong> trust <strong>of</strong> the medical center sponsoring<br />
the trial. Many had referred patients to a medical<br />
center but received little to no information about them<br />
thereafter; hence, this generated fear <strong>of</strong> losing their patients.<br />
Moreover, this lack <strong>of</strong> communication was viewed<br />
as a lack <strong>of</strong> respect for minority physicians. The ECOGaffiliated<br />
physicians cited these barriers less frequently.<br />
Regarding patient barriers, both physician groups cited<br />
patient or family suspicion and fear. All agreed to the need<br />
for patient education materials specifically designed for<br />
minority patients. ECOG subsequently developed a culturally<br />
sensitive patient educational brochure. In addition, a<br />
physician newsletter and laminated pocket cards were developed<br />
to provide clinical trial information, information<br />
regarding the importance <strong>of</strong> minority participation in cancer<br />
clinical trials, how to enroll patients in cancer clinical trials,<br />
and other information that would enhance communication<br />
between the NMA and ECOG physicians.<br />
Additional Strategies to Enhance Accrual <strong>of</strong><br />
Black Patients<br />
In addition to addressing patient and physician barriers,<br />
it is crucial to examine clinical trial design and patient<br />
155
eligibility, particularly as cancer clinical trials increase in<br />
complexity. Comorbidities are a substantial barrier to black<br />
patient participation in clinical trials. It remains unclear<br />
whether inferior treatment outcomes for minority patients<br />
are caused by the comorbidity or drug and/or dose alterations<br />
because <strong>of</strong> the comorbidity. Lee et al 33 provided a<br />
review <strong>of</strong> chemotherapy trials to determine the effect <strong>of</strong><br />
comorbidities on treatment outcomes and survival. Most<br />
studies included in the review were based on retrospective<br />
cancer registry and administrative data. Patients with comorbidities<br />
were less likely to receive chemotherapy regardless<br />
<strong>of</strong> cancer type and stage. Most trials reported decreased<br />
survival for patients who had comorbidities compared with<br />
those who did not. However, both chemotherapy use and<br />
survival was not evaluated in any <strong>of</strong> the studies; hence, a<br />
correlation between the two cannot be made. Major limitations<br />
<strong>of</strong> the review were the heterogeneity <strong>of</strong> the studies and<br />
the poor quality <strong>of</strong> the data.<br />
Collection <strong>of</strong> comorbidity data lacks uniformity and is<br />
<strong>of</strong>ten insufficiently detailed. 33 Furthermore, there is a lack<br />
<strong>of</strong> assessment tools, heavy reliance on performance status,<br />
and lack <strong>of</strong> management guidelines for specific comorbidities.<br />
There may be comorbidities that are modifiable before<br />
enrollment and patients for whom modifications <strong>of</strong> eligibility<br />
criteria is appropriate to increase enrollment. It may be that<br />
patients with certain comorbidities could be included in cancer<br />
clinical trials if the patient is closely monitored and with<br />
strict management guidelines in place. However, the appropriate<br />
collection <strong>of</strong> data is essential to support these efforts.<br />
In addition to comorbidity data, there is an overriding<br />
need to uniformly collect data that will encourage crossstudy<br />
analyses and be used to develop a databank for future<br />
queries. In a conversation with K. Flaherty, he indicated<br />
that recommendations include the collection and inclusion<br />
<strong>of</strong> race/ethnicity, age, educational level, income, zip code,<br />
marital status, smoking, obesity, spoken language, and<br />
insurance (October 2011). These data should be selfreported<br />
and collected in a standardized manner. However,<br />
there are several challenges, including who collects the data,<br />
the format used, commitment level from institutional administrators,<br />
resources and support services, staff training,<br />
implementation planning, and data quality assurance. 34<br />
Consideration must be given to the feasibility <strong>of</strong> enrolling<br />
minority and other underserved populations to a clinical<br />
trial during its design. Factors that must be considered are<br />
identifying unnecessarily restrictive eligibility criteria, identifying<br />
important end points that would address research<br />
questions geared toward a minority population, and determining<br />
additional data to collect that may inform future<br />
clinical care <strong>of</strong> minority populations. Additional considerations<br />
include participant burden, such as frequent and<br />
inconvenient clinic visits and assessments that prohibit<br />
enrollment by those patients with transportation issues,<br />
work, or family commitments. Finally, engaging minority<br />
physicians and community representatives during design<br />
development is also key to the viability <strong>of</strong> minority accrual.<br />
A Novel Approach: Addressing Culturally Specific<br />
Attitudes Toward <strong>Clinical</strong> Trials<br />
Lack <strong>of</strong> awareness or understanding <strong>of</strong> clinical trials is<br />
arguably one <strong>of</strong> the easier barriers to address. In a NCI’s<br />
156<br />
2007 Health Information National Trends Survey, black and<br />
Hispanic respondents placed the greatest trust in information<br />
from radio, television, newspapers, and religious organizations<br />
compared with white respondents, who placed<br />
greater trust in physicians. 26 Hence, the use <strong>of</strong> media that<br />
is culturally and ethnically sensitive to enhance awareness<br />
and promote clinical trials is a viable approach. In their<br />
review, Lai et al 19 reported on three studies that effectively<br />
used media campaigns and church-based project sessions.<br />
For black patients with cancer, the literature identifies a<br />
range <strong>of</strong> attitudinal barriers to trial participation. 3,6,25,31<br />
With such well-documented attitudinal barriers, it seems<br />
reasonable to assume the importance <strong>of</strong> integrating and<br />
addressing such sociocultural concerns in any effective approach<br />
to increasing accrual.<br />
A recent pilot study aimed to address the specific attitudes<br />
<strong>of</strong> black patients toward clinical trials. 32 The goal was to<br />
affect specific attitudes and increase the intention <strong>of</strong> patients<br />
to enroll in a therapeutic trial. The approach involved<br />
developing and testing a 15-minute, culturally targeted,<br />
narrative video to address six documented attitudinal barriers<br />
to trial participation by black patients. These six<br />
attitudes were the concern about the ethical misconduct <strong>of</strong><br />
investigators, poor treatment for being a minority or economically<br />
disadvantaged, fear and distrust <strong>of</strong> the medical<br />
establishment, worry about loss <strong>of</strong> autonomy, concern about<br />
privacy, and a general lack <strong>of</strong> knowledge and awareness<br />
about clinical trials. 3,6,25,31<br />
One hundred eight black patients with cancer in active<br />
treatment were enrolled in the pilot intervention at a large,<br />
urban cancer institute. Patients had never previously participated<br />
in any type <strong>of</strong> research study or clinical trial.<br />
Attitudes and self-reported intention to enroll in a clinical<br />
trial were assessed before and immediately after viewing the<br />
video. Results showed changes in all six attitudinal barriers<br />
from before the video to after the video. Patients’ likelihood<br />
<strong>of</strong> enrolling in a clinical trial increased after watching the<br />
video, with 34% <strong>of</strong> the patients showing positive changes in<br />
intention. 32<br />
The results <strong>of</strong> this study suggest that effectively addressing<br />
myths and strongly held attitudes toward trials, particularly<br />
among black patients, is a key patient-level approach<br />
to improving accrual. Despite the modest sample size, patients<br />
with more than 12 different primary tumor types were<br />
included, suggesting the ability to generalize the effectiveness<br />
<strong>of</strong> this video and its potential utility to increase willingness<br />
to participate across all cancer types and varied<br />
therapeutic trials. This intervention presents a practical<br />
approach for delivering clinical trial information in a culturally<br />
sensitive and effective format.<br />
Conclusion<br />
BANDA ET AL<br />
The extra time needed to gain individual and community<br />
acceptance <strong>of</strong> clinical trials among minorities is a known<br />
prerequisite to affect their participation in clinical trials. 34<br />
It is crucial to devote resources and continue research efforts<br />
to accrue minorities to clinical trials, particularly in the face<br />
<strong>of</strong> the changing demographics in the United States, a changing<br />
health care system, and a shifting economic climate.<br />
With an anticipated increase in cancer incidence for minor-
BLACK PATIENTS AND CANCER CLINICAL TRIALS<br />
ity populations far greater than the corresponding increase<br />
for white patients, 35 it is imperative that the inherent selection<br />
bias and objectivity <strong>of</strong> clinical trial design be addressed. 9<br />
Authors’ Disclosures <strong>of</strong> Potential Conflicts <strong>of</strong> Interest<br />
Employment or<br />
Leadership<br />
Positions<br />
Consultant or<br />
Advisory Role<br />
Author<br />
Deliya R. Banda*<br />
Diane St. Germain*<br />
Worta McCaskill-Stevens*<br />
Jean G. Ford*<br />
Sandra M. Swain BiPar Sciences<br />
(U); Eisai (U);<br />
Nektar (U);<br />
Novartis (U);<br />
Roche/Genentech<br />
(U); San<strong>of</strong>i (U)<br />
*No relevant relationships to disclose.<br />
1. Baseline Study <strong>of</strong> Patient Accrual onto Publicly Sponsored US Cancer<br />
<strong>Clinical</strong> Trials: An Analysis Conducted for the Global Access Project <strong>of</strong> the<br />
National Patient Advocate Foundation. Philadelphia, PA: Coalition <strong>of</strong> Cancer<br />
Cooperative Groups; 2006.<br />
2. Comis RL, Miller JD, Aldigé CR, Krebs L, Stoval E. Public attitudes<br />
toward participation in cancer clinical trials. J Clin Oncol. 2003;21:830-835.<br />
3. Advani AS, Atkeson B, Brown CL, et al. Barriers to the participation <strong>of</strong><br />
African-<strong>American</strong> patients with cancer in clinical trials: a pilot study. Cancer.<br />
2003;97:1499-1506.<br />
4. Lara PN Jr, Paterniti DA, Chiechi C, et al. Evaluation <strong>of</strong> factors affecting<br />
awareness <strong>of</strong> and willingness to participate in cancer clinical trials. J Clin<br />
Oncol. 2005;23:9282-9289.<br />
5. Simon MS, Du W, Flaherty L, et al. Factors associated with breast<br />
cancer clinical trials participation and enrollment at a large academic medical<br />
center. J Clin Oncol. 2004;22:2046-2052.<br />
6. Corbie-Smith G. The continuing legacy <strong>of</strong> the Tuskegee Syphilis Study:<br />
considerations for clinical investigation. Am J Med Sci. 1999;317:5-8.<br />
7. Corbie-Smith G, Thomas SB, Williams MV, Moody-Ayers S. Attitudes<br />
and beliefs <strong>of</strong> African <strong>American</strong>s toward participation in medical research.<br />
J Gen Intern Med. 1999;14:537-546.<br />
8. Swanson GM, Ward AJ. Recruiting minorities into clinical trials: toward<br />
a participant-friendly system. J Natl Cancer Inst. 1995;87:1747-1759.<br />
9. Adams-Campbell LL, Ahaghotu C, Gaskins M, et al. Enrollment <strong>of</strong><br />
African <strong>American</strong>s onto clinical treatment trials: study design barriers. J Clin<br />
Oncol. 2004;22:730-734.<br />
10. Shavers-Hornaday VL, Lynch CF, Burmeister LF, Torner JC. Why are<br />
African <strong>American</strong>s under-represented in medical research studies? Impediments<br />
to participation. Ethn Health. 1997;2:31-45.<br />
11. Baquet CR, Commiskey P, Daniel Mullins C, Mishra SI. Recruitment<br />
and participation in clinical trials: socio-demographic, rural/urban, and<br />
health care access predictors. Cancer Detect Prev. 2006;30(1):24-33.<br />
12. <strong>American</strong> Cancer <strong>Society</strong>. Cancer Facts & Figures for African <strong>American</strong>s<br />
2011-<strong>2012</strong>. Chicago, IL: <strong>American</strong> Cancer <strong>Society</strong>; 2011.<br />
13. Ford JG, Howerton MW, Lai GY, et al. Barriers to recruiting underrepresented<br />
populations to cancer clinical trials: a systematic review. Cancer.<br />
2008;112:228-242.<br />
14. Shavers VL, Brown ML. Racial and ethnic disparities in the receipt <strong>of</strong><br />
cancer treatment. J Natl Cancer Inst. 2002;94:334-357.<br />
15. Broadening Recruitment for Minorities, the Elderly. Cancer Discovery.<br />
2011;1:461.<br />
16. Branson RD, Davis K Jr, Butler KL. African <strong>American</strong>s’ participation<br />
in clinical research: importance, barriers, and solutions. Am J Surg. 2007;<br />
193:32-39.<br />
17. Townsley CA, Selby R, Siu LL. Systematic review <strong>of</strong> barriers to the<br />
recruitment <strong>of</strong> older patients with cancer onto clinical trials. J Clin Oncol.<br />
2005;23:3112-3124.<br />
18. Mouton CP, Harris S, Rovi S, Solorzano P, Johnson MS. Barriers to<br />
black women’s participation in cancer clinical trials. J Natl Med Assoc.<br />
1997;89:721-727.<br />
19. Lai GY, Gary TL, Tilburt J, et al. Effectiveness <strong>of</strong> strategies to recruit<br />
underrepresented populations into cancer clinical trials. Clin Trials. 2006;3:<br />
133-141.<br />
Acknowledgment<br />
Supported by National Center on Minority Health and Health<br />
Disparities: 1RC1MD004185-01.<br />
Stock<br />
Ownership Honoraria<br />
REFERENCES<br />
Research<br />
Funding<br />
Agendia; BiPar<br />
Sciences;<br />
Bristol-Myers<br />
Squibb; Novartis;<br />
Pfizer; Roche/<br />
Genentech;<br />
San<strong>of</strong>i<br />
Expert<br />
Testimony<br />
Other<br />
Remuneration<br />
San<strong>of</strong>i<br />
20. Brewster WR, Anton-Culver H, Ziogas A, et al. Recruitment strategies<br />
for cervical cancer prevention study. Gynecol Oncol. 2002;85:250-254.<br />
21. Ford ME, Havstad SL, Davis SD. A randomized trial <strong>of</strong> recruitment<br />
methods for older African <strong>American</strong> men in the Prostate, Lung, Colorectal<br />
and Ovarian (PLCO) Cancer Screening Trial. Clin Trials. 2004;1:343-351.<br />
22. Linnan LA, Emmons KM, Klar N, Fava J, LaForge R, Abrams D.<br />
Challenges to improving the impact <strong>of</strong> worksite cancer prevention programs:<br />
comparing reach, enrollment, and attrition using active versus passive<br />
recruitment strategies. Ann Behav Med. 2002;24:157-166.<br />
23. Moinpour CM, Atkinson JO, Thomas SM, et al. Minority recruitment in<br />
the prostate cancer prevention trial. Ann Epidemiol. 2000;10(8, suppl 1):S85-<br />
S91.<br />
24. Paskett ED, Cooper MR, DeGraffinreid CR, Spurr CL, Dolecek TA.<br />
Community <strong>Clinical</strong> <strong>Oncology</strong> Program as a recruitment vehicle for cancer<br />
control research. The Southeast Cancer Control Consortium experience. NC<br />
Med J. 1995;56:283-286.<br />
25. Scharff DP, Mathews KJ, Jackson P, H<strong>of</strong>fsuemmer J, Martin E,<br />
Edwards D. More than Tuskegee: understanding mistrust about research<br />
participation. J Health Care Poor Underserved. 2010;21:879-897.<br />
26. Katz RV, Kegeles SS, Kressin NR, et al. Awareness <strong>of</strong> the Tuskegee<br />
Syphilis Study and the US presidential apology and their influence on<br />
minority participation in biomedical research. Am J Public Health. 2008;98:<br />
1137-1142.<br />
27. Katz RV, Green BL, Kressin NR, et al. The legacy <strong>of</strong> the Tuskegee<br />
Syphilis Study: assessing its impact on willingness to participate in biomedical<br />
studies. J Health Care Poor Underserved. 2008;19:1168-1180.<br />
28. Linden HM, Reisch LM, Hart A Jr, et al. Attitudes toward participation<br />
in breast cancer randomized clinical trials in the African <strong>American</strong> community:<br />
a focus group study. Cancer Nursing. 2007;30:261-269.<br />
29. Shavers VL, Lynch CF, Burmeister LF. Factors that influence African-<br />
<strong>American</strong>s’ willingness to participate in medical research studies. Cancer.<br />
2001;91(1 suppl):233-236.<br />
30. McCaskill-Stevens W, Pinto H, Marcus AC, et al. Recruiting minority<br />
cancer patients into cancer clinical trials: a pilot project involving the Eastern<br />
Cooperative <strong>Oncology</strong> Group and the National Medical Association. J Clin<br />
Oncol. 1999;17:1029.<br />
31. Chandra A, Paul DP 3rd. African <strong>American</strong> participation in clinical<br />
trials: recruitment difficulties and potential remedies. Hosp Top. 2003;81:33-<br />
38.<br />
32. Banda D, Wang H, Libin AV, Swain SM. A clinical trial with culturally<br />
tailored video to increase participation <strong>of</strong> African <strong>American</strong>s in cancer clinical<br />
trials. Oncologist. <strong>2012</strong>;(In Press).<br />
33. Lee L, Cheung WY, Atkinson E, Krzyzanowska MK. Impact <strong>of</strong> comorbidity<br />
on chemotherapy use and outcomes in solid tumors: a systematic<br />
review. J Clin Oncol. 2011;29:106-117.<br />
34. Probstfield JL, Frye RL. Strategies for recruitment and retention <strong>of</strong><br />
participants in clinical trials. JAMA. 2011;306:1798-1799.<br />
35. Smith BD, Smith GL, Hurria A, Hortobagyi GN, Buchholz TA. Future<br />
<strong>of</strong> cancer incidence in the United States: burdens upon an aging, changing<br />
nation. J Clin Oncol. 2009;27:2758-2765.<br />
157
ANTIBODY CONJUGATES: TARGETED THERAPY<br />
MEETS CHEMOTHERAPY IN ONE DRUG<br />
CHAIR<br />
Howard A. Burris III, MD<br />
The Sarah Cannon Research Institute<br />
Nashville, TN<br />
SPEAKERS<br />
John P. Leonard, MD<br />
Weill Cornell Medical College<br />
New York, NY<br />
James D. Marks, MD, PhD<br />
University <strong>of</strong> California, San Francisco<br />
San Francisco, CA
Trastuzumab Emtansine (T-DM1): Hitching a<br />
Ride on a Therapeutic Antibody<br />
Overview: The treatment <strong>of</strong> cancers with chemotherapy is<br />
frequently limited by side effects. The effectiveness may be<br />
improved by the use <strong>of</strong> monoclonal antibodies to deliver<br />
cytotoxic agents to cancer cells while limiting exposure to<br />
normal tissues. The use <strong>of</strong> antibody-drug conjugates (ADCs) is<br />
one such strategy: a drug connected by a linker to an antibody<br />
specific for a tumor antigen is the basic makeup <strong>of</strong> an ADC.<br />
Overexpression and amplification <strong>of</strong> HER2 is associated with<br />
clinically aggressive breast cancers, and the use <strong>of</strong> trastuzumab<br />
to target HER2 has been highly effective. That said,<br />
most patients with HER2-positive metastatic breast cancer<br />
will eventually have disease progression during targeted therapy.<br />
Trastuzumab emtansine (T–DM1) is a novel ADC that<br />
combines the humanized antibody trastuzumab and the potent<br />
antimicrotubule agent T-DM1 (derivative <strong>of</strong> maytansine) using<br />
a unique and highly stable linker. The potential <strong>of</strong> maytansine<br />
was found in the 1970s with clinical responses noted against<br />
breast cancer; however, substantial toxicity prohibited further<br />
development. DM1 possesses in vitro cytotoxicity 10 to 200<br />
times greater than that <strong>of</strong> taxanes and vinca alkaloids. A phase<br />
I trial <strong>of</strong> T-DM1 for patients with heavily pretreated HER2positive<br />
breast cancer determined a recommended dose <strong>of</strong> 3.6<br />
THE TREATMENT <strong>of</strong> many cancers is <strong>of</strong>ten based on the<br />
use <strong>of</strong> standard chemotherapy agents and regimens.<br />
The effectiveness <strong>of</strong> this approach is frequently limited by<br />
substantial systemic toxicities associated with these agents.<br />
The therapeutic index can be markedly improved with the<br />
use <strong>of</strong> monoclonal antibodies to deliver potent cytotoxic<br />
agents to cancer cells, thus minimizing the exposure <strong>of</strong> the<br />
agents to normal tissues. The use <strong>of</strong> antibody-drug conjugates<br />
(ADCs) is one such strategy; a cytotoxic drug connected<br />
by a chemical linker to a monoclonal antibody specific<br />
for a tumor antigen is the basic makeup <strong>of</strong> an ADC. More<br />
stable linkers and different cytotoxic agents have enabled a<br />
new generation <strong>of</strong> ADCs to enter the clinic. First and<br />
foremost, the design <strong>of</strong> an ADC centers on the selection <strong>of</strong> an<br />
antigen that is tumor-specific and accessible to antibody<br />
binding at the tumor cell.<br />
Overexpression and amplification <strong>of</strong> HER2 is associated<br />
with clinically aggressive breast cancers that have historically<br />
had an overall poor prognosis and therapeutic resistance<br />
to traditional drugs. The use <strong>of</strong> trastuzumab to target<br />
the extracellular domain <strong>of</strong> HER2 has been highly effective<br />
in the treatment <strong>of</strong> this type <strong>of</strong> breast cancer. Multiple<br />
mechanisms for the efficacy <strong>of</strong> trastuzumab have been<br />
proposed, including inhibition <strong>of</strong> the PI3K signal transduction<br />
pathway, antibody-dependent cell-mediated cytotoxicity<br />
(ADCC), and induction <strong>of</strong> apoptosis. When combined with<br />
chemotherapy, trastuzumab improves the time to disease<br />
progression and overall survival for patients with HER2positive<br />
metastatic breast cancer. Furthermore, mature<br />
data from four large phase III trials in which trastuzumab<br />
was evaluated in the adjuvant setting, have demonstrated<br />
marked improvements in both disease-free and overall survival.<br />
That said, most patients with HER2-positive metastatic<br />
breast cancer will eventually have disease progression<br />
during targeted therapy while the cancer continues to both<br />
By Howard A. Burris III, MD<br />
mg per kilogram delivered every 3 weeks. Responses were<br />
seen in multiple patients. T-DM1 was then studied in phase II<br />
trials <strong>of</strong> patients with HER2-positive metastatic breast cancer.<br />
In a studies <strong>of</strong> 112 and 110 patients in whom disease had<br />
progressed during HER2-directed therapy, T-DM1 was associated<br />
with objective response rates <strong>of</strong> 26% and 34%, respectively.<br />
The agent was well tolerated in both trials, with most<br />
toxicities being grade 1 and 2, and no bleeding episodes or<br />
cardiac events occurring. Additional phase II and III trials are<br />
now evaluating T-DM1 in the first-line setting. In one such<br />
trial, T-DM1 was compared with standard dosing <strong>of</strong> trastuzumab<br />
every 3 weeks plus docetaxel every 3 weeks. Objective<br />
response rates were comparable and grade 3 or4 adverse<br />
events were substantially reduced in the T-DM1 arm. The<br />
anticipated selective activity and reduction in side effects<br />
were thus noted. Randomized multicenter phase III trials are<br />
ongoing, including the EMILIA trial, an open-label trial <strong>of</strong><br />
T-DM1 compared with the U.S. Food & Drug Administrationapproved<br />
regimen <strong>of</strong> capecitabine plus lapatinib. The results<br />
<strong>of</strong> studies completed to date suggest T-DM1 is active in<br />
patients who have cancer resistant to trastuzumab-based<br />
combinations.<br />
express HER2 and demonstrate sensitivity to antimicrotubule<br />
agents.<br />
T-DM1 is a novel ADC that combines the humanized<br />
antibody trastuzumab and the potent antimicrotubule agent<br />
DM1 (derivative <strong>of</strong> maytansine) using a unique and highly<br />
stable linker. 1 T-DM1, with its ability to bind HER2 with the<br />
same affinity as trastuzumab, maintains the activity <strong>of</strong><br />
trastuzumab in addition to providing intracellular delivery<br />
<strong>of</strong> the antimicrotubule agent DM1. It is hypothesized that<br />
when T-DM1 binds to HER2 receptors, a portion <strong>of</strong> them<br />
undergo receptor internalization, followed by lysosomal degradation.<br />
Activated DM1 is then released from lysosome into<br />
the cellular cytoplasm after antibody degradation, inhibiting<br />
microtubule assembly and causing cell death. Potent cytotoxic<br />
agents are needed to maximize the role <strong>of</strong> drug conjugates.<br />
In addition, the drug must be inactive and nontoxic in<br />
the conjugated form to avoid systemic toxicities. Few agents<br />
are able to fulfill these characteristics, including the inhibitors<br />
<strong>of</strong> tubulin polymerization (the maytansinoids and the<br />
auristatins).<br />
The potential <strong>of</strong> maytansine as an anticancer agent was<br />
originally discovered in the 1970s with clinical responses<br />
noted against breast cancer. However, substantial and random<br />
toxicities <strong>of</strong> neuropathy and myelosuppression were<br />
prohibitive <strong>of</strong> further clinical development. Recently, an<br />
attempt at improving the therapeutic index through conjugation<br />
with trastuzumab was undertaken, leading to the<br />
From the Sarah Cannon Research Institute, Nashville, TN.<br />
Author’s disclosures <strong>of</strong> potential conflicts <strong>of</strong> interest are found at the end <strong>of</strong> this article.<br />
Address reprint requests to Howard A. Burris, MD, Tennessee <strong>Oncology</strong>/Sarah Cannon<br />
Research Institute, 250 25 th Avenue N., Suite 110, Nashville, TN 37203; email:<br />
Howard.burris@scresearch.net<br />
© <strong>2012</strong> by <strong>American</strong> <strong>Society</strong> <strong>of</strong> <strong>Clinical</strong> <strong>Oncology</strong>.<br />
1092-9118/10/1-10<br />
159
development <strong>of</strong> DM1 (derivative <strong>of</strong> maytansine 1). DM1<br />
possesses in vitro cytotoxicity that is 10 to 200 times greater<br />
than that <strong>of</strong> other tubulin inhibitors, such as taxanes and<br />
vinca alkaloids. A suitable linker is critical to this process,<br />
and it must have a higher degree <strong>of</strong> stability in the circulation<br />
and allow efficient release <strong>of</strong> the potent cytotoxic agent<br />
once inside the tumor cell. The cytotoxic drug DM1 is<br />
conjugated to lysine residues <strong>of</strong> trastuzumab using a unique<br />
hetero-bifunctional reagent, N-succinimidyl 4-(Nmaleimidomethyl)<br />
cyclohexane-1-carboxylate (SMCC), in a<br />
two-step process. Trastuzumab is initially reacted with<br />
SMCC to form trastuzuamb-MCC. Next, T-MCC is then<br />
conjugated to DM1 to make T-DM1. The thioether linker<br />
was developed to provide a bond between trastuzumab and<br />
KEY POINTS<br />
Table 1. Summary <strong>of</strong> <strong>Clinical</strong> Efficacy Data from Trastuzumab Emtasine (T-DM1) <strong>Clinical</strong> Trials<br />
Trial and Reference Treatment Regimens<br />
TDM3569g 2 T-DM1 0.3 tp 4.8 mg/kg q3w for previously treated HER2 �<br />
MBC after previous chemotherapy and disease progression<br />
on trastuzumab<br />
TDM4258g3 T-DM1 3.6 mg/kg q3w for HER2-positive MBC after previous<br />
chemotherapy and disease progression on HER2-targeted<br />
therapy<br />
TDM4374g4 T-DM1 3.6 mg/kg q3w for HER2-positive MBC after previous<br />
exposure to an anthracycline, a taxane, capecitabine and 2<br />
HER2-directed therapies in the metastatic setting<br />
TDM4373g 5 T-DM1 3.6 mg/kg q3w � pertuzumab 840 mg loading dose<br />
then 420 mg q3w, for HER2 � MBC first-line treatment or<br />
after previous chemotherapy and HER2-directed therapy<br />
● The therapeutic index for treating patients can be<br />
improved by the use <strong>of</strong> monoclonal antibodies to<br />
deliver potent cytotoxic agents to cancer cells while<br />
minimizing exposure <strong>of</strong> the agents to normal tissues.<br />
● The use <strong>of</strong> antibody-drug conjugates (ADCs) is one<br />
such strategy; a cytotoxic drug connected by a chemical<br />
linker to a monoclonal antibody specific for a<br />
tumor antigen is the basic makeup <strong>of</strong> an ADC.<br />
● Trastuzumab emtansine (T–DM1) is a novel ADC<br />
that combines the humanized antibody trastuzumab<br />
and the potent antimicrotubule agent T-DM1 (derivative<br />
<strong>of</strong> maytansine) with use <strong>of</strong> a unique and highly<br />
stable linker.<br />
● In phase II studies <strong>of</strong> patients in whom metastatic<br />
breast cancer previously progressed during multiple<br />
HER2-directed therapies, T-DM1 was associated<br />
with response rates <strong>of</strong> 26% to 34% and was reasonably<br />
well tolerated.<br />
● The results <strong>of</strong> phase II studies suggest T-DM1 can<br />
improve outcomes for patients with cancers that are<br />
resistant to trastuzumab-based combinations, and<br />
phase III trials are underway.<br />
Number <strong>of</strong><br />
Patients<br />
ORR, Number <strong>of</strong><br />
Patients (%)<br />
24 6/24 (25.0)<br />
112 29 (26)<br />
110 38 (34.5)<br />
67 First line: 9/22 (41),<br />
Relapsed: 19/45 (42)<br />
TDM4450g6 T-DM1 3.6 mg/kg q3w<br />
Or 67 32 (47.8)<br />
Trastuzumab 8 mg/kg loading dose then 6 mg/kg q3w �<br />
docetaxel 75 mg/m2 or 100 mg/m2 q3w<br />
70 29 (41.4)<br />
Abbreviations: MBC, metastatic breast cancer; MTD, maximum tolerated dose; NR, not reported; ORR, objective response rate; q3w, every 3 weeks.<br />
160<br />
HOWARD A. BURRIS III<br />
DM1 that is more stable than hydrazone or disulfide linkers.<br />
The therapeutic index <strong>of</strong> DM1 is thus enhanced by minimizing<br />
systemic exposure to free DM1 and improving exposure<br />
to T-DM1. Before development in the clinic, the conjugate<br />
was extensively studied in preclinical models. The effectiveness<br />
<strong>of</strong> T-DM1 was established in three murine models <strong>of</strong><br />
HER2-expressing human breast carcinoma. In contrast,<br />
little activity was seen in the normal human cells or breast<br />
cancer cells not overexpressing HER2, demonstrating the<br />
specificity <strong>of</strong> the ADC. 1<br />
T-DM1 was the first HER2-targeted ADC with this unique<br />
SMCC linker to be studied in patients. A phase I trial in<br />
patients with HER2-positive breast cancer that had progressed<br />
during prior trastuzumab-based therapy determined<br />
a maximum tolerated dose <strong>of</strong> 3.6 mg per kilogram,<br />
delivered every 3 weeks. 2 The dose-limiting toxicity was<br />
grade 4 thrombocytopenia that was rapidly reversible and<br />
not associated with clinically meaningful bleeding events.<br />
No cardiac events or left ventricular ejection fraction declines<br />
were noted. In addition, no alopecia greater than<br />
grade 1 was noted, further evidence for the lack <strong>of</strong> systemic<br />
toxicity.<br />
Six <strong>of</strong> the 24 patients had an objective partial response.<br />
All patients had previously been treated with trastuzumab,<br />
with a median exposure <strong>of</strong> approximately 2 years, as well as<br />
microtubulin inhibiting agents. Of the six responses, four<br />
occurred in the nine patients treated at the maximum<br />
tolerated dose. The trial pharmacokinetics demonstrated<br />
peak free (unconjugated) DM1 plasma concentrations immediately<br />
after dosing which were low on all time points,<br />
suggesting that any systemic toxicity was unrelated to<br />
circulating unconjugated DM1. Weekly dosing was also<br />
explored and was both active and well tolerated, but it<br />
showed no particular advantage from either a dose intensity<br />
or density standpoint.<br />
After these results, T-DM1 was studied in phase II trials<br />
<strong>of</strong> patients with HER2-positive metastatic breast cancer at<br />
the recommended dose <strong>of</strong> 3.6 mg per kilogram every 3 weeks.<br />
In a study <strong>of</strong> 112 patients who had disease progression<br />
during HER2-directed therapy, T-DM1 was associated with<br />
an objective response rate <strong>of</strong> 26% based on independent<br />
review, and progression-free survival <strong>of</strong> 4.6 months. 3 The
T-DM1 AND THERAPEUTIC ANTIBODIES<br />
agent was well tolerated, with most toxicities being grade 1<br />
and 2, and no bleeding episodes or cardiac events.<br />
A second trial was conducted in 110 patients with HER2positive<br />
metastatic breast cancer who had received prior<br />
treatment with an anthracycline, a taxane, capecitabine,<br />
trastuzumab, and lapatinib (and had had disease progression<br />
during treatment with the most recent regimen). 4<br />
T-DM1 demonstrated an objective response rate <strong>of</strong> 34.5%<br />
and a median progression-free survival <strong>of</strong> 6.9 months,<br />
based on independent reviews. The agent was again well<br />
tolerated in this heavily pretreated population, and no<br />
cardiac toxicity signals were noted. During these phase II<br />
trials, central review for HER2-positivity was required.<br />
Response rates were higher amongpatients with centrally<br />
verified HER2-positive tumors, whereas few responses were<br />
noted among patients with tumors that tested negatively on<br />
central review, confirming the relationship with drug activity.<br />
Additional trials are now being done to evaluat T-DM1<br />
asfirst-line treatment. One such trial involved 137 patients<br />
who received either T-DM1 or standard dosing <strong>of</strong> trastuzumab<br />
every 3 weeks plus docetaxel 75 or 100 mg/m 2 every<br />
3 weeks. 5 Objective response rates were comparable, at 48%<br />
and 41%, respectively. Of note, grade 3 or 4 adverse events<br />
were substantially reduced in the T-DM1 arm (37% compared<br />
with 75%). The selective activity and proposed reduction<br />
in side effects were demonstrated in this randomized<br />
phase II trial.<br />
Two randomized multicenter phase III trials are ongoing<br />
to evaluate the role <strong>of</strong> T-DM1 in earlier lines <strong>of</strong> therapy.<br />
MARIANNE is designed to compare the efficacy and safety<br />
<strong>of</strong> single-agent T-DM1, alone or in combination with pertu-<br />
Author’s Disclosures <strong>of</strong> Potential Conflicts <strong>of</strong> Interest<br />
Author<br />
Howard A. Burris III*<br />
*No relevant relationships to disclose.<br />
Employment or<br />
Leadership<br />
Positions<br />
Consultant or<br />
Advisory Role<br />
1. Burris HA. Trastuzumab emtansine: A novel antibody-drug conjugate<br />
for HER2-positive breast cancer. Expert Opin Bio Ther. 2011;11:807-819.<br />
2. Krop IE, Beeram M, Modi S, et al. Phase I study <strong>of</strong> trastuzumab-<br />
DM1, a HER2 antibody-drug conjugate, given every 3 weeks to patients<br />
with HER2-positive metastatic breast cancer. J Clin Oncol. 2010;28:2698-<br />
2704.<br />
3. Burris HA 3rd, Rugo HS, Vukelja SJ, et al. Phase II study <strong>of</strong> the antibody<br />
drug conjugate trastuzumab-DM1 for the treatment <strong>of</strong> human epidermal<br />
growth factor receptor 2 (HER2)-positive breast cancer after prior HER2directed<br />
therapy. J Clin Oncol. 2011;29:398-405.<br />
4. Krop I, Lorusso P, Miller KD, et al. A phase II study <strong>of</strong> trastuzumab-<br />
DM1 (T-DM1), a novel HER2 antibody-drug conjugate, in patients with<br />
HER2-positive metastatic breast cancer who were previously treated with an<br />
zumab, with the standard trastuzumab plus a taxane (paclitaxel<br />
or docetaxel). The exposure to pertuzumab is blinded,<br />
and the standard arm is open-label. 6<br />
Another study, EMILIA, is an open-label trial in which<br />
T-DM1 is being compared with the U.S. Food and Drug<br />
Administration (FDA)-approved regimen <strong>of</strong> capecitabine<br />
plus lapatinib in patients previously treated with both a<br />
taxane and trastuzumab. Interestingly, in another trial,<br />
T-DM1 is being compared with physician’s choice <strong>of</strong> treatment<br />
in patients who had disease progression after multiple<br />
prior regimens. Trastuzumab alone or with chemotherapy is<br />
allowed in the physician’s-choice arm.<br />
Conclusion<br />
The discovery <strong>of</strong> HER2 gene amplification and the subsequent<br />
development <strong>of</strong> trastuzumab has markedly improved<br />
the prognosis for patients with HER2-positive breast cancer.<br />
Unfortunately, not all patients have a response to trastuzumab,<br />
and disease will progress in most patients with<br />
metastatic HER2-positive disease. T-DM1 meets the criteria<br />
for a successful ADC by combining the targeted effect <strong>of</strong><br />
trastuzumab with the cytotoxic potency <strong>of</strong> DM1 using a<br />
stable linker and minimizing systemic toxicity. In addition,<br />
other tumor histologies, such as gastrointestinal cancers<br />
that are HER2-positive may be sensitive to this agent. The<br />
results <strong>of</strong> phase II studies suggest T-DM1 can improve<br />
outcomes for patients with cancers that are resistant to<br />
trastuzumab-based combinations. An aggressive portfolio <strong>of</strong><br />
phase II and III clinical trials will help determine the role<br />
<strong>of</strong> T-DM1 in earlier lines <strong>of</strong> therapy or with combinations <strong>of</strong><br />
other targeted agents.<br />
Stock<br />
Ownership Honoraria<br />
REFERENCES<br />
Research<br />
Funding<br />
Expert<br />
Testimony<br />
Other<br />
Remuneration<br />
anthracycline, a taxane, capecitabine, lapatinib, and trastuzumab. Paper<br />
presented at: the European <strong>Society</strong> for Medical <strong>Oncology</strong>. October 10, 2010,<br />
Milan, Italy, abstract 277O.<br />
5. Perez EA, Dirix L, Kocsis J, et al. Efficacy and safety <strong>of</strong> trastuzumab-<br />
DM1 versus trastuzumab plus docetaxel in HER2-positive metastatic breast<br />
cancer patients with no prior chemotherapy for metastatic disease: Preliminary<br />
results <strong>of</strong> a randomized, multicenter open-label phase 2 (TDM4450g).<br />
Ann Oncol. 2010; 21 Suppl 8:LBA3.<br />
6. A study <strong>of</strong> trastuzumab-MCC-DM1 (T-DM1) in combination with pertuzumab<br />
administered to patients with HER2-positive locally advanced or<br />
metastatic breast cancer who have previously received trastuzumab. http://<br />
www.clinicaltrials.gov/ct2/show/NCT00875979?term�t-dm1�4373&rank�1.<br />
Accessed August 3, 2010.<br />
161
Targeting CD30 in Hodgkin Lymphoma:<br />
Antibody-Drug Conjugates Make a Difference<br />
By Catherine S. M. Diefenbach, MD, and John P. Leonard, MD<br />
Overview: CD30 expression is characteristic <strong>of</strong> the malignant<br />
Reed-Sternberg cell in Hodgkin lymphoma (HL) and several<br />
other lymphoid malignancies, such as anaplastic large-cell<br />
lymphoma (ALCL). Although unconjugated anti-CD30 antibodies<br />
have had minimal therapeutic activity in patients with HL<br />
as single agents, the CD30-directed antibody-drug conjugate<br />
(ADC) brentuximab vedotin has demonstrated activity that has<br />
resulted in its recent regulatory approval for the treatment <strong>of</strong><br />
patients with relapsed HL and ALCL. Approximately 75% <strong>of</strong><br />
patients with recurrent HL achieve objective responses, with<br />
HODGKIN LYMPHOMA (HL) is an uncommon B-cell<br />
lymphoid neoplasm, which accounts for approximately<br />
10% <strong>of</strong> all lymphomas and 0.6% <strong>of</strong> all cancers diagnosed in<br />
the developed world each year. In 2009, approximately 8,500<br />
cases <strong>of</strong> HL were diagnosed in the United States. 1 The<br />
median age at diagnosis is 38 years, and at least 40% <strong>of</strong><br />
patients are younger than age 35 at the time <strong>of</strong> diagnosis. 2<br />
Over the past 30 years, advances in the management <strong>of</strong><br />
HL have led to generally successful clinical outcomes, with<br />
roughly 80% <strong>of</strong> patients cured with chemotherapy or<br />
combined-modality therapy. Despite this high cure rate,<br />
approximately 5% to 10% <strong>of</strong> patients have primary refractory<br />
disease, and 20% to 30% <strong>of</strong> patients will experience<br />
relapse after attaining initial complete remission. Radiation<br />
therapy has an important role in treatment, although primarily<br />
in limited-stage and bulky disease. Second-line chemotherapy<br />
and autologous stem cell transplantation (ASCT)<br />
approaches are curative for only approximately 50% <strong>of</strong><br />
patients with relapsed or refractory disease. Patients whose<br />
disease recurs after second-line therapy or who have primary<br />
refractory disease have generally poor outcomes with<br />
conventional approaches. For patients whose disease recurs<br />
after ASCT, the median time-to-progression with subsequent<br />
therapy is reported to be 3.8 months, and median<br />
survival is 26 months. 3,4 Allogeneic stem cell transplantation<br />
(alloSCT) can induce durable remissions in some patients<br />
with relapsed and primary refractory HL; however,<br />
the use <strong>of</strong> this modality is limited in part by the challenges<br />
<strong>of</strong> achieving adequate disease control in this patient population<br />
before transplantation. Despite the successes associated<br />
with the modern management <strong>of</strong> HL, an estimated<br />
1,300 deaths from HL still occur each year in the United<br />
States, 5 many <strong>of</strong> which are young adults. Novel therapies for<br />
these patients are clearly needed.<br />
Regarding disease pathology, HL is a lymphoid neoplasm,<br />
likely derived from preapoptotic germinal-center B cells,<br />
characterized by the presence <strong>of</strong> large binucleated or multinucleated<br />
cells with prominent nucleoli termed Hodgkin and<br />
Reed-Sternberg (HRS) cells. Immunostains are characteristically<br />
positive for CD15 and CD30, and this pattern confirms<br />
diagnosis. The malignant HRS cells, typically<br />
comprise a small fraction <strong>of</strong> the total cellular population<br />
(0.1% to 10%), and reside in an environment <strong>of</strong> reactive<br />
inflammatory cells that produce soluble and membranebound<br />
factors that promote tumor cell growth, evasion <strong>of</strong><br />
self-immunity, and survival, 6,7 and typically comprise a<br />
162<br />
the principal toxicity being peripheral neuropathy. Ongoing<br />
studies are evaluating treatment with this agent as part <strong>of</strong><br />
first-line therapy, for patients with relapsed disease, and for<br />
patients with resistant disease and limited other options.<br />
Brentuximab vedotin demonstrates the therapeutic value <strong>of</strong><br />
antibody-drug conjugation and serves as a model <strong>of</strong> how a<br />
novel, targeted approach can be employed to potentially<br />
further improve outcomes in settings where curative chemotherapeutic<br />
regimens are already available.<br />
small fraction <strong>of</strong> the total cellular population (0.1% to 10%).<br />
HRS cells may constitutively express nuclear factor kappa B<br />
(NF-�B) and other antiapoptotic proteins that inhibit both<br />
the intrinsic and extrinsic pathways <strong>of</strong> apoptosis, such as<br />
c-FLICE and X-linked inhibitor <strong>of</strong> apoptosis protein. 8 A<br />
therapy targeted to the malignant cells in HL must therefore<br />
effectively disrupt a small but key component <strong>of</strong> the<br />
actual tumor mass, the malignant HRS cells.<br />
CD30: A Novel Therapeutic Target<br />
CD30 is a 120-KDa type I transmembrane glycoprotein<br />
belonging to the tumor necrosis factor (TNF) superfamily.<br />
The cytoplasmic tail <strong>of</strong> CD30 contains TNF receptor–associated<br />
(i.e., TRAF) binding sequences that can mediate the<br />
activation <strong>of</strong> NF-�B. Overexpression <strong>of</strong> CD30 may cause<br />
NF-�B activation, independent <strong>of</strong> its ligand CD30L. CD30<br />
may also have a role in maintaining CD4 memory, but the<br />
precise function <strong>of</strong> CD30 signaling in healthy individuals<br />
has not been well clarified. In normal lymphocytes, CD30 is<br />
expressed by only a small percentage <strong>of</strong> activated B cells and<br />
on naïve T and B cells where CD30 expression is enhanced<br />
by the presence <strong>of</strong> T helper 2 cytokines, costimulatory<br />
signals from interleukin-4, and CD28 during T-cell activation.<br />
CD30 signaling is associated with pleotropic downstream<br />
effects including cellular survival, differentiation,<br />
and lymphocyte activation. Although CD30 signaling may<br />
induce apoptosis in some cell types, HL cells are protected<br />
from CD30-mediated apopotosis by concomitant constitutive<br />
activation <strong>of</strong> NF-�B.<br />
A preferred antigen target for tumor-directed antibody<br />
therapy should both be differentially expressed and have<br />
strong immunogenicity (if immune mechanisms are the<br />
dominant mechanism <strong>of</strong> action). The high expression <strong>of</strong><br />
CD30 on the malignant HRS cells combined with its limited<br />
expression on normal activated T and B cells makes CD30<br />
an appealing target for directed therapy. In addition, CD30<br />
From the New York University Cancer Institute, NYU Langone Medical Center, New<br />
York, NY; Weill Cornell Cancer Center, Weill Cornell Medical College and New York<br />
Presbyterian Hospital, New York, NY.<br />
Authors’ disclosures <strong>of</strong> potential conflicts <strong>of</strong> interest are found at the end <strong>of</strong> this article.<br />
Address reprint requests to John P. Leonard MD, Center for Lymphoma and Myeloma,<br />
Division <strong>of</strong> Hematology and Medical <strong>Oncology</strong>, Weill Cornell Cancer Center, Weill Cornell<br />
Medical College and New York Presbyterian Hospital, 1305 York Avenue, New York, NY;<br />
email: jpleonar@med.cornell.edu.<br />
© <strong>2012</strong> by <strong>American</strong> <strong>Society</strong> <strong>of</strong> <strong>Clinical</strong> <strong>Oncology</strong>.<br />
1092-9118/10/1-10
TARGETING CD30 IN HODGKIN LYMPHOMA<br />
protein may be shed into a soluble form in the serum<br />
(sCD30). Elevated levels <strong>of</strong> sCD30 have been detected in the<br />
serum <strong>of</strong> patients with CD30-positive malignancies, and<br />
there is some suggestion that this might correlate with<br />
tumor burden, clinical outcome, or both. 9 CD30 expression is<br />
present in many conditions <strong>of</strong> dysregulated immunity, including<br />
cutaneous T-cell lymphoma, diffuse large B-cell<br />
lymphomas, follicular lymphoma, post-transplantation lymphoproliferative<br />
disorder, and virally infected T cells (e.g.,<br />
HIV). 10-12<br />
Unconjugated CD30-Directed Monoclonal Antibodies<br />
Several groups have evaluated unconjugated monoclonal<br />
antibodies against CD30 as therapeutic agent. Recent advances<br />
in antibody engineering have optimized two unconjugated<br />
antibodies: SGN-30 (cAC10), a chimerized<br />
immunoglobulin G1 (IgG1) monoclonal antibody, and MDX-<br />
060, a fully human monoclonal antibody that has shown<br />
enhanced antibody-dependent cell-mediated cytotoxicity<br />
(ADCC) compared with SGN-30. SGN-30 was evaluated in a<br />
phase II trial <strong>of</strong> patients with HL and systemic CD30positive<br />
anaplastic large-cell lymphoma (sALCL) (Table<br />
1). 13 The treatment was well tolerated, however objective<br />
responses were not observed in patients with HL. Yet, seven<br />
<strong>of</strong> the 41 patients with sALCL (17%) attained therapeutic<br />
KEY POINTS<br />
Table 1. Summary <strong>of</strong> <strong>Clinical</strong> Response Data to CD30-Targeted Antibody Therapy<br />
Drug Investigated<br />
n<br />
Treatment Dosage<br />
ORR (%)<br />
(reference) <strong>Clinical</strong> trial HL ALCL Total (mg/kg body weight)<br />
HL ALCL Overall<br />
SGN-30 (13) Phase II 38 41 79 6–12 weekly 0 17 (CR 5, PR 12) 9<br />
MDX-060 (14) Phase I/II 63 7 72* 0.1–15 weekly 6 (CR 3, PR 3) 29 (CR 29) 8*<br />
Brentuximabvedotin<br />
(19)<br />
Brentuximabvedotin<br />
(20)<br />
Brentuximabvedotin<br />
(21,22)<br />
Brentuximab vedotin<br />
(23)<br />
Initial phase I 42 2 45† 0.1–3.6 every 3 wk 36 (CR 21) 100 (CR 100) 38† (CR 24; ORR 50 for<br />
patients treated at the<br />
MTD)<br />
Phase I 44 5 38 0.4–1.4 weekly 59 (CR 34)<br />
Pivotal phase II, in<br />
patients with HL<br />
Phase II, in patients<br />
with sALCL<br />
● CD30 expression is characteristic <strong>of</strong> Reed-Sternberg<br />
cells in Hodgkin lymphoma.<br />
● Efforts to target CD30 with unconjugated or “naked”<br />
monoclonal antibodies have had limited therapeutic<br />
success.<br />
● Brentuximab vedotin is a CD30-directed antibody<br />
conjugated to a synthetic taxane (MMAE) with substantial<br />
therapeutic activity in recurrent Hodgkin<br />
lymphoma.<br />
● The toxicity pr<strong>of</strong>ile <strong>of</strong> brentuximab vedotin suggests<br />
that it may be combined successfully with several<br />
standard chemotherapeutic agents.<br />
● Ongoing efforts are directed toward integrating brentuximab<br />
vedotin into Hodgkin lymphoma therapeutic<br />
regimens to maximally improve outcomes.<br />
102 – 102 1.8 every 3 wk 75 (CR 34) – –<br />
– 58 58 1.8 every 3 wk – 86 (CR 53) –<br />
Abbreviations: ALCL: anaplastic large-cell lymphoma; CR: complete remission; HL: Hodgkin lymphoma; MTD: maximum-tolerated dose; ORR: overall response rate;<br />
PR: partial remission.<br />
* Including two patients with CD30 � T cell lymphoma. † Including one patient with CD30 � angioimmunoblastic T cell lymphoma. ‡ Including one patient with peripheral<br />
T cell lymphoma.<br />
responses: two (5%) attained complete response (CR) and<br />
five (12%) attained partial remission (PR), which provides<br />
some pro<strong>of</strong> <strong>of</strong> concept. A phase I/II trial <strong>of</strong> MDX-060 was<br />
performed in 72 patients with relapsed or refractory CD30positive<br />
lymphomas. 14 The overall response rate (ORR) was<br />
6% in the patients with HL and 29% in the patients with<br />
ALCL. With preclinical data suggesting synergy between<br />
cytotoxic chemotherapy and SGN-30, this agent was combined<br />
with GVD (gemcitabine, vinorelbine, and liposomal<br />
doxorubicin) chemotherapy by the Cancer and Leukemia<br />
Group B (CALGB) in a phase II trial for patients with<br />
relapsed HL. 15 This trial was prematurely closed because <strong>of</strong><br />
grades 3 to 5 pneumonitis that occurred in five <strong>of</strong> the 30<br />
patients enrolled. Interestingly, all five patients with pneumonitis<br />
were noted to have a valine/phenylalanine (V/F)<br />
polymorphism in the FcgammaRIIIa gene, suggesting a<br />
possible immune component. These pulmonary issues, including<br />
the potential for worsened risk <strong>of</strong> lung toxicity in<br />
combination with bleomycin (which can cause lung injury<br />
itself) are <strong>of</strong> importance as the development <strong>of</strong> CD30directed<br />
agents moves forward. Given the relatively low<br />
response rate seen with both SGN-30 and MDX-060 (i.e.,<br />
unconjugated anti-CD30 antibodies), their continued development<br />
has been limited, and the assessment <strong>of</strong> novel,<br />
conjugated agents has taken higher priority.<br />
Brentuximab-Vedotin: A Novel CD30-Targeted<br />
Antibody-Drug Conjugate<br />
Antibody-drug conjugates (ADC) can provide specific,<br />
preferential targeting <strong>of</strong> potent chemotherapeutic agents or<br />
toxins to differentially expressed antibody targets. Regarding<br />
binding to cell-surface targets, ADCs are generally<br />
internalized primarily via clathrin-mediated endocytosis<br />
and subsequent lysosomal proteolysis. This feature allows<br />
the therapeutic agent to be delivered into the target cell,<br />
ideally without affecting bystander and normal cells because<br />
they lack the antigen providing specificity. By allowing more<br />
<strong>of</strong> the drug or toxin to reach the target cell, therapeutic<br />
benefit should result, and because normal cells would be<br />
spared, toxicity should be minimized. In addition, if the<br />
antibody itself and the drug conjugated to it both have<br />
antitumor activity, multiple mechanisms <strong>of</strong> therapeutic action<br />
may be in effect.<br />
Brentuximab vedotin is a CD30-targeted ADC composed<br />
<strong>of</strong> the chimeric antibody cAC10 (SGN-30) modified by the<br />
addition <strong>of</strong> a valine-citrulline peptide linker to monomethyl<br />
163
auristatin E (MMAE), a synthetic analog <strong>of</strong> the naturally<br />
occurring antimitotic agent dolastatin 10. Brentuximab vedotin<br />
appears to have a multistep mechanism <strong>of</strong> action that<br />
is initiated by binding to CD30 on the cell surface, followed<br />
by clathrin-mediated endocytosis. MMAE is released in the<br />
lysosome from its conjugate through proteolytic degradation<br />
<strong>of</strong> the peptide linker. 16 The binding <strong>of</strong> released MMAE to<br />
tubulin disrupts the microtubule network, leading to G2/M<br />
phase cell-cycle arrest and apoptosis. 17 Overall efficacy <strong>of</strong><br />
the ADC may be enhanced by the fact that a small fraction<br />
<strong>of</strong> drug diffuses out <strong>of</strong> the targeted tumor cell and exerts<br />
cytotoxic effects on the surrounding tumor microenvironment.<br />
Preclinical in vitro and in vivo studies demonstrated that<br />
brentuximab vedotin was highly selective for its target<br />
CD30, was highly potent, and had low toxicity. 18 Brentuximab<br />
vedotin displayed a half maximal inhibitory concentration<br />
(IC 50) <strong>of</strong> less than 10 ng/mL against CD30-positive<br />
tumor cell lines, but was 300-fold less active against CD30negative<br />
cells. In severe combined immunodeficiency (SCID)<br />
mouse xenograft models <strong>of</strong> HL, brentuximab vedotin was<br />
efficacious at doses as low as 1 mg/kg. SCID mice treated<br />
with doses as high as 30 mg/kg showed no evidence <strong>of</strong><br />
toxicity. Enhanced in vitro antitumor activity against HL<br />
was reported for the combination <strong>of</strong> brentuximab vedotin<br />
with cytotoxic chemotherapy agents such as bleomycin,<br />
dacarbazine, vinblastine, doxorubicin, or gemcitabine. Collectively,<br />
these experiences suggested that brentuximab<br />
vedotin has the potential to be a highly effective agent<br />
against CD30-expressing malignancies.<br />
<strong>Clinical</strong> Data<br />
In the initial phase I trial <strong>of</strong> brentuximab vedotin in<br />
patients with relapsed or refractory CD30-positive malignancies,<br />
45 patients were enrolled with histologies including<br />
HL (42 patients), sALCL (two patients), and CD30-positive<br />
angioimmunoblastic T-cell lymphoma (one patient). 19 The<br />
primary objective <strong>of</strong> the study was to determine the maximum<br />
tolerated dose (MTD), and the secondary objectives<br />
were to evaluate overall response rate (ORR) and pharmacokinetics.<br />
Patients had a median age <strong>of</strong> 36 (range, 20 to 87),<br />
had Eastern Cooperative <strong>Oncology</strong> Group (ECOG) performance<br />
status <strong>of</strong> 0 or 1 (93%), and were heavily pretreated,<br />
with a median <strong>of</strong> three prior chemotherapy regimens (range,<br />
one to seven). Seventy-three percent had recurrent disease<br />
after prior ASCT. Brentuximab vedotin was administered as<br />
outpatient intravenous infusions at doses ranging from 0.1<br />
mg/kg to 3.6 mg/kg, every 21 days.<br />
Brentuximab vedotin was well tolerated at doses less than<br />
1.8 mg/kg, with the most common adverse events being<br />
fatigue (16 patients [36%]), pyrexia (15 patients [33%]), and<br />
diarrhea, nausea, neutropenia, or peripheral neuropathy (10<br />
patients [l22%] for each). No grade 3 or 4 adverse events<br />
occurred at the 1.2-mg/kg dose level (one level below the<br />
MTD). Grade 3 neutropenia, limb pain, and back pain each<br />
occurred in one patient <strong>of</strong> the 12 receiving the 1.8-mg/kg<br />
dose. At doses higher than 1.8 mg/kg, grade 3 neutropenia<br />
and pyrexia were seen in two <strong>of</strong> the 12 patients (17%)<br />
receiving the 2.7-mg/kg dose. One patient receiving the<br />
3.6-mg/kg dose developed febrile neutropenia and died <strong>of</strong><br />
infectious complications. Besides the low level <strong>of</strong> hematologic<br />
toxicity, the most important side effect was peripheral<br />
neuropathy, which occurred in 16 patients (36%), 13 <strong>of</strong><br />
164<br />
DIEFENBACH AND LEONARD<br />
whom were receiving the 1.8-mg/kg or 2.7-mg/kg doses. This<br />
observation was typically characterized by grade 1 or 2<br />
parasthesias in the hands or feet and prompted the discontinuation<br />
<strong>of</strong> treatment in three patients. Importantly, subsequent<br />
resolution <strong>of</strong> peripheral neuropathy was noted in 10<br />
<strong>of</strong> the 16 patients (63%) at the last safety assessment for the<br />
study.<br />
Of greatest note in this initial trial was that 17 patients<br />
demonstrated objective responses, including 11 CRs. Six <strong>of</strong><br />
the 12 patients (50%) treated at the MTD <strong>of</strong> 1.8 mg/kg had<br />
an objective response with a median response duration <strong>of</strong> 9.7<br />
months (ranging from 0.6 month to more than 19.5 months).<br />
Some reduction <strong>of</strong> tumor size was noted in 36 <strong>of</strong> the 42<br />
evaluable patients (86%), and 13 <strong>of</strong> the 16 patients (81%)<br />
with disease-related symptoms at baseline (e.g., pruritis and<br />
night sweats) had their symptoms resolve upon receiving<br />
therapy, irrespective <strong>of</strong> their antitumor response status. 19<br />
To explore a weekly dosing schedule (compared with the<br />
3-week schedule in the prior study), a second phase I trial<br />
was conducted in a similar patient population. 20 A total <strong>of</strong> 44<br />
patients—38 with HL, five with sALCL, and one with<br />
peripheral T-cell lymphoma—were enrolled. Patients had a<br />
median age <strong>of</strong> 33 (range, 12 to 82) with a median <strong>of</strong> three<br />
prior therapies (range, one to eight). Sixty-eight percent <strong>of</strong><br />
patients had received prior ASCT. Patients received weekly<br />
brentuximab vedotin for 3 <strong>of</strong> 4 weeks, with doses ranging 0.4<br />
mg/kg to 1.4 mg/kg. The most common treatment-related<br />
adverse events again included peripheral neuropathy, nausea,<br />
fatigue, neutropenia, diarrhea, and dizziness. The MTD<br />
for weekly dosing was 1.2 mg/kg. The ORR evaluated across<br />
all dose levels was 59%, with 34% achieving CR. The median<br />
duration <strong>of</strong> response had not been reached at the follow-up<br />
at 45 weeks. Data from the phase 1 trials demonstrated<br />
that across the dosing range <strong>of</strong> 1.2 to 2.7mg/kg ADC exposures<br />
were dose proportional, and steady state levels were<br />
achieved within 21 days on an every 3 week dosing schedule.<br />
On this schedule there was minimal to no accumulation <strong>of</strong><br />
ADC observed with multiple treatments.<br />
Given the high response rates in both <strong>of</strong> these phase I<br />
trials, a pivotal phase II trial was conducted in patients with<br />
relapsed or refractory HL, and outcomes were reported at<br />
the 2010 <strong>American</strong> <strong>Society</strong> <strong>of</strong> Hematology (ASH) annual<br />
meeting 21 with updates presented at the 2011 <strong>American</strong><br />
<strong>Society</strong> <strong>of</strong> <strong>Clinical</strong> <strong>Oncology</strong> (ASCO) annual meeting. 22 In<br />
the study, 102 patients with relapsed or refractory HL were<br />
treated every 3 weeks with brentuximab vedotin at a dose <strong>of</strong><br />
1.8 mg/kg. Demographics <strong>of</strong> the study population included a<br />
median age <strong>of</strong> 31 and a median <strong>of</strong> four prior therapies. All<br />
participants had undergone ASCT. The ORR was 75% with<br />
a CR rate <strong>of</strong> 34%, and reductions in tumor volume were<br />
reported in up to 95% <strong>of</strong> patients. The median duration <strong>of</strong><br />
response for patients achieving CR had not been reached at<br />
the time <strong>of</strong> the presentation.<br />
The data from a second phase II clinical trial <strong>of</strong> patients<br />
with sALCL were also reported at ASH 2011. 24 A total <strong>of</strong> 58<br />
patients with sALCL were treated with the same dose <strong>of</strong><br />
brentuximab vedotin (1.8 mg/kg) on the 3-week schedule.<br />
The spectrum <strong>of</strong> toxicity was similar to that observed in<br />
patients with HL. The ORR was 86% with a CR rate <strong>of</strong> 53%,<br />
and response duration also had not been reached at the time<br />
<strong>of</strong> the presentation. The study was updated at ASH 2011. 24<br />
At the time <strong>of</strong> the updated analysis, all but two patients had<br />
discontinued therapy, and the median number <strong>of</strong> treatment
TARGETING CD30 IN HODGKIN LYMPHOMA<br />
cycles with brentuximab vedotin was seven (range, one to<br />
16). The median duration <strong>of</strong> objective response was 13.0<br />
months (range 0.1 month to more than 19.1 months), the<br />
median PFS was 14.6 months, and the median OS had not<br />
yet been reached. Peripheral sensory neuropathy was observed<br />
in 41% <strong>of</strong> patients, which in 10 patients (17%) was<br />
grade 3. Among the patients with neuropathy <strong>of</strong> any grade,<br />
26 (79%) had achieved either resolution or improvement in<br />
symptoms with a median time-to-improvement <strong>of</strong> 13.3<br />
weeks (range 0.3 week to 48.7 weeks).<br />
These results led to the unanimous recommendation <strong>of</strong><br />
the <strong>Oncology</strong> Drug Advisory Committee <strong>of</strong> the United States<br />
Food and Drug Administration for accelerated approval <strong>of</strong><br />
brentuximab vedotin (Adcetris; Seattle Genetics, Inc, Bothell,<br />
WA) for the treatment <strong>of</strong> relapsed or refractory HL in<br />
patients with progressive disease after either ASCT or two<br />
chemotherapy regimens in patients ineligible for transplantation.<br />
Brentuximab vedotin was also approved for patients<br />
with sALCL who had at least one prior treatment fail.<br />
Brentuximab vedotin is the first new therapy approved for<br />
the treatment <strong>of</strong> HL in more than 30 years.<br />
Ongoing Trials<br />
Although brentuximab vedotin therapy demonstrated<br />
substantial single-agent activity in patients with relapsed or<br />
refractory HL, one would anticipate that even greater benefits<br />
could be achieved if it is integrated with or substituted<br />
for components <strong>of</strong> standard treatment. In addition, aspects<br />
<strong>of</strong> the optimal use <strong>of</strong> single-agent therapy need clarification.<br />
A variety <strong>of</strong> planned and ongoing studies are attempting to<br />
answer several key questions:<br />
Can patients who achieved CR with brentuximab vedotin<br />
who then stopped therapy in remission benefit from retreatment<br />
at a time <strong>of</strong> subsequent relapse? Small case series<br />
suggest that indeed patients may display sensitivity to<br />
retreatment with brentuximab vedotin at the time <strong>of</strong> relapse.<br />
25 Bartlett and colleagues reported a study <strong>of</strong> seven<br />
patients retreated after relapse (one patient twice), for a<br />
total <strong>of</strong> eight retreatment experiences. Responses included<br />
two patients with CR, four with PR, and two with stable<br />
disease. 26<br />
Can brentuximab vedotin be safely combined with and<br />
improve the results <strong>of</strong> first-line treatment regimens? An<br />
ongoing phase I trial <strong>of</strong> brentuximab vedotin in combination<br />
with either ABVD (doxorubicin, bleomycin, vinblastine,<br />
and dacarbazine) or AVD (doxorubicin, vinblastine, and<br />
dacarbazine) chemotherapy in untreated patients with<br />
HL is evaluating this issue (<strong>Clinical</strong>Trials.gov identifier:<br />
NCT01060904). Interim data for the first 31 patients treated<br />
were presented at ASH 2011. 27 Patient demographics included<br />
an international prognostic index greater than 4 in<br />
29% <strong>of</strong> patients, stage IV disease in 55%, and a median age<br />
<strong>of</strong> 35 (range, 19 to 59). Brentuximab vedotin at either 0.6<br />
mg/kg, 0.9 mg/kg, or 1.2 mg/kg was administered with<br />
standard doses <strong>of</strong> ABVD or at 1.2 mg/kg with AVD, on days<br />
1 and 15 <strong>of</strong> each 28-day cycle for six cycles. Overall, adverse<br />
events included peripheral neuropathy (48%), fatigue (45%),<br />
and neutropenia (77%). Grades 3 or 4 adverse events included<br />
neutropenia (74%), febrile neutropenia (16%), and<br />
anemia (13%). Importantly, in the ABVD cohort <strong>of</strong> 25<br />
patients, seven patients (28%) developed pulmonary toxicity,<br />
dyspnea, and interstitial lung disease, which led to the<br />
discontinuation <strong>of</strong> bleomycin. Of the 10 patients who were<br />
available for response assessment, all achieved CR. Based<br />
on these data, the ABVD/brentuximab vedotin arm has been<br />
closed to further enrollment and AVD/brentuximab vedotin<br />
is undergoing further evaluation.<br />
Does brentuximab vedotin have a role in targeting minimal<br />
residual disease as a maintenance strategy to prevent<br />
relapse in high-risk patients? This question is currently<br />
being investigated in patients who are at high risk after<br />
ASCT in a placebo-controlled, randomized trial <strong>of</strong> brentuximab<br />
vedotin maintenance, referred to as AETHERA (ADC<br />
Empowered Trial for Hodgkin to Evaluate Progression after<br />
ASCT; <strong>Clinical</strong>Trials.gov identifier: NCT01100502).<br />
Future Directions<br />
Determining the optimal timing for brentuximab vedotin<br />
therapy and whether brentuximab vedotin will ultimately<br />
be best used alone as a single agent or as a component <strong>of</strong> a<br />
novel treatment platform remain ongoing issues. Many <strong>of</strong><br />
the patients treated in both <strong>of</strong> the original phase I studies<br />
and the pivotal phase II study were heavily pretreated, <strong>of</strong>ten<br />
with relapse after autologous or allogeneic bone marrow<br />
transplantation. The response rate to brentuximab vedotin<br />
in treatment-naïve patients is not known, but because this is<br />
a patient population with chemotherapy-sensitive disease,<br />
one might expect that the activity <strong>of</strong> the drug would be<br />
greater as part <strong>of</strong> initial or second-line treatment. First-line<br />
ABVD continues to be generally well tolerated, based on 30<br />
years <strong>of</strong> safety data and an extremely high response rate.<br />
Thus an improvement in outcome for low-risk patients<br />
might not be easily discernible without very large studies.<br />
However, a meaningful relapse rate occurs, and morbidity<br />
and (albeit rare) mortality with this regimen is measurable.<br />
Improvements in both efficacy and tolerability may be<br />
achievable. For high-risk patients, the issue <strong>of</strong> whether<br />
including brentuximab vedotin would obviate the need for<br />
more intensive chemotherapy, such as augmented BEA-<br />
COPP, remains quite relevant.<br />
In addition, the role <strong>of</strong> brentuximab vedotin as part <strong>of</strong> a<br />
dose-intensive transplantation approach is important. Maximal<br />
cytoreduction before ASCT is associated with a more<br />
favorable outcome. The current standard salvage chemotherapy<br />
regimens, such as ifosfamide, carboplatin, and etoposide<br />
(i.e., ICE), have a low CR rate (26%) despite a high<br />
ORR. 28 Improving the CR rate might allow more patients<br />
with relapsed or refractory HL to proceed to and have<br />
successful results from ASCT. A retrospective analysis reported<br />
at ASH 2011 suggests that brentuximab vedotin<br />
therapy before reduced-intensity alloSCT for patients with<br />
HL may result in prolonged disease control without an<br />
increase in engraftment delay, post-transplant infectious<br />
complications, acute or chronic graft-versus-host disease, or<br />
nonrelapse-related transplant mortality. 29<br />
The retreatment data, albeit on a small number <strong>of</strong> patients,<br />
suggest that some patients with relapsed HL may<br />
maintain sensitivity to brentuximab vedotin. This issue<br />
warrants more detailed characterization in larger numbers<br />
<strong>of</strong> patients to better identify which patients might be most<br />
likely to benefit from retreatment. A phase II retreatment<br />
trial in patients who have previously achieved objective<br />
responses to brentuximab vedotin is currently ongoing<br />
(<strong>Clinical</strong>Trials.gov identifier: NCT00947856).<br />
The challenges for this highly active and novel therapy<br />
165
also include the assessment <strong>of</strong> potential long-term toxicities<br />
in patients, which might be more relevant in individuals<br />
receiving brentuximab vedotin as part <strong>of</strong> curative therapy<br />
at earlier time points. The incidence <strong>of</strong> bleomycin-induced<br />
pulmonary toxicity in patients receiving ABVD/brentuximab<br />
vedotin (28%) is unusually high for this population. The<br />
standard risk <strong>of</strong> bleomycin-induced pulmonary toxicity<br />
with six cycles <strong>of</strong> ABVD is approximately 2% in patients<br />
without other pulmonary risk factors. These data, combined<br />
with the earlier data from Blum and the CALGB indicating<br />
a specific V/F polymorphism in the FcgammaRIIIa gene<br />
underlying susceptibility, 15 suggest that more research on<br />
intrinsic risk factors for pulmonary toxicity should be undertaken<br />
and that the ideal combination platform for brentuximab<br />
vedotin must also avoid concurrent use <strong>of</strong> other<br />
Authors’ Disclosures <strong>of</strong> Potential Conflicts <strong>of</strong> Interest<br />
Employment or<br />
Leadership Consultant or Stock<br />
Author<br />
Positions Advisory Role Ownership Honoraria<br />
Catherine S. M.<br />
Diefenbach<br />
Seattle Genetics Seattle Genetics<br />
John P. Leonard Seattle Genetics<br />
1. <strong>American</strong> Cancer <strong>Society</strong>. Cancer Facts & Figures 2009. Atlanta: <strong>American</strong><br />
Cancer <strong>Society</strong>, 2009.<br />
2. SEER Cancer Statistics Review 2001-2008. National Cancer Institute.<br />
www.seer.cancer.gov.<br />
3. Bonfante V, Santoro A, Viviani S, et al. Outcome <strong>of</strong> patients with<br />
Hodgkin’s disease failing after primary MOPP-ABVD. J Clin Oncol. 1997;15:<br />
528-534.<br />
4. Longo DL, Duffey PL, Young RC, et al. Conventional-dose salvage<br />
combination chemotherapy in patients relapsing with Hodgkin’s disease after<br />
combination chemotherapy: the low probability for cure. J Clin Oncol.<br />
1992;10:210-218.<br />
5. Jemal A, Siegel R, Ward E, et al. Cancer statistics, 2009. CA Cancer<br />
J Clin. 2009;59:225-249.<br />
6. Aldinucci D, Gloghini A, Pinto A, et al. The classical Hodgkin’s lymphoma<br />
microenvironment and its role in promoting tumour growth and<br />
immune escape. J Pathol. 2010;221:248-263.<br />
7. Steidl C, Connors JM, Gascoyne RD. Molecular pathogenesis <strong>of</strong> Hodgkin’s<br />
lymphoma: increasing evidence <strong>of</strong> the importance <strong>of</strong> the microenvironment.<br />
J Clin Oncol. 2011;29:1812-1826.<br />
8. Farrell K, Jarrett RF. The molecular pathogenesis <strong>of</strong> Hodgkin lymphoma.<br />
Histopathology. 2011;58:15-25.<br />
9. Nadali G, Vinante F, Ambrosetti A, et al. Serum levels <strong>of</strong> soluble CD30<br />
are elevated in the majority <strong>of</strong> untreated patients with Hodgkin’s disease and<br />
correlate with clinical features and prognosis. J Clin Oncol. 1994;12:793-797.<br />
10. Younes A, Carbone A. CD30/CD30 ligand and CD40/CD40 ligand in<br />
malignant lymphoid disorders. Int J Biol Markers. 1999;14:135-143.<br />
11. Gardner LJ, Polski JM, Evans HL, et al. CD30 expression in follicular<br />
lymphoma. Arch Pathol Lab Med. 2001;125:1036-1041.<br />
12. Horie R, Watanabe T. CD30: expression and function in health and<br />
disease. Semin Immunol. 1998;10:457-470.<br />
13. Forero-Torres A, Leonard JP, Younes A, et al. A Phase II study <strong>of</strong><br />
SGN-30 (anti-CD30 mAb) in Hodgkin lymphoma or systemic anaplastic large<br />
cell lymphoma. Br J Haematol. 2009;146:171-179.<br />
14. Ansell SM, Horwitz SM, Engert A, et al. Phase I/II study <strong>of</strong> an<br />
anti-CD30 monoclonal antibody (MDX-060) in Hodgkin’s lymphoma and<br />
anaplastic large-cell lymphoma. J Clin Oncol. 2007;25:2764-2769.<br />
15. Blum KA, Jung SH, Johnson JL, et al. Serious pulmonary toxicity in<br />
patients with Hodgkin’s lymphoma with SGN-30, gemcitabine, vinorelbine,<br />
and liposomal doxorubicin is associated with an Fc�RIIIa-158 V/F polymorphism.<br />
Ann Oncol. 2010;21:2246-2254.<br />
16. Sutherland MS, Sanderson RJ, Gordon KA, et al. Lysosomal trafficking<br />
and cysteine protease metabolism confer target-specific cytotoxicity by<br />
peptide-linked anti-CD30-auristatin conjugates. J Biol Chem. 2006;281:<br />
10540-10547.<br />
166<br />
REFERENCES<br />
agents with overlapping toxicities. Given the relatively high<br />
incidence <strong>of</strong> peripheral sensory neuropathy seen in phase II<br />
trials, attention should also be paid to this side effect during<br />
long-term therapy, although the data thus far suggest that<br />
this symptom improves in most patients with the cessation<br />
<strong>of</strong> therapy.<br />
As we gain greater scientific understanding <strong>of</strong> the role <strong>of</strong><br />
CD30 signaling, the immunologic microenvironment, and<br />
mechanisms <strong>of</strong> tumor resistance in CD30-positive lymphomas,<br />
one may anticipate that even more rationally designed<br />
treatment regimens will emerge. Despite the relative success<br />
achieved for most patients with HL with standard,<br />
largely empirically designed regimens, brentuximab vedotin<br />
use provides an example <strong>of</strong> the direction <strong>of</strong> future collaborative<br />
and translational efforts.<br />
Research<br />
Funding<br />
DIEFENBACH AND LEONARD<br />
Expert<br />
Testimony<br />
Other<br />
Remuneration<br />
17. Francisco JA, Cerveny CG, Meyer DL, et al. cAC10-vcMMAE, an<br />
anti-CD30-monomethyl auristatin E conjugate with potent and selective<br />
antitumor activity. Blood. 2003;102:1458-1465.<br />
18. McEarchern JA, Kennedy D, McCormick R, et al. Activity <strong>of</strong> SGN-35 in<br />
preclinical models <strong>of</strong> combination therapy and relapse prevention. Haematol.<br />
2010;95 (suppl; abstr 49)<br />
19. Younes A, Bartlett NL, Leonard JP, et al. Brentuximab vedotin<br />
(SGN-35) for relapsed CD30-positive lymphomas. N Engl J Med. 2010;363:<br />
1812-1821.<br />
20. Fanale MA, Forero-Torres A, Rosenblatt JD, et al. A phase I weekly<br />
dosing study <strong>of</strong> brentuximab vedotin in patients with relapsed/refractory<br />
CD30-positive hematologic malignancies. Clin Cancer Res. <strong>2012</strong>;18:248-255.<br />
21. Chen R, Gopal AK, Smith SE, et al. Results <strong>of</strong> a pivotal Phase 2 study<br />
<strong>of</strong> brentuximab vedotin (SGN-35) in patients with relapsed or refractory<br />
Hodgkin lymphoma. Blood . 2010;116:21 (suppl; abstr 283).<br />
22. Chen RW, Gopal AK, Smith SE, et al. Results from a pivotal phase II<br />
study <strong>of</strong> brentuximab vedotin (SGN-35) in patients with relapsed or refractory<br />
Hodgkin lymphoma (HL). J Clin Oncol. 2011;29:15 (suppl; abstr 8031).<br />
23. Pro B, Advani R, Brice P, et al. Durable remissions with brentuximab<br />
vedotin (SGN-35): Updated results <strong>of</strong> a phase II study in patients with<br />
relapsed or refractory systemic anaplastic large cell lymphoma (sALCL).<br />
J Clin Oncol. 2011;29;15 (suppl; abstr 8032).<br />
24. Advani RH, Shustov AR, Brice P, et al. Brentuximab vedotin (SGN-35)<br />
in patients with relapsed or refractory systemic anaplastic large cell lymphoma:<br />
A Phase 2 study update. Blood. 2011;118;21 (suppl; abstr 443).<br />
25. Foyil KV, Kennedy DA, Grove LE, et al. Extended retreatment with<br />
brentuximab vedotin (SGN-35) maintains complete remission in patient with<br />
recurrent systemic anaplastic large-cell lymphoma. Leuk Lymphoma. <strong>2012</strong>;<br />
53:506-507<br />
26. Bartlett N, Grove LE, Kennedy DA, et al. Objective responses with<br />
brentuximab vedotin (SGN-35) retreatment in CD30-positive hematologic<br />
malignancies: A case series. J Clin Oncol. 2010;28:15 (suppl; abstr 8062).<br />
27. Younes A, Connors JM, Park SI, et al. Frontline therapy with brentuximab<br />
vedotin combined with ABVD or AVD in patients with newly diagnosed<br />
advanced stage Hodgkin lymphoma). Blood. 2011;118;21 (suppl; abstr 955).<br />
28. Moskowitz C. Risk-adapted therapy for relapsed and refractory lymphoma<br />
using ICE chemotherapy. Cancer Chemother Pharmacol. 2002;49:1<br />
(suppl; S9-12)<br />
29. Chen RW, Forman SJ, Palmer J, et al. Brentuximab vedotin (SGN-35)<br />
enables successful reduced intensity allogeneic hematopoietic cell transplantation<br />
in relapsed/refractory Hodgkin lymphoma. Blood. 2011;118 (supple;<br />
abstr 664).
EARLY DRUG DEVELOPMENT: CASTING A WIDE<br />
NET VERSUS PRESELECTING A NARROW AUDIENCE<br />
CHAIR<br />
Anthony W. Tolcher, MD<br />
South Texas Accelerated Research Therapeutics (START)<br />
San Antonio, TX<br />
SPEAKERS<br />
Josep Tabernero, MD<br />
Vall d’Hebron University Hospital<br />
Barcelona, Spain<br />
Bruce Chabner, MD<br />
Massachusetts General Hospital<br />
Boston, MA
Approval <strong>of</strong> New Agents after Phase II Trials<br />
Overview: Cancer drug approval has evolved as the understanding<br />
<strong>of</strong> cancer biology, and the ability to select patients<br />
for trials <strong>of</strong> targeted agents, has matured. The longstanding<br />
reliance on Phase III trials to prove drug efficacy and positive<br />
impact on patient survival may no longer be necessary, as<br />
early trials, particularly the expansion phase <strong>of</strong> a Phase I trial,<br />
may provide convincing evidence <strong>of</strong> a high response rate to a<br />
targeted drug in a patient population who has been poorly<br />
CANCER AND AIDS represent unique diseases and<br />
public health challenges. Both have the potential <strong>of</strong> a<br />
fatal outcome and affect large populations <strong>of</strong> patients. Cure<br />
or long-term control is difficult when patients present with<br />
advanced disease. Both engender a sense <strong>of</strong> urgency related<br />
to efforts to develop better therapies. In particular, the AIDS<br />
epidemic prompted a rethinking <strong>of</strong> traditional paths <strong>of</strong> drug<br />
testing and approval. For the past 50 years, since the<br />
Kefauver legislation in 1962, drug approval by the United<br />
States Food and Drug Administration (FDA) has required<br />
pro<strong>of</strong> <strong>of</strong> safety and efficacy through the vehicle <strong>of</strong> “wellcontrolled”<br />
clinical trials. 1 Phase III comparisons <strong>of</strong> new and<br />
standard treatments, with a survival end point, were the<br />
gold standard for approval <strong>of</strong> new agents. As a viral infectious<br />
disease that spread rapidly throughout the Western<br />
world in the early 1980s, AIDS brought forth a new paradigm:<br />
advocates campaigned for immediate access to promising<br />
agents. This advocacy led to an early implementation<br />
<strong>of</strong> compassionate release <strong>of</strong> azidothymidine (AZT) before its<br />
FDA approval in 1987 and the establishment in 1992 <strong>of</strong> a<br />
new category <strong>of</strong> marketing approval referred to as accelerated<br />
approval, based on “surrogate endpoints” <strong>of</strong> drug efficacy.<br />
2 No longer was it necessary for patients to wait for<br />
phase III trials to provide pro<strong>of</strong> <strong>of</strong> an extension <strong>of</strong> survival.<br />
Intermediate endpoints, such as viral load for AIDS patients<br />
and clinical response for patients with cancer, could become<br />
useful markers <strong>of</strong> benefit for randomized, or single-arm<br />
phase II studies.<br />
For the past 20 years, the accelerated approval category<br />
has allowed early access to cancer drugs that showed promise<br />
in addressing unmet needs for treatment <strong>of</strong> potentially<br />
fatal disease, and more than 35 new cancer drugs have<br />
received marketing permission by this new route. 3 Of the<br />
agents approved by this mechanism, full approval has subsequently<br />
been granted to most (the exact number is not<br />
available on the FDA website), although a small number <strong>of</strong><br />
agents have been withdraw for safety (gemtuzumab) or<br />
efficacy (gefitinib in lung cancer and bevicizumab in breast<br />
cancer) reasons. This mechanism has not been without<br />
controversy, as both the FDA and its critics have noted a<br />
failure to complete the required confirmatory trials postapproval<br />
in perhaps a third to one-half <strong>of</strong> accelerated approvals.<br />
3<br />
Recent developments in cancer drug discovery have only<br />
heightened interest in earlier approval. The rapid expansion<br />
<strong>of</strong> knowledge regarding the genetic basis <strong>of</strong> cancer has<br />
revealed specific molecular targets that underlie common<br />
malignancies. At the same time, major categories <strong>of</strong> disease,<br />
such as lung and colorectal cancer, once thought to be<br />
relatively homogeneous, are now recognized as encompass-<br />
By Bruce Chabner, MD<br />
responsive to conventional therapy. If the new drug produces<br />
no safety signals <strong>of</strong> great concern, and if a validated biomarker<br />
for patient selection has been established and is<br />
readily available, accelerated approval may be achievable<br />
prior to completion <strong>of</strong> a randomized trial. The advantages, and<br />
potential downside, <strong>of</strong> rapid approval scenarios will be discussed<br />
in this article.<br />
ing a number <strong>of</strong> unique molecular entities, and in selected<br />
cases, these entities respond to treatments that target the<br />
underlying mutations. Thus, mutated receptor tyrosine kinases<br />
and activated intracellular signaling pathways have<br />
become rational targets for experimental cancer therapies<br />
(Table 1), and their early application in carefully selected<br />
subsets <strong>of</strong> patients has met with impressive success. The<br />
widespread adoption <strong>of</strong> this “targeted” approach, called<br />
personalized or precision medicine, has had immediate consequences<br />
for cancer drug development and approval.<br />
Recent examples <strong>of</strong> successful targeted development include<br />
crizotinib for non–small cell carcinoma <strong>of</strong> the lung<br />
(NSCLC) with EML-4-ALK translocation 4 and vemurafenib<br />
for BRAF-mutated (V600E) melanoma. 5 In both cases, specific<br />
selection <strong>of</strong> patients with the appropriate molecular<br />
lesion allowed investigators to establish impressive response<br />
rates during the “expansion” phase <strong>of</strong> phase I trials,<br />
encompassing 30 to 80 patients treated at the maximum<br />
tolerated dose. In these initial patient populations, the<br />
majority <strong>of</strong> patients received “clinical benefit,” response<br />
rates exceeded 50 percent, and the clinical value <strong>of</strong> the new<br />
agent was clear even at this early stage <strong>of</strong> development, as<br />
alternative standard therapies were largely ineffective and<br />
seriously toxic. The results <strong>of</strong> the phase I studies, including<br />
the remarkable trial <strong>of</strong> GDC-0449 in patients with basal cell<br />
cancer, are shown in Table 1 and illustrate the power <strong>of</strong><br />
initiating trials in rationally selected patients. In both cases,<br />
an assay for a reliable biomarker for patient selection was<br />
developed and implemented during the earliest phases <strong>of</strong><br />
the first-in-human trials and was validated by the positive<br />
result. The biomarker assay, an essential requirement for<br />
rapid drug development and early approval, must be reliable,<br />
reproducible, and ultimately, available nationwide.<br />
The strategy for drug approval differed in the cases <strong>of</strong><br />
these two drugs, and these differences are instructive in a<br />
discussion <strong>of</strong> early approval strategies. In the case <strong>of</strong> crizotinib,<br />
the sponsor undertook two additional trials, even as<br />
the expanded phase I trial continued; the first was a phase<br />
II in previously treated patients with NSCLC and the<br />
EML-4/ALK translocation, 6 and the second trial was a<br />
randomized trial comparing the new medication to standard<br />
combination chemotherapy, with crossover at the time <strong>of</strong><br />
From the Massachusetts General Hospital Cancer Center, Boston, MA.<br />
Author’s disclosures <strong>of</strong> potential conflicts <strong>of</strong> interest are found at the end <strong>of</strong> this article.<br />
Address reprint requests to Bruce Chabner, MD, Massachusetts General Hospital Cancer<br />
Center, 10 North Grove Street, Boston, MA 02114; email: bchabner@partners.org.<br />
© <strong>2012</strong> by <strong>American</strong> <strong>Society</strong> <strong>of</strong> <strong>Clinical</strong> <strong>Oncology</strong>.<br />
1092-9118/10/1-10<br />
e1
Table 1. Response Rates in Expanded Cohorts <strong>of</strong> Recent Phase I Trials <strong>of</strong> Targeted Drugs in Selected Patient Populations<br />
Response<br />
Response<br />
Tumor Mutation Drug No. Patients Complete Partial Rate (%) Ref.<br />
Basal Cell Cancer smoothed in hedgehog pathway GDC-0449 33 2 16 54 15<br />
NSCLC EML4-ALK translocation Crizotinib 82 1 46 57 4<br />
Melanoma BRAF (V600E) Vemurafenib 32 2 24 81 5<br />
progression. The endpoint for the latter trial was response<br />
rate, time to progression (TTP), and survival (despite the<br />
crossover design).<br />
Crizotinib received accelerated approval in August <strong>of</strong><br />
2011, 3 years after the first patient with EML-4/ALK<br />
NSCLC entered the phase I trial, based on the results <strong>of</strong> the<br />
82 patient expansion cohort <strong>of</strong> the phase I trial and the<br />
confirmation <strong>of</strong> response rates and safety in the 173 phase II<br />
trial. The phase III trial <strong>of</strong> crizotinib versus chemotherapy<br />
has completed enrollment but has not yet been analyzed.<br />
For vemurafenib, the strategy for obtaining drug approval<br />
was somewhat different. The key registration trial was a<br />
phase III comparing the new agent to a standard, but<br />
ineffective drug, DTIC, which has a response rate <strong>of</strong> 10 percent.<br />
7 In fact, DTIC is the only approved chemotherapy<br />
agent for metastatic melanoma but does not prolong survival.<br />
The endpoint was overall survival, and crossover to<br />
the experimental drug at the time <strong>of</strong> progression was not<br />
allowed. Vemurafenib was approved in September <strong>of</strong> 2011,<br />
4 years after the first patient with BRAF V600E melanoma<br />
entered the phase I trial for that drug; full marketing<br />
approval for first-line treatment <strong>of</strong> metastatic melanoma<br />
was based on a survival advantage versus DTIC in the<br />
randomized phase III trial, which reached an early conclusion<br />
1 year after receiving its first patient. The rapid<br />
appreciation <strong>of</strong> success <strong>of</strong> the experimental drug in the<br />
phase III trial led to amendment to allow crossover <strong>of</strong><br />
patients on the DTIC arm midway through the trial. The<br />
decision to conduct a phase III trial <strong>of</strong> vemurafenib, without<br />
crossover, provoked substantial criticism in the lay press<br />
and editorial comment in leading medical journals, where<br />
the ethics <strong>of</strong> comparing a clearly active agent with a drug <strong>of</strong><br />
minimal activity (basically a placebo) were questioned. The<br />
decision to do so was attributed to the sponsor, which wished<br />
to obtain exclusivity for first-line use <strong>of</strong> its drug. Interestingly,<br />
crizotinib’s accelerated approval specified its use for<br />
both first-line or later treatment <strong>of</strong> patients with stage III<br />
(not the subject <strong>of</strong> a trial) or stage IV disease.<br />
What are we to conclude from these recent decisions and<br />
e2<br />
KEY POINTS<br />
● A patient selection strategy is essential to rapid drug<br />
approval.<br />
● Expansion phases <strong>of</strong> phase I trials allow early establishment<br />
<strong>of</strong> disease response rates.<br />
● Biomarkers for patient selection can be validated in<br />
expanded phase I.<br />
● One confirmatory phase II may be sufficient for<br />
accelerated approval.<br />
● Postapproval trials may refine the dose, schedule,<br />
sequence, and combination therapies.<br />
BRUCE CHABNER<br />
strategies? It seems apparent that with a valid biomarker<br />
test in hand and appropriate patient selection, an appropriately<br />
targeted therapy can provide strong indications <strong>of</strong> its<br />
ultimate value in an expanded cohort <strong>of</strong> subjects during a<br />
phase I trial. A confirmatory phase II trial, which need not<br />
be randomized if an active control is not available, can<br />
provide sufficient evidence to convince regulatory authorities<br />
to grant accelerated approval, and the process can be<br />
completed in three years or less. In many types <strong>of</strong> cancer, a<br />
strong indication <strong>of</strong> activity, such as a 50 percent or greater<br />
partial and complete response rate coupled with a time to<br />
progression <strong>of</strong> 4 to 6 months or greater, would represent a<br />
significant step forward and need not require preapproval<br />
confirmation in a phase III trial, particularly if no standard<br />
therapy provides benefit. 8 One can imagine many such disease<br />
entities, including grade glioblastomas, metastatic pancreatic<br />
cancer, metastatic cholangiocarcinoma, hepatoma, and various<br />
subsets <strong>of</strong> s<strong>of</strong>t tissue sarcoma, in which standard therapies<br />
are minimally active and significantly toxic. Given the<br />
modest toxicity <strong>of</strong> most targeted agents, if a new agent<br />
demonstrates impressive activity in its early trials, there is<br />
minimal safety risk in granting accelerated approval for these<br />
agents. Confirmatory trials postapproval would be required<br />
and might take the form <strong>of</strong> randomized comparison <strong>of</strong> two<br />
dose levels, intermittent compared with continuous therapy,<br />
randomized discontinuation in patients with stable disease,<br />
or randomization to continued therapy beyond progression<br />
in combination with an alternative therapy. Such trials<br />
would provide a better definition <strong>of</strong> safety, and survival<br />
benefit, although in most cases, it would not be ethical to<br />
deny the new agent to patients on any arm <strong>of</strong> the trial.<br />
As an example <strong>of</strong> the potential strategy for a new targeted<br />
agent, we might consider the development <strong>of</strong> a new drug for<br />
patients with IDH 1 or IDH 2 mutations (Fig. 1). IDH is a<br />
key enzyme in the intermediary metabolism <strong>of</strong> glucose in the<br />
Krebs cycle, in which isocitrate to alpha ketoglutarate<br />
(AKG). AKG is further utilized as a substrate for the Krebs<br />
cycle and for more than 60 dioxygenase reactions. 9,10 The<br />
dioxygenases require AKG as a substrate in reactions that<br />
hydroxylate methylated bases, such as methyl cytosine<br />
residues in DNA, and methyl groups in the histones that<br />
control gene expression; hydroxylation <strong>of</strong> these sites leads to<br />
demethylation and changes in gene expression. Dioxygenases<br />
have other important functions. They participate in<br />
the repair <strong>of</strong> DNA alkylation through removal <strong>of</strong> methyl or<br />
ethyl adducts. Recent work has shown that mutations in<br />
IDH 1 and 2 reduce AKG to a new enzymatic product,<br />
2 hydroxyglutarate (2HG), the isomers <strong>of</strong> which function<br />
either as potent inhibitors <strong>of</strong> dioxygenases or, in the case <strong>of</strong><br />
the R-isomer, activators <strong>of</strong> the same enzymes. 11 Inhibition<br />
<strong>of</strong> dioxygenase function leads to hypermethylation <strong>of</strong> DNA<br />
and <strong>of</strong> histone H3K27. 9,10 The IDH mutations, and their<br />
product, 2HG, have transforming activity in preclinical<br />
systems. Experimental and human tumors with underlying
APPROVAL OF NEW AGENTS<br />
IDH mutations show hypermethylation <strong>of</strong> both DNA and<br />
histone marks. Human tumors that display mutations in<br />
IDH 1 or 2 include a subset <strong>of</strong> human leukemias (AML),<br />
thyroid cancers, gliomas, chondrosarcomas, and intrahepatic<br />
cholangiocarcinomas. 12<br />
Two kinds <strong>of</strong> drugs might be effective in tumors driven by<br />
IDH 1 or 2 mutation. Analogs <strong>of</strong> AKG/2HG that inhibit<br />
mutant IDH enzymatic activity are an obvious choice, and<br />
are under development by Agios and other pharmaceutical<br />
companies. 13 A second strategy would be to develop inhibitors<br />
<strong>of</strong> H3K27 methylation, and such drugs are also nearing<br />
the clinic from Epizyme and others. 14 A third strategy would<br />
be to use inhibitors <strong>of</strong> DNA methylation, although it is not<br />
clear whether the primary transforming event is methylation<br />
<strong>of</strong> DNA or histones. A tightly focused phase I strategy<br />
aimed at relapsed or refractory IDH mutant AML, grade 4<br />
gliomas, chondrosarcomas, or cholangiocarcinoma might<br />
produce convincing early results with these agents. None <strong>of</strong><br />
these settings are curable with standard therapy (with the<br />
possible exception <strong>of</strong> high dose chemotherapy/stem cell<br />
transplant for relapsed AML). Such a trial will require the<br />
Author’s Disclosures <strong>of</strong> Potential Conflicts <strong>of</strong> Interest<br />
Author<br />
Bruce Chabner*<br />
*No relevant relationships to disclose.<br />
Employment or<br />
Leadership<br />
Positions<br />
Consultant or<br />
Advisory Role<br />
1. United States Food And Drug Administration. Significant Dates in U.S.<br />
Food and Drug Law History. http://www.fda.gov/AboutFDA/WhatWeDo/<br />
History/Milestones/ucm128305.htm. Last accessed 3/7/12. Accessed October<br />
14, 2010.<br />
2. Broder S. The development <strong>of</strong> antiretroviral therapy and its impact on<br />
the HIV-1/AIDS pandemic. Antiviral Res. 2010;85:1-18.<br />
3. Johnson JR, Ning YM, Farrell A, et al.Accelerated approval <strong>of</strong> oncology<br />
products: The food and drug administration experience. J Natl Cancer Inst.<br />
2011;103:636-44.<br />
4. Kwak EL, Bang YJ, Camidge DR, et al. Anaplastic Lymphoma Kinase<br />
Inhibition in Non-Small-Cell Lung Cancer. N Engl J Med. 2010;363:1693-<br />
1703.<br />
5. Flaherty KT, Puzanov I, Kim KB, et al. Inhibition <strong>of</strong> mutated, activated<br />
BRAF in metastic melanoma. N Engl J Med. 2010;26:809-19.<br />
6. FDA approves Xalkori with companion diagnostic for type <strong>of</strong> late-stage<br />
cancer. http://www.fda.gov/NewsEvents/Newsroom/PressAnnouncements/<br />
ucm269856.htm. Last Accessed March 7, <strong>2012</strong>.<br />
7. Chapman PB, Hauschild A, Robert C, , et al. Improved survival with<br />
vemurafenim in melanoma with BRAF V600E mutation. N Engl J Med.<br />
2011;364:2507-2516.<br />
participation <strong>of</strong> multiple institutions to identify the relatively<br />
uncommon subjects for these studies. Mutational<br />
analysis has been developed to identify candidates for accrual<br />
to these trials. Confirmation <strong>of</strong> target engagement will<br />
require evaluation <strong>of</strong> 2HG levels in tumor cells, serum or<br />
urine, or by imaging, and analysis <strong>of</strong> DNA and histone<br />
methylation before and after treatment. If response rates<br />
and TTP reach impressive levels in any <strong>of</strong> these subject<br />
tumor subsets, and provided that the safety pr<strong>of</strong>ile <strong>of</strong> the<br />
drug is acceptable, there would be an obvious path to<br />
accelerated approval.<br />
Despite the early success <strong>of</strong> targeted therapies in various<br />
settings, it is apparent that single agent treatment will not<br />
cure metastatic disease. Combination therapies aimed at<br />
mechanisms <strong>of</strong> resistance will be required to derive longterm<br />
benefit. Proving the value <strong>of</strong> combination trials will<br />
likely require randomized phase III studies to achieve marketing<br />
approval. The phase III trial is not antiquated or<br />
obsolete. It simply needs to be adapted to the new paradigm<br />
<strong>of</strong> rapid, single agent approval. Its continued place in the<br />
sun is assured.<br />
Stock<br />
Ownership Honoraria<br />
REFERENCES<br />
Research<br />
Funding<br />
Expert<br />
Testimony<br />
Other<br />
Remuneration<br />
8. Chabner BA. Early accelerated approval for highly targeted cancer<br />
drugs. N Engl J Med. 2011;364:1087-1089.<br />
9. Chao Lu, Ward PS, Kapoor GS, et al. IDH Mutation impairs histone<br />
demethylation and results in a block to cell differentiation. Nature. Epub <strong>2012</strong><br />
Feb 12.<br />
10. Turcan S, Rohle D, Goenka A, , et al. IDH1 mutation is sufficient to<br />
establish the glioma hypermethylator phenotype. Nature. Epub <strong>2012</strong> Feb. 15.<br />
11. Koivunen P, Lee S, Duncan CG, Lopez G, Lu G, et al. Transformation<br />
by the (R)-enantiomer <strong>of</strong> 2-hudroxyglutarate linked EGLN activation. Nature.<br />
Epub <strong>2012</strong> Mar 22.<br />
12. Borger DR, Tanabe KK, Fran KC, et al. Frequent mutation <strong>of</strong> isocitrate<br />
dehydrogenase (IDH)1 and IDH2 in cholangiocarcinoma identified through<br />
broad-based tumor genotyping. Oncologist. <strong>2012</strong>;17:72-79.<br />
13. Yen KE, Schenkein DP. Cancer-associated isocitrate dehydrogenase<br />
mutations. Oncologist. <strong>2012</strong>;17:72-79.<br />
14. Copeland RA, Solomon ME, Richon VM. Protein methyltransferases as<br />
a target class for drug discovery. Nat Rev Drug Discov. 2009;8:724-732.<br />
15. Von H<strong>of</strong>f DD, LoRusso PM, Rudin CM, et al. Inhibition <strong>of</strong> the hedgehog<br />
pathway in advanced basal-cell carcinoma. N Engl J Med. 2009;361:1164-<br />
1172.<br />
e3
Drug Development in the Era <strong>of</strong> Personalized<br />
<strong>Oncology</strong>: From Population-Based Trials to<br />
Enrichment and Prescreening Strategies<br />
By Rodrigo Dienstmann, MD, Jordi Rodon, MD, and Josep Tabernero, MD<br />
Overview: Recent advances in tumor biology and human<br />
genetics along with the development <strong>of</strong> drugs for specific<br />
targets hold promise for an era <strong>of</strong> personalized oncology<br />
treatment. Routine use <strong>of</strong> modern technologies, such as<br />
large-scale genome sequencing, will help to unravel the specific<br />
biology <strong>of</strong> each tumor. Adding a rigorous genomic view<br />
could determine key genetic events, critical dependencies,<br />
and stratification <strong>of</strong> patients in early clinical trials. Integrating<br />
biomarker development into the early testing <strong>of</strong> novel agents<br />
might provide clinically relevant therapeutic opportunities for<br />
DRUG DEVELOPMENT in oncology remains a slow,<br />
costly, and inefficient process. Taking advantage <strong>of</strong><br />
new science in the design and execution <strong>of</strong> clinical investigation,<br />
with a switch from histology-driven therapy to molecularly<br />
driven clinical oncology, is one way <strong>of</strong> addressing<br />
the high failure rate <strong>of</strong> developing an innovative anticancer<br />
agent. 1 Targeted therapeutics is optimal when applied in the<br />
appropriate molecular context: it indicates a drug that<br />
interferes with the function <strong>of</strong> a molecule readily identified<br />
in cancer cells but not healthy tissues. This target plays a<br />
critical role in the abnormal growth and/or invasiveness <strong>of</strong><br />
the tumor. A relative specificity <strong>of</strong> the drug exists, leading to<br />
improvement in the therapeutic index over standard cancer<br />
chemotherapy, higher efficacy, and decreased unnecessary<br />
toxicity. Accomplishments <strong>of</strong> modern technologies in genomics<br />
and biomarker detection are providing information that<br />
enables clinical researchers to investigate selective therapies<br />
for individual patients with cancer. Hundreds <strong>of</strong> new<br />
experimental drugs and biologic agents target the products<br />
<strong>of</strong> aberrant genes that are mutated or abnormally expressed<br />
in human cancers. 2 Many <strong>of</strong> these approaches are based on<br />
concepts such as oncogene addiction, non-oncogene addiction,<br />
and synthetic lethality, which target distinctive and<br />
complementary capabilities that enable tumor growth and<br />
metastatic dissemination. 3,4<br />
Given this perspective, phase I trials are providing an<br />
arena for early hypothesis testing. 5 Molecular biomarkerbased<br />
patient selection in early clinical trials has many<br />
possible advantages over an unselected population-based<br />
approach and include: (a) clinically qualify potential predictive<br />
biomarkers for molecularly targeted agents (MTAs)<br />
as early as possible; (b) generate valuable information regarding<br />
cancer biology; (c) accelerate patient benefit; and<br />
(d) affect the drug development process by assisting in the<br />
“go–versus-no-go” decisions and changing drug approval<br />
registration strategies <strong>of</strong> promising agents. 6,7 In this manuscript,<br />
we describe a new trend in early drug development<br />
with enrichment and prescreening strategies for personalized<br />
oncology.<br />
Success Stories<br />
The success <strong>of</strong> several targeted agents has shown that<br />
personalized cancer treatments can have an enormous effect.<br />
Development <strong>of</strong> trastuzumab in HER2-amplified breast<br />
cancers, with conversion <strong>of</strong> a biomarker for negative prog-<br />
168<br />
patients with advanced-stage cancer and also accelerate the<br />
drug-approval process. After recent success stories <strong>of</strong> therapies<br />
targeting driver molecular aberrations in genetically<br />
defined tumor subtypes, innovative clinical trials based on a<br />
strong biologic hypothesis are expected to bring further<br />
excitement to the field. In this article, we describe a new trend<br />
in biomarker-driven early drug development using enrichment<br />
and prescreening strategies. Technical and logistical obstacles<br />
that may hinder progress <strong>of</strong> this approach will be discussed,<br />
along with ethical and economic concerns.<br />
nosis into one that was predictive <strong>of</strong> benefit from the drug,<br />
was a major advance in the field. 8 More recently, patients<br />
with advanced gastric and gastroesophageal junction adenocarcinomas<br />
overexpressing HER2 have also been shown to<br />
achieve improved survival when trastuzumab is combined<br />
with chemotherapy. Another example is the PARP inhibitor<br />
olaparib, developed for patients with BRCA1/2-mutant cancers,<br />
with antitumor activity observed in different malignancies,<br />
including breast, ovarian, and prostate. 9 Further<br />
excitement for this approach came with the recent regulatory<br />
approvals <strong>of</strong> vemurafenib for advanced melanoma that<br />
harbors BRAF V600E mutation and crizotinib for non–small<br />
cell lung cancer (NSCLC) with EML4-anaplastic lymphoma<br />
kinase (ALK) translocation. 10,11 Interestingly, responses<br />
with selective BRAF inhibitors were seen in other tumor<br />
types carrying BRAF V600E mutations, such as thyroid<br />
cancer. The same is true for crizotinib in inflammatory<br />
my<strong>of</strong>ibroblastic tumor and anaplastic large-cell lymphomas<br />
harboring ALK translocation. The hedgehog inhibitor vismodegib<br />
has also been developed with enrichment for specific<br />
target patient populations in early clinical trials. Clear signs<br />
<strong>of</strong> activity were seen both in basal cell carcinomas and<br />
medulloblastomas known to have activating mutations in<br />
the pathway components Patched or Smoothened. 12 This<br />
highlights that cancers should likely be treated with targeted<br />
agents based on their specific molecular alterations<br />
rather than organ <strong>of</strong> origin.<br />
The impressive response rate in the selected population <strong>of</strong><br />
patients included in the particular trials listed in Table 1 8-16<br />
is in contrast with the approximate 5% response observed<br />
in early-phase trials conducted in unselected patient populations.<br />
17,18 Based on preclinical data, potential enrichment<br />
biomarkers already under investigation include PIK3CAactivating<br />
mutations and PTEN loss <strong>of</strong> expression in early<br />
trials with PI3K pathway inhibitors, with focus on breast<br />
and gynecologic tumors. <strong>Clinical</strong> development <strong>of</strong> chaperone<br />
(HSP90) inhibitors is now focused on ALK-rearranged<br />
From the Molecular Therapeutics Research Unit, Vall d’Hebron University Hospital,<br />
Barcelona, Spain.<br />
Authors’ disclosures <strong>of</strong> potential conflicts <strong>of</strong> interest are found at the end <strong>of</strong> this article.<br />
Address reprint requests to Josep Tabernero, P. Vall d’Hebron 119-129, Barcelona, Spain<br />
08035; email: jtabernero@vhebron.net.<br />
© <strong>2012</strong> by <strong>American</strong> <strong>Society</strong> <strong>of</strong> <strong>Clinical</strong> <strong>Oncology</strong>.<br />
1092-9118/10/1-10
DRUG DEVELOPMENT AND PERSONALIZED ONCOLOGY<br />
NSCLC and HER2-amplified breast tumors. Mesenchymal<br />
epithelial transition (MET) amplification, mutations, or receptor<br />
overexpression have been linked to responses with<br />
MET inhibitors in different tumor types, such as gastric<br />
adenocarcinomas, papillary renal cell carcinoma, and<br />
NSCLC. The same approach is being implemented in phase<br />
I trials evaluating fibroblast growth factor receptor (FGFR)<br />
inhibitors in FGFR amplified tumors, such as breast cancer<br />
and squamous NSCLC.<br />
Challenges<br />
Table 1. Response Rate <strong>of</strong> Successful Targeted Therapies in Selected Populations Evaluated in Early <strong>Clinical</strong> Trials<br />
Many investigators have raised concerns over the use <strong>of</strong><br />
predictive biomarkers in early clinical trials, including inac-<br />
KEY POINTS<br />
Marker/Population Agent Mechanism <strong>of</strong> Action Response Reference<br />
HER2-overexpressed/amplified breast cancer Trastuzumab Anti-HER2 antibody 12% 8<br />
Trastuzumab-DM1 Anti-HER2 antibody � drug conjugate 44% 13<br />
CD117-overexpressed GIST Imatinib c-KIT, PDGFR inhibitor 54% 14<br />
BRCA1/2 mutant breast, ovarian and prostate cancer Olaparib PARP inhibitor 47% 9<br />
BRAF V600E-mutant melanoma Vemurafenib BRAF inhibitor 75% 10<br />
Dabrafenib* BRAF inhibitor 60% 15<br />
ALK-rearranged NSCLC Crizotinib ALK, MET inhibitor 57% 11<br />
Basal cell carcinomas (majority have inactivating<br />
mutations in PTCH1 or activation <strong>of</strong> SMO)<br />
Vismodegib SMO inhibitor (Hh pathway) 58% 12<br />
Medullary thyroid cancer (known to have RET mutations,<br />
MET expression and VEGF activation)<br />
Cabozantinib MET, VEGFR2, RET inhibitor 29% 16<br />
* GSK2118436.<br />
Abbreviations: GIST, gastrointestinal stromal tumor; PDGFR, platelet-derived growth factor receptor; PARP, poly(ADP-ribose) polymerase; ALK, anaplastic lymphoma<br />
kinase; NSCLC, non–small cell lung cancer; MET, mesenchymal epithelial transition; SMO, smoothened; Hh, hedgehog; VEGFR, vascular endothelial growth factor<br />
receptor.<br />
● The “one-size-fits-all” approach for drug development<br />
<strong>of</strong> molecularly targeted agents in solid tumors has<br />
shown disappointing results.<br />
● With the switch from histology-driven to molecularly<br />
driven therapy, phase I trials in clinical oncology are<br />
providing an arena for clinical trial testing.<br />
● Genomics-driven early clinical trial enrollment has<br />
many possible advantages over an unselected<br />
population-based approach, including clinical qualification<br />
<strong>of</strong> predictive biomarkers, accelerated patient<br />
benefit, and a positive impact on the drug development<br />
process.<br />
● Some concerns over the use <strong>of</strong> biomarkers in early<br />
clinical trials have been raised, including the use <strong>of</strong><br />
assays that are not validated or certified, incorrect<br />
patient selection, increased cost, logistical issues related<br />
to the prescreening strategies, and ethical issues<br />
regarding mandatory tumor biopsies.<br />
● For the accelerated approach <strong>of</strong> drug development <strong>of</strong><br />
molecularly targeted agents to be possible, a cooperative<br />
environment is necessary, with experienced<br />
academic institutions, regulatory authorities, and<br />
pharmaceutical and diagnostic companies working<br />
together to promote data sharing, standardized procedures,<br />
technology exchange, and avoidance <strong>of</strong> duplicative<br />
efforts.<br />
curate assays, high cost, and ethical issues regarding tumor<br />
biopsies. Most importantly, the potential beneficial effects<br />
<strong>of</strong> the MTA in a more broadly defined population could be<br />
missed on the basis <strong>of</strong> incorrect patient selection. 6 Nevertheless,<br />
unselected patient evaluation has a higher risk <strong>of</strong><br />
failure in late-stage trials because <strong>of</strong> the tumor’s molecular<br />
heterogeneity, unless the prevalence <strong>of</strong> the predictive biomarker<br />
is already known to be high in the disease overall. 5<br />
It is clear that many technical, laboratory, and clinical<br />
challenges still remain to be resolved before the widespread<br />
implementation <strong>of</strong> molecular biomarker-based patient selection<br />
in early clinical trials. At first, investigators need to be<br />
aware <strong>of</strong> the possible reasons for failure <strong>of</strong> MTAs matched to<br />
specific biomarkers <strong>of</strong> vulnerability in advanced disease.<br />
The presence <strong>of</strong> the biomarker may not be representative <strong>of</strong><br />
the disease (tumor heterogeneity) or its biology (“driver” vs.<br />
“passenger” genetic alteration). Analyzing the tumor sample<br />
from disease diagnosis may not reflect the molecular alterations<br />
that drive the metastatic disease as a result <strong>of</strong> further<br />
molecular alterations (clonal evolution), selection pressure<br />
from previous treatments, and tumor heterogeneity. 19<br />
Driver mutations that promote growth and metastasis are<br />
considered the primary targets for therapeutic intervention.<br />
Nevertheless, ‘‘passenger’’ molecular dysregulations that<br />
arise through the rapid growth and genetic instability <strong>of</strong><br />
tumor cells can confer resistance to specific therapies.<br />
Secondly, the predictive test may not be accurate and the<br />
subpopulation most likely to benefit from MTAs may not be<br />
readily and reliably identified. Methodological problems,<br />
including a threshold (cut<strong>of</strong>f point) for defining the status<br />
<strong>of</strong> the potential biomarker, may still be missing. It is also<br />
important to note that the strategy <strong>of</strong> matching a predictive<br />
biomarker with MTAs will not always be applicable to all<br />
novel therapies, for example, broad-spectrum inhibitors targeting<br />
multiple signaling pathways and antiangiogenic<br />
agents. In addition, some MTAs may have substantial activity<br />
but no identifiable predictive marker, such as histone<br />
deacetylase (HDAC) and proteasome inhibitors. Therefore,<br />
a strong biologic hypothesis, solid preclinical data, and<br />
tumor models that mimic the clinical disease accurately are<br />
needed in order to assist in the drug development process. 5<br />
Thirdly, rapid acquisition <strong>of</strong> resistance to single agents<br />
is a problem. Complex interacting networks with adaptive<br />
negative feedback and compensatory receptor tyrosinekinase<br />
stimulation are imperative in cancer cells. In addition<br />
to cross-pathways escape, there is always the possibility<br />
169
that the core pathway is not completely blocked with the<br />
selected MTA. Furthermore, in tumors that have similar<br />
molecular aberrations, variability in patient outcomes has<br />
been observed, such as selective BRAF inhibitors in BRAFmutant<br />
melanoma and colorectal cancer. Single genomic<br />
markers are <strong>of</strong>ten found to be unsuitable as biomarkers in<br />
clinical studies, and probably a biomarker signature will<br />
eventually be required to predict a response. 20 This points<br />
toward a broader approach when analyzing tumor samples<br />
to include molecular pr<strong>of</strong>iling <strong>of</strong> several cancer genes, such<br />
as mutations, rearrangements, and somatic copy-number<br />
alterations. In addition, for true personalized medicine, one<br />
should also take into consideration not only characteristics<br />
<strong>of</strong> the tumor but also the relationship host-tumor (tumor<br />
microenvironment and immune response) and host-drug<br />
(metabolism genes and pharmacogenomics). 19<br />
Additional Issues to Consider for Genomics-Driven<br />
Early <strong>Clinical</strong> Trial Enrollment<br />
Do We Need Certified and Validated Biomarkers?<br />
An increasingly large number <strong>of</strong> putative biomarkers<br />
using innovative sophisticated technologies are being used<br />
in phase I trials. The lack <strong>of</strong> fully validated and reproducible<br />
assays that can be conducted in appropriately certified<br />
laboratories operating according to <strong>Clinical</strong> Laboratory Improvement<br />
Amendments (CLIA) or Good <strong>Clinical</strong> Laboratory<br />
Practice (GCLP) standards is a major concern. In order to<br />
deal with this obstacle, De Bono and colleagues propose a<br />
parallel and simultaneous predictive biomarker/clinical<br />
anticancer drug development process: in early clinical trials,<br />
when patient selection is many times based on the best<br />
guess, pharmacodynamic biomarkers must follow very rigorous<br />
criteria in order to define “pro<strong>of</strong>-<strong>of</strong>-mechanism.” Conversely,<br />
predictive biomarkers for selecting patients could<br />
be explored according to less strict standards. 21 Successfully<br />
enriching phase I trials with patients whose tumors harbor<br />
specific molecular aberrations that may predict response<br />
can demonstrate “pro<strong>of</strong>-<strong>of</strong>-concept” and encourage further<br />
research with a given drug or target. Lack <strong>of</strong> anticancer<br />
effect in the best-case scenario, provided that sufficient<br />
target inhibition is achieved, may ultimately redefine drug<br />
development strategies. Importantly, delineating a selected<br />
patient population does not restrict the late development <strong>of</strong><br />
a specific drug to that subpopulation. If predictive biomarkers<br />
are proven robust and potentially useful in early clinical<br />
trials, they can be further clinically qualified through prospective<br />
evaluation in large randomized controlled trials<br />
before regulatory approval. In other cases in which the<br />
predictive value is not very high, and, therefore, the clinical<br />
utility <strong>of</strong> the biomarker is not obvious, a randomized and<br />
stratified phase II trial is needed to assess both the new drug<br />
and the corresponding biomarker. 20<br />
How Can We Identify Suitable Candidates?<br />
Targeted agents are developed to treat small subsets <strong>of</strong><br />
patient populations, and the operations to perform trials<br />
with the old methods become inefficient. Many difficulties in<br />
recruiting suitable patients have to be addressed. The most<br />
evident is the time to perform molecular analysis to identify<br />
targetable aberrations in the context <strong>of</strong> an early drug development<br />
program.<br />
The traditional approach is to prescreen patients by send-<br />
170<br />
DIENSTMANN, RODON, AND TABERNERO<br />
ing their tumor tissue to a central laboratory for analysis<br />
(either the sponsor <strong>of</strong> the trial or a vendor) just before<br />
considering the enrollment <strong>of</strong> the patient in a trial with an<br />
MTA. This strategy usually demands availability <strong>of</strong> large<br />
amounts <strong>of</strong> tumor tissue, which may be a limitation when<br />
scarce material was used for diagnostic purposes, patients<br />
were enrolled in previous trials, or there is competing<br />
in-house academic research. In addition, because <strong>of</strong> the<br />
delay during the prescreening process and clinical deterioration<br />
at the time <strong>of</strong> recruitment, patients may miss the<br />
opportunity to receive a promising agent.<br />
The alternative strategy calls for local prescreening at<br />
academic institutions and analyzing tumor samples <strong>of</strong> patients<br />
who are still receiving standard treatment for advanced<br />
disease. Molecular selection can be performed at any<br />
time during the disease course. Advantages include the<br />
requirement <strong>of</strong> only one consent form (based on a protocol<br />
approved by the local institutional review board) and the<br />
reduction in the time from disease progression with standard<br />
treatment until enrollment in a clinical trial with an<br />
MTA. Nevertheless, upfront tumor analysis has one big<br />
hurdle: its cost is usually not covered by health care providers,<br />
neither national health systems nor private insurance<br />
companies. Building the cost <strong>of</strong> a broad molecular analysis<br />
into the budget <strong>of</strong> a specific trial is not feasible because it<br />
precedes inclusion in any given trial. Finally, there is always<br />
the possibility <strong>of</strong> not having a slot at the time <strong>of</strong> progression<br />
or the patient not being eligible for a particular trial<br />
with matched MTA as a result <strong>of</strong> inclusion/exclusion criteria<br />
being too strict.<br />
Another important dilemma is the magnitude and type <strong>of</strong><br />
molecular analysis that is likely to be informative for definition<br />
<strong>of</strong> tumor vulnerability and corresponding targeted<br />
therapy. Usually, targetable molecular aberrations are examined,<br />
including specific mutations, common gene amplifications/translocations,<br />
and selected protein-expression<br />
levels. Screening for single mutations has now been replaced<br />
by multiassay platforms, and several are now available both<br />
commercially and at academic centers that evaluate simultaneously<br />
for mutations in 40 to 200 genes (Sequenom,<br />
SNaP shot, among others). Such platforms require very little<br />
tissue, can be performed quickly, and are less expensive.<br />
Academic institutions are already starting to run pilot<br />
projects, including comparative genomic hybridization,<br />
microarray-based pr<strong>of</strong>iling <strong>of</strong> gene expression, massively<br />
parallel sequencing, exome, and whole-genome sequencing.<br />
22,23 In this regard, there is the issue <strong>of</strong> whether crossvalidation<br />
<strong>of</strong> molecular pr<strong>of</strong>iling results should be performed<br />
at each single institution or at a centralized laboratory<br />
serving many institutions in the same region. In addition,<br />
translating high-throughput sequencing for biomarkerdriven<br />
clinical trials for personalized oncology presents<br />
unique logistical challenges, including the development <strong>of</strong><br />
an informed-consent process that addresses how to handle<br />
incidental findings, the selection <strong>of</strong> the results that should<br />
be disclosed to patients, and the implementation <strong>of</strong> efficient<br />
and integrative computational pipelines for data analysis. 22<br />
Should Tumor Biopsies Be Mandatory?<br />
Tissue that is collected as part <strong>of</strong> a research protocol (at<br />
baseline, during treatment, or at the time <strong>of</strong> tumor progression)<br />
has great potential to advance scientific knowledge<br />
by determining how well a drug is affecting the target tissue,
DRUG DEVELOPMENT AND PERSONALIZED ONCOLOGY<br />
Fig 1. Matching tumor type, molecular<br />
aberration, screening platform, and<br />
early clinical trial with targeted agent.<br />
In the first scenario (left to right), a patient<br />
with ovarian cancer (whose tumor<br />
may present PI3K pathway aberrations,<br />
RAS, RAF or BRCA mutations) will<br />
have molecular prescreening with multiple<br />
assays and will be ultimately referred<br />
to a corresponding phase I trial.<br />
On the other hand (right to left), if the<br />
screening is performed according to<br />
the molecular target agent available<br />
at the site (e.g., FGFR inhibitor), screening<br />
will be focused on different tumor<br />
types, such as HR positive breast cancer<br />
and squamous NSCLC.<br />
Abbreviations: HR, hormone receptor;<br />
NSCLC, non-small cell lung cancer; CRC,<br />
colorectal cancer; IHC, immunohistochemistry;<br />
FISH, fluorescence in situ hybridization;<br />
CGH, comparative genomic<br />
hybridization; PCR, polymerase chain<br />
reaction.<br />
identifying mechanisms <strong>of</strong> drug resistance, describing new<br />
oncologic signaling pathways, or establishing predictors <strong>of</strong><br />
a favorable or unfavorable outcome. 24 To date, many assays<br />
still require fresh, frozen samples for technical reasons.<br />
Biomarker analysis can sometimes directly benefit the patient<br />
(for example, in breast cancer relapses presenting a<br />
change in hormonal receptor and HER2 status) and also<br />
future patients with cancer. Mandatory biopsies are acceptable<br />
when investigators appropriately weigh the risks<br />
against the necessity <strong>of</strong> the correlative question. However,<br />
various hurdles must be faced when protocols mandate<br />
biopsies. Ethical concerns have been raised when the participant’s<br />
enrollment in a clinical trial depends on his or her<br />
consent to undergo a research biopsy. Investigators may feel<br />
forced to find suitable on-trial or <strong>of</strong>f-trial options for patients<br />
whose tumors have been biopsied and undergone molecular<br />
pr<strong>of</strong>iling. The risks <strong>of</strong> the procedure also need to be taken<br />
into consideration. However, according to the experience <strong>of</strong><br />
large centers, biopsies in early phase clinical trials are safe,<br />
with less than 1.5% risk <strong>of</strong> serious complications. 25<br />
Ongoing Innovative Trials<br />
<strong>Clinical</strong> trial design needs to be adjusted to a new era <strong>of</strong><br />
personalized oncology. Novel approaches include histologyindependent<br />
trials (for example, patients with BRCA1/2<br />
mutated tumors, regardless <strong>of</strong> histology), window-<strong>of</strong>opportunity<br />
trials (when standard anticancer therapy is<br />
delayed, and patients receive first a matched MTA allowing<br />
chemotherapy-free intervals) or trials that subclassify a<br />
specific disease into discrete molecular categories (such as<br />
the Investigation <strong>of</strong> Serial Studies to Predict Your Therapeutic<br />
Response with Imaging and Molecular Analysis [I-<br />
SPY 2] trial in breast cancer and the Biomarker-integrated<br />
Approaches <strong>of</strong> Targeted Therapy for Lung Cancer Elimination<br />
[BATTLE] trial in NSCLC). 26,27 For these approaches<br />
to be possible, a cooperative environment is necessary, with<br />
experienced academic institutions, regulatory authorities,<br />
and pharmaceutical and diagnostic companies working together<br />
in order to promote data sharing, standardized procedures,<br />
technology exchange, and avoidance <strong>of</strong> duplicative<br />
efforts. 2 Initiatives, such as Worldwide Innovative Network<br />
(WIN) Consortium, are a great example <strong>of</strong> this collaboration<br />
toward a common goal. Together, investigators can utilize<br />
structured methods to define when a new agent should cease<br />
further study and when it has demonstrated sufficient<br />
likelihood <strong>of</strong> potential benefit to proceed to the next phase <strong>of</strong><br />
testing.<br />
Preliminary results from personalized medicine programs<br />
suggest that molecular analysis <strong>of</strong> tumors and the use <strong>of</strong><br />
MTAs to counteract the effects <strong>of</strong> specific aberrations improves<br />
the outcomes <strong>of</strong> affected patients. In a seminal trial,<br />
Von H<strong>of</strong>f and colleagues used molecular pr<strong>of</strong>iling (based<br />
on immunohistochemistry, immun<strong>of</strong>luorescence, and microarray<br />
analysis) to select a targeted agent and were able to<br />
prove that this approach is feasible. In 27% <strong>of</strong> the patients,<br />
treatment based on molecular analysis resulted in longer<br />
progression-free survival than the prior unmatched therapy.<br />
28 Similar results were seen in an innovative model for<br />
phase I clinical trials performed at the University <strong>of</strong> Texas<br />
M. D. Anderson Cancer Center. Matching patients with an<br />
MTA in early clinical trials, mainly based on mutation<br />
status <strong>of</strong> frequently activated genes in cancer (KRAS,<br />
NRAS, BRAF, PIK3CA, EGFR, CKIT) and PTEN protein<br />
expression levels, resulted in longer time to treatment<br />
failure (5.3 vs. 3.2 months) compared with their prior systemic<br />
antitumor therapy. In addition, for patients with one<br />
molecular aberration, the response rate was 29% with<br />
matched targeted therapy versus 8% without matching,<br />
highlighting the potential improved outcomes <strong>of</strong> heavily<br />
pretreated patients enrolled in genomic-driven phase I trials<br />
<strong>of</strong> MTAs. 18 Based on these and other observations, our<br />
institute, the Massachusetts General Hospital, the Royal<br />
Marsden Cancer Center, and the Institute Gustave Roussy,<br />
among others, are already building flexible programs<br />
(adapting the putative predictive biomarkers and the clinical<br />
trials available on-site) to serve their phase I trials.<br />
Figure 1 exemplifies different ways <strong>of</strong> matching tumor type,<br />
molecular aberration, screening platform, and early clinical<br />
trials with MTAs.<br />
Conclusion<br />
Phase I trials in which an MTA can be matched precisely<br />
with a population <strong>of</strong> patients whose tumors are driven<br />
predominantly by the target <strong>of</strong> interest are still uncommon.<br />
171
This situation is not surprising, given that most advanced<br />
human malignancies have complex signaling networks. Nevertheless,<br />
the one-size-fits-all approach for drug development<br />
<strong>of</strong> MTAs in solid tumors has shown disappointing<br />
results in recent years. It is no longer viable to develop<br />
targeted drugs without a reliable indicator <strong>of</strong> a potential<br />
benefit derived from them. The genomic-driven accelerated<br />
approach <strong>of</strong> drug development is applicable when the appropriate<br />
patient population can be evaluated in the phase I<br />
trial. This conceptual shift <strong>of</strong> clinical development <strong>of</strong> pairing<br />
therapies and biomarkers will require a change in the design<br />
and execution <strong>of</strong> clinical trials to one where antitumor<br />
Authors’ Disclosures <strong>of</strong> Potential Conflicts <strong>of</strong> Interest<br />
Employment or<br />
Leadership<br />
Positions<br />
Consultant or<br />
Advisory Role<br />
Author<br />
Rodrigo Dienstmann*<br />
Jordi Rodon*<br />
Josep Tabernero Amgen; Bristol-<br />
Myers Squibb;<br />
Genentech;<br />
Merck KGaA;<br />
Novartis; Onyx;<br />
Pfizer; Roche;<br />
San<strong>of</strong>i<br />
*No relevant relationships to disclose.<br />
1. de Bono JS, Ashworth A. Translating cancer research into targeted<br />
therapeutics. Nature. 2010;467:543-549.<br />
2. Mendelsohn J. A national cancer clinical trials system for targeted<br />
therapies. Sci Transl Med. 2011;3:75cm8.<br />
3. Luo J, Solimini NL, Elledge SJ. Principles <strong>of</strong> cancer therapy: oncogene<br />
and non-oncogene addiction. Cell. 2009;136:823-837.<br />
4. Hanahan D, Weinberg RA. Hallmarks <strong>of</strong> cancer: the next generation.<br />
Cell. 2011;144:646-674.<br />
5. Yap TA, Sandhu SK, Workman P, et al. Envisioning the future <strong>of</strong> early<br />
anticancer drug development. Nat Rev Cancer. 2010;10:514-523.<br />
6. Carden CP, Sarker D, Postel-Vinay S, et al. Can molecular biomarkerbased<br />
patient selection in Phase I trials accelerate anticancer drug development?<br />
Drug Discov Today. 2010;15:88-97.<br />
7. Garrido-Laguna I, Hidalgo M, Kurzrock R. The inverted pyramid <strong>of</strong><br />
biomarker-driven trials. Nat Rev Clin Oncol. 2011;8:562-566.<br />
8. Baselga J, Tripathy D, Mendelsohn J, et al. Phase II study <strong>of</strong> weekly<br />
intravenous recombinant humanized anti-p185HER2 monoclonal antibody<br />
in patients with HER2/neu-overexpressing metastatic breast cancer. J Clin<br />
Oncol. 1996;14:737-744.<br />
9. Fong PC, Boss DS, Yap TA, et al. Inhibition <strong>of</strong> poly(ADP-ribose)<br />
polymerase in tumors from BRCA mutation carriers. N Engl J Med. 2009;<br />
361:123-134.<br />
10. Flaherty KT, Puzanov I, Kim KB, Ribas A, et al. Inhibition <strong>of</strong> mutated,<br />
activated BRAF in metastatic melanoma. N Engl J Med. 2010;363:809-819.<br />
11. Kwak EL, Bang YJ, Camidge DR, et al. Anaplastic lymphoma kinase<br />
inhibition in non-small-cell lung cancer. N Engl J Med. 2010;363:1693-1703.<br />
12. LoRusso PM, Rudin CM, Reddy JC, et al. Phase I trial <strong>of</strong> hedgehog<br />
pathway inhibitor vismodegib (GDC-0449) in patients with refractory, locally<br />
advanced or metastatic solid tumors. Clin Cancer Res. 2011;17:2502-2511.<br />
13. Krop IE, Beeram M, Modi S, Jones SF, et al. Phase I study <strong>of</strong><br />
trastuzumab-DM1, an HER2 antibody-drug conjugate, given every 3 weeks to<br />
patients with HER2-positive metastatic breast cancer. J Clin Oncol. 2010;28:<br />
2698-2704.<br />
14. Demetri GD, von Mehren M, Blanke CD, et al. Efficacy and safety <strong>of</strong><br />
imatinib mesylate in advanced gastrointestinal stromal tumors. N Engl<br />
J Med. 2002;347:472-480.<br />
15. Kefford R, Arkenau H, Brown MP, et al. Phase I/II study <strong>of</strong><br />
GSK2118436, a selective inhibitor <strong>of</strong> oncogenic mutant BRAF kinase, in<br />
patients with metastatic melanoma and other solid tumors. J Clin Oncol.<br />
2010;28:abstract 8503.<br />
172<br />
activity <strong>of</strong> MTAs is evaluated. Hopefully, the decreasing cost<br />
<strong>of</strong> high-throughput molecular technologies will enable more<br />
comprehensive characterization <strong>of</strong> the numerous additional<br />
tumor genome alterations, and, therefore, facilitate a personalized<br />
oncology approach. 29 In addition, alternative techniques<br />
to biopsy <strong>of</strong> the metastatic disease might be<br />
considered in the near future, such as molecular imaging<br />
and genomic analysis <strong>of</strong> circulating tumor cells, or tumor<br />
DNA detected in peripheral blood. 30 In conclusion, biologic<br />
evidence, clinical data, and previous experience propose that<br />
it is time for a change in the current way drug development<br />
<strong>of</strong> targeted agents is done.<br />
Stock<br />
Ownership Honoraria<br />
REFERENCES<br />
Amgen; Merck<br />
KGaA; Novartis;<br />
Roche; San<strong>of</strong>i<br />
DIENSTMANN, RODON, AND TABERNERO<br />
Research<br />
Funding<br />
Expert<br />
Testimony<br />
Other<br />
Remuneration<br />
16. Kurzrock R, Sherman SI, Ball DW, et al. Activity <strong>of</strong> XL184 (Cabozantinib),<br />
an oral tyrosine kinase inhibitor, in patients with medullary thyroid<br />
cancer. J Clin Oncol. 2011;29:2660-2666.<br />
17. Horstmann E, McCabe MS, Grochow L, et al. Risks and benefits <strong>of</strong><br />
phase 1 oncology trials, 1991 through 2002. N Engl J Med. 2005;352:895-904.<br />
18. Tsimberidou AM, Iskander NG, Hong DS, et al. Personalized medicine<br />
in a phase I clinical trials program: The MD Anderson Cancer Center<br />
Initiative. J Clin Oncol. 2011;29:abstract CRA2500.<br />
19. Gasparini G, Longo R. The paradigm <strong>of</strong> personalized therapy in<br />
oncology. Expert Opin Ther Targets. 2011 Nov 11. [Epub ahead <strong>of</strong> print]<br />
20. Coyle VM, Johnston PG. Genomic markers for decision making: what is<br />
preventing us from using markers? Nat Rev Clin Oncol. 2010;7:90-97.<br />
21. Garcia VM, Cassier PA, De Bono JS. Parallel anticancer drug development<br />
and molecular stratification to qualify predictive biomarkers: Dealing<br />
with obstacles hindering progress. Cancer Discovery. 2011;1:207-212.<br />
22. Roychowdhury S, Iyer MK, Robinson DR, et al. Personalized oncology<br />
through integrative high-throughput sequencing: A pilot study. Sci Transl<br />
Med. 2011;3:111ra121.<br />
23. Wagle N, Berger MF, Davis MJ, et al. High-throughput detection <strong>of</strong><br />
actionable genomic alterations in clinical tumor samples by targeted, massively<br />
parallel sequencing. Cancer Discovery. <strong>2012</strong>;2:82-93.<br />
24. Olson EM, Lin NU, Krop IE, Winer EP. The ethical use <strong>of</strong> mandatory<br />
research biopsies. Nat Rev Clin Oncol. 2011;8:620-625.<br />
25. El-Osta H, Hong D, Wheler J, et al. Outcomes <strong>of</strong> research biopsies in<br />
phase I clinical trials: the M. D. Anderson Cancer Center experience.<br />
Oncologist. 2011;16:1292-1298.<br />
26. Barker AD, Sigman CC, Kell<strong>of</strong>f GJ, et al. I-SPY 2: an adaptive breast<br />
cancer trial design in the setting <strong>of</strong> neoadjuvant chemotherapy. Clin Pharmacol<br />
Ther. 2009;86:97-100.<br />
27. Kim ES, Herbst RS, Wistuba II, et al. The BATTLE Trial: Personalizing<br />
therapy for lung cancer. Cancer Discovery. 2011;1:44-53.<br />
28. Von H<strong>of</strong>f DD, Stephenson JJ Jr., Rosen P, et al. Pilot study using<br />
molecular pr<strong>of</strong>iling <strong>of</strong> patients’ tumors to find potential targets and select<br />
treatments for their refractory cancers. J Clin Oncol. 2010;28:4877-4883.<br />
29. MacConaill LE, Garraway LA. <strong>Clinical</strong> implications <strong>of</strong> the cancer<br />
genome. J Clin Oncol. 2010;28:5219-5228.<br />
30. Devriese LA, Voest EE, Beijnen JH, Schellens JH. Circulating tumor<br />
cells as pharmacodynamic biomarker in early clinical oncological trials.<br />
Cancer Treat Rev. 2011;37:579-589.
IMMUNOTHERAPY WITH THE T-CELL CHECKPOINT<br />
ANTIBODIES: IMPLICATIONS FOR THE<br />
PRACTITIONER<br />
CHAIR<br />
Mario Sznol, MD<br />
Yale Cancer Center<br />
New Haven, CT<br />
SPEAKERS<br />
Jeffrey S. Weber, MD, PhD<br />
H. Lee M<strong>of</strong>fitt Cancer Center & Research Institute<br />
Tampa, FL<br />
F. Stephen Hodi, MD<br />
Dana-Farber Cancer Institute<br />
Boston, MA
Practical Management <strong>of</strong> Immune-Related<br />
Adverse Events from Immune Checkpoint<br />
Protein Antibodies for the Oncologist<br />
Overview: Monoclonal antibodies directed against immune<br />
checkpoint proteins, such as cytotoxic T-lymphocyte<br />
antigen-4 (CTLA-4) or programmed death-1 (PD-1), can boost<br />
endogenous immune responses directed against tumor cells.<br />
Recently, ipilimumab was approved by the U.S. Food and Drug<br />
Administration (FDA) for the treatment <strong>of</strong> metastatic melanoma,<br />
and the anti-PD-1 antibody BMS-936558 has shown<br />
promising results in patients with melanoma, non-small cell<br />
lung cancer, and renal cell cancer. During treatment with<br />
these antibodies, a unique set <strong>of</strong> toxicities occur called<br />
immune-related adverse events (irAEs). These irAEs may occur<br />
at any time during treatment and include colitis characterized<br />
by a mild to moderate but occasionally severe and<br />
persistent diarrhea. Hypophysitis, hepatitis, pancreatitis, iri-<br />
THE DOCUMENTED efficacy <strong>of</strong> the anti-CTLA-4 antibody<br />
ipilimumab that led to its FDA approval for<br />
treatment <strong>of</strong> metastatic melanoma suggests that dissemination<br />
<strong>of</strong> knowledge about and familiarity with its unique side<br />
effects is important for the oncology community. 1,2 Treating<br />
physicians must be able to quickly detect and adequately<br />
treat the toxicities <strong>of</strong> ipilimumab. The management <strong>of</strong> these<br />
toxicities requires the collaborative efforts <strong>of</strong> a multidisciplinary<br />
team led by a physician and including the nursing/<br />
midlevel staff who are routinely in contact with the patient.<br />
In addition, understanding irAE management is important<br />
for consulting specialists, such as gastroenterologists, endocrinologists,<br />
hepatologists, dermatologists, surgeons, and<br />
others. Therapeutic algorithms have been developed for the<br />
management <strong>of</strong> irAEs and are part <strong>of</strong> the package insert for<br />
ipilimumab, yet it is felt that a review <strong>of</strong> the management<br />
<strong>of</strong> these novel toxicities might be useful for those who treat<br />
patients with melanoma. 3<br />
The irAE: What Is It? When Does It Happen?<br />
Where Does It Happen?<br />
Early in ipilimumab’s development, it became clear that it<br />
induced dose-related, immune-related, or inflammatory side<br />
effects. The most common systems affected were the skin,<br />
gut, liver, and pituitary. Immunohistochemistry <strong>of</strong> affected<br />
skin and gut revealed infiltration by CD4 and CD8 T cells,<br />
and highly activated effector cells correlated with side effect<br />
intensity. 4 Elevated inflammatory cytokines in the sera, as<br />
well as rapid resolution <strong>of</strong> some irAE symptoms with use <strong>of</strong><br />
the tumor necrosis factor-alpha antibody infliximab, suggested<br />
that cytokine release by activated T cells was associated<br />
with irAEs. 5 Allergic and anaphylactic reactions during<br />
infusions were rare, since ipilimumab is a human antibody.<br />
In a pooled analysis <strong>of</strong> 325 patients treated with ipilimumab<br />
10 mg/kg four times every three weeks, drug-related overall<br />
adverse events were observed in 84.6% <strong>of</strong> patients, <strong>of</strong> which<br />
immune-related adverse events <strong>of</strong> any grade accounted<br />
for 72.3%. 6 Grade 3 or 4 irAEs were observed in 25.2% <strong>of</strong><br />
patients, mainly in the gastrointestinal (GI) tract (12%),<br />
liver (7%), skin (3%), and endocrine organs (3%). These<br />
adverse events exhibited a characteristic timing. Skin-<br />
174<br />
By Jeffrey S. Weber, MD, PhD<br />
docyclitis, lymphadenopathy, neuropathies, and nephritis<br />
have also been reported with ipilimumab, and a subset <strong>of</strong><br />
those side effects has also been observed with BMS-936558.<br />
Patient and physician education as well as good patient–<br />
caretaker communication are keys to limiting the morbidity <strong>of</strong><br />
irAEs. Early recognition <strong>of</strong> these irAEs and initiation <strong>of</strong> treatment<br />
are critical to reduce the risk <strong>of</strong> complications, since<br />
virtually all irAEs are reversible with the use <strong>of</strong> steroids and<br />
other immune suppressants. The onset <strong>of</strong> grade 3 to 4 irAEs<br />
correlated with treatment response in some ipilimumab studies.<br />
This article provides detailed description and recommendations<br />
for practicing oncologists to manage the common<br />
irAEs associated with antibodies against immune checkpoint<br />
blockade.<br />
related adverse events occurred after a period <strong>of</strong> 2 to 3 weeks<br />
from treatment (1 dose). Thereafter, GI and hepatic adverse<br />
events began to occur within 6 to 7 weeks (2 to 3 doses), and<br />
finally endocrinologic side effects were noted after an average<br />
<strong>of</strong> 9 weeks (3 to 4 doses) since the last dose <strong>of</strong> ipilimumab.<br />
The frequencies <strong>of</strong> grade 3 to 4 ipilimumab-related<br />
irAEs increased with dose and were not observed at 0.3<br />
mg/kg. Few irAEs were seen at 1 mg/kg, and they were<br />
present in 10% <strong>of</strong> patients at the 3 mg/kg dose level, rising to<br />
over 20% <strong>of</strong> patients at 10 mg/kg. 7 The long half life <strong>of</strong><br />
ipilimumab may be consistent with slow resolution <strong>of</strong> many<br />
irAEs. The primary approach to grade 1 to 2 irAEs is<br />
supportive and symptomatic care; for higher grade irAEs,<br />
steroids by mouth or parenterally are given, and either<br />
skipping a dose or stopping therapy is appropriate. Recurrent<br />
grade 2 irAEs may also mandate skipping a dose or the<br />
use <strong>of</strong> steroids. Algorithms for toxicity described below,<br />
available online in the ipilimumab package insert, describe<br />
detailed management guidelines that are tailored to the<br />
severity and nature <strong>of</strong> the particular irAE.<br />
Colitis<br />
Diarrhea occurs in up to 44% <strong>of</strong> patients receiving<br />
ipilimumab. 2,8-12 Diarrhea <strong>of</strong> grade 3 or 4 (7 diarrheal bowel<br />
movements above baseline in 24 hours) was reported in 18%<br />
<strong>of</strong> patients at the dose <strong>of</strong> 10 mg/kg. 2 Diarrhea <strong>of</strong> grades 3/4<br />
can cause weakness, imbalanced electrolytes, and weight<br />
loss. Diarrhea can also be associated with colitis, which, if<br />
unattended, can lead to obstruction and bowel perforation.<br />
To date, very few cases <strong>of</strong> bowel perforation leading to<br />
colectomy have been reported with ipilimumab therapy. 8,13<br />
Unlike chronic inflammatory bowel diseases (IBD), such<br />
as Crohn’s disease or ulcerative colitis, colitis associated<br />
with ipilimumab predominantly involves the descending<br />
From the M<strong>of</strong>fitt Cancer Center, Tampa, FL.<br />
Author’s disclosure <strong>of</strong> potential conflicts <strong>of</strong> interest are found at the end <strong>of</strong> this article.<br />
Address reprint requests to Jeffrey S. Weber, MD, PhD, M<strong>of</strong>fitt Cancer Center, 12902<br />
Magnolia Dr., SRB-2, Tampa, FL; email: jeffrey.weber@m<strong>of</strong>fitt.org.<br />
© <strong>2012</strong> by <strong>American</strong> <strong>Society</strong> <strong>of</strong> <strong>Clinical</strong> <strong>Oncology</strong>.<br />
1092-9118/10/1-10
CHECKPOINT PROTEIN INHIBITOR TOXICITY MANAGEMENT<br />
colon. 12 Neutrophilic infiltrates are seen in 46% <strong>of</strong> patients,<br />
lymphocytic infiltrates in 15%, and a mixed neutrophiliclymphocytic<br />
infiltrate in 38%. 12,14<br />
Low-grade diarrhea (grade 1, an increase <strong>of</strong> 2 over baseline<br />
in 24 hours) should be treated symptomatically using<br />
loperamide, oral hydration, and electrolyte substitution. The<br />
<strong>American</strong> Dietary Association colitis diet is useful. With<br />
persistent or higher-grade diarrhea, bacterial or parasitic<br />
infection, viral gastroenteritis, or the first manifestation <strong>of</strong><br />
an IBD must be ruled out by examination for stool leukocytes,<br />
stool cultures, and a Clostridium difficile titer. Grade<br />
2 diarrhea can be treated with the addition <strong>of</strong> oral diphenoxylate<br />
hydrochloride and atropine sulfate four times daily<br />
and budesonide 9 mg daily. Endoscopy is recommended to<br />
confirm or rule out colitis with persistent grade 2 diarrhea or<br />
grades 1 to 2 diarrhea with bleeding. Endoscopy can generally<br />
be done safely in patients with the GI side effects from<br />
ipilimumab, although there is the risk <strong>of</strong> perforation with<br />
biopsies. The presence and nature <strong>of</strong> colitis may change<br />
one’s therapeutic management, since diffuse ulceration and<br />
bleeding in the setting <strong>of</strong> grade 2 diarrhea may mandate a<br />
course <strong>of</strong> oral steroids and skipping a dose and can also<br />
represent an increased risk for the development <strong>of</strong> a bowel<br />
perforation.<br />
For grade 3 or 4 diarrhea (7 or more increase over baseline<br />
in 24 hours), treatment with ipilimumab should be permanently<br />
discontinued and intravenous steroids and replenishment<br />
<strong>of</strong> fluid and electrolytes intravenously should be<br />
instituted. Intravenous methylprednisolone 125 mg should<br />
be given. Oral dexamethasone 4 mg every four hours or<br />
prednisone 1 to 2 mg/kg/daily can be given thereafter,<br />
followed by a taper and discontinuation over the next 6<br />
weeks. Generally, there is a significant improvement <strong>of</strong> GI<br />
symptoms within 1 to 2 weeks. Steroid therapy must be<br />
tapered over at least 4 weeks to ensure complete resolution<br />
<strong>of</strong> symptoms. In patients with diffuse and severe ulceration<br />
and/or bleeding, a tapering <strong>of</strong> up to 6 to 8 weeks is appropriate<br />
since rapid tapering can result in recurrence or<br />
worsening <strong>of</strong> symptoms. The use <strong>of</strong> opiates and other analgesics<br />
may mask pain associated with colitis induced by<br />
ipilimumab. A low threshold for concern about GI complications,<br />
including perforation and obstruction, should be maintained.<br />
A low threshold for imaging with plain films or<br />
computed tomography (CT), as well as a surgical consult—<br />
especially in any patients admitted to the hospital for GI<br />
irAE management—is appropriate.<br />
KEY POINTS<br />
● Immune-related adverse events (irAEs) represent a<br />
unique spectrum <strong>of</strong> toxicities with checkpoint protein<br />
inhibition.<br />
● Colitis, hepatitis, endocrinopathies, and dermatologic<br />
toxicities are seen with ipilimumab and the key to<br />
proper management <strong>of</strong> these irAEs is good patientcaretaker<br />
communication.<br />
● irAEs are generally managed successfully with steroids<br />
and other immune suppressants.<br />
● irAEs are also observed with the anti-PD-1 antibody<br />
BMS-936558.<br />
If intravenous steroids followed by high-dose oral steroids<br />
does not decrease symptoms within 48 to 72 hours, treatment<br />
with infliximab at 5 mg/kg every 2 weeks is an<br />
alternative. 5,15 Once relief <strong>of</strong> symptoms is achieved, which<br />
can be very rapid and dramatic, it should be discontinued<br />
and a prolonged steroid taper over 45 to 60 days should be<br />
instituted. There may be a waxing and waning <strong>of</strong> the GI<br />
toxicities <strong>of</strong> ipilimumab during the course <strong>of</strong> steroids. As<br />
steroids are tapered, symptoms may worsen, mandating a<br />
retapering <strong>of</strong> steroids starting at a higher dose <strong>of</strong> 80 or 100<br />
mg, a more prolonged taper, and additional use <strong>of</strong> infliximab.<br />
Ipilimumab should be permanently discontinued for<br />
grade 3 or 4 diarrhea or colitis.<br />
Dermatologic Toxicity<br />
A diffuse, erythematous maculopapular rash that can be<br />
intensely pruritic was observed in 47% to 68% <strong>of</strong> patients,<br />
starting an average <strong>of</strong> 3 to 4 weeks after ipilimumab. 1-3,7-9<br />
In 4% <strong>of</strong> patients, it was severe. Ipilimumab-induced rashes<br />
can be quite focal or even patchy. Microscopic examination<br />
shows a perivascular lymphocytic infiltrate that extends<br />
deep into the dermis in most cases. 4 Immunohistochemical<br />
staining showed that CD4-positive and Melan-A-specific<br />
CD8-positive T cells were in close proximity to apoptotic<br />
melanocytes, suggesting that an immune response was<br />
directed against melanocytes. 4 This is consistent with a<br />
reported 11% rate <strong>of</strong> vitiligo with ipilimumab. 7 Most skin<br />
eruptions and pruritus associated with ipilimumab are generally<br />
managed symptomatically and usually do not require<br />
skipping a dose or discontinuation. Topical glucocorticosteroids<br />
(e.g., betamethasone 0.1% cream) or urea-containing<br />
creams in combination with oral antipruritics (e.g., diphenhydramine<br />
HCl or hydroxyzine HCl) are recommended. 3 For<br />
grade 3 dermatologic irAEs, one should hold a dose and treat<br />
witha3to4-week tapering course <strong>of</strong> oral steroids, starting<br />
at 1 mg/kg prednisone or dexamethasone 4 mg four times<br />
orally daily. Ipilimumab can be held for moderate to severe<br />
skin toxicity but should be permanently discontinued for<br />
severe, life-threatening skin toxicity and steroids initiated<br />
at 1 to 2 mg/kg prednisone orally or its equivalent tapering<br />
over not less than 30 days. Rare cases <strong>of</strong> toxic epidermal<br />
necrolysis, as well as Stevens-Johnson syndrome, both in less<br />
than 1% <strong>of</strong> patients, have been reported with ipilimumab,<br />
and several patients with those conditions have died.<br />
Endocrinopathies<br />
Hypophysitis is an uncommon complication <strong>of</strong> treatment<br />
with ipilimumab with an incidence <strong>of</strong> approximately<br />
1.5%. 16,17 Behavioral change, fatigue, severe headaches,<br />
blurring or diplopia, myalgias, loss <strong>of</strong> appetite, or nausea<br />
and vomiting have been seen, but symptoms can be vague.<br />
Visual changes and/or headaches in patients with metastatic<br />
melanoma should also prompt concern for possible<br />
central nervous system or ocular metastases and should be<br />
evaluated by magnetic resonance imaging (MRI) <strong>of</strong> the brain<br />
with gadolinium. Hypophysitis can present as a diffuse,<br />
heterogenous enlargement <strong>of</strong> the pituitary on a brain MRI<br />
but can be completely normal. 18 When hypophysitis with<br />
pituitary dysfunction is suspected, blood tests including<br />
thyroid-stimulating hormone (TSH), free T4, adrenocorticotropic<br />
stimulating hormone, cortisol, leutinizing hormone,<br />
and follicle-stimulating hormone should be obtained in<br />
women, and the first four plus testosterone in men. Typically<br />
175
the anterior pituitary axis is involved, affecting thyroid,<br />
gonadal, and adrenal function, but isolated axis dysfunction<br />
can be seen. Hypophysitis will cause low free T4 as well as<br />
TSH, but with peripheral thyroiditis leading to hypothyroidism,<br />
the TSH will be elevated even though T3 and free T4<br />
will be low.<br />
Hypophysitis with clinically significant adrenal insufficiency<br />
and hypotension, dehydration, and electrolyte abnormalities—such<br />
as hyponatremia and hyperkalemia—<br />
constitutes adrenal crisis. Hospitalization and intravenous<br />
steroids with mineralocorticoid activity, such as methylprednisolone,<br />
should be initiated while waiting for laboratory<br />
results. Infection and sepsis should be ruled out with<br />
appropriate cultures and imaging. Prednisone 1 mg/kg by<br />
mouth should be administered if patients are clinically<br />
stable. Steroids can usually be tapered over 30 days to<br />
achieve physiologic replacement levels. Thyroid hormone<br />
and/or testosterone replacement therapy may not be permanent,<br />
as the need for those hormones may wane over months<br />
in some patients. Cortisone replacement may also not be<br />
permanent in a modest portion <strong>of</strong> patients. With grade 2<br />
endocrinopathies, ipilimumab therapy can be safely continued,<br />
in our experience, but proper informed consent should<br />
be obtained from the patient before continuing treatment<br />
since continued deterioration <strong>of</strong> endocrine function may<br />
occur. Grade 3 to 4 endocrinopathies require discontinuation<br />
<strong>of</strong> ipilimumab.<br />
Isolated thyroid dysfunction can also be seen with ipilimumab,<br />
initially presenting as autoimmune thyroiditis<br />
with elevation <strong>of</strong> free T4, evolving to a state <strong>of</strong> “burnt-out”<br />
hypopthyroidism. Current recommendations in the ipilimumab<br />
package insert include checking baseline TSH and<br />
free T4 and subsequent monitoring every 3 weeks during<br />
induction ipilimumab treatment and every 3 months for 6<br />
months following completion <strong>of</strong> therapy. Isolated adrenal<br />
insufficiency unrelated to hypophysitis can also be seen. 19<br />
Hepatitis<br />
Hepatotoxicity has been observed in 3% to 9% <strong>of</strong> patients<br />
receiving ipilimumab. 1,2,6-9 It usually presents as an asymptomatic<br />
increase <strong>of</strong> transaminases and bilirubin, although<br />
some patients also have fevers and malaise. Liver biopsies<br />
have shown a diffuse T-cell infiltrate consistent with<br />
immune-related hepatitis. This must be differentiated from<br />
progressive metastases in the liver, as well as other etiologies<br />
such as viral hepatitis or another drug-specific toxic<br />
reaction. One should perform a standard workup to rule out<br />
viral hepatitis, disease progression, and other drug-related<br />
causes for abnormal liver functions. The current algorithm<br />
for the management <strong>of</strong> a hepatotoxicity irAE contains the<br />
recommendation that for grades 3 to 5 toxicity, one should<br />
use high-dose intravenous glucocorticosteroids for 24 to 48<br />
hours, followed by an oral steroid taper with dexamethasone<br />
in a dosage <strong>of</strong> 4 mg every 4 hours or prednisone at 1 to 2<br />
mg/kg tapered over not less than 30 days. Liver function<br />
blood tests, if elevated beyond eight times normal, should<br />
be checked every other day until they begin to drop, then<br />
weekly until they are normal and mandate stopping ipilimumab.<br />
If serum transaminase levels do not decrease 48<br />
hours after initiation <strong>of</strong> systemic steroids, consideration<br />
should be given to the use <strong>of</strong> oral mycophenolate m<strong>of</strong>etil 500<br />
mg every 12 hours. Infliximab—because <strong>of</strong> its potential for<br />
hepatotoxicity—should be avoided in this setting. One re-<br />
176<br />
JEFFREY S. WEBER<br />
port describes a patient treated with antithymocyte globulin<br />
after developing severe hepatotoxicity from ipilimumab despite<br />
high-dose steroids and mycophenylate m<strong>of</strong>etil. 20 A<br />
waxing and waning picture may be seen with anti-CTLA-4–<br />
induced hepatitis, and several courses <strong>of</strong> tapering steroids<br />
may be required. A very significant rate <strong>of</strong> hepatotoxicity<br />
was seen when ipilimumab was combined with dacarbazine<br />
and may be related to hepatic toxicity associated with<br />
dacarbazine that is exacerbated by ipilimumab. 2 Ipilimumab<br />
should be permanently discontinued for grade 3 to 4<br />
hepatotoxicity.<br />
Current guidelines recommend evaluation <strong>of</strong> serum markers<br />
<strong>of</strong> hepatic function at baseline and before each dose, as<br />
well as periodically (every three months) after completion <strong>of</strong><br />
induction ipilimumab.<br />
The Role <strong>of</strong> Surgery in Managing irAEs<br />
With proper management <strong>of</strong> colitis, the likelihood that<br />
obstruction or perforation <strong>of</strong> the bowel will occur is low, but<br />
rarely colitis cannot be controlled by conservative measures<br />
such as high-dose intravenous steroids, hydration, or repeated<br />
injections <strong>of</strong> infliximab. 21 If colitis is resistant to<br />
those measures, patients must be made nil per os (NPO)<br />
except for required oral medicines, have a central venous<br />
catheter inserted, and have complete bowel rest with total<br />
parenteral nutrition (TPN) <strong>of</strong>ten for a period <strong>of</strong> 2 to 3<br />
months. 3 Rarely diarrhea and symptoms <strong>of</strong> colitis may<br />
persist despite a trial <strong>of</strong> NPO and TPN while steroids are<br />
unable to be tapered. In that case, if a colonoscopy confirms<br />
recurring inflammation, then surgery to perform an ileostomy<br />
and put the colon at complete bowel rest is a reasonable<br />
option. In the experience <strong>of</strong> the author, a3to4month<br />
period suffices to allow inflammation to resolve, the colonoscopic<br />
picture to improve, and diarrhea to disappear. At that<br />
time, the ileostomy may be surgically reversed with some<br />
assurance that the colitis has completely resolved after<br />
CT scanning <strong>of</strong> the abdomen and repeat colonoscopy. A<br />
partial or even complete colectomy may be necessary in very<br />
rare cases <strong>of</strong> ipilimumab-induced colitis refractory to all<br />
measures. Bowel perforation is rare—occurring in less<br />
than 1% <strong>of</strong> patients—but must be considered. 13 The risk <strong>of</strong><br />
death with ipilimumab is low, at 1% or less, <strong>of</strong>ten related to<br />
complications <strong>of</strong> GI irAEs. 1,2 It should be noted that patients<br />
previously treated with high-dose interleukin-2 appear to<br />
have an increased risk <strong>of</strong> intestinal perforation with ipilimumab,<br />
based on a small experience at one institution.<br />
irAEs Observed with PD-1 Antibodies<br />
In phase I and ongoing phase II trials, the spectrum <strong>of</strong> side<br />
effects <strong>of</strong> PD-1 antibody BMS-936558 has been shown to<br />
be qualitatively similar to ipilimumab but with fewer doselimiting<br />
immune-related side effects. In an initial singledose,<br />
phase I trial, no dose-limiting toxicity up to 10 mg/kg<br />
was observed, and the side effects ranged from lymphopenia<br />
to myalgias to fatigue <strong>of</strong> grades 1 and 2. 22 One patient<br />
developed grade 3 inflammatory colitis after five doses,<br />
which responded to steroids and infliximab. One patient<br />
developed grade 2 hypothyroidism requiring hormone replacement.<br />
Two patients developed grade 2 polyarticular<br />
arthropathies requiring oral steroids; all three <strong>of</strong> the latter<br />
could be considered irAEs. In an ongoing phase I/II trial <strong>of</strong><br />
every two weekly BMS-936558 with a peptide vaccine, no<br />
maximum tolerated dose was reached up to 10 mg/kg and
CHECKPOINT PROTEIN INHIBITOR TOXICITY MANAGEMENT<br />
hypothyroidism <strong>of</strong> grade 2 was observed; one patient developed<br />
grade 3 colitis, and one patient had dose-limiting optic<br />
neuritis, both <strong>of</strong> which were felt to be irAEs.<br />
Another PD-1 antibody, CT-011, has been tested in patients<br />
with hematologic malignancies at doses up to 6 mg/kg<br />
with dose-limiting side effects. 23 Only diarrhea was observed<br />
in more than one patient, along with drug-induced<br />
weakness, rash, and flushing <strong>of</strong> grades 1 and 2.<br />
Are irAEs Associated with <strong>Clinical</strong> Benefit?<br />
A number <strong>of</strong> investigators have suggested that the occurrence<br />
<strong>of</strong> grade 3 to 4 irAEs is associated with overall<br />
responses to ipilimumab. In patients with melanoma and<br />
Author’s Disclosure <strong>of</strong> Potential Conflicts <strong>of</strong> Interest<br />
Author<br />
Employment or<br />
Leadership<br />
Positions<br />
Consultant or<br />
Advisory Role<br />
Jeffrey S. Weber Altor<br />
BioScience;<br />
Celldex; Genesis<br />
Biopharma<br />
renal cancer who received ipilimumab at doses ranging from<br />
3 to 9 mg/kg every 3 weeks, there was a clear association<br />
between development <strong>of</strong> irAEs, chiefly GI- and endocrinerelated,<br />
as well as response to ipilimumab. 24 In several<br />
small adjuvant trials with ipilimumab, dose-limiting grade<br />
2, 3, and 4 irAEs were both dose-related and associated with<br />
prolonged relapse-free survival. 25 The results <strong>of</strong> phase II<br />
studies in metastatic melanoma have been mixed, with<br />
several trials showing a modest association <strong>of</strong> dose-limiting<br />
irAEs and overall response rate, but others contradicting<br />
that observation. Interestingly, the use <strong>of</strong> steroids to treat<br />
irAEs has not been shown to impact on the benefit <strong>of</strong><br />
ipilimumab. 8,9<br />
Stock<br />
Ownership Honoraria<br />
Altor<br />
BioScience;<br />
Celldex; Genesis<br />
Biopharma<br />
1. Hodi FS, O’Day SJ, McDermott DF, et al. Improved survival with<br />
ipilimumab in patients with metastatic melanoma. N Engl J Med. 2010;363:<br />
711-723.<br />
2. Robert C, Thomas L, Bondarenko I, et al. Ipilimumab plus dacarbazine<br />
for previously untreated metastatic melanoma. N Engl J Med. 2011;364:2517-<br />
2526.<br />
3. Weber J. Ipilimumab: Controversies in its development, utility, and<br />
autoimmune adverse events. Cancer Immunol Immunother. 2009;58:823-830.<br />
4. Hodi FS, Mihm MC, Soiffer RJ, et al. Biologic activity <strong>of</strong> cytotoxic T<br />
lymphocyte-associated antigen 4 antibody blockade in previously vaccinated<br />
metastatic melanoma and ovarian carcinoma patients. Proc Natl Acad Sci<br />
USA.2003;100:4712-4717.<br />
5. Johnston RL, Lutzky J, Chodhry A, et al. Cytotoxic T-lymphocyteassociated<br />
antigen 4 antibody-induced colitis and its management with<br />
infliximab. Dig Dis Sci. 2009;54:2538-2540.<br />
6. Lebbé C,O�Day S, Chiarion Sileni V, et al. Analysis <strong>of</strong> the onset and<br />
resolution <strong>of</strong> immune-related adverse events during treatment with ipilimumab<br />
in patients with metastatic melanoma. Presented at Perspectives in<br />
Melanoma XII, The Hague, Netherlands, October 2-4, 2008.<br />
7. Wolchok JD, Neyns B, Linette G, et al. Ipilimumab monotherapy in<br />
patients with pretreated advanced melanoma: A randomised, double-blind,<br />
multicentre, phase 2, dose-ranging study. Lancet Oncol. 2010;11:155-164.<br />
8. Phan GQ, Yang JC, Sherry RM, et al. Cancer regression and autoimmunity<br />
induced by cytotoxic T lymphocyte-associated antigen 4 blockade in<br />
patients with metastatic melanoma. Proc Natl Acad Sci USA.2003;100:<br />
8372-8377.<br />
9. Attia P, Phan GQ, Maker AV, et al. Autoimmunity correlates with tumor<br />
regression in patients with metastatic melanoma treated with anticytotoxic<br />
T-lymphocyte antigen-4. J Clin Oncol. 2005;23:6043-6053.<br />
10. Beck KE, Blansfield JA, Tran KQ, et al. Enterocolitis in patients with<br />
cancer after antibody blockade <strong>of</strong> cytotoxic T-lymphocyte-associated antigen<br />
4. J Clin Oncol. 2006;24:2283-2289.<br />
11. Weber J, Thompson JA, Hamid O, et al. A randomized, double-blind,<br />
placebo-controlled, phase II study comparing the tolerability and efficacy <strong>of</strong><br />
ipilimumab administered with or without prophylactic budesonide in patients<br />
with unresectable stage III or IV melanoma. Clin Cancer Res. 2009;15:5591-<br />
5598.<br />
12. Oble DA, Mino-Kenudson M, Goldsmith J, et al. Alpha-CTLA-4 mAbassociated<br />
panenteritis: A histologic and immunohistochemical analysis.<br />
Am J Surg Pathol. 2008;32:1130-1137.<br />
13. Smith FO, G<strong>of</strong>f SL, Klapper JA, et al. Risk <strong>of</strong> bowel perforation in<br />
patients receiving interleukin-2 after therapy with anti-CTLA 4 monoclonal<br />
antibody. J Immunother. 2007;30:130.<br />
REFERENCES<br />
Bristol-Myers<br />
Squibb; Celldex;<br />
Genentech;<br />
Merck;<br />
Prometheus<br />
Research<br />
Funding<br />
Novartis<br />
Expert<br />
Testimony<br />
Other<br />
Remuneration<br />
14. Berman D, Parker SM, Siegel J, et al. Blockade <strong>of</strong> cytotoxic<br />
T-lymphocyte antigen-4 by ipilimumab results in dysregulation <strong>of</strong> gastrointestinal<br />
immunity in patients with advanced melanoma. Cancer Immun.<br />
2010;10:11.<br />
15. O’Day SJ, Maio M, Chiarion-Sileni V, et al. Efficacy and safety <strong>of</strong><br />
ipilimumab monotherapy in patients with pretreated advanced melanoma:<br />
A multicenter single-arm phase II study. Ann Oncol. 2010;21:1712-1717.<br />
16. Blansfield JA, Beck KE, Tran K, et al. Cytotoxic T-lymphocyteassociated<br />
antigen-4 blockage can induce autoimmune hypophysitis in patients<br />
with metastatic melanoma and renal cancer. J Immunother. 2005;28:<br />
593-598.<br />
17. Dillard T, Yedinak CG, Alumkal J, et al. Anti-CTLA-4 antibody therapy<br />
associated autoimmune hypophysitis: Serious immune related adverse events<br />
across a spectrum <strong>of</strong> cancer subtypes. Pituitary. 2010;13:29-38.<br />
18. Carpenter KJ, Murtagh RD, Lilienfeld H, et al. Ipilimumab-induced<br />
hypophysitis: MR imaging findings. AJNR Am J Neuroradiol. 2009;30:1751-<br />
1753.<br />
19. Min L, Vaidya A, Becker C. Ipilimumab therapy for advanced melanoma<br />
is associated with secondary adrenal insufficiency: A case series.<br />
Endocr Pract. Epub 2011 Dec 2.<br />
20. Chmiel KD, Suan D, Liddle C, et al. Resolution <strong>of</strong> severe ipilimumabinduced<br />
hepatitis after antithymocyte globulin therapy. J Clin Oncol. 2011;<br />
29:e237-e240.<br />
21. Chin K, Ibrahim R, Berman D, et al. Treatment guidelines for the<br />
management <strong>of</strong> immune-related adverse events in patients treated with<br />
ipilimumab, an anti-CTLA4 therapy. Ann Oncol. 2008;19 (suppl 8):787.<br />
22. Brahmer JR, Drake CG, Wollner I, et al. Phase I study <strong>of</strong> single-agent<br />
anti-programmed death-1 (MDX-1106) in refractory solid tumors: Safety,<br />
clinical activity, pharmacodynamics, and immunologic correlates. J Clin<br />
Oncol. 2010;28:3167-3175.<br />
23. Berger R, Rotem-Yehudar R, Slama G, et al. Phase I safety and<br />
pharmacokinetic study <strong>of</strong> CT-011, a humanized antibody interacting with<br />
PD-1, in patients with advanced hematologic malignancies. Clin Cancer Res.<br />
2008;14:3044-3051.<br />
24. Downey SG, Klapper JA, Smith FO, et al. Prognostic factors related to<br />
clinical response in patients with metastatic melanoma treated by CTLassociated<br />
antigen-4 blockade. Clin Cancer Res. 2007;13:6681-6688.<br />
25. Sarnaik AA, Yu B, Yu D, et al. Extended dose ipilimumab with a<br />
peptide vaccine: Immune correlates associated with clinical benefit in patients<br />
with resected high-risk stage IIIc/IV melanoma. Clin Cancer Res.<br />
2011;17:896-906.<br />
177
TARGETING CRITICAL MOLECULAR ABERRATIONS<br />
EARLY IN THE COURSE OF SOLID TUMORS:<br />
IS IT ABOUT TIME?<br />
CHAIR<br />
Razelle Kurzrock, MD<br />
University <strong>of</strong> Texas M. D. Anderson Cancer Center<br />
Houston, TX<br />
SPEAKERS<br />
Hagop Kantarjian, MD<br />
University <strong>of</strong> Texas M. D. Anderson Cancer Center<br />
Houston, TX<br />
Laura Esserman, MD, MBA<br />
University <strong>of</strong> California, San Francisco<br />
San Francisco, CA
Treatment <strong>of</strong> Chronic Myelogenous Leukemia<br />
as a Paradigm for Solid Tumors: How<br />
Targeted Agents in Newly Diagnosed Disease<br />
Transformed Outcomes<br />
By Jason R. Westin, MD, Hagop Kantarjian, MD, and Razelle Kurzrock, MD<br />
Overview: Although chronic myelogenous leukemia (CML) is<br />
rare, with approximately 5000 new cases in the United States<br />
annually, it may be the poster child for the future <strong>of</strong> oncology.<br />
Imatinib mesylate, a selective Bcr-Abl tyrosine kinase inhibitor<br />
(TKI), transformed the course <strong>of</strong> CML from a rapidly fatal<br />
disease (median survival, 3 to 6 years) to a functionally<br />
curable, indolent disease with an estimated median survival <strong>of</strong><br />
more than 25 years. This transformation can be attributed to<br />
several key factors: the identification <strong>of</strong> a causal and actionable<br />
molecular aberration—BCR-ABL; the development <strong>of</strong> a<br />
potent and selective Bcr-Abl TKI—imatinib; and, importantly<br />
the application <strong>of</strong> imatinib in the earliest phase <strong>of</strong> CML. In<br />
contrast, imatinib, if used in CML blastic phase, improves<br />
median survival to only about 1 year. Similar to CML blastic<br />
TARGETED THERAPY in cancer started long before the<br />
imatinib era in chronic myelogenous leukemia (CML).<br />
Examples <strong>of</strong> targeted therapies include hormone therapy<br />
(tamoxifen) in estrogen- and progesterone-receptor positive<br />
breast cancers, monoclonal antibodies in leukemias and<br />
lymphomas (rituximab, gemtuzumab ozogamycin), and others.<br />
However, the era <strong>of</strong> targeted therapy in cancer is<br />
identified with imatinib as a result <strong>of</strong> several factors: 1) deciphering<br />
the Philadelphia chromosome (Ph)-associated<br />
and BCR-ABL-associated molecular abnormalities in CML;<br />
2) defining a causal association between the BCR-ABL<br />
molecular events and the development <strong>of</strong> CML in animal<br />
models, thus proving that Bcr-Abl is a legitimate target for<br />
therapeutic interventions; 3) discovering a relatively nontoxic,<br />
orally available selective Bcr-Abl inhibitor, imatinib<br />
mesylate; and 4) demonstrating the clinical efficacy <strong>of</strong> imatinib.<br />
After positive clinical trials were reported, media<br />
outlets suggested this new “magic bullet” could represent “A<br />
New Hope For Cancer” and signal the dawn <strong>of</strong> a new phase<br />
in the war on cancer. 1<br />
Left untreated, CML will ineluctably progress from the<br />
chronic phase to the accelerated phase, and eventually<br />
transform to CML blastic phase. 2 This evolution is analogous<br />
to the natural disease course believed to occur in most<br />
cancers. In the time before imatinib, therapies for CML<br />
improved blood counts but achieved complete cytogenetic<br />
responses (0% Ph-positive [Ph�] metaphases in the bone<br />
marrow) in only 15% <strong>of</strong> patients and improved median<br />
survival from 3 years with busulphan or hydroxyurea to only<br />
5–7 years with interferon alpha-based therapies. In contrast,<br />
when treating the earliest phases <strong>of</strong> CML with a<br />
therapy—imatinib—specific for the critical genetic abnormality,<br />
the cumulative complete cytogenetic response rate<br />
was greater than 80%, and the estimated 10-year survival<br />
increased from less than 20% to 85% or more. Indeed, at<br />
present, the estimated median survival <strong>of</strong> patients with<br />
CML exceeds 25 years. 3-7 Because the median age <strong>of</strong> patients<br />
with CML at the time <strong>of</strong> diagnosis is nearly 60 years,<br />
it is now estimated that most patients with CML have a<br />
normal life expectancy if treated appropriately with Bcr-Abl<br />
phase, metastatic solid malignancies have undergone genetic<br />
evolution, and their molecular aberrations are complex. As a<br />
result, resistance is common and eradication is difficult. The<br />
key to the dramatic improvement in the survival <strong>of</strong> patients<br />
with CML involved using imatinib in newly diagnosed disease,<br />
before blastic transformation. We hypothesize that metastatic<br />
solid tumors are analogous to CML blastic phase, and that to<br />
achieve improvements in solid tumor outcomes similar to<br />
those seen in CML, application <strong>of</strong> targeted agents to newly<br />
diagnosed disease may be required to prevent disease transformation<br />
(i.e., metastases). Targeting driver mutations at the<br />
time <strong>of</strong> diagnosis may be critical to the goal <strong>of</strong> markedly<br />
changing the outlook for patients with cancer.<br />
tyrosine kinase inhibitors (TKIs). These patients are therefore<br />
“functionally cured.” Imatinib changed the course <strong>of</strong><br />
CML from a previously fatal disease (median survival, 3 to 6<br />
years) to a functionally and molecularly curable disorder.<br />
Patients diagnosed with chronic-phase CML today may be<br />
reassured, with our current knowledge, that most can expect<br />
to live decades and will likely die with, and not <strong>of</strong>, CML—<br />
provided they comply with TKI therapy, there is a reasonable<br />
observation for progression <strong>of</strong> the disease, and an<br />
appropriate change is made to second- and third-generation<br />
TKIs if disease resistance evolves.<br />
Imatinib was approved by the Food and Drug Agency in<br />
2001 for treatment <strong>of</strong> Ph� CML. During the decade to<br />
follow, a sea change occurred in oncology with the approval<br />
<strong>of</strong> dozens <strong>of</strong> agents to be used for targetable cancerassociated<br />
molecular aberrations in both hematologic and<br />
solid tumor malignancies. Unfortunately, in solid tumors,<br />
these new targeted therapies have to date not achieved the<br />
expectations set by imatinib in CML. Therefore, we believe a<br />
review <strong>of</strong> the CML experience is required to potentially<br />
apply the lessons learned to solid tumors.<br />
Lessons from the CML Experience<br />
In hindsight, the positive results <strong>of</strong> imatinib therapy in<br />
CML seem now preordained. However, at the start <strong>of</strong> the<br />
imatinib trials, a common prediction shared by some CML<br />
experts was that 1) CML is a heterogeneous disease with<br />
multiple molecular abnormalities with increasing prevalence<br />
in advanced CML; 2) although early trials appeared<br />
favorable, CML cells will ultimately develop resistance and<br />
eventual late disease transformation even with early Bcr-<br />
From the Department <strong>of</strong> Leukemia; Department <strong>of</strong> Lymphoma and Myeloma; Department<br />
<strong>of</strong> Investigational Cancer Therapeutics, University <strong>of</strong> Texas M. D. Anderson Cancer Center,<br />
Houston, TX.<br />
Authors’ disclosures <strong>of</strong> potential conflicts <strong>of</strong> interest are found at the end <strong>of</strong> this article.<br />
Address reprint requests to Razelle Kurzrock, MD, M. D. Anderson Cancer Center, 1400<br />
Holcombe Blvd., Houston, TX 77030; email: rkurzro@mdanderson.org.<br />
© <strong>2012</strong> by <strong>American</strong> <strong>Society</strong> <strong>of</strong> <strong>Clinical</strong> <strong>Oncology</strong>.<br />
1092-9118/10/1-10<br />
179
Abl suppression; and 3) the early enthusiasm regarding the<br />
positive results with imatinib should be tempered because<br />
historic experiences in oncology have <strong>of</strong>ten resulted in disappointing<br />
long-term results.<br />
The long-term positive experience with imatinib in CML<br />
has in fact been beyond our original expectations. Currently,<br />
the annual mortality related to CML is reduced from the<br />
historic rate <strong>of</strong> 10% to 20% to less than 1%. We also learned<br />
multiple important lessons that were unanticipated based<br />
on previous cancer treatment experience: 1) the importance<br />
<strong>of</strong> early therapy with targeted agents to achieve optimal<br />
outcome; 2) the importance <strong>of</strong> an optimal biologic dose <strong>of</strong> the<br />
targeting agent, rather than a classical maximum-tolerated<br />
dose; 3) the rate <strong>of</strong> transformation was low, and if it<br />
occurred, it did so early rather than late; 4) the development<br />
<strong>of</strong> mutations in the targeted kinase as a resistant mechanism<br />
to TKIs; and 5) rationally designed and more potent<br />
KEY POINTS<br />
Table 1. Comparison <strong>of</strong> Response Rates with Matched Targeted Agents in CML and Solid Tumors<br />
● Chronic myelogenous leukemia (CML) is the poster<br />
child for the future <strong>of</strong> oncology, having been transformed<br />
from a disease that was rapidly fatal (median<br />
survival <strong>of</strong> 3 to 6 years) to one that is functionally<br />
cured (estimated survival <strong>of</strong> 25 years).<br />
● In CML blastic phase (advanced stage), responses to<br />
imatinib are transient, development <strong>of</strong> resistance<br />
common, and survival is still only 1 year.<br />
● Targeted therapies in solid tumors are typically evaluated<br />
in patients with advanced disease and achieve<br />
responses similar to imatinib in CML blastic phase.<br />
● The effectiveness <strong>of</strong> targeted therapies in solid tumors<br />
may be underestimated by evaluation at a stage<br />
analogous to CML blastic phase.<br />
● In order to realize the same response outcomes in<br />
solid tumors as have occurred in CML, it may be<br />
necessary to use the high successful strategy <strong>of</strong> treating<br />
the earliest stage <strong>of</strong> the disease with targeted<br />
therapy.<br />
Complete Cytogenetic<br />
Response Rate (%) with<br />
Bcr-Abl TKIs* References<br />
CML blastic phase 0 to 30 8–11, 17<br />
CML accelerated phase 17 to 24 12<br />
CML relapsed chronic phase 13–49 3, 24<br />
CML Newly-diagnosed chronic phase 65 to �80 3,14,15<br />
Response Rate (%)<br />
with vemurafenib<br />
(BRAF inhibitor)<br />
Partial Complete<br />
Melanoma: relapsed metastatic V600E BRAF mutation 75 6 26<br />
Melanoma: untreated metastatic V600E BRAF mutation 47# 1# 32<br />
PR rate with crizotinib<br />
(ALK inhibitor)<br />
CR rate with crizotinib<br />
(ALK inhibitor)<br />
Lung Cancer: relapsed, advanced, ALK-rearranged 56% 1% 27<br />
Abbreviations: CML, chronic myelogenous leukemia; TKI, tyrosine kinase inhibitor.<br />
* Bcr-ABl TKIs include imatinib, dastatinib and nilotinib, #: Response rate reported with median follow up <strong>of</strong> 3.8 months, tumor regression rate was approximately<br />
95%.<br />
180<br />
WESTIN, KANTARJIAN, AND KURZROCK<br />
TKIs can be developed to overcome some <strong>of</strong> the mechanisms<br />
<strong>of</strong> CML resistance to imatinib.<br />
Importance <strong>of</strong> Treating Newly Diagnosed CML<br />
The use <strong>of</strong> targeted agents in the earliest phases <strong>of</strong> the<br />
disease is likely the most important lesson from the CML<br />
experience that could be directly applied to other solid<br />
tumors. With imatinib and other TKIs (nilotinib, dasatinib),<br />
therapeutic outcome is most strongly associated with the<br />
phase <strong>of</strong> CML (ie, chronic versus blastic phase) during which<br />
TKI therapy is introduced (Table 1). In patients with blastic<br />
phase CML, TKI therapy produces complete cytogenetic<br />
response rates <strong>of</strong> only 10% to 20%, with transient cytogenetic<br />
responses and a median survival that increased from a<br />
median <strong>of</strong> 3–6 months to 9–18 months. Invariably, in most<br />
patients with blastic-phase CML resistance develops rapidly<br />
to TKI therapy and patients die from complications <strong>of</strong> their<br />
disease. 8-11 These poor responses to TKI are assumed to be<br />
related to the multiple additional resistance mechanisms<br />
(cytogenetic clonal evolution, mutations, and multiple parallel<br />
and additional molecular abnormalities causing CML<br />
transformation).<br />
Even delaying therapy moderately so that it is given in<br />
late chronic phase or accelerated phase markedly compromises<br />
complete cytogenetic response rates. For instance,<br />
patients with newly diagnosed chronic phase CML have a<br />
complete cytogenetic response rate <strong>of</strong> over 80% with Bcr-Abl<br />
TKI treatment. 7 The cytogenetic response rates drop to<br />
approximately 40% in patients with previously treated<br />
chronic phase CML. 3 By the time disease is in acceleratedphase<br />
CML, complete cytogenetic responses occur in approximately<br />
20% <strong>of</strong> cases. 12,13<br />
In addition, deeper levels <strong>of</strong> (molecular) responses are<br />
achieved at a greater frequency in early chronic-phase CML<br />
compared with more advanced disease, attesting to the<br />
increased efficacy <strong>of</strong> TKI therapies in early chronic-phase<br />
CML. Achieving a major molecular response (defined as<br />
BCR-ABL transcript levels less than 0.1%) is noted in 40% to<br />
70% <strong>of</strong> patients with newly diagnosed CML, with higher<br />
molecular response rates observed with the more potent<br />
second-generation TKIs (eg, dasatinib and nilotinib) compared<br />
with imatinib. 14,15 The major molecular response rate<br />
rates were lower in accelerated-phase CML (20% to 30%),
TARGETED TREATMENT OF NEWLY DIAGNOSED DISEASE<br />
and less than 10% in blastic-phase CML. Similarly, achievement<br />
<strong>of</strong> undetectable BCR-ABL levels (more than 4–4.5 log<br />
reduction <strong>of</strong> CML burden) was observed in 20% to 50% <strong>of</strong><br />
patients with newly diagnosed CML treated with TKI therapy<br />
but was uncommon in advanced CML.<br />
Optimal Dose <strong>of</strong> Targeted Therapy<br />
Historically, most anticancer agents have been tested at<br />
their maximum-tolerated dose, typically defined as a dose<br />
below the one that produced severe toxicities in less than<br />
one third <strong>of</strong> the patients. The assumption that the<br />
maximum-tolerated dose is optimal is based on observations<br />
that the highest possible tolerable dose would result in the<br />
greatest antitumor results. This assumption may not apply<br />
to targeted therapies, as was observed with both TKIs in<br />
CML and the response rates <strong>of</strong> targeted therapies in phase<br />
I trials. 16 The phase I study <strong>of</strong> imatinib noted rare complete<br />
cytogenetic responses at doses below 300 mg orally daily and<br />
frequent complete cytogenetic responses at doses ranging<br />
from 400 to 800 mg per day (on average, there were higher<br />
responses with 800 mg than 400 mg). 17 Interestingly, no<br />
further improvement in complete cytogenetic responses<br />
were seen at doses <strong>of</strong> 1000 mg daily and above, although<br />
these doses were reasonably tolerated by patients. The<br />
phase II study proceeded with imatinib doses <strong>of</strong> 400 mg<br />
orally daily based on the imatinib plasma levels that were<br />
anticipated to produce maximum anti-CML activity. Higher<br />
doses <strong>of</strong> imatinib (800 mg daily) were later observed to<br />
induce higher rates <strong>of</strong> complete cytogenetic and major molecular<br />
responses, but the additional benefit (as measured by<br />
survival) was controversial. The experience with nilotinib<br />
indicated that despite the long half-life <strong>of</strong> the drug, gastrointestinal<br />
absorption plateaued at a dose <strong>of</strong> 400 mg, resulting<br />
in schedules <strong>of</strong> 400 mg orally twice daily, implemented<br />
into phase II studies in late chronic-phase CML.<br />
The randomized studies <strong>of</strong> nilotinib compared with imatinib<br />
in newly diagnosed CML (Evaluating Nilotinib Efficacy<br />
and Safety in clinical Trials-Newly Diagnosed patients<br />
[ENEST-nd] trial) also compared a lower dose schedule <strong>of</strong><br />
nilotinib, 300 mg twice daily, with the approved standard<br />
dose <strong>of</strong> 400 mg twice daily, demonstrating the equal efficacy<br />
<strong>of</strong> the lower dose schedule and the benefit <strong>of</strong> a lower cost<br />
and toxicity. 14 Nilotinib 300 mg orally twice daily is now the<br />
standard dose schedule for newly diagnosed CML. The early<br />
studies with dasatinib suggested twice-daily schedules<br />
(based on the short half-life <strong>of</strong> the drug) and doses <strong>of</strong> 140 mg<br />
daily to be optimal. Subsequent randomized trials <strong>of</strong> four<br />
different schedules established the single-daily 100 mg dose<br />
<strong>of</strong> dasatinib to be the optimal schedule (equal efficacy, lower<br />
toxicity), which is now used as the standard <strong>of</strong> care in<br />
CML. 15 These studies suggest that treatment with optimal<br />
biologic dosages, as opposed to maximum-tolerated doses, <strong>of</strong><br />
TKIs in CML is best.<br />
This finding has subsequently been validated by the<br />
experience with epigenetic therapy (decitabine, azacitidine)<br />
in myelodysplastic syndrome. Following early studies <strong>of</strong><br />
these drugs in the 1980s at the maximum-tolerated clinical<br />
dosages, the drugs were almost abandoned until the 1990s,<br />
when the concept <strong>of</strong> epigenetic therapy was developed. 18<br />
Using optimal biologic dosages <strong>of</strong> the drugs, decitabine 50 to<br />
100 mg per m 2 per course (instead <strong>of</strong> 1000 to 2500 mg per m 2<br />
per course) and azacitidine 250 to 525 mg per m 2 per course<br />
(instead <strong>of</strong> higher dosages), resulted in high efficacy results<br />
and acceptable toxicities and established these agents at the<br />
optimal biologic dosages as new standards <strong>of</strong> care in myelodysplastic<br />
syndrome. 19<br />
Progression to Blastic Phase <strong>of</strong> CML Is Rare with<br />
TKI Treatment and Occurs Early Rather than Late<br />
The expectation during the International Randomized<br />
Study <strong>of</strong> Interferon and STI571 (IRIS) trial was that leukemic<br />
cells would develop resistance and transform into<br />
blastic-phase CML either at a uniform annual rate or later<br />
in the course <strong>of</strong> therapy. This expectation was based on<br />
the assumption that resistance develops under a timeassociated<br />
selective pressure from TKIs. Surprisingly, the<br />
opposite was in fact observed. 3 The annual transformation<br />
rate during imatinib therapy was low, but if transformation<br />
occurred, it happened early (2% to 4% annually in the first 3<br />
years, and less than 1% annually thereafter). These data are<br />
now interpreted to suggest that some patients with newly<br />
diagnosed chronic-phase CML may harbor a small subset(s)<br />
<strong>of</strong> CML clones with a priori TKI resistance, silently displaying<br />
the molecular resistance mechanisms <strong>of</strong> transformation<br />
that would not be influenced by imatinib therapy. In these<br />
patients transformation will develop within the first 2–3<br />
years <strong>of</strong> therapy. In contrast, most patients with newly<br />
diagnosed chronic-phase CML have clones that do not harbor<br />
intrinsic transformation mechanisms at diagnosis. In<br />
these patients, imatinib therapy will ultimately reduce the<br />
transformation rate to less than 1% per year.<br />
Mutations as a Mechanism <strong>of</strong> Resistance in CML<br />
During imatinib therapy, clinical resistance will develop<br />
at a rate <strong>of</strong> 2% to 4% annually. Mutations at the BCR-ABL<br />
kinase domain were noted in 30% to 40% <strong>of</strong> patients with<br />
resistance in chronic phase and in more than 50% <strong>of</strong> patients<br />
with resistance in transformation. 20 Of note, a portion <strong>of</strong><br />
these mutations would alone be insufficient to result in<br />
resistance to imatinib therapy (suggesting that other mechanisms<br />
<strong>of</strong> resistance were operational). Other mutations had<br />
intermediate to high resistance to imatinib but were sensitive<br />
to second-generation TKIs. A pan-TKI resistant mutation,<br />
T315I (also referred to as “gatekeeper” mutation), was<br />
detected in 20% <strong>of</strong> mutations (5% to 10% <strong>of</strong> resistant cases),<br />
was resistant to second-generation TKIs (eg, dasatinib,<br />
nilotinib, bosutinib), but was sensitive to several thirdgeneration<br />
TKIs including ponatinib. 21 This experience suggests<br />
that in CML, and possibly in solid tumors, resistance<br />
to targeted therapy may develop through the selective pressure<br />
<strong>of</strong> the targeted agent, resulting in mutations at the<br />
target site that prevent the targeted agent from exerting<br />
its effect. This phenomenon has now been observed with<br />
FLT3 inhibitors (development <strong>of</strong> FLT3 point mutations<br />
associated with resistance to FLT3 inhibitors) and in patients<br />
with solid tumors treated with targeted agents (eg,<br />
development <strong>of</strong> mutations in the epidermal growth factor<br />
receptor [EGFR] associated with resistance to EGFR TKI<br />
therapy). 22,23<br />
Rationally Designed More Potent TKIs Overcome CML<br />
Resistance to Imatinib<br />
The development <strong>of</strong> mutations at the Bcr-Abl kinase<br />
domain sites in TKI-treated CML had been predicted in<br />
vitro from selection pressure models that identified several<br />
181
50%<br />
50%<br />
100%<br />
Change in Tumor Size<br />
mutations later confirmed in vivo. This discovery resulted in<br />
the development <strong>of</strong> a novel generation <strong>of</strong> TKIs that are<br />
highly effective in suppressing the mutated CML clones<br />
and have demonstrated clinical efficacy in patients with<br />
imatinib-resistant CML. Second-generation TKIs resulted<br />
in complete cytogenetic response rates <strong>of</strong> 50% or more in<br />
such patients; these responses were durable and reduced<br />
the post-imatinib resistance mortality rate from 10% to 15%<br />
annually to 5% or less. 24,25 These results encouraged the<br />
front-line use <strong>of</strong> these agents, which when compared with<br />
imatinib, produced higher rates <strong>of</strong> complete cytogenetic<br />
response, major molecular response, and complete molecular<br />
response. 14,15 More importantly, they were associated with<br />
lower rates <strong>of</strong> transformation to the accelerated and blastic<br />
phases <strong>of</strong> CML. 14,15 Together, these data suggest that the<br />
earlier use <strong>of</strong> more potent TKIs could further suppress the<br />
inherent mechanisms <strong>of</strong> resistance <strong>of</strong> CML clones at diagnosis<br />
in at least some patients. Furthermore, the successful<br />
development <strong>of</strong> potent T315I inhibitors demonstrated their<br />
high efficacy among patients with T315I-mutated CML.<br />
Such experiences should also be considered in the context <strong>of</strong><br />
research and solid tumors.<br />
Why Are We Not Achieving Similar Results in<br />
Solid Tumors?<br />
The development <strong>of</strong> imatinib for the treatment <strong>of</strong> CML is<br />
considered a paradigm for targeted therapy. Despite numerous<br />
promising drugs and herculean efforts by expert oncologists,<br />
to date the CML/imatinib success remains<br />
unmatched. At least two possibly valid conclusions can be<br />
reached: 1) the CML/imatinib experience is singular (i.e.,<br />
CML is so genetically simple and homogeneous that it is not<br />
a relevant comparison to solid tumors) or 2) the use <strong>of</strong><br />
targeted therapy in solid tumors has not yet followed the<br />
highly successful strategy used by CML investigators.<br />
Although targeted therapies in solid tumors have not yet<br />
transformed disease outcomes to match a normal life span,<br />
they certainly cannot be described as ineffective. Therapies<br />
targeting the driving molecular abnormalities in solid tu-<br />
???<br />
Survival Outcome<br />
Fig 1. Left panel shows typical waterfall plot for successful targeted agent in metastatic solid tumors. Many patients show tumor regression<br />
but few achieve complete remission.<br />
Upper right hand panel shows typical survival curve for patients with metastatic solid tumors treated with an active targeted agent. Even with<br />
improved survival, most patients die <strong>of</strong> their disease, with survival gains generally measured in weeks to months. Survival curves for patients<br />
with CML blast phase remain similar, even in the imatinib era.<br />
Lower right hand panel shows typical survival curve for patients with newly-diagnosed CML chronic phase treated with imatinib or other<br />
Bcr-ABl TKIs. High rates <strong>of</strong> long-term survival are achieved.<br />
182<br />
% Surviving<br />
Time<br />
mors, such as vemurafenib for V600E BRAF mutations in<br />
melanoma and crizotinib in ALK-rearranged lung cancer,<br />
have resulted in substantial tumor response rates in preliminary<br />
trials. In a landmark phase I trial, Flaherty and<br />
colleagues 26 demonstrated an overall response rate <strong>of</strong> 81%<br />
(26 <strong>of</strong> 32; two complete responses and 24 partial responses)<br />
with the BRAF inhibitor vemurafenib given at a dose <strong>of</strong> 960<br />
mg twice daily in the dose-expansion cohort <strong>of</strong> heavily<br />
pretreated patients with mutated V600E BRAF metastatic<br />
melanoma. Perhaps even more impressive, the doseescalation<br />
phase demonstrated a 69% response rate (11 <strong>of</strong><br />
16; one complete response and 10 partial responses) with<br />
vemurafenib at doses <strong>of</strong> 240 mg or more twice daily. Unfortunately,<br />
disease control was transient in six <strong>of</strong> the 11<br />
patients with response, and disease progressed within 9<br />
months after the initiation <strong>of</strong> therapy. In the expansion<br />
phase <strong>of</strong> the phase I trial <strong>of</strong> the ALK inhibitor crizotinib, the<br />
response rate for patients with advanced, pretreated ALKrearranged<br />
lung cancer was 57% (47 <strong>of</strong> 82; one complete<br />
response and 46 partial responses). 27 These therapies were<br />
largely well tolerated, especially when compared with standard<br />
cytotoxic chemotherapy. Although exciting, the data<br />
indicate that these targeted therapies do not commonly<br />
result in complete responses, that “cures” are rare, and that<br />
patients who have response will likely have relapse. At first<br />
glance, these results may be perceived as disappointing<br />
compared with the CML experience. However, on closer<br />
observation, they have striking parallels. A portion <strong>of</strong> patients<br />
with advanced disease, either CML or solid tumors,<br />
benefit from targeted therapy, but resistance inevitably<br />
develops (Fig. 1).<br />
Importance <strong>of</strong> Early Therapy<br />
WESTIN, KANTARJIAN, AND KURZROCK<br />
In CML, the timing <strong>of</strong> targeted therapy is critical: patients<br />
with newly-diagnosed CML do much better than those<br />
with advanced or transformed disease (Table 1). As in CML,<br />
it is generally accepted that newly-diagnosed solid tumors<br />
are more responsive to therapy than previously treated or<br />
relapsed cancers. Because <strong>of</strong> genomic instability, faulty
TARGETED TREATMENT OF NEWLY DIAGNOSED DISEASE<br />
DNA repair, or various selection pressures, advanced cancers<br />
accumulate multiple genetic and/or molecular derangements<br />
over time, conferring various phenotypic advantages<br />
on each subpopulation. 28,29 Metastatic tumors from breast,<br />
lung, and gastrointestinal cancers have more genetic abnormalities<br />
than do their primary tumors, and <strong>of</strong>ten display<br />
considerable heterogeneity compared with primary lesions,<br />
including discordant clinically utilized biomarkers. 30,31<br />
A counter argument to the conventional wisdom that solid<br />
tumors are genetically more complex and heterogenous than<br />
CML is the efficacy <strong>of</strong> targeted agents (e.g., vemurafenib and<br />
crizotinib) in selected patients with advanced disease. In the<br />
phase III vemurafenib trial, the overwhelming majority <strong>of</strong><br />
previously untreated patients with advanced melanoma had<br />
tumor regression, although 48% <strong>of</strong> the evaluable patients<br />
met criteria for response. 32<br />
An example <strong>of</strong> response rates <strong>of</strong> targeted therapy in<br />
advanced solid tumors degrading with increasing stage <strong>of</strong><br />
disease is lapatinib, an oral dual EGFR-erbB2 TKI. When<br />
evaluated in patients with newly diagnosed metastatic<br />
HER2-positive breast cancer, lapatinib achieved an response<br />
rate <strong>of</strong> 24%. 33 Among patients with previously<br />
treated, metastatic HER2-positive breast cancer, the response<br />
rate declined to 7%. 34 These trials cannot be directly<br />
compared but do suggest an improvement in response rates<br />
with earlier use <strong>of</strong> targeted therapy, similar to the CML<br />
experience.<br />
Thus, although it appears likely that targeted agents<br />
perform better in untreated patients than patients with<br />
relapsed metastatic disease, we believe that this experience<br />
cannot be compared with newly diagnosed CML, as untreated<br />
metastatic solid tumors are analogous to untreated<br />
CML blastic phase. Interestingly, the results attained in<br />
metastatic solid tumors (either untreated or previously<br />
treated) are remarkably similar to those achieved in patients<br />
with CML blastic phase (either untreated or previously<br />
treated): 1) many patients have some regression; 2) a<br />
partial or complete response is achieved in a small subset <strong>of</strong><br />
patients ; 3) resistance develops in most patients; and 4) the<br />
increase in survival can be measured in months, not years.<br />
Mutations as a Mechanism <strong>of</strong> Resistance in<br />
Solid Tumors<br />
Although other pathways may be relevant for disease<br />
progression, much <strong>of</strong> the resistance to imatinib in CML is<br />
due to BCR-ABL kinase domain mutations. 35,36 Resistance<br />
to targeted therapy <strong>of</strong> solid tumors is less well characterized,<br />
but preliminary evidence suggests a similar phenomenon<br />
may occur, i.e., mutation <strong>of</strong> the original critical target. 37,38 It<br />
is not clear if this dependence on the initial oncologic driver<br />
is more pronounced early, or is constant throughout the<br />
disease course. In CML blastic phase, the minimal and<br />
transient efficacy <strong>of</strong> Bcr-Abl TKIs, including those able to<br />
overcome many common mutations, would be an arguement<br />
for a substantial change in disease biology away from initial<br />
driver dependence with disease progression.<br />
When a crizotinib-resistant model <strong>of</strong> ALK-rearranged<br />
lung cancer was generated with serial exposure, the most<br />
common mechanisms <strong>of</strong> resistance were amplification <strong>of</strong><br />
the EML4-ALK gene and a “gatekeeper” mutation within<br />
the kinase domain. 37 As with CML, this finding suggests<br />
that the driving abnormality remains critical to the survival<br />
<strong>of</strong> the cancer cells, even in disease resistant to targeted<br />
therapy.<br />
In contrast, early studies <strong>of</strong> vemurafenib resistance in<br />
V600E mutant melanoma cell lines have not identified<br />
mutations in BRAF. 39 In lieu <strong>of</strong> mutating the kinase domain,<br />
these vemurafenib-resistant cells were able to activate<br />
the MAPK pathway in a BRAF-independent fashion,<br />
but remained sensitive to targeting the pathway farther<br />
downstream.<br />
The use <strong>of</strong> EGFR inhibitors in EGFR-mutated lung cancer<br />
provides corroborating evidence <strong>of</strong> both resistance mechanisms.<br />
Mutations in exon 19 and 21 <strong>of</strong> EGFR are known to<br />
be associated with increased sensitivity to EGFR TKIs. In<br />
previously untreated patients with advanced lung cancer<br />
who have an EGFR mutation, treatment with an EGFR<br />
inhibitor typically yields median progression-free survivals<br />
<strong>of</strong> approximately 9 months. 40 When resistance develops, the<br />
most common mechanisms are the development <strong>of</strong> a secondary<br />
point mutation in EGFR (T790M), which reduces the<br />
ability <strong>of</strong> TKIs to inhibit EGFR, or amplification <strong>of</strong> the MET<br />
gene. 38,40,41<br />
Together, these preliminary data suggest that a portion <strong>of</strong><br />
solid tumors may gave resistance mechanisms to targeted<br />
therapy similar to those in CML, even while there is continued<br />
dependence on the initial genetic driver. If so, a CMLlike<br />
approach <strong>of</strong> developing second-generation targeted<br />
agents with activity in resistant populations with mutated,<br />
yet functional target kinase, or in targeting the critical<br />
pathway at multiple points, could demonstrate efficacy by<br />
either resensitizing resistant disease or blocking development<br />
<strong>of</strong> resistance.<br />
Conclusion<br />
Cancer claims more than half a million lives annually and<br />
is now the leading cause <strong>of</strong> death under age 85 in the United<br />
States. 42 Globally, an estimated 12.7 million new cancer<br />
diagnoses and 7.6 million cancer deaths occurred in 2008,<br />
and these numbers are expected to nearly double in the next<br />
25 years. 43,44 In order to make a substantial impact on this<br />
global health crisis, we believe it is beneficial to re-examine<br />
one <strong>of</strong> oncology’s greatest success stories–imatinib in CML.<br />
Like solid tumors, CML accumulates new genetic aberrations<br />
throughout its course and eventually reaches a stage<br />
where therapy targeted against its driving aberration has<br />
markedly reduced effectiveness. It is our contention that<br />
CML blastic phase is the equivalent <strong>of</strong> metastatic disease in<br />
solid tumors. Both CML blastic phase and metastatic solid<br />
tumors are difficult to eradicate because <strong>of</strong> their molecular<br />
complexity. Deciphering this complexity and determining<br />
how to best treat it is perceived as a staggeringly complicated<br />
task. Indeed, although the survival <strong>of</strong> patients with<br />
CML blastic phase has increased in the imatinib era, it<br />
remains at about 1 year. Yet, outcomes in chronic phase<br />
CML have been transformed, with overall survival from<br />
diagnosis increasing from approximately 4 years to more<br />
than 25 years. As seen in CML, the true leap forward in<br />
outcomes occurred only after investigators used Bcr-Abl<br />
TKIs in the earliest stage <strong>of</strong> the disease. Importantly,<br />
delaying therapy even slightly, to later chronic phase and<br />
certainly to accelerated phase CML, is associated with a<br />
precipitous fall in response rates and much poorer outcomes.<br />
It is therefore plausible, if the CML paradigm holds true,<br />
that administering targeted therapies to patients with ad-<br />
183
vanced solid tumors, even if they are untreated, will continue<br />
to result in only incremental improvements. Newly<br />
diagnosed metastatic disease is already genetically complex<br />
and likely analogous to newly diagnosed CML blastic phase,<br />
and not to newly diagnosed chronic phase CML. Timing may<br />
therefore be a crucial missing component in the optimal<br />
application <strong>of</strong> targeted therapies to solid tumors. The les-<br />
Authors’ Disclosures <strong>of</strong> Potential Conflicts <strong>of</strong> Interest<br />
Author<br />
Employment or<br />
Leadership<br />
Positions<br />
Consultant or<br />
Advisory Role<br />
sons <strong>of</strong> CML suggest that once a targeted therapy has shown<br />
some activity in advanced disease, it may have the potential<br />
for a high rates <strong>of</strong> functional cures but only if given at the<br />
time <strong>of</strong> diagnosis <strong>of</strong> early stage disease. The current strategy<br />
<strong>of</strong> using targeted therapy for patients with metastatic solid<br />
tumors may therefore underestimate, perhaps dramatically,<br />
the ultimate effectiveness <strong>of</strong> active targeted agents.<br />
Stock<br />
Ownership Honoraria<br />
Research<br />
Funding<br />
Jason R. Westin*<br />
Hagop Kantarjian Novartis Bristol-Myers<br />
Squibb; Cyclacel;<br />
Genzyme;<br />
Novartis; Pfizer<br />
Razelle Kurzrock AstraZeneca;<br />
Health<br />
Advances;<br />
Johnson &<br />
Johnson; Merck;<br />
SAIC-Frederick<br />
*No relevant relationships to disclose.<br />
1. Lemonick MD, Park A. New hope for cancer. TIME. May, 28, 2001.<br />
2. Kurzrock R, Gutterman JU, Talpaz M. The molecular genetics <strong>of</strong><br />
Philadelphia chromosome-positive leukemias. N Engl J Med. 1988 Oct<br />
13;319:990-998.<br />
3. Deininger M, O’Brien, SG, Guilhot F, et al. International randomized<br />
study <strong>of</strong> interferon vs STI571 (IRIS) 8-year follow up: sustained survival and<br />
low risk for progression or events in patients with newly diagnosed chronic<br />
myeloid leukemia in chronic phase (CML-CP) treated with imatinib. Blood.<br />
2009;114:22 (suppl; abstr 1126).<br />
4. Druker BJ, Guilhot F, O’Brien SG, et al. Five-year follow-up <strong>of</strong> patients<br />
receiving imatinib for chronic myeloid leukemia. N Engl J Med. 2006;355:<br />
2408-2417.<br />
5. Kantarjian HM, Talpaz M, O’Brien S, et al. Survival benefit with<br />
imatinib mesylate versus interferon-alpha-based regimens in newly diagnosed<br />
chronic-phase chronic myelogenous leukemia. Blood. 2006;108:1835-<br />
1840.<br />
6. Björkholm M, Ohm L, Eloranta S, et al. Success story <strong>of</strong> targeted therapy<br />
in chronic myeloid leukemia: a population-based study <strong>of</strong> patients diagnosed<br />
in Sweden from 1973 to 2008. J Clin Oncol. 2011;29:2514-2520.<br />
7. Kantarjian H, O’Brien S, Jabbour E. Improved survival in chronic<br />
myeloid leukemia since the introduction <strong>of</strong> imatinib therapy: a singleinstitution<br />
historical experience. Blood. <strong>2012</strong>;119:1981-1987.<br />
8. Kantarjian HM, Cortes J, O’Brien S, et al. Imatinib mesylate (STI571)<br />
therapy for Philadelphia chromosome-positive chronic myelogenous leukemia<br />
in blast phase. Blood. 2002;99:3547-3553.<br />
9. Cortes J, Kim DW, Raffoux E, et al. Efficacy and safety <strong>of</strong> dasatinib in<br />
imatinib-resistant or -intolerant patients with chronic myeloid leukemia in<br />
blast phase. Leukemia. 2008;22:2176-2183.<br />
10. Giles FJ, Kantarjian HM, le Coutre PD, et al. Nilotinib is effective in<br />
imatinib-resistant or -intolerant patients with chronic myeloid leukemia in<br />
blastic phase. Leukemia. Epub 2011 Dec 13.<br />
11. Sawyers CL, Hochhaus A, Feldman E, et al. Imatinib induces hematologic<br />
and cytogenetic responses in patients with chronic myelogenous leukemia<br />
in myeloid blast crisis: results <strong>of</strong> a phase II study. Blood. 2002;99:3530-<br />
3539.<br />
12. Talpaz M, Silver RT, Druker BJ, et al. Imatinib induces durable<br />
hematologic and cytogenetic responses in patients with accelerated phase<br />
184<br />
REFERENCES<br />
Amgen;<br />
AstraZeneca;<br />
Enzon; Exelixis;<br />
Genentech;<br />
Johnson &<br />
Johnson;<br />
Maxygen;<br />
Merck; Myriad;<br />
Nereus; Pfizer;<br />
Pharmion;<br />
Roche; SAIC-<br />
Frederick<br />
Amgen;<br />
AmpliMed;<br />
AstraZeneca;<br />
Callisto;<br />
Centocor Ortho<br />
Biotech;<br />
CuraGen; Enzon;<br />
Exelixis;<br />
GlaxoSmithKline;<br />
Globomax;<br />
Merck; Myriad<br />
Genetics;<br />
Nereus;<br />
Novartis; Pfizer;<br />
Roche<br />
WESTIN, KANTARJIAN, AND KURZROCK<br />
Expert<br />
Testimony<br />
Other<br />
Remuneration<br />
Exelixis; Nova<br />
Research Co.;<br />
Pfizer; Roche<br />
chronic myeloid leukemia: results <strong>of</strong> a phase 2 study. Blood. 2002;99:1928-<br />
1937.<br />
13. Guilhot F, Apperley J, Kim DW, et al. Dasatinib induces significant<br />
hematologic and cytogenetic responses in patients with imatinib-resistant or<br />
-intolerant chronic myeloid leukemia in accelerated phase. Blood. 2007;109:<br />
4143-4150.<br />
14. Kantarjian HM, Hochhaus A, Saglio G, et al. Nilotinib versus imatinib<br />
for the treatment <strong>of</strong> patients with newly diagnosed chronic phase, Philadelphia<br />
chromosome-positive, chronic myeloid leukaemia: 24-month minimum<br />
follow-up <strong>of</strong> the phase 3 randomised ENESTnd trial. Lancet Oncol. 2011;12:<br />
841-851.<br />
15. Kantarjian HM, Shah NP, Cortes JE, et al. Dasatinib or imatinib in<br />
newly diagnosed chronic phase chronic myeloid leukemia: 2-year follow-up<br />
from a randomized phase 3 trial (DASISION). Blood. Epub 2011 Dec 9.<br />
16. Jain RK, Lee JJ, Hong D, et al. Phase I oncology studies: evidence that<br />
in the era <strong>of</strong> targeted therapies patients on lower doses do not fare worse. Clin<br />
Cancer Res. 2010;16:1289-1297.<br />
17. Druker BJ, Talpaz M, Resta DJ, et al. Efficacy and safety <strong>of</strong> a specific<br />
inhibitor <strong>of</strong> the BCR-ABL tyrosine kinase in chronic myeloid leukemia.<br />
N Engl J Med. 2001;344:1031-1037.<br />
18. Jabbour E, Issa JP, Garcia-Manero G, et al. Evolution <strong>of</strong> decitabine<br />
development: accomplishments, ongoing investigations, and future strategies.<br />
Cancer. 2008;112:2341-2351.<br />
19. Fenaux P, Mufti GJ, Hellstrom-Lindberg E, et al. Efficacy <strong>of</strong> azacitidine<br />
compared with that <strong>of</strong> conventional care regimens in the treatment <strong>of</strong><br />
higher-risk myelodysplastic syndromes: a randomised, open-label, phase III<br />
study. Lancet Oncol. 2009;10:223-232.<br />
20. Shah NP, Nicoll JM, Nagar B, et al. Multiple BCR-ABL kinase domain<br />
mutations confer polyclonal resistance to the tyrosine kinase inhibitor imatinib<br />
(STI571) in chronic phase and blast crisis chronic myeloid leukemia.<br />
Cancer Cell. 2002;2:117-125.<br />
21. Santos FP, Quintás-Cardama A. New drugs for chronic myelogenous<br />
leukemia. Curr Hematol Malig Rep. 2011;6:96-103.<br />
22. Heidel F, Solem FK, Breitenbuecher F, et al. <strong>Clinical</strong> resistance to the<br />
kinase inhibitor PKC412 in acute myeloid leukemia by mutation <strong>of</strong> Asn-676<br />
in the FLT3 tyrosine kinase domain. Blood. 2006;107:293-300.<br />
23. Wheeler DL, Dunn EF, Harari PM. Understanding resistance to EGFR
TARGETED TREATMENT OF NEWLY DIAGNOSED DISEASE<br />
inhibitors-impact on future treatment strategies. Nat Rev Clin Oncol. 2010;<br />
7:493-507.<br />
24. Hochhaus A, Baccarani M, Deininger M, et al. Dasatinib induces<br />
durable cytogenetic responses in patients with chronic myelogenous leukemia<br />
in chronic phase with resistance or intolerance to imatinib. Leukemia.<br />
2008;22:1200-1206.<br />
25. Kantarjian HM, Giles FJ, Bhalla KN, et al. Nilotinib is effective in<br />
patients with chronic myeloid leukemia in chronic phase after imatinib<br />
resistance or intolerance: 24-month follow-up results. Blood. 2011;117:1141-<br />
1145.<br />
26. Flaherty KT, Puzanov I, Kim KB, et al. Inhibition <strong>of</strong> mutated, activated<br />
BRAF in metastatic melanoma. N Engl J Med. 2010:363:809-819.<br />
27. Kwak EL, Bang YJ, Camidge DR, et al. Anaplastic lymphoma kinase<br />
inhibition in non-small-cell lung cancer. N Engl J Med. 2010;363:1693-1703.<br />
28. Bown NP. Chromosome studies <strong>of</strong> solid tumours. J Clin Pathol.<br />
1992;45:556-560.<br />
29. Nowell PC. Cytogenetics <strong>of</strong> tumor progression. Cancer. 1990: 65:2172-<br />
2177.<br />
30. Barlogie B, Johnston DA, Smallwood L, et al. Prognostic implications <strong>of</strong><br />
ploidy and proliferative activity in human solid tumors. Cancer Genet Cytogenet.<br />
1982;6:17-28.<br />
31. Curigliano G, Bagnardi V, Viale G, et al. Should liver metastases <strong>of</strong><br />
breast cancer be biopsied to improve treatment choice? Ann Oncol. 2011;<br />
22:2227-2233.<br />
32. Chapman PB, Hauschild A, Robert C, et al. Improved survival with<br />
vemurafenib in melanoma with BRAF V600E mutation. N Engl J Med.<br />
2011;364:2507-2516.<br />
33. Gomez HL, Doval DC, Chavez MA, et al. Efficacy and safety <strong>of</strong> lapatinib<br />
as first-line therapy for ErbB2-amplified locally advanced or metastatic<br />
breast cancer. J Clin Oncol. 2008;26:2999-3005.<br />
34. Blackwell KL, Burstein HJ, Storniolo AM, et al. Randomized study <strong>of</strong><br />
Lapatinib alone or in combination with trastuzumab in women with ErbB2-<br />
positive, trastuzumab-refractory metastatic breast cancer. J Clin Oncol.<br />
2010;28:1124-1130.<br />
35. Gorre ME, Mohammed M, Ellwood K, et al. <strong>Clinical</strong> resistance to<br />
STI-571 cancer therapy caused by BCR-ABL gene mutation or amplification.<br />
Science. 2001;293:876-880.<br />
36. Shah NP, Nicoll JM, Nagar B, et al. Multiple BCR-ABL kinase domain<br />
mutations confer polyclonal resistance to the tyrosine kinase inhibitor imatinib<br />
(STI571) in chronic phase and blast crisis chronic myeloid leukemia.<br />
Cancer Cell. 2002;2:117-125.<br />
37. Katayama R, Khan TM, Benes C, et al. Therapeutic strategies to<br />
overcome crizotinib resistance in non-small cell lung cancers harboring the<br />
fusion oncogene EML4-ALK. Proc Natl Acad Sci USA. 2011:108:7535-7540.<br />
38. Pao W, Miller VA, Politi KA, et al. Acquired resistance <strong>of</strong> lung<br />
adenocarcinomas to gefitinib or erlotinib is associated with a second mutation<br />
in the EGFR kinase domain. PLoS Med. 2005;2:e73. doi:10.1371/journal.<br />
pmed.0020073.<br />
39. Villanueva J, Vultur A, Herlyn M. Resistance to BRAF inhibitors:<br />
unraveling mechanisms and future treatment options. Cancer Res. 2011;71:<br />
7137-7140.<br />
40. Mok TS, Wu YL, Thongprasert S, et al. Gefitinib or carboplatinpaclitaxel<br />
in pulmonary adenocarcinoma. N Engl J Med. 2009:361:947-957.<br />
41. Sequist LV, Martins RG, Spigel D, et al. First-line gefitinib in patients<br />
with advanced non, small-cell lung cancer harboring somatic EGFR mutations.<br />
J Clin Oncol. 2008;26:2442-2449.<br />
42. Siegel R, Ward E, Brawley O, et al. Cancer statistics, 2011: the impact<br />
<strong>of</strong> eliminating socioeconomic and racial disparities on premature cancer<br />
deaths.. CA Cancer J Clin. 2011: 61:212-236.<br />
43. Jemal A, Bray F, Center MM, et al. Global cancer statistics. CA Cancer<br />
J Clin. 2011:61:69-90.<br />
44. Ferlay J, Shin HR, Bray F,et al. GLOBOCAN 2008 v1.2, Cancer<br />
incidence and mortality worldwide: IARC CancerBase No. 10 [Internet].<br />
Lyon, France: International Agency for Research on Cancer; 2010. http://<br />
globocan.iarc.fr. Accessed Dec. 15, 2011.<br />
185
Targeting Molecular Aberrations in Breast<br />
Cancer: Is It about Time?<br />
By Laura Esserman, MD, MBA, Christopher Benz, MD, and Angela DeMichele, MD<br />
Overview: Breast cancer is not one homogeneous disease; it<br />
is a heterogeneous cancer with remarkable genomic variation<br />
and variable risk for systemic spread, time to recurrence, and<br />
response to treatment. Although it is increasingly clear that a<br />
substantial proportion <strong>of</strong> newly diagnosed patients with<br />
breast cancer carry only minimal risk <strong>of</strong> developing metastatic<br />
disease, our inability to accurately prognosticate leads to<br />
systemic overtreatment. The recent introduction <strong>of</strong> multigene<br />
predictors for baseline risk assessment and treatment response<br />
in breast cancer subsets heralds the beginning <strong>of</strong><br />
BREAST CANCER is the first solid tumor to demonstrate<br />
the effectiveness <strong>of</strong> targeted therapy. The discovery<br />
more than three decades ago that the steroid<br />
hormone receptors (HRs), estrogen receptor (ER) and progesterone<br />
receptor (PR), are overexpressed in the majority <strong>of</strong><br />
human breast cancers led to the eventual understanding<br />
that endocrine therapy only benefits women with either ERor<br />
PR-positive disease. Initially thought to provide clinical<br />
benefit only to postmenopausal patients with breast cancer,<br />
ER-targeted agents such as tamoxifen were later shown to<br />
reduce disease burden, decrease recurrence rates, and improve<br />
disease-free survival in premenopausal patients, 1<br />
demonstrating that targeting the ER pathway is a biologically<br />
based—not an age-based—therapeutic approach.<br />
Another successful example <strong>of</strong> breast cancer receptor<br />
targeting is the HER2/neu story. The discovery by Slamon<br />
and colleagues 25 years ago that a substantial proportion<br />
(approximately 20%) <strong>of</strong> breast cancers possess genomic amplification<br />
<strong>of</strong> the HER2/neu oncogene and overexpress its<br />
membrane-bound protein receptor, and that this is associated<br />
with worse outcome, 2 spurred development <strong>of</strong> HER2<br />
receptor–targeted therapies. The first <strong>of</strong> these anti-HER2<br />
therapeutics was the humanized monoclonal antibody, trastuzumab,<br />
that was approved for the treatment <strong>of</strong> metastatic<br />
HER2-positive breast cancer 3 after 13 years, and later<br />
approved for treatment in the adjuvant setting on the basis<br />
<strong>of</strong> its significant improvement in disease-free survival and<br />
overall survival in patients with HER2-positive disease. 4<br />
The total time from identification that the receptor was a<br />
critical driver to drug approval in the adjuvant setting was<br />
18 years.<br />
The subsequent introduction <strong>of</strong> prognostic and predictive<br />
breast cancer biomarkers and multigene assays has had<br />
some impact on our ability to determine who is most at risk<br />
for developing metastatic recurrence and who would benefit<br />
most from interventions such as systemic adjuvant chemotherapy.<br />
Bernie Fisher put forward the hypothesis that, for<br />
most women, breast cancer is a systemic disease. 5 A series <strong>of</strong><br />
large randomized clinical trials and meta-analyses have<br />
proven conclusively that adjuvant chemotherapy can reduce<br />
future recurrence <strong>of</strong> metastatic disease and markedly improve<br />
patient survival, verifying Fisher’s hypothesis. The<br />
same data, however, have provided us with the sober realization<br />
that only some women benefit from this aggressive<br />
systemic therapy. Many women do not need such aggressive<br />
therapy because their tumors have indolent growth charac-<br />
186<br />
personalized cancer care and the transition to precision<br />
medicine. At the same time, rapid clinical adoption <strong>of</strong> these<br />
predictors illustrates the voracious unmet need for better and<br />
more finely tuned prognostic and predictive tools. Recent<br />
advances in clinical trial designs have enabled the testing <strong>of</strong><br />
novel agents in combination with standard therapy in the<br />
neoadjuvant setting, with a goal <strong>of</strong> identifying the specific<br />
biomarker-drug pairs that should be advanced for confirmatory<br />
trials and accelerated approval on the basis <strong>of</strong> response<br />
to therapy.<br />
teristics and would not likely recur in their lifetime; for<br />
others, their tumors are not innately sensitive to the empirically<br />
determined drug combinations typically used in adjuvant<br />
treatment.<br />
With the advent <strong>of</strong> widespread mammographic screening,<br />
some aggressive breast cancers are detected at an earlier<br />
(and more curable) clinical stage, but more common is the<br />
detection <strong>of</strong> indolent tumors that are less likely to present a<br />
life-threatening problem. 6,7 We once provided treatment<br />
with chemotherapy and endocrine therapy for any woman<br />
with a tumor larger than 1 cm; 1 there are now multigene<br />
predictors that have enabled a more tailored approach. 8,9<br />
Chemotherapy is actually more effective against highly<br />
proliferative tumors and less effective against low-grade,<br />
slowly proliferating tumors, and commercially approved<br />
multigene predictors that commonly interrogate proliferation<br />
genes have recently been introduced into clinical practice<br />
to tailor individual therapy. 10 Interestingly, currently<br />
approved multigene predictors have been unable to prognosticate<br />
the outcome <strong>of</strong> women diagnosed with HR-negative<br />
tumors, which are usually more proliferative than ER- or<br />
PR-positive tumors, but whose risk <strong>of</strong> dissemination may be<br />
more dependent on yet-to-be-understood host factors. This is<br />
just one <strong>of</strong> several obvious areas <strong>of</strong> unmet clinical need, in<br />
which challenges and opportunities exist to expand the use<br />
<strong>of</strong> gene pr<strong>of</strong>iling for developing novel predictors, and to tailor<br />
existing and emerging therapies for those who will receive<br />
the greatest clinical impact. 11<br />
The translational research community has produced several<br />
validated gene signatures, and many more gene signature<br />
candidates that can potentially push the field forward<br />
have been described. Such signatures may answer questions<br />
about response to our current medications—including powerful<br />
biostatistical approaches that combine and order several<br />
predictors, such as hormone therapy sensitivity and<br />
chemotherapy sensitivity—ultimately allowing us to iden-<br />
From the Carol Franc Buck Breast Care Center, University <strong>of</strong> California, San Francisco;<br />
Cancer and Developmental Therapeutics Program, Buck Institute for Research on Aging,<br />
Novato, CA; Breast Cancer Program, Abramson Cancer Center, University <strong>of</strong> Pennsylvania,<br />
Philadelphia, PA.<br />
Authors’ disclosures <strong>of</strong> potential conflicts <strong>of</strong> interest are found at the end <strong>of</strong> this article.<br />
Address reprint requests to Laura Esserman, MD, MBA, University <strong>of</strong> California, San<br />
Francisco, 1600 Divisadero St., 2nd Floor, Box 1710, San Francisco, CA 94115; email:<br />
laura.esserman@ucfsmedctr.org.<br />
© <strong>2012</strong> by <strong>American</strong> <strong>Society</strong> <strong>of</strong> <strong>Clinical</strong> <strong>Oncology</strong>.<br />
1092-9118/10/1-10
MOLECULAR ABERRATIONS IN BREAST CANCER<br />
tify patients who may be refractory to therapy and at<br />
highest risk for metastatic progression. 12 In addition, we<br />
now have access to powerful assay platforms and informatics<br />
tools that can generate and integrate tens <strong>of</strong> thousands <strong>of</strong><br />
gene expression data from very small slices <strong>of</strong> archival<br />
tumor. This, in turn, creates the opportunity to validate<br />
promising gene signatures and integrate them into a userfriendly,<br />
personalized clinical breast cancer application that<br />
provides prognostic and predictive tools to all patients with<br />
all types <strong>of</strong> breast cancer.<br />
There are several important questions that have the<br />
potential to be addressed by using new and emerging candidate<br />
signatures. We discuss them in the following sections.<br />
Evidence <strong>of</strong> Very-Low-Risk Tumors and Tools<br />
for Prediction<br />
Screening has clearly led to an increase in the number <strong>of</strong><br />
detected cancers. 13 Many <strong>of</strong> these tumors are likely to be<br />
indolent, and the term “IDLE tumors” has been suggested to<br />
differentiate tumors that are life-threatening and those that<br />
are not. 13 As shown by Esserman and van’t Veer, it is<br />
possible to identify molecular evidence <strong>of</strong> such tumors by<br />
defining an “ultra-low” threshold for the 70-gene pr<strong>of</strong>ile that<br />
could identify tumors that, in the absence <strong>of</strong> any adjuvant<br />
therapy, would not result in distant progression. By using<br />
two large European data sets, they showed that with the<br />
advent <strong>of</strong> screening, the incidence <strong>of</strong> these IDLE tumors has<br />
increased, and that in women with screen-detected cancers,<br />
30% could fit the definition <strong>of</strong> IDLE tumors and potentially<br />
not need additional therapy. 14 Validation studies are currently<br />
in progress. A tool that successfully predicted indolent<br />
disease would have an enormous and beneficial impact<br />
KEY POINTS<br />
● Breast cancer is made up <strong>of</strong> several different diseases<br />
with remarkable variation in the risk and timing <strong>of</strong><br />
recurrence.<br />
● There are targeted treatments in breast cancer that<br />
are established on the basis <strong>of</strong> inhibiting overexpressed<br />
estrogen (ER) or HER2 receptors, and these<br />
treatment strategies are being further refined by<br />
using multigene subclassifiers.<br />
● Proliferative lesions are most likely to benefit from<br />
chemotherapy, and the genes turned on by activated<br />
ER are emerging as predictors for sensitivity to<br />
endocrine therapy. Additional signatures are being<br />
developed to predict resistance to either or both<br />
endocrine and chemotherapies.<br />
● Existing and emerging predictive and prognostic biomarkers<br />
are being integrated into clinical trials to<br />
help focus treatments on patient subpopulations<br />
most likely to benefit from systemic therapy.<br />
● New and adaptive clinical trial designs can improve<br />
the efficiency <strong>of</strong> new drug evaluation, particularly in<br />
the context <strong>of</strong> biomarkers, and have already begun to<br />
define a new regulatory path that will accelerate the<br />
ability <strong>of</strong> targeted drug and biomarker pairs to be<br />
used in the adjuvant setting.<br />
for women, reducing the use <strong>of</strong> toxic therapies in women who<br />
do not need them.<br />
Evidence that Timing <strong>of</strong> Risk Is Related to<br />
Tumor Biology<br />
There is now a body <strong>of</strong> literature supporting the notion<br />
that the timing <strong>of</strong> recurrence is a critical feature <strong>of</strong> the<br />
biology <strong>of</strong> HR-positive breast cancer. By using a data set<br />
from the late 1970s from Guy’s Hospital in London, before<br />
the advent <strong>of</strong> adjuvant therapy and screening, we were able<br />
to newly characterize 349 cases using more modern scoring<br />
<strong>of</strong> ER and PR (HR), HER2, and grade to develop new tools to<br />
understand the best order <strong>of</strong> integrating predictors <strong>of</strong> outcome.<br />
The routinely used staging, histologic and immunohistochemical<br />
features <strong>of</strong> primary breast cancers, provide<br />
important information about overall relapse risk and timing<br />
<strong>of</strong> future metastatic recurrence. By using a risk-partitioning<br />
algorithm, we can identify the order in which the standard<br />
clinical information can separate out groups <strong>of</strong> patients by<br />
the degree and timing <strong>of</strong> risk. By using a classification and<br />
regression tree approach, which identifies the order <strong>of</strong> importance<br />
<strong>of</strong> the clinical variables, the first split in the data is<br />
by node status (0 vs. any). For the node-negative patients,<br />
the groups then split by HR-positive versus triple-negative<br />
(both ER and PR negative) and HER2–positive disease.<br />
HR-positive disease was split by low grade versus others.<br />
The low-grade tumors appear to have late but not early<br />
recurrence risk. 15 The risk-partitioning method identifies<br />
the critical dependencies among variables and opportunities<br />
where molecular information can clearly improve our ability<br />
to fine-tune predictions <strong>of</strong> both early and late risk, as well as<br />
the benefits <strong>of</strong> chemotherapy. Curated and analyzed published<br />
breast cancer data sets can serve as an important<br />
resource to validate candidate predictors for use in <strong>Clinical</strong><br />
Laboratory Improvement Amendments (CLIA)-certified<br />
laboratories. 16-18 Although our analyses have shown that<br />
tumor proliferative capacity is predictive <strong>of</strong> early breast<br />
cancer metastasis—but only for HR-positive breast cancer<br />
15-17 —we have also uncovered new predictors <strong>of</strong> late<br />
metastatic recurrence by HR-positive breast cancer that<br />
need additional validation with respect to routine adjuvant<br />
endocrine therapy use. 15<br />
We have clearly established that the timing <strong>of</strong> metastatic<br />
recurrence is a critical feature <strong>of</strong> tumor biology. For HRnegative<br />
tumors and HER2-positive tumors, recurrence risk<br />
is early (within the first 5 years), but for HR-positive breast<br />
cancer, the risk can extend for many years. Others have<br />
demonstrated this as well. The conclusion that emerges is<br />
that HR-positive tumors have a recurrence risk that can<br />
span more than 20 years from the time <strong>of</strong> diagnosis. For<br />
these patients, signatures that are based largely on the<br />
expression <strong>of</strong> genes driving cell proliferation are predictive<br />
only <strong>of</strong> early recurrence. Fortunately, investigators are now<br />
focusing on finding predictors that can be used to identify<br />
HR-positive breast tumors with higher likelihood <strong>of</strong> recurrence.<br />
Dr. Liu and colleagues have described a 91-gene<br />
signature to look for early recurrence as well as late recurrence,<br />
between 5 and 20 years after diagnosis. 19,20<br />
Sensitivity to Endocrine Therapy<br />
Prediction <strong>of</strong> HR-positive patients most likely to benefit<br />
from endocrine therapy remains a major clinical challenge to<br />
187
date. With the advent <strong>of</strong> gene expression pr<strong>of</strong>iling, several<br />
gene signatures <strong>of</strong> endocrine sensitivity, including the<br />
HOXB13:IL17RB ratio 21,22 and the Sensitivity to Endocrine<br />
Therapy (SET index), 23 have been identified. In addition,<br />
the estrogen-regulated gene GREB1 has been shown to be a<br />
critical regulator <strong>of</strong> hormone-dependent breast cancer<br />
growth and may serve as a marker <strong>of</strong> endocrine therapy<br />
sensitivity 24 Bianchi and colleagues from Italy recently<br />
presented evidence that a signature <strong>of</strong> four estrogen-related<br />
genes and 12 mitotic kinase genes has the potential to<br />
predict resistance/sensitivity to hormone therapy. 20 The fact<br />
that many signatures are emerging on the basis <strong>of</strong> the same<br />
biologic principles, and can be generated from the tissue<br />
sample at the time <strong>of</strong> diagnosis, suggests that these types <strong>of</strong><br />
signatures are likely to validate and find their way into<br />
clinical trials and practice. By using the neoadjuvant approach,<br />
even more discriminatory signatures are likely to<br />
emerge. These predictors will not only identify women at<br />
risk, but we hope will also suggest targeted strategies, based<br />
on timing <strong>of</strong> treatments as well as novel agent combinations,<br />
that will substantially improve outcomes specifically for the<br />
groups at risk.<br />
Identifying Risk and Response to Therapy in<br />
HR-Negative Tumors<br />
For HR-negative tumors and HER2-positive tumors, the<br />
risk <strong>of</strong> recurrence is early, largely in the first 5 years. In<br />
these tumors, proliferation does not seem to be a driver, as<br />
the majority <strong>of</strong> tumors are indeed proliferative. In this<br />
population, immune-based signatures seem to have the most<br />
predictive signatures, and they can predict risk within the<br />
first 5 years. 15 Importantly, there is evidence that in some<br />
patients with node-negative, HR-negative disease, risk is<br />
quite low as a result <strong>of</strong> high immune responsiveness. This<br />
has been shown independently in several studies. 25 We have<br />
also recently identified immune- and cytokine-based gene<br />
expression levels capable <strong>of</strong> identifying early-stage HRnegative/triple-negative<br />
breast cancer at the lowest risk for<br />
developing early or late metastatic recurrence. 18 Both <strong>of</strong><br />
these signatures have been translated to CLIA-certified<br />
RNA-based assay platforms and preliminarily validated<br />
with our retrospective collection <strong>of</strong> formalin-fixed paraffinembedded<br />
breast tumor–derived RNA samples, 18,25 and<br />
have been shown to be able to predict HR-negative and<br />
triple-negative metastatic outcome. 15,18 There is clearly a<br />
spectrum <strong>of</strong> risk that is modified by the expression <strong>of</strong><br />
immune-related genes, suggesting that there is an important<br />
component <strong>of</strong> immune responsiveness that drives outcome<br />
in HR-negative tumors. This conclusion is further<br />
supported by the identification <strong>of</strong> the T-cell/macrophage<br />
signature that confers chemoresistance, described in the<br />
following section.<br />
Immunity and Sensitivity to Chemotherapy: T-Cell and<br />
Macrophage Signatures<br />
Leukocyte infiltration into breast tumors has been observed<br />
in the context <strong>of</strong> both protumorigenic inflammation<br />
and anticancer immunosurveillance. Thus, beyond just the<br />
composition <strong>of</strong> the immune infiltrate, the immune context <strong>of</strong><br />
these leukocytes (their distribution, density, architecture,<br />
and interactions) must be analyzed to understand their<br />
prognostic significance. Although the mechanisms involved<br />
in leukocyte infiltration into tumors are poorly character-<br />
188<br />
ESSERMAN, BENZ, AND DEMICHELE<br />
ized, a recent study found high endothelial venules (HEVs),<br />
blood vessels normally found in lymphoid tissues specialized<br />
in lymphocyte recruitment, in a variety <strong>of</strong> solid tumors. 26<br />
High densities <strong>of</strong> HEVs in breast cancers correlated with a<br />
more pronounced infiltration by T cells and conferred a<br />
lower risk <strong>of</strong> relapse and significantly longer metastasisfree,<br />
disease-free, and overall survival rates. 26<br />
T lymphocytes—in particular CD8-positive T lymphocytes—are<br />
a crucial component <strong>of</strong> cell-mediated immunity.<br />
Several studies have examined the prognostic value <strong>of</strong><br />
tumor-infiltrating lymphocytes (TILs) and/or CD8-positive<br />
TILs in breast cancer. 27,28 Interestingly, high TILs are<br />
associated with better responses to chemotherapy and improved<br />
disease-specific survival, particularly in HR-negative<br />
tumors.<br />
We, and others, have shown that high numbers <strong>of</strong> tumorassociated<br />
macrophages (TAMs) are associated with high<br />
grade, HR-negative breast cancers and poor outcomes. 29,30<br />
Recently, we proposed a gene expression signature related to<br />
cytotoxic T-cell (Tc) responses and major histocompatibility<br />
complex class II expression that might serve as a surrogate<br />
for an anticancer immune microenvironment in breast cancer.<br />
31 Others have proposed combinations <strong>of</strong> macrophage<br />
and T-cell markers as markers <strong>of</strong> poor outcomes on the basis<br />
<strong>of</strong> both mouse and human studies. 32,33 These results illustrate<br />
the importance <strong>of</strong> understanding the immune context<br />
<strong>of</strong> leukocytes within the tumor microenvironment, and the<br />
opportunity we may have to improve outcomes by targeting<br />
the immune microenvironment as part <strong>of</strong> our therapeutic<br />
strategy.<br />
Trials that Focus on Development <strong>of</strong> Drug-Biomarker<br />
Pairs and Response to Treatment in the<br />
Neoadjuvant Setting<br />
Increasingly, it is clear that all drugs will not benefit all<br />
patients with breast cancer. As new targeted drugs are being<br />
tested, more and more trials are incorporating biopsies<br />
before and after treatment to both determine predictors to<br />
therapy response and also assess pharmacodynamics. Biopsy<br />
at the time <strong>of</strong> presentation <strong>of</strong> metastatic disease is<br />
increasingly becoming the standard <strong>of</strong> care, given that 20%<br />
to 35% <strong>of</strong> metastatic lesions are discordant from the primary<br />
tumor on the basis <strong>of</strong> HR or HER2 status. 34,35 However, this<br />
is problematic on several fronts. First, it is challenging to<br />
perform biopsies <strong>of</strong> metastatic lesions: There are associated<br />
risks depending on the location <strong>of</strong> these lesions. Second, the<br />
standard setting for testing new drugs is in patients with<br />
metastatic disease, <strong>of</strong>ten when they have experienced failure<br />
with first- and second-line treatments. Novel therapies<br />
may be less effective once tumors have progressed through<br />
several treatment regimens, and serial biopsies performed<br />
with subsequent lines <strong>of</strong> therapy compound the risk. Finally,<br />
effective agents have resulted in longer control <strong>of</strong> metastatic<br />
disease, but rarely cure. Those same agents, when administered<br />
at the time <strong>of</strong> primary diagnosis, have resulted in cure.<br />
Trastuzumab is the prime example <strong>of</strong> this phenomenon. 4,5<br />
Determining appropriate end points in the metastatic setting<br />
that predict curative potential when administered at<br />
the time <strong>of</strong> diagnosis has been challenging.<br />
One approach to determining which biologically defined<br />
tumors are most likely to benefit from which therapy is to<br />
introduce new agents at the time <strong>of</strong> diagnosis. This approach<br />
is feasible because neoadjuvant therapy is increasingly
MOLECULAR ABERRATIONS IN BREAST CANCER<br />
Fig. 1. Recurrence-free survival (RFS) stratified<br />
by response to therapy, complete pathologic<br />
response (pCR) for the population <strong>of</strong> 172<br />
patients, excluding all patients receiving neoajuvant<br />
trastuzumab overall and by hormone<br />
receptor (HR)-positive/HER2-negative subset,<br />
HR-negative/HER2-negative (triple-negative)<br />
subset, and HER2-positive subset. All patients<br />
received neoadjuvant chemotherapy with a<br />
doxorubicin regimen followed by a taxane.<br />
RFS for all patients (excluding those receiving<br />
trastuzumab, 37 patients) is shown on the left<br />
and on the right for the receptor subsets. The<br />
solid line represents patients who achieved a<br />
pCR and the dotted lines represent patients<br />
who did not. HR-positive/HER2-negative tumors<br />
are shown in blue, triple negative tumors<br />
are shown in red, and HER2-positive tumors<br />
are shown in green. Reprinted with permission.<br />
© <strong>American</strong> <strong>Society</strong> <strong>of</strong> <strong>Clinical</strong> <strong>Oncology</strong>.<br />
All rights reserved. 40<br />
being used as a standard <strong>of</strong> care. 36 The order <strong>of</strong> therapy<br />
(surgery or systemic therapy first) does not affect the efficacy<br />
<strong>of</strong> therapy. 37,38 However, sequencing systemic therapy first<br />
allows the measurement <strong>of</strong> response to therapy and the safe<br />
capture <strong>of</strong> tissue specimens during the course <strong>of</strong> treatment.<br />
The I-SPY TRIAL and others have demonstrated that response<br />
to therapy is predictive <strong>of</strong> ultimate outcome and that<br />
those with either complete or near-complete response to<br />
chemotherapy in fact have much better outcomes than those<br />
who have large residual cancer burdens. 39-41 Importantly,<br />
this relationship between response to treatment and outcome<br />
is even stronger by tumor subset (Fig. 1)—whether<br />
looking by receptor subtype 40 or by molecular subset. 42<br />
We now have the opportunity to test emerging new agents<br />
in the neoadjuvant setting for women at high risk for<br />
recurrence. To do this efficiently, we need clinical trial<br />
designs that take the heterogeneity <strong>of</strong> disease into account,<br />
a paradigm for identifying the most effective agents and<br />
their associated predictive biomarkers and a more rapid<br />
path to getting such drugs to market. The I-SPY 2 TRIAL<br />
design is an adaptive clinical trial focused on women at the<br />
time <strong>of</strong> primary diagnosis with tumors 2.5 cm or larger. The<br />
first step <strong>of</strong> the process is to generate a molecular pr<strong>of</strong>ile<br />
from core biopsy <strong>of</strong> the primary tumor; those women found to<br />
be at high risk for early recurrence, on the basis <strong>of</strong> a<br />
high-risk classification by the 70-gene pr<strong>of</strong>ile, are then<br />
randomly assigned to either standard therapy or standard<br />
therapy in combination with a novel agent. The philosophy<br />
<strong>of</strong> the I-SPY 2 TRIAL is to accelerate the testing <strong>of</strong> promising<br />
new targeted agents by enabling their first phase II<br />
assessment to be in the neoadjuvant setting. To accomplish<br />
this, the I-SPY 2 TRIAL has established a precompetitive<br />
189
consortium to facilitate the testing <strong>of</strong> agents against a<br />
backdrop <strong>of</strong> comprehensive tumor pr<strong>of</strong>iling, with the ability<br />
to use the adaptive design process to enrich enrollment on<br />
the basis <strong>of</strong> standard markers. The biomarker design <strong>of</strong> the<br />
trial allows the evaluation <strong>of</strong> three levels <strong>of</strong> biomarkers. The<br />
first are standard or investigational device exemption (IDE)<br />
approved, which are integral markers used in the adaptive<br />
randomization. These include hormone and HER2 receptors,<br />
the U.S. Food and Drug Administration (FDA)—approved<br />
70-gene pr<strong>of</strong>ile, other measures <strong>of</strong> HER2 overexpression<br />
(fluorescent in situ hybridization and expression array) and<br />
volume on magnetic resonance imaging. The second level<br />
comprises “qualifying” biomarkers. These are biomarkers<br />
that have promise in predicting response to a specific agent,<br />
but require validation in a CLIA-approved testing environment.<br />
This includes cell line predictors that predict response<br />
and can be read by expression array, or other specific<br />
phosphoprotein that is thought to be a key target <strong>of</strong> a<br />
particular agent. These markers, if they prove to be predictive<br />
<strong>of</strong> response, would potentially be used as integral<br />
markers in future trials <strong>of</strong> the agent and the data generated<br />
from their evaluation in the I-SPY 2 TRIAL can be used to<br />
support an IDE application. The last layer is the exploratory<br />
biomarkers. These include hypothesis-generating markers<br />
from new platforms such as RNAseq. The platforms being<br />
Authors’ Disclosures <strong>of</strong> Potential Conflicts <strong>of</strong> Interest<br />
Author<br />
Employment or<br />
Leadership<br />
Positions<br />
Consultant or<br />
Advisory Role<br />
used to characterize tumors in the trial include Agilent<br />
Technologies (Palo Alto, CA) and Affymetrix (Santa Clara,<br />
CA) expression arrays, reverse phase protein arrays,<br />
genome-wide association study (GWAS), and RNAseq.<br />
The goal <strong>of</strong> the trial is to graduate agents and biomarker<br />
pairs with a predicted likelihood <strong>of</strong> success in phase III<br />
trials. 7 Going forward, if an agent is shown to improve the<br />
chance <strong>of</strong> obtaining a pathologic complete response and this<br />
observation can be confirmed in a separate neoadjuvant<br />
trial, the FDA is willing to consider awarding accelerated<br />
approval on the basis <strong>of</strong> these data. 43 This provides a<br />
regulatory path forward for accelerating successful agents to<br />
the neoadjuvant or adjuvant settings, in which they are<br />
likely to provide the greatest benefit. Indeed, it is about time<br />
that we design our trials around the biomarkers that characterize<br />
risk and response to treatment. The I-SPY 2 TRIAL<br />
is one such effort directed at more rapidly and efficiently<br />
finding better options for patients with breast cancer. Furthermore,<br />
this platform provides the opportunity to validate<br />
several <strong>of</strong> the emerging markers described in the first<br />
portion <strong>of</strong> this article, and to help us move faster toward the<br />
time when we will be able to provide precision medicine<br />
approaches to reduce the suffering and death associated<br />
with a diagnosis <strong>of</strong> breast cancer.<br />
Stock<br />
Ownership Honoraria<br />
Laura Esserman Agendia<br />
Christopher Benz*<br />
Angela DeMichele*<br />
*No relevant relationships to disclose.<br />
1. Early Breast Cancer Trialists’ Collaborative Group (EBCTCG). Effects <strong>of</strong><br />
chemotherapy and hormonal therapy for early breast cancer on recurrence<br />
and 15-year survival: an overview <strong>of</strong> the randomised trials. Lancet. 2005;365:<br />
1687-1717.<br />
2. Slamon DJ. Proto-oncogenes and human cancers. N Engl J Med. 1987;<br />
317:955-957.<br />
3. Slamon DJ, Leyland-Jones B, Shak S, et al. Use <strong>of</strong> chemotherapy plus a<br />
monoclonal antibody against HER2 for metastatic breast cancer that overexpresses<br />
HER2. N Engl J Med. 2001;344:783-792.<br />
4. Romond EH, Perez EA, Bryant J, et al. Trastuzumab plus adjuvant<br />
chemotherapy for operable HER2-positive breast cancer. N Engl J Med<br />
2005;353:1673-1684.<br />
5. Fisher B. Breast-cancer management: alternatives to radical mastectomy.<br />
N Engl J Med. 1979;301:326-328.<br />
6. Esserman LJ, Shieh Y, Rutgers EJ, et al. Impact <strong>of</strong> mammographic<br />
screening on the detection <strong>of</strong> good and poor prognosis breast cancers. Breast<br />
Cancer Res Treat. 2011;130:725-734.<br />
7. Barker AD, Sigman CC, Kell<strong>of</strong>f GJ, et al. I-SPY 2: an adaptive breast<br />
cancer trial design in the setting <strong>of</strong> neoadjuvant chemotherapy. Clin Pharmacol<br />
Ther. 2009;86:97-100.<br />
8. van de Vijver MJ, He YD, van’t Veer LJ, et al. A gene-expression<br />
signature as a predictor <strong>of</strong> survival in breast cancer. New Engl J Med.<br />
2002;347:1999-2009.<br />
9. Paik S, Shak S, Tang G, et al. A multigene assay to predict recurrence <strong>of</strong><br />
tamoxifen-treated, node-negative breast cancer. New Engl J Med. 2004;351:<br />
2817-2826.<br />
10. Paik S, Tang G, Shak S, et al. Gene expression and benefit <strong>of</strong><br />
chemotherapy in women with node-negative, estrogen receptor-positive<br />
breast cancer. J Clin Oncol. 2006;24:3726-3734.<br />
11. van’t Veer LJ, Bernards R. Enabling personalized cancer medicine<br />
through analysis <strong>of</strong> gene-expression patterns. Nature. 2008;452:564-570.<br />
190<br />
REFERENCES<br />
Research<br />
Funding<br />
ESSERMAN, BENZ, AND DEMICHELE<br />
Expert<br />
Testimony<br />
Other<br />
Remuneration<br />
12. Hatzis C, Pusztai L, Valero V, et al. A genomic predictor <strong>of</strong> response<br />
and survival following taxane-anthracycline chemotherapy for invasive<br />
breast cancer. JAMA. 2011;305:1873-1881.<br />
13. Esserman L, Shieh Y, Thompson I. Rethinking screening for breast<br />
cancer and prostate cancer. JAMA. 2009;302:1685-1692.<br />
14. Kok M, Linn SC, Van Laar RK, et al. Comparison <strong>of</strong> gene expression<br />
pr<strong>of</strong>iles predicting progression in breast cancer patients treated with tamoxifen.<br />
Breast Cancer Res Treat. 2009;113:275-283.<br />
15. Esserman LJ, Moore DH, Tsing PJ, et al. Biologic markers determine<br />
both the risk and the timing <strong>of</strong> recurrence in breast cancer. Breast Cancer Res<br />
Treat. 2011;129:607-616.<br />
16. Tutt A, Wang A, Rowland C, et al. Risk estimation <strong>of</strong> distant metastasis<br />
in node-negative, estrogen receptor-positive breast cancer patients using an<br />
RT-PCR based prognostic expression signature. BMC Cancer. 2008;8:339.<br />
17. Yau C, Wang Y, Zhang Y, et al. Young age, increased tumor proliferation<br />
and FOXM1 expression predict early metastatic relapse only for<br />
endocrine-dependent breast cancers. Breast Cancer Res Treat. 2011;126:803-<br />
810.<br />
18. Yau C, Esserman L, Moore DH, et al. A multigene predictor <strong>of</strong><br />
metastatic outcome in early stage hormone receptor-negative and triplenegative<br />
breast cancer. Breast Cancer Res. 2010;12:R85.<br />
19. Liu MC, Dixon JM, Xuan JJ, et al. Molecular signaling distinguishes<br />
early ER positive breast cancer recurrences despite adjuvant tamoxifen.<br />
Cancer Res. 2011;71 (suppl; abstr S1-8).<br />
20. Bianchini G, Pusztai L, Iwamoto T, et al. Molecular Tumor Characteristics<br />
Influence Adjuvant Endocrine Treatment Outcome. Cancer Res. 2011;71<br />
(suppl; abstr S1-7).<br />
21. Sgroi DC, Haber DA, Ryan PD, et al. RE: A two-gene expression ratio<br />
predicts clinical outcome in breast cancer patients treated with tamoxifen.<br />
Cancer Cell. 2004;6:445.<br />
22. Ma XJ, Wang Z, Ryan PD, et al. A two-gene expression ratio predicts
MOLECULAR ABERRATIONS IN BREAST CANCER<br />
clinical outcome in breast cancer patients treated with tamoxifen. Cancer<br />
Cell. 2004;5:607-616.<br />
23. Symmans WF, Hatzis C, Sotiriou C, et al. Genomic index <strong>of</strong> sensitivity<br />
to endocrine therapy for breast cancer. J Clin Oncol. 2010;28:4111-4119.<br />
24. Rae JM, Johnson MD, Scheys JO, et al. GREB 1 is a critical regulator<br />
<strong>of</strong> hormone dependent breast cancer growth. Breast Cancer Res Treat.<br />
2005;92:141-149.<br />
25. Teschendorff AE, Caldas C. A robust classifier <strong>of</strong> high predictive value<br />
to identify good prognosis patients in ER-negative breast cancer. Breast<br />
Cancer Res. 2008; 10:R73.<br />
26. Martinet L, Garrido I, Filleron T, et al. Human solid tumors contain<br />
high endothelial venules: association with T- and B-lymphocyte infiltration<br />
and favorable prognosis in breast cancer. Cancer Res. 2011;71:5678-5687.<br />
27. Denkert C, Loibl S, Noske A, et al. Tumor-associated lymphocytes as an<br />
independent predictor <strong>of</strong> response to neoadjuvant chemotherapy in breast<br />
cancer. J Clin Oncol. 2010;28:105-113.<br />
28. Mahmoud SM, Lee AH, Paish EC, et al. The prognostic significance <strong>of</strong><br />
B lymphocytes in invasive carcinoma <strong>of</strong> the breast. Breast Cancer Res Treat.<br />
Epub 2011 Jun 14.<br />
29. Mukhtar RA, Moore AP, Nseyo O, et al. Elevated PCNA� tumorassociated<br />
macrophages in breast cancer are associated with early recurrence<br />
and non-Caucasian ethnicity. Breast Cancer Res Treat. 2011;130:635-644.<br />
30. Mahmoud SM, Lee AH, Paish EC, et al. Tumour-infiltrating macrophages<br />
and clinical outcome in breast cancer. J Clin Pathol. <strong>2012</strong>;65:159-163.<br />
31. Mukhtar R, Wolf D, Tandon V, et al. PCNA� tumor associated<br />
macrophages are associated with M1 and M2 gene expression, and confer poor<br />
prognosis in the absence <strong>of</strong> anti-tumor immune environment. Cancer Res.<br />
2011;71 (suppl; abstr P4-09-19).<br />
32. DeNardo DG, Barreto JB, Andreu P, et al. CD4(�) T cells regulate<br />
pulmonary metastasis <strong>of</strong> mammary carcinomas by enhancing protumor<br />
properties <strong>of</strong> macrophages. Cancer Cell. 2009;16:91-102.<br />
33. DeNardo DG, Brennan DJ, Rexhepaj E, et al. Leukocyte complexity in<br />
breast cancer predicts overall survival and functionally regulates response to<br />
chemotherapy. Cancer Discov. 2011;1:54-67.<br />
34. Wilking U, Karlsson E, Skoog L, et al. HER2 status in a populationderived<br />
breast cancer cohort: discordances during tumor progression. Breast<br />
Cancer Res Treat. 2011;125:553-561.<br />
35. Amir E, Miller N, Geddie W, et al. Prospective study evaluating the<br />
impact <strong>of</strong> tissue confirmation <strong>of</strong> metastatic disease in patients with breast<br />
cancer. J Clin Oncol. <strong>2012</strong>;30:587-592.<br />
36. Buchholz TA, Lehman CD, Harris JR, et al. Statement <strong>of</strong> the science<br />
concerning locoregional treatments after preoperative chemotherapy for<br />
breast cancer: a National Cancer Institute conference. J Clin Oncol 2008;26:<br />
791-797.<br />
37. Bear HD, Anderson S, Smith RE, et al. Sequential preoperative or<br />
postoperative docetaxel added to preoperative doxorubicin plus cyclophosphamide<br />
for operable breast cancer: National Surgical Adjuvant Breast and<br />
Bowel Project Protocol B-27. J Clin Oncol. 2006;24:2019-2027.<br />
38. Fisher B, Brown A, Mamounas E, et al. Effect <strong>of</strong> preoperative chemotherapy<br />
on local-regional disease in women with operable breast cancer:<br />
findings from National Surgical Adjuvant Breast and Bowel Project B-18.<br />
J Clin Oncol. 1997;15:2483-2493.<br />
39. Carey LA, Dees EC, Sawyer L, et al. The triple negative paradox:<br />
primary tumor chemosensitivity <strong>of</strong> breast cancer subtypes. Clin Cancer Res.<br />
2007;13:2329-2334.<br />
40. Esserman LJ, Berry DA, DeMichele A, et al. Pathologic complete<br />
response predicts recurrence-free survival more effectively by cancer subset:<br />
results from the I-SPY 1 TRIAL (CALGB 150007/150012; ACRIN 6657).<br />
J Clin Oncol. In Press.<br />
41. Symmans WF, Peintinger F, Hatzis C, et al. Measurement <strong>of</strong> residual<br />
breast cancer burden to predict survival after neoadjuvant chemotherapy.<br />
J Clin Oncol, 2007;25:4414-4422.<br />
42. Esserman LJ, Berry DA, Cheang MC, et al. Chemotherapy response<br />
and recurrence-free survival in neoadjuvant breast cancer depends on biomarker<br />
pr<strong>of</strong>iles: Results from the I-SPY 1 TRIAL (CALGB 150007/150012;<br />
ACRIN 6657). Breast Cancer Res Treat. Epub 2011 Dec 25.<br />
43. Esserman LJ. Woodcock J. Accelerating identification and regulatory<br />
approval <strong>of</strong> investigational cancer drugs. JAMA. 2011;306:2608-2609.<br />
191
ETHICAL CHALLENGES OF HEALTH CARE REFORM<br />
FOR ONCOLOGY PROVIDERS: A DEBATE<br />
CHAIR<br />
Beverly Moy, MD, MPH<br />
Massachusetts General Hospital<br />
Boston, MA<br />
SPEAKERS<br />
Jeffrey M. Peppercorn, MD, MPH<br />
Duke University Medical Center<br />
Durham, NC<br />
Amy P. Abernethy, MD<br />
Duke University Medical Center<br />
Durham, NC
Core Elements <strong>of</strong> the Patient Protection and<br />
Affordable Care Act and Their Relevance to<br />
the Delivery <strong>of</strong> High-Quality Cancer Care<br />
By Beverly Moy, MD, MPH, Amy P. Abernethy, MD,<br />
and Jeffrey M. Peppercorn, MD, MPH<br />
Overview: The Affordable Care Act (ACA) contains many<br />
provisions that affect cancer care. The provisions <strong>of</strong> health<br />
care reform aim to improve access to quality cancer care,<br />
particularly among the most vulnerable <strong>American</strong>s. However,<br />
health care reform also <strong>of</strong>fers many challenges and opportu-<br />
IN MARCH 2010, President Obama signed into law the<br />
Patient Protection and Affordable Care Act (ACA). 1,2 The<br />
ACA contains provisions that have important implications<br />
for cancer care. 3 Its implications for patients are sweeping<br />
and have the potential to expand access to care and improve<br />
cancer care among vulnerable groups. The ACA, although<br />
not perfect, aims to make cancer care more comprehensive,<br />
affordable, and accessible. This article provides an overview<br />
<strong>of</strong> the ACA in relation to oncology, discusses the ethical<br />
challenges for the oncology provider, and summarizes issues<br />
relevant to various stakeholders in the oncology community.<br />
Overview <strong>of</strong> Cancer Care and the<br />
Affordable Care Act<br />
Insurance Reforms<br />
In 2008, there were an estimated 46.3 million uninsured<br />
<strong>American</strong>s, equaling 15% <strong>of</strong> the United States population,<br />
plus an additional 25 million underinsured <strong>American</strong>s. 4,5<br />
According to the Congressional Budget Office, the ACA is<br />
expected to expand health insurance coverage to 32 million<br />
individuals by 2019. 6<br />
Medicaid<br />
The ACA expands Medicaid to individuals with incomes<br />
up to 133% <strong>of</strong> the federal poverty level (FPL), thereby adding<br />
16 million to 20 million individuals to the Medicaid roster.<br />
The ACA also standardizes Medicaid benefits by guaranteeing<br />
a minimum package <strong>of</strong> essential services. However,<br />
oncology provider participation in Medicaid is in jeopardy,<br />
particularly in states that are economically poorer as a<br />
result <strong>of</strong> low reimbursement levels. Currently, some states<br />
are reporting additional cuts in Medicaid payments, further<br />
reducing the value <strong>of</strong> this coverage. Also <strong>of</strong> major concern is<br />
that there is convincing evidence that adult patients with<br />
cancer with Medicaid have similarly poor clinical outcomes<br />
compared with those <strong>of</strong> uninsured patients. 7-10 Therefore,<br />
Medicaid expansion would not necessarily be expected to<br />
improve cancer outcomes among vulnerable populations as<br />
the system currently stands. Given the anticipated expansion<br />
in the Medicaid population, the quality <strong>of</strong> care under<br />
the Medicaid program may be further compromised.<br />
Health Insurance Exchanges<br />
By 2014, health insurance exchanges must be established<br />
in all states. These exchanges are designed to provide<br />
individuals or small businesses the opportunity to shop for<br />
health insurance.<br />
e4<br />
nities that affect every stakeholder in oncology. This article<br />
summarizes the ACA provisions relevant to oncology, discusses<br />
the ethical implications for the oncology caregiver,<br />
and describes the effects on specific oncology stakeholders.<br />
Elimination <strong>of</strong> Coverage Barriers<br />
The ACA prohibits insurers from denying coverage to<br />
children with preexisting medical conditions and allows<br />
young adults up to age 26 to remain on their parents’ health<br />
plans. Starting in 2014, all insurers will have to accept<br />
all applicants, irrespective <strong>of</strong> preexisting conditions such as<br />
cancer, and renew coverage. The law prohibits canceling<br />
coverage, eliminates the lifetime amount insurance will<br />
pay for certain conditions, and restricts annual limits. Patients<br />
with cancer and survivors <strong>of</strong> cancer will not be denied<br />
coverage on the basis <strong>of</strong> their preexisting diagnosis <strong>of</strong><br />
cancer.<br />
Medicare “Donut Hole”<br />
The ACA helps to close the so-called Medicare donut hole<br />
so that seniors do not face a costly gap in prescription drug<br />
coverage. This is especially relevant in cancer because many<br />
modern anticancer therapies are now in oral form and can be<br />
prohibitively expensive. Each eligible senior will receive a<br />
one-time, tax-free $250 rebate check.<br />
Pediatric Cancer<br />
In aggregate, childhood cancer is the sixth most common<br />
cancer in the United States. The ACA specifically mandates<br />
that Medicaid-eligible children who voluntarily opt for hospice<br />
services will no longer be required to forego curative<br />
services, such as active cancer treatment, to receive hospice<br />
care coverage.<br />
Cancer Prevention and Survivorship<br />
Early detection <strong>of</strong> cancer through adherence to recommended<br />
screening examinations leads to decreased mortality<br />
from most cancers. The ACA requires all health plans to<br />
cover preventive services that receive an “A” or “B” rating<br />
from the U.S. Preventive Services Task Force (USPSTF).<br />
These treatments must be covered with no deductibles or<br />
copays and with no maximums allowed. However, the ACA<br />
does not expressly require insurers to cover follow-up testing<br />
<strong>of</strong> abnormalities found in a cancer screening examination.<br />
The ACA also does not comment on reimbursement for<br />
From the Massachusetts General Hospital Cancer Center, Boston, MA; and Duke<br />
Comprehensive Cancer Center, Durham, NC.<br />
Authors’ disclosures <strong>of</strong> potential conflicts <strong>of</strong> interest are found at the end <strong>of</strong> this article.<br />
Address reprint requests to Beverly Moy, MD, MPH, Massachusetts General Hospital,<br />
55 Fruit St., YAW 9A, Boston, MA 02114; email: bmoy@partners.org.<br />
© <strong>2012</strong> by <strong>American</strong> <strong>Society</strong> <strong>of</strong> <strong>Clinical</strong> <strong>Oncology</strong>.<br />
1092-9118/10/1-10
HEALTH CARE REFORM AND ONCOLOGY<br />
development and implementation <strong>of</strong> cancer survivorship<br />
care plans. 11<br />
Cancer <strong>Clinical</strong> Trials<br />
The ACA mandates coverage <strong>of</strong> routine costs for patients<br />
who participate in cancer clinical trials. Insurers are prohibited<br />
from dropping or limiting coverage for participants in<br />
cancer clinical trials.<br />
Improving Quality and Lowering Costs<br />
The ACA establishes a national pilot program to encourage<br />
hospitals, doctors, and other providers to work together<br />
to improve the coordination and quality <strong>of</strong> patient care.<br />
Under payment bundling, hospitals, doctors, and providers<br />
are paid a flat rate for an episode <strong>of</strong> care such as a cancer<br />
diagnosis, rather than the current fee-for-service system.<br />
The entire oncology team is compensated with a “bundled”<br />
payment that provides incentives to deliver health care<br />
services more efficiently while maintaining or improving<br />
quality <strong>of</strong> care. Bundled payments may be the most striking<br />
change resulting from the ACA that oncologists will<br />
experience.<br />
ACA Impact on <strong>Oncology</strong><br />
The stated goals <strong>of</strong> ACA are to expand access and control<br />
health care costs. However, although the expansion <strong>of</strong> access<br />
is relatively easy to quantify, the impact on health care costs<br />
is less clear. On the surface, increasing demand for any good<br />
or service typically results in higher costs in a free market<br />
unless supply is similarly increased. Both expanded coverage<br />
through the ACA and the projected increase in cancer<br />
cases as a result <strong>of</strong> demographic shifts in the U.S. population<br />
(e.g., aging) are likely to increase demand for oncology goods<br />
and services over time. In contrast, on the supply side, an<br />
oncology physician workforce shortage is projected, 12 novel<br />
drugs with complex production requirements may be scarce<br />
after initial approval, 13 and shortages <strong>of</strong> commonly used<br />
generic drugs are already widespread and projected to<br />
increase over time. 14 One <strong>of</strong> the key questions then is how,<br />
in this context, the impetus to control costs will affect care in<br />
the clinic.<br />
To the extent that the ACA provides oncologists with more<br />
resources to care for patients, the ACA may ease ethical<br />
tensions. An oncologist faced with a patient in clinic with a<br />
treatable disease or symptoms has an ethical obligation to<br />
provide care or to help the patient obtain care elsewhere.<br />
<strong>Oncology</strong> practices are currently faced with the dilemma <strong>of</strong><br />
how to care for patients without adequate resources. 15 Few<br />
would argue that such differences in access or outcomes on<br />
the basis <strong>of</strong> ability to pay alone are morally acceptable.<br />
However, there is legitimate debate regarding how society<br />
KEY POINTS<br />
● The ACA contains provisions that are relevant to<br />
oncology.<br />
● Health care reform poses ethical implications for<br />
oncology providers.<br />
● Specific stakeholders in the oncology community will<br />
face a variety <strong>of</strong> relevant issues.<br />
can best address such disparities and the extent to which the<br />
ACA will really help. At the center <strong>of</strong> the decades-old<br />
controversy over health care reform in the United States<br />
is the question <strong>of</strong> whether the goals <strong>of</strong> providing access<br />
and improving health outcomes are best achieved through<br />
a) government-provided health care, as is the dominant<br />
model in the British National Health Service, b) governmentfinanced<br />
health care, as in Canada, c) increased government<br />
regulation <strong>of</strong> health care and health insurance, as established<br />
by the ACA, or d) deregulation and greater freedom<br />
within the medical care and medical insurance market<br />
place, as proposed by some critics <strong>of</strong> the recent health care<br />
law.<br />
At the level <strong>of</strong> the individual provider, regardless <strong>of</strong> the<br />
basis for insurance coverage, any increase in the amount <strong>of</strong><br />
resources available to care for an individual patient presenting<br />
with cancer should provide a greater opportunity to meet<br />
our ethical obligation to provide care. There are several<br />
other features <strong>of</strong> the new law that will enhance our opportunity<br />
to provide cancer care. As described herein, insurers<br />
will be unable to refuse coverage to patients with preexisting<br />
conditions and there will be no lifetime cap on coverage—<br />
critical issues for patients with cancer. There will be no cost<br />
sharing for recommended cancer screenings such as mammography<br />
and cervical cancer screening. In addition, in a<br />
provision that has received little attention, the law establishes<br />
a federal right to timely independent external appeal<br />
<strong>of</strong> adverse coverage decisions. This could be particularly<br />
important for patients with rare cancers or rare presentations<br />
<strong>of</strong> cancer requiring <strong>of</strong>f-label therapy.<br />
Ethical Challenges in Cost Control<br />
The predominant ethical challenges facing oncologists will<br />
likely come from increasing pressure on oncologists to act<br />
both in their patient’s interest and simultaneously, in the<br />
interest <strong>of</strong> society, the government, the insurer, or the<br />
accountable care organization (ACO) in helping control<br />
costs. These pressures are likely to be most acute in the<br />
newly established ACOs, in which there are direct financial<br />
incentives for controlling the cost <strong>of</strong> health care, but will<br />
likely extend to consideration <strong>of</strong> care for all patients. Not<br />
only is the entire success <strong>of</strong> this experiment in expanded<br />
health care coverage dependent on financial sustainability,<br />
but oncologists are likely to directly feel the impact <strong>of</strong> any<br />
ongoing failure to shift the cost curve.<br />
The law establishes a new Independent Advisory Board<br />
charged with controlling Medicare spending. Given that<br />
most potential remedies for rising costs are explicitly prohibited,<br />
such as rationing, cuts in services, cost sharing, or<br />
changes in hospital reimbursement, it appears likely that<br />
rising costs will be met with cuts in reimbursement to<br />
providers. One <strong>of</strong> the issues likely to face providers is simply<br />
how (and whether) to practice in an era <strong>of</strong> lowered reimbursement.<br />
Will we protect revenue at the expense <strong>of</strong> time<br />
with patients or provision <strong>of</strong> ancillary services? Will this<br />
affect individual treatment decisions? There may be very<br />
personal answers to these questions for each oncologist.<br />
The goal, <strong>of</strong> course, is to provide higher-quality care at<br />
lower cost by focusing on the value <strong>of</strong> the care we provide. 16<br />
In one <strong>of</strong> few studies to demonstrate this potential, Neubauer<br />
and colleagues found that adherence to lung cancer<br />
guidelines can reduce costs substantially with no detriment<br />
in survival. 17 The potentially high costs associated with<br />
e5
non–evidence-based cancer care have also been documented.<br />
18 However, it is likely that in some settings we will<br />
face real trade-<strong>of</strong>fs between marginal benefits <strong>of</strong> an intervention<br />
and additional cost. For example, one year <strong>of</strong> trastuzumab<br />
for HER2-positive breast cancer reduces the risk <strong>of</strong><br />
recurrence by close to 50% compared with chemotherapy<br />
alone, but at an additional cost <strong>of</strong> approximately $50,000 per<br />
patient. 19 Recent data in both the neoadjuvant and metastatic<br />
breast cancer settings suggest that additional interventions<br />
above and beyond trastuzumab, such as lapatinib<br />
and pertuzumab, may further improve outcomes for some<br />
patients. 20,21 If additional benefit is confirmed in large<br />
adjuvant randomized trials, one can imagine a scenario in<br />
which we are forced to decide on further improving outcomes<br />
versus doubling or tripling the cost <strong>of</strong> therapy. In fact,<br />
throughout oncology, novel interventions are improving outcomes,<br />
but in many cases such progress comes at a substantial<br />
price. 13,22<br />
If, as is currently the case, interventions continue to be<br />
approved on the basis <strong>of</strong> marginal benefits in efficacy, how<br />
will oncologists decide which interventions truly bring<br />
meaningful clinical benefit? What role will patient preference<br />
or shared decision making have in an era <strong>of</strong> increased<br />
pressure to control costs? If nothing else, the recent decision<br />
by the U.S. Food and Drug Administration (FDA) to withdraw<br />
approval for bevacizumab in breast cancer demonstrated<br />
lack <strong>of</strong> consensus on what constitutes value in<br />
oncology among both clinicians and patients.<br />
Many <strong>of</strong> these tensions exist within the current health<br />
care system. They may be magnified if there is an assumption<br />
<strong>of</strong> universal access to cancer care that is not matched<br />
by the actual reimbursement for services. Will patients with<br />
different forms <strong>of</strong> insurance within the new system be<br />
treated differently? Will treatment decisions vary for patients<br />
who are within or outside <strong>of</strong> an ACO?<br />
In the setting <strong>of</strong> expanded access and pressure to control<br />
costs, oncologists may face ethical challenges related both to<br />
how they practice and whom they are willing to treat. With<br />
more insured patients there may be greater demand for<br />
oncology services, but as noted above, it is expected that<br />
important differences in the adequacy <strong>of</strong> coverage for cancer<br />
care will persist. Will the expanded cohort <strong>of</strong> patients with<br />
Medicaid and continued low reimbursement levels paradoxically<br />
push a greater number <strong>of</strong> oncologists to close their<br />
practices to Medicaid? Where low reimbursement for a small<br />
percentage <strong>of</strong> patients might be subsidized within a large<br />
practice, the potential economic impact <strong>of</strong> a larger share <strong>of</strong><br />
such patients might prove to be too great a risk in some<br />
settings.<br />
In addition, the ACO model <strong>of</strong> rewarding physicians and<br />
practices for improving the quality, coordination, and cost <strong>of</strong><br />
care may create a disincentive to treat more complex patients<br />
who may require care outside <strong>of</strong> the expected norms.<br />
On the basis <strong>of</strong> unusual disease presentation, tolerance <strong>of</strong><br />
standard therapies, or comorbidities, some patients may<br />
require oncologists to either step outside <strong>of</strong> the standard<br />
care plans and consider options that might be more expensive<br />
than the norm. As noted herein, this will even be<br />
facilitated under the new law through a strengthened right<br />
to appeal coverage decisions. There would seem to be a clear<br />
incentive to both provide treatment for patients with less<br />
complex disease, whose expected costs <strong>of</strong> care are likely to<br />
e6<br />
be at or below the average, and to constrain our decision<br />
making for individual patients.<br />
There are clearly some pitfalls ahead, primarily in the<br />
area <strong>of</strong> cost containment, and no easy solutions. The question<br />
will be how to find the balance in our practices between<br />
advocacy for our patient’s interests and preferences and<br />
some effort to consider wise and efficient use <strong>of</strong> resources.<br />
Failure to strike this balance may leave us with a “tragedy<br />
<strong>of</strong> the commons” in which individual decisions to ignore<br />
societal costs <strong>of</strong> care results in unsustainability <strong>of</strong> the entire<br />
system, and threatens our ability to guarantee high-quality<br />
cancer care to all patients.<br />
The ACA and <strong>Oncology</strong> Stakeholders<br />
As demonstrated in the preceding section, the ACA promises<br />
to change the landscape <strong>of</strong> health care for virtually<br />
every stakeholder. Insurance reforms will inevitably lead<br />
to changes that will dramatically affect the composition <strong>of</strong><br />
populations <strong>of</strong> patients with cancer around the country. A<br />
focus on health service delivery innovation, comparative<br />
effectiveness research (CER), and reducing health care disparities<br />
will shift the pr<strong>of</strong>ile <strong>of</strong> cancer research.<br />
<strong>Oncology</strong> Workforce<br />
MOY, ABERNETHY, AND PEPPERCORN<br />
In addition to the ethical considerations and the projected<br />
increased demand for oncology care outlined above, improvements<br />
in access to health care will affect the oncology<br />
workforce. Support for new models <strong>of</strong> health care delivery<br />
under the ACA will help meet escalating demand for cancer<br />
care. The ACA established the Center for Medicare &<br />
Medicaid Innovation (CMI) to test innovative payment and<br />
service delivery models to reduce Medicare and Medicaid<br />
expenditures while improving quality <strong>of</strong> care. Beginning<br />
with its Health Care Innovation Challenge program in <strong>2012</strong>,<br />
CMI will provide funding to demonstrate rapidly deployable<br />
new models for reducing total costs <strong>of</strong> care while improving<br />
quality and health outcomes; proposed models will likely<br />
address access and volume in the context <strong>of</strong> resource constraints.<br />
ACA encourages models, such as the medical home and<br />
ACO, which are designed to optimize care and largely focus<br />
on community-based delivery through which the majority<br />
<strong>of</strong> patients access cancer care. A common theme among<br />
the various emerging delivery models is coordination <strong>of</strong><br />
care. Models for which ACA authorizes funding include<br />
(1) community-based collaborative care networks—consortia<br />
<strong>of</strong> health care providers that include a safety-net hospital,<br />
have a joint governance structure, and provide comprehensive<br />
coordinated and integrated services to low-income<br />
populations; (2) interdisciplinary, interpr<strong>of</strong>essional teams <strong>of</strong><br />
health care providers who work with primary care providers<br />
to provide integrated community-based care; and (3) community<br />
health centers, which serve an estimated one in<br />
three low-income people and one in four low-income minority<br />
individuals. Provisions pertaining to medical homes<br />
allow states, under Medicaid, to make medical assistance<br />
payments at an enhanced federal match to teams <strong>of</strong> providers.<br />
Clearly, under the ACA, coordinated care and clinician<br />
teamwork will increasingly become the norm.<br />
Provider organizations are encouraged, through the ACA,<br />
to expand and diversify their workforce. Through the Centers<br />
for Disease Control and Prevention (CDC), the law
HEALTH CARE REFORM AND ONCOLOGY<br />
authorizes support <strong>of</strong> community health workers at hospitals,<br />
federally qualified health centers, and other public or<br />
nonpr<strong>of</strong>it entities. The ACA also authorizes grants for various<br />
workforce diversity measures including recruitment <strong>of</strong><br />
individuals from under-represented, disadvantaged, or rural<br />
backgrounds into health pr<strong>of</strong>essions; community-based<br />
training and education, with an emphasis on primary care<br />
in underserved areas; and community-based field placements<br />
or preceptorships. 3 The law encourages development<br />
<strong>of</strong> curricula for cultural competency programs, and inclusion<br />
<strong>of</strong> competency measures within quality measurement systems.<br />
3<br />
Hospital and Private Practice Administration<br />
The ACA will have substantial administrative impact on<br />
providers (both organizations and individual practitioners).<br />
Hospitals and physicians will be required to collect and<br />
report quality data, with penalties imposed for not reporting<br />
or not meeting quality standards. These requirements, although<br />
initially burdensome, may encourage provider participation<br />
in important quality initiatives such as ASCO’s<br />
national quality improvement program and clinical registry,<br />
the Quality <strong>Oncology</strong> Practice Initiative (QOPI). The ACA<br />
reflects a national trend toward linking health care financing<br />
to quality metrics. For example, under ACA, providers <strong>of</strong><br />
pediatric medical care who meet specified requirements can<br />
be recognized by states as ACOs, and will receive incentive<br />
payments based on metrics specified by the U.S. Department<br />
<strong>of</strong> Health and Human Services.<br />
Insurers<br />
Under the ACA, payers such as health insurers will lose a<br />
substantial degree <strong>of</strong> selectivity. As stated previously, they<br />
will be required to accept all new applications for health<br />
insurance and to renew individuals’ coverage, regardless <strong>of</strong><br />
preexisting conditions. They will no longer be allowed to<br />
cancel coverage or to set a lifetime amount that a policy will<br />
pay for certain conditions, and annual limits will be restricted.<br />
Health insurance exchanges will create an environment<br />
<strong>of</strong> open competition, although the extent to which<br />
plans will differ remains to be seen.<br />
Cancer Researchers<br />
Following on the heels <strong>of</strong> the <strong>American</strong> Reinvestment and<br />
Recovery Act (ARRA) <strong>of</strong> 2009, the ACA extends the impact <strong>of</strong><br />
health care reform on medical research. Academic oncologists<br />
will continue to see federal funds designated for CER,<br />
which is characterized by direct comparison <strong>of</strong> existing<br />
interventions, enrollment <strong>of</strong> “real world” populations (i.e.,<br />
patients as they are typically seen in day-to-day clinical<br />
care), and diversity <strong>of</strong> study design. The mantra <strong>of</strong> “getting<br />
the right care to the right patient at the right time” guides<br />
CER. Substantial ARRA funding has already supported<br />
CER studies. Building on CER momentum ignited by ARRA,<br />
the ACA established the Patient-Centered Outcomes Research<br />
Institute (PCORI); this nonpr<strong>of</strong>it, nongovernmental<br />
entity promotes CER by identifying research priorities,<br />
establishing and maintaining a research agenda, and providing<br />
funding for research.<br />
Further, the CER agenda will be facilitated by a national<br />
transition to electronic medical records, data availability,<br />
and data use, supported by the Health Information Technology<br />
for Economic and <strong>Clinical</strong> Health (HITECH) Act <strong>of</strong> 2009,<br />
a part <strong>of</strong> ARRA. Critical to the Office <strong>of</strong> the National<br />
Coordinator’s strategy is to develop the information substrate<br />
needed to support a system <strong>of</strong> rapid learning health<br />
care, a goal well aligned with the rapid learning cancer<br />
care agenda <strong>of</strong> the oncology community. 23 The HITECH Act<br />
will affect the cancer community broadly, predominantly<br />
through the requirement to transition to electronic medical<br />
records and the use <strong>of</strong> data for quality monitoring, and will<br />
affect the cancer research community quite specifically<br />
through the provision <strong>of</strong> clinical data to support CER,<br />
annotation <strong>of</strong> biospecimens, and clinical trials.<br />
Aspects <strong>of</strong> the ACA will aid oncology researchers by<br />
helping overcome barriers to clinical research. As noted<br />
above, to remove an economic barrier and thereby facilitate<br />
participation in clinical trials, ACA prohibits health insurers<br />
from denying coverage <strong>of</strong> routine costs associated with<br />
clinical trial participation and from discriminating against<br />
patients participating in clinical trials. Because rates <strong>of</strong><br />
racial and ethnic minority participation in cancer clinical<br />
trials are disproportionately low, 24 this provision may prove<br />
particularly beneficial to the quality and generalizability <strong>of</strong><br />
cancer research. Health services research will be facilitated<br />
by a requirement that all federally funded public health and<br />
health care programs collect data on race, ethnicity, sex,<br />
primary language, disability, and geographic data at the<br />
lowest level (i.e., state, local, institutional) where they can<br />
be aggregated; oncologists seeking to study disparities will<br />
benefit from this provision.<br />
Academic oncologists will see further evidence <strong>of</strong> federal<br />
commitment to reducing disparities as the National Center<br />
on Minority Health and Health Disparities (NCMHHD)<br />
becomes a new institute within the National Institutes <strong>of</strong><br />
Health (NIH). Created by the ACA, this institute will hold<br />
responsibility for planning and coordinating all NIHsupported<br />
health disparities research. The law also moves<br />
the Office <strong>of</strong> Minority Health from the Public Health Service<br />
to HHS, and establishes parallel <strong>of</strong>fices in several HHS<br />
agencies including the CDC, FDA, and Agency for Healthcare<br />
Research and Quality (AHRQ)—further highlighting<br />
health disparity research as a national priority while endeavoring<br />
to coordinate approaches to studying disparities.<br />
Conclusion<br />
Health care reform has the potential to improve cancer<br />
care in the United States by making it more accessible,<br />
affordable, and comprehensive. The law contains many<br />
provisions that may meaningfully change clinical outcomes,<br />
especially for the most vulnerable <strong>American</strong>s diagnosed<br />
with cancer. The vast majority <strong>of</strong> patients with cancer will<br />
now have insurance coverage and will no longer face the risk<br />
<strong>of</strong> losing coverage or reaching a lifetime cap on benefits,<br />
regardless <strong>of</strong> changes in employment or personal wealth.<br />
However, the ACA may exacerbate existing practical and<br />
ethical challenges in oncology practice as we attempt to<br />
deliver the highest-quality cancer care to an expanded pool<br />
<strong>of</strong> patients on a financially sustainable basis. This law, and<br />
the opportunities and challenges that accompany it, affect<br />
every stakeholder in cancer care, from patient to clinicians<br />
to payers to researchers. Going forward, it is imperative that<br />
the entire oncology community has a clear vision for maintaining<br />
and improving the quality <strong>of</strong> care within the new<br />
framework for access, coordination, and cost control established<br />
under ACA.<br />
e7
Authors’ Disclosures <strong>of</strong> Potential Conflicts <strong>of</strong> Interest<br />
Author<br />
Employment or<br />
Leadership<br />
Positions<br />
Consultant or<br />
Advisory Role<br />
Beverly Moy GlaxoSmithKline<br />
(U); Pfizer (U)<br />
Amy P. Abernethy Bristol-Myers<br />
Squibb; Helsinn<br />
Jeffrey M. Peppercorn GlaxoSmithKline (I) AVEO; Bayer;<br />
Genentech<br />
1. Patient Protection and Affordable Care Act. Public Law 111-148. http://<br />
www.gpo.gov/fdsys/pkg/PLAW-111publ148/html/PLAW-111publ148.htm. Accessed<br />
February 17, <strong>2012</strong>.<br />
2. Health Care and Education Reconciliation Act <strong>of</strong> 2010. Public Law<br />
111-152. http://www.gpo.gov/fdsys/pkg/PLAW-111publ152/html/PLAW-111<br />
publ152.htm. Accessed February 17, <strong>2012</strong>.<br />
3. Moy B, Polite BN, Halpern MT, et al. <strong>American</strong> <strong>Society</strong> <strong>of</strong> <strong>Clinical</strong><br />
<strong>Oncology</strong> policy statement: opportunities in the patient protection and affordable<br />
care act to reduce cancer care disparities. J Clin Oncol. 2011;29:3816-<br />
3824<br />
4. DeNavas-Walt C, Proctor BD, Smith JC. US Census Bureau, Current<br />
Population Reports. Income, Poverty, and Health Insurance Coverage in the<br />
United States: 2008. Washington, DC: U.S. Government Printing Office,<br />
Economics and Statistics Administration, 2009.<br />
5. Schoen C, Collins SR, Kriss JL, et al. How many are underinsured?<br />
trends among U.S. adults, 2003 and 2007. Health Aff (Milwood). 2008;27:<br />
w298-w309.<br />
6. Elemendorf DW. Letter to the Honorable Nancy Pelosi. http://www.<br />
cbo.gov/ftpdocs/113xx/doc11379/AmendReconProp.pdf. Accessed February 16,<br />
2011.<br />
7. Ayanian JZ, Kohler BA, Abe T, et al. The relation between health<br />
insurance coverage and clinical outcomes among women with breast cancer.<br />
N Engl J Med. 1993;329:326-331.<br />
8. Kelz RR, Gimotty PA, Polsky D, et al. Morbidity and mortality <strong>of</strong><br />
colorectal carcinoma surgery differs by insurance status. Cancer. 2004;101:<br />
2187-2194.<br />
9. Roetzheim RG, Gonzalez EC, Ferrante JM, et al. Effects <strong>of</strong> health<br />
insurance and race on breast carcinoma treatments and outcomes. Cancer.<br />
2000;89:2202-2213.<br />
10. Roetzheim RG, Pal N, Gonzalez EC, et al. Effects <strong>of</strong> health insurance<br />
and race on colorectal cancer treatments and outcomes. Am J Pub Health.<br />
2000;90:1746-1754.<br />
11. From Cancer Patient to Cancer Survivor: Lost in Translation. Washington,<br />
DC: The National Academies Press, 2005.<br />
e8<br />
Stock<br />
Ownership Honoraria<br />
REFERENCES<br />
Research<br />
Funding<br />
Amgen; Novartis Amgen; BioVex;<br />
Bristol-Myers<br />
Squibb; DARA;<br />
Eisai; Helsinn;<br />
KangLaiTe; Lilly;<br />
Pfizer<br />
Genentech;<br />
Novartis<br />
MOY, ABERNETHY, AND PEPPERCORN<br />
Expert<br />
Testimony<br />
Other<br />
Remuneration<br />
12. Hortobagyi GN, <strong>American</strong> <strong>Society</strong> <strong>of</strong> <strong>Clinical</strong> <strong>Oncology</strong>. A shortage <strong>of</strong><br />
oncologists? The <strong>American</strong> <strong>Society</strong> <strong>of</strong> <strong>Clinical</strong> <strong>Oncology</strong> workforce study.<br />
J Clin Oncol. 2007;25:1468-1469.<br />
13. Peppercorn J, Armstrong A, Zaas DW, et al. Rationing in urologic<br />
oncology: lessons from sipuleucel-T for advanced prostate cancer. Urol Oncol.<br />
Epub <strong>2012</strong> Feb 3.<br />
14. Gatesman ML, Smith TJ. The shortage <strong>of</strong> essential chemotherapy<br />
drugs in the United States. N Engl J Med. 2011;365:1653-1655.<br />
15. Ward E, Jemal A, Cokkinides V, et al. Cancer disparities by race/<br />
ethnicity and socioeconomic status. CA Cancer J Clin. 2004;54:78-93.<br />
16. Schnipper LE, Meropol NJ, Brock DW. Value and cancer care: toward<br />
an equitable future. Clin Cancer Res. 2010;16:6004-6008.<br />
17. Neubauer MA, Hoverman JR, Kolodziej M, et al. Cost effectiveness <strong>of</strong><br />
evidence-based treatment guidelines for the treatment <strong>of</strong> non-small-cell lung<br />
cancer in the community setting. J Oncol Pract. 2010;6:12-18.<br />
18. de Souza JA, Polite BN, Zhu S, et al. Utilization and costs <strong>of</strong> nonevidence-based<br />
(non-EBM) antineoplastic agents in patients with metastatic<br />
colon cancer (mCC). J Clin Oncol 2009;29 (suppl; abstr 6002)<br />
19. Kurian AW, Thompson RN, Gaw AF, et al. A cost-effectiveness analysis<br />
<strong>of</strong> adjuvant trastuzumab regimens in early HER2/neu-positive breast cancer.<br />
J Clin Oncol. 2007;25:634-641.<br />
20. Baselga J, Bradbury I, Eidtmann H, et al. Lapatinib with trastuzumab<br />
for HER2-positive early breast cancer (NeoALTTO): a randomised, openlabel,<br />
multicentre, phase 3 trial. Lancet, Epub <strong>2012</strong> Jan 16.<br />
21. Baselga J, Cortés J, Kim SB, et al. Pertuzumab plus trastuzumab plus<br />
docetaxel for metastatic breast cancer. N Engl J Med. <strong>2012</strong>;366:109-119.<br />
22. Fojo T, Grady C. How much is life worth: cetuximab, non-small cell lung<br />
cancer, and the $440 billion question. J Natl Cancer Inst. 2009;101:1044-<br />
1048.<br />
23. Abernethy AP, Etheredge LM, Ganz PA, et al. Rapid-learning system<br />
for cancer care. J Clin Oncol 28:4268-4274<br />
24. Park ER, Weiss ES, Moy B. Recruiting and enrolling minority patients<br />
to cancer clinical trials. Community <strong>Oncology</strong>. 2007;4:254-257.
INTERNATIONAL VARIATION IN UNDERSTANDING<br />
ONCOLOGISTS’ PROFESSIONAL DUTIES TOWARD<br />
PATIENTS, FAMILIES, AND THEMSELVES<br />
CHAIR<br />
Amy P. Abernethy, MD<br />
Duke University Medical Center<br />
Durham, NC<br />
SPEAKERS<br />
Antonella Surbone, MD, PhD<br />
New York University<br />
New York, NY<br />
Simon Wein, MD<br />
David<strong>of</strong>f Cancer Center, Rabin Medical Center<br />
Petah Tikva, Israel
A Balanced Approach to Physician<br />
Responsibilities: Oncologists’ Duties<br />
toward Themselves<br />
Overview: Although critical to the provision <strong>of</strong> best patient<br />
care, physician self-care is an underattended aspect <strong>of</strong> responsibility<br />
in the medical pr<strong>of</strong>essions, including oncology.<br />
Neglecting self-care bears negative consequences for the<br />
individual oncologist, ranging from burnout and fatigue to interpersonal<br />
and relationship stress, addiction, and disruptive<br />
behavior. It may also contribute to medical errors, disinterest<br />
in or depersonalization <strong>of</strong> patient care, and lower quality <strong>of</strong><br />
care. Because <strong>of</strong> its effect on physicians, patients, and the<br />
health care environment, physician self-care is increasingly<br />
recognized as an important pr<strong>of</strong>essional responsibility.<br />
OFTEN CONSTRUED as one’s obligation toward others<br />
or used interchangeably with the concepts <strong>of</strong> reliability<br />
and dependability, responsibility is a quality that defines<br />
both our competence and the caliber and character <strong>of</strong> the<br />
care we provide.<br />
The notion <strong>of</strong> responsibility is inherent in medicine. We<br />
enter the pr<strong>of</strong>ession taking the Hippocratic Oath, which,<br />
whether in its original or modernized form, lays out a set <strong>of</strong><br />
responsibilities to which the new physician agrees to abide.<br />
Medical training inculcates in us a sense <strong>of</strong> responsibility<br />
first and foremost toward our patients and secondarily toward<br />
our colleagues, discipline, institution, and scientific field.<br />
Less well defined, if even recognized, have been the physician’s<br />
responsibilities to himself or herself. Selfless service<br />
has been held up as an ideal in health care, whereas a focus<br />
on one’s self has been viewed with skepticism, even stigma.<br />
Is a definition <strong>of</strong> responsibility that heavily emphasizes<br />
duties toward others, <strong>of</strong>ten at the expense <strong>of</strong> the self and its<br />
reasonable needs, realistic or even beneficial for all involved?<br />
A disproportionate focus on others, without adequate<br />
attention to self, results in negative consequences for<br />
not only the oncologist but also patients and other stakeholders<br />
in the health care environment.<br />
Evidence and Consequences <strong>of</strong> Inadequate Self-Care<br />
Much as a computer that is run continuously, without<br />
periodic shut-downs and reboots to allow for system updates,<br />
does not run at optimal efficiency using the latest programs,<br />
or much as an athlete who trains continuously, without<br />
taking rest days for muscle recovery and mental recharge,<br />
fails to achieve or sustain peak performance, the physician<br />
who neglects to take care <strong>of</strong> himself or herself will not<br />
function at his/her best with respect to patient care, research,<br />
education and training, or administrative tasks. This neglect<br />
ultimately manifests in burnout and/or impairment.<br />
A familiar concept, burnout results from tension over time<br />
between what the individual is doing and is capable <strong>of</strong> doing<br />
and what he or she is expected to do. It is experienced as a<br />
progressive “erosion <strong>of</strong> the soul.” In physicians, burnout has<br />
been called “the silent anguish <strong>of</strong> healers”—felt as anguish,<br />
perhaps, because <strong>of</strong> a painful awareness that what we feel<br />
called to do as medical pr<strong>of</strong>essionals diverges from what we<br />
are actually accomplishing and because <strong>of</strong> an escalating<br />
inability to care about, or even to meet, our responsibilities<br />
as pr<strong>of</strong>essionals and healers.<br />
By Amy P. Abernethy, MD<br />
Nonetheless, pr<strong>of</strong>essional obligations, competing demands<br />
on time, and personal priorities conspire to prevent a large<br />
proportion <strong>of</strong> oncologists from adequately attending to selfcare<br />
in even simple ways, such as getting sufficient exercise<br />
and sleep. This chapter discusses the need for physician<br />
self-care and the repercussions <strong>of</strong> not meeting this fundamental<br />
responsibility. Self-care is described in the context <strong>of</strong> three<br />
life domains: pr<strong>of</strong>essional, personal (physical, psychological,<br />
mental, and spiritual), and interpersonal (relationships, family,<br />
social, and community). Strategies are provided for caring for<br />
the self in each domain.<br />
Physician burnout is characterized by three main dimensions:<br />
(1) emotional exhaustion, cynicism, and/or depersonalization<br />
in relationships with coworkers and/or patients, 1<br />
manifesting as detachment from one’s job; (2) low sense <strong>of</strong><br />
personal accomplishment; and (3) perceived clinical ineffectiveness.<br />
2 Any or all <strong>of</strong> these factors might be present in the<br />
affected physician. 3 The gradual deterioration that leads to<br />
physician burnout typically starts in medical school, residency,<br />
or fellowship, although its symptoms may not present<br />
until midcareer. 4 Signs and symptoms <strong>of</strong> burnout are<br />
many 3,5 : overwhelming physical and emotional exhaustion,<br />
overidentification or overinvolvement with patients, irritability<br />
and hypervigilance, sleep problems, social withdrawal,<br />
pr<strong>of</strong>essional and personal boundary violations, poor judgment,<br />
perfectionism and rigidity, questioning the meaning<br />
<strong>of</strong> life, interpersonal conflicts, avoidance <strong>of</strong> emotionally difficult<br />
clinical situations, numbness and detachment, difficulty<br />
in concentrating, frequent illness (e.g., headaches and<br />
gastrointestinal disturbances), and immune system impairment.<br />
Although attention to these signs and symptoms, in one’s<br />
self and in colleagues, is the first-line approach to identifying<br />
burnout, this symptom can be diagnosed with validated<br />
assessment instruments. The most commonly used and<br />
well-recognized tool for this purpose is the 22-item Maslach<br />
Burnout Inventory–Human Services Survey, which generates<br />
subscales for emotional exhaustion, depersonalization,<br />
and personal accomplishment. 6 Using this instrument as<br />
the criterion for burnout assessment, studies have documented<br />
that half to two-thirds <strong>of</strong> oncologists experience<br />
emotional exhaustion. 7,8 Among practicing physicians more<br />
generally, studies have reported a 46% to 80% prevalence <strong>of</strong><br />
moderate to high levels <strong>of</strong> emotional exhaustion, a 22% to<br />
93% prevalence <strong>of</strong> moderate to high levels <strong>of</strong> depersonalization,<br />
and a 16% to 79% prevalence <strong>of</strong> low to moderate levels<br />
<strong>of</strong> personal achievement. 9 Burnout does not only affect the<br />
From the Division <strong>of</strong> Medical <strong>Oncology</strong>, Department <strong>of</strong> Medicine, and the Duke Cancer<br />
Institute, Duke University Medical Center, Durham, NC.<br />
Author’s disclosures <strong>of</strong> potential conflicts <strong>of</strong> interest are found at the end <strong>of</strong> this article.<br />
Address reprint requests to Amy P. Abernethy, MD, Duke University Medical Center, Box<br />
3436, Durham, NC 27710; e-mail: amy.abernethy@duke.edu.<br />
© <strong>2012</strong> by <strong>American</strong> <strong>Society</strong> <strong>of</strong> <strong>Clinical</strong> <strong>Oncology</strong>.<br />
1092-9118/10/1-10<br />
e9
midcareer physician; in one cross-sectional study, more than<br />
three-fourths <strong>of</strong> internal medicine residents exhibited burnout.<br />
10<br />
Risk factors for physician burnout have been described<br />
across medical disciplines. Demographic factors associated<br />
with increased risk <strong>of</strong> burnout include younger age and<br />
junior status, female sex, and unmarried status. 11 Personality<br />
factors predisposing physicians to burnout include high<br />
level <strong>of</strong> motivation and intense investment in one’s pr<strong>of</strong>ession<br />
12 ; personality traits, such as intensity, impulsivity, and<br />
compulsiveness; previous mental health problems, particularly<br />
depression; and wishful thinking as one’s coping style.<br />
Environmental or contextual risk factors include work overload;<br />
perception <strong>of</strong> a lack <strong>of</strong> control over one’s workload,<br />
when that workload exceeds capacity; lack <strong>of</strong> social support<br />
from colleagues; dissatisfaction with resources at work;<br />
practice within a high-risk medical specialty (such as oncology);<br />
work stressors, including reimbursement issues,<br />
complex regulations, and interactions with insurers; organizational<br />
concerns and management style; and family stress.<br />
The risk <strong>of</strong> burnout increases proportionally with time spent<br />
in direct patient care; time away and a mix <strong>of</strong> job responsibilities<br />
are protective. 8 Care for people who are dying and<br />
disease that is not responsive to aggressive treatment, both<br />
prominent in oncology, are sources <strong>of</strong> burnout that accumulate<br />
over time, especially when the physician feels illequipped<br />
to provide expert end-<strong>of</strong>-life care or experiences<br />
a sense <strong>of</strong> failure when disease progresses. Spirituality and<br />
religiosity have been found to exert a protective effect<br />
against burnout.<br />
Multiple dire consequences result from physician burnout<br />
and from inadequate care <strong>of</strong> the self more generally. Among<br />
physicians, substance abuse rates <strong>of</strong> 10% to 15% have been<br />
reported, 13 although physicians tend to use alcohol and<br />
prescription medications rather than illicit substances. 14<br />
Physician fatigue is cited in the Institute <strong>of</strong> Medicine’s<br />
seminal report, To Err is Human, as a factor contributing to<br />
the tens <strong>of</strong> thousands <strong>of</strong> medical errors occurring in the<br />
United States each year. 15 A prospective, longitudinal study<br />
found increased burnout and reduced empathy to be associated<br />
with increased likelihood <strong>of</strong> committing a self-perceived<br />
KEY POINTS<br />
● Self-care encompasses all strategies implemented to<br />
support one’s own optimal health and well-being.<br />
● Because failure to attend to self-care negatively affects<br />
patients both directly and indirectly (e.g., depersonalization<br />
leading to lower empathy or physician<br />
exit from the pr<strong>of</strong>ession), self-care is a vital responsibility<br />
<strong>of</strong> every oncologist.<br />
● Self-care can be conceptualized as a set <strong>of</strong> activities<br />
that one uses on a regular basis to promote fulfillment<br />
and wellness in the pr<strong>of</strong>essional, personal, and<br />
interpersonal domains <strong>of</strong> life.<br />
● A vast array <strong>of</strong> options for self-care, and supporting<br />
facilitators, are available in each domain.<br />
● Self-care strategies must be personalized, with sensitivity<br />
to cultural nuances and other individuallyspecific<br />
determinants <strong>of</strong> best approach.<br />
e10<br />
medical error in the subsequent 3 months. 16 Failure <strong>of</strong><br />
physician self-care to establish well-being is evidenced in<br />
high suicide rates. A meta-analysis that included 25 studies<br />
reported that male physicians commit suicide at a rate 40%<br />
higher than that <strong>of</strong> men in the general public, and female<br />
physicians do so at 130% the rate <strong>of</strong> women in general. 17<br />
Among physicians 55 years and younger, higher perceived<br />
stress is associated with (1) lower satisfaction levels, which<br />
in turn are related to intention to quit, decrease work hours,<br />
change specialty, or leave direct patient care, and (2) poorer<br />
mental health, which in turn is related to intention to leave<br />
direct patient care. 18 Other potential consequences <strong>of</strong><br />
inadequate physician self-care include impaired job performance,<br />
poor physical health (e.g., headaches, sleep disturbance,<br />
hypertension, and cardiac events), poor emotional<br />
health and well-being (e.g., irritability, fatigue, anxiety, and<br />
depression), and relational and marital difficulties.<br />
Ultimately, it is not only the oncologist’s self that is<br />
affected. Physically, emotionally, mentally, or spiritually<br />
depleted physicians can neither take good care <strong>of</strong> patients<br />
nor model good health or a lifestyle conducive to health for<br />
their patients. Physicians tend to counsel patients in ways<br />
that are consistent with their own health habits, 19 and if<br />
these habits are not sound, patients receive bad advice.<br />
Furthermore, as oncologists, we are not just caretakers <strong>of</strong><br />
people with cancer, we are also members <strong>of</strong> our own families<br />
and circles <strong>of</strong> care; physicians experiencing burnout cannot<br />
contribute as well to family life and other valued relationships,<br />
and consequences can be disastrous for all involved.<br />
Pr<strong>of</strong>essional Recognition <strong>of</strong> Physician Burnout and<br />
Need for Self-Care<br />
The <strong>American</strong> Medical Association (AMA) Code <strong>of</strong> Medical<br />
Ethics, Opinion 9.0305, states, “To preserve the quality <strong>of</strong><br />
their performance, physicians have a responsibility to maintain<br />
their health and wellness, construed broadly as preventing<br />
or treating acute or chronic diseases, including<br />
mental illness, disabilities, and occupational stress.” A variety<br />
<strong>of</strong> resources are available to support physicians’ attention<br />
to their own health. The AMA produces the Physician’s<br />
Guide to Personal Health Toolkit, which covers healthy<br />
eating, physical activity, reduced alcohol consumption, and<br />
smoking cessation, as well as the AMA Physician Health<br />
e-Letter. The biennial International Conference on Physician<br />
Health, sponsored by the AMA, Canadian Medical Association,<br />
and British Medical Association, presents research and<br />
original presentations on topics such as burnout and peer<br />
support, physician health as linked to quality and patient<br />
safety, resilience and work-life balance, and physical and<br />
mental health.<br />
Despite the availability <strong>of</strong> supportive resources, backed by<br />
injunctions from respected agencies, many oncologists remain<br />
unclear about, neglect, or postpone active strategies to<br />
take optimal care <strong>of</strong> themselves. In a 2009 survey <strong>of</strong> California<br />
physicians (n � 1,875), 53% reported moderate to<br />
severe stress, 35% reported no or occasional exercise, 34%<br />
reported getting 6 hours or less <strong>of</strong> sleep per night, and 27%<br />
never or occasionally ate breakfast. 20<br />
Conceptual Framework<br />
AMY P. ABERNETHY<br />
Clarity in conceptualizing what, specifically, is entailed in<br />
physician self-care may facilitate individuals’ responsible
SELF-CARE FOR ONCOLOGISTS<br />
Fig. 1. Consequences <strong>of</strong> self-care or no<br />
self-care for self and patients.<br />
planning and decisions. The World Health Organization<br />
defines wellness as “the optimal state <strong>of</strong> health <strong>of</strong> individuals<br />
and groups” and explains that there are “two focal<br />
concerns: the realization <strong>of</strong> the fullest potential <strong>of</strong> an individual<br />
physically, psychologically, socially, spiritually, and<br />
economically, and the fulfillment <strong>of</strong> one’s role expectations in<br />
the family, community, place <strong>of</strong> worship, workplace, and<br />
other settings.” 21 Thus defined, a physician’s duties to himself<br />
or herself span the totality <strong>of</strong> life experience and affect<br />
not only himself or herself but also those with whom he or<br />
she interacts.<br />
The oncologist’s responsibilities to self can be conceptualized<br />
in three domains: pr<strong>of</strong>essional, personal, (physical,<br />
psychological, mental, and spiritual), and interpersonal (relational,<br />
family, social, and community) (Fig. 1). The following<br />
sections briefly describe each domain and provide<br />
pragmatic approaches to self-care within each. Because<br />
many <strong>of</strong> the basic elements <strong>of</strong> self-care are straightforward<br />
and likely familiar to most oncologists, only simple guidelines<br />
and principles are presented, with discussion <strong>of</strong> facilitators<br />
for their implementation. There is no one-size-fits-all<br />
approach; each oncologist, knowing himself or herself best,<br />
must craft an approach that fits well with his or her own<br />
preferences, characteristics, and circumstances.<br />
Duties to Self within the Three Domains <strong>of</strong><br />
Responsibility<br />
Pr<strong>of</strong>essional. Most, if not all, oncologists enter medicine<br />
with a strong sense <strong>of</strong> pr<strong>of</strong>essional responsibility. Upholding<br />
this initial commitment is integral to positive self-esteem<br />
and a sense <strong>of</strong> personal integrity. Activities to maintain and<br />
further develop one’s pr<strong>of</strong>essional competence, such as conferences<br />
and symposia, and to pursue one’s own pr<strong>of</strong>essional<br />
interests, such as committee work and research, <strong>of</strong>fer mechanisms<br />
for cultivating pr<strong>of</strong>essional competence. Ensuring a<br />
mix <strong>of</strong> responsibilities and some time away from patient care<br />
is important. Goal setting in the pr<strong>of</strong>essional domain is a<br />
method <strong>of</strong> articulating personal standards <strong>of</strong> excellence and<br />
holding one’s self accountable to personally meaningful<br />
progress—a form <strong>of</strong> self-care.<br />
Because caring for people at or near the end <strong>of</strong> life can be<br />
especially burdensome, exacting a toll in terms <strong>of</strong> the individual’s<br />
mental and emotional resources, specific strategies<br />
to build palliative care skills can help with self-care. Some<br />
strategies are to hone one’s skills and confidence in delivering<br />
bad news; obtain palliative care–specific training<br />
through conferences and training videos; develop relationships<br />
with palliative care colleagues and call on them<br />
frequently; debrief after difficult cases and losses, and take<br />
time to acknowledge that “this is hard”; recognize that some<br />
patients affect us more than others, and that this is normal;<br />
and finally, when a death occurs, take time to reflect, allow<br />
yourself to grieve, attend the funeral, reconnect with the<br />
family, write a letter, and smile at the gift <strong>of</strong> knowing the<br />
person when alive.<br />
In the pr<strong>of</strong>essional setting, we must care for ourselves and<br />
the teams within which we work. A few strategies for care <strong>of</strong><br />
self and others in the work environment are to manage the<br />
volume <strong>of</strong> time spent in patient care; ensure that the team<br />
has mechanisms in place to debrief and share stressful<br />
experiences; allow staff time to get together outside the<br />
workplace; role model good personal and pr<strong>of</strong>essional habits<br />
(e.g., time <strong>of</strong>f for important family events and prioritization<br />
<strong>of</strong> exercise) and promote these for all members <strong>of</strong> the team;<br />
and be attuned to any signs <strong>of</strong> burnout among team members<br />
and educate them to be alert to these signs in self and<br />
others. A well-functioning, personally healthy team provides<br />
an important support network for the oncologist which can<br />
undergird self-care for each individual and for the team as a<br />
whole.<br />
Personal. Table 1 presents recommendations for personal<br />
self-care in the domains <strong>of</strong> physical, psychological and cognitive,<br />
and spiritual wellness. Although most oncologists<br />
will agree to the needs listed and to the corresponding<br />
recommendations and although it is important to promote<br />
these to patients, far fewer actually succeed in integrating<br />
these or other self-care strategies into daily living. (Table 1<br />
is intended not to provide a comprehensive list <strong>of</strong> self-care<br />
needs, recommendations, or strategies but rather to <strong>of</strong>fer<br />
an initial set <strong>of</strong> practical suggestions. Knowing himself or<br />
e11
Table 1. Recommendations for Personal Self-Care<br />
Domain Need Recommendations Facilitating Strategies<br />
Physical Sleep Aim for 7–9 hours per night Set a “black out” time beyond which you don’t<br />
answer the telephone or email<br />
Try sleep hygiene before resorting to sleep<br />
medications<br />
Discontinue work at a set time each day<br />
Sleep in a totally darkened room Keep electronics (e.g., television, laptop, and<br />
Maintain consistent bedtime and waking hours<br />
Minimize bedroom stimulation<br />
Avoid working in the bedroom<br />
mobile telephone) out <strong>of</strong> the bedroom<br />
Psychological and<br />
cognitive<br />
Nutrition Eat a diet featuring lean proteins (e.g., poultry, fish,<br />
eggs, and unsweetened yogurt), vegetables, fruits,<br />
nuts and seeds, legumes (e.g., black beans), and<br />
moderate portions <strong>of</strong> whole grains<br />
Buy a week’s lunches from the salad bar on the<br />
weekend or have healthy lunches delivered<br />
Avoid trans and hydrogenated fats Set up a healthy breakfast the night before<br />
Avoid stimulants and substances that affect<br />
Pack healthy snacks (e.g., nuts and fruit)<br />
metabolism (e.g., sugars, caffeine, and alcohol)<br />
Supplement to ensure adequate vitamins and<br />
Carry a water bottle<br />
antioxidants<br />
Spread caloric intake across the day, beginning with<br />
breakfast<br />
Make use <strong>of</strong> grocery stores’ list-filling services,<br />
including online shopping<br />
Drink at least eight 8-oz glasses <strong>of</strong> water per day Monitor your c<strong>of</strong>fee and alcohol intake<br />
Exercise Get at least 30 minutes <strong>of</strong> exercise on 3 days or<br />
more per week<br />
Work out with a personal trainer<br />
Mix types <strong>of</strong> exercise to maintain strength (e.g.,<br />
bodyweight exercises), flexibility (e.g., yoga), and<br />
cardiovascular conditioning (e.g., lap swimming)<br />
Exercise before work<br />
Choose the outdoors, when possible Make standing exercise dates with friends<br />
Take a refreshing walk or jog outside<br />
Balance Be aware <strong>of</strong> the valued domains in your life (e.g.,<br />
work, family, friends, and external commitments)<br />
Work with a life coach<br />
Take stock <strong>of</strong> what fulfills you in each domain Write down realistic goals; post them prominently<br />
and review frequently<br />
Stress reduction and<br />
relaxation<br />
Spiritual Connection to a higher<br />
power<br />
e12<br />
Allocate time to (and prioritize) outside interests<br />
(e.g., creative, intellectual, athletic, and community)<br />
Set aside 10 minutes before bedtime or in the<br />
early morning to journal, meditate, or reflect<br />
Practice a relaxation technique (e.g., meditation and Schedule a regular massage<br />
deep breathing)<br />
Use complementary alternative medicine approaches Make friend/family activities fun and light (e.g.,<br />
watch a comedy)<br />
Engage in a relaxing sport or hobby Walk (even briefly) under the stars at day’s end<br />
Spend time outdoors<br />
Mental rejuvenation Feed your sense <strong>of</strong> humor<br />
Play games<br />
Take regular vacations<br />
Activate your sense <strong>of</strong> humor Find a “brain game” (e.g., sudoko or cross-word<br />
puzzle) that appeals to you and make it a<br />
regular practice<br />
Take time away from work—completely away<br />
Engage in activities that transport you to another<br />
realm (e.g., reading fiction)<br />
Join a drop-in or online book club with a realistic<br />
schedule<br />
Sense <strong>of</strong> meaning and<br />
purpose<br />
If you are religious, attend services at a church,<br />
temple, or synagogue<br />
Pray, meditate, or engage in a meaningful spiritual<br />
practice regularly<br />
Sign up to usher to ensure your attendance at<br />
services<br />
Take minibreaks at points throughout the day to<br />
“listen” within<br />
Spend time outdoors Visit a gallery or art museum<br />
Find sources <strong>of</strong> beauty to appreciate Listen to classical music on your commute<br />
Read a poem at bedtime<br />
Consider what makes you feel most fulfilled During 1 week, take notes <strong>of</strong> times when you feel<br />
fulfilled and purposeful<br />
Identify ways that you can attend to these aspects <strong>of</strong><br />
your life<br />
AMY P. ABERNETHY<br />
Plan to incorporate this/these things into daily life<br />
on a regular basis<br />
Prioritize these commitments Share this plan with someone who cares about<br />
Reward yourself for living your values<br />
you to support accountability to self
SELF-CARE FOR ONCOLOGISTS<br />
herself best, each oncologist is encouraged to adapt this sort<br />
<strong>of</strong> list into a realistic and effective self-care plan.)<br />
Given oncologists’ hectic schedules and competing demands,<br />
these recommendations can be challenging to implement.<br />
It is useful, therefore, to identify and enlist<br />
facilitators. For example, having a regularly scheduled personal<br />
trainer can help ensure that strength training occurs;<br />
making a standing weekly commitment to walk, run, or bike<br />
with a friend can help ensure cardiovascular exercise. Engagement<br />
<strong>of</strong> a housecleaning service can liberate weekend<br />
time for hobbies and other personally fulfilling activities. A<br />
life coach can help establish time management practices<br />
that support balance and identify pragmatic ways to reach<br />
lifestyle and balance goals. Something as simple as purchasing<br />
a water bottle and carrying it at all times can help<br />
remind a busy physician to stay well hydrated and set a<br />
positive example for staff and patients.<br />
Stress reduction warrants special mention as key to avoiding<br />
burnout. There are many options for alleviating or<br />
reducing stress, defined by individual character and interests.<br />
Spending time outdoors; meditating; participating in a<br />
favorite sport; pursuing a hobby, such as playing a musical<br />
instrument; simple relaxation, such as reading a novel in the<br />
evening; and activating one’s sense <strong>of</strong> humor all represent<br />
effective approaches to stress reduction while also meeting<br />
self-fulfillment needs. A complementary and alternative<br />
medicine practitioner, such as a good acupuncturist or<br />
massage therapist, can become a valued ally in stress<br />
reduction. Resources such as meditation, relaxation, or<br />
guided imagery CDs are readily available and can make<br />
stress reduction both convenient and enjoyable. There are<br />
many forms <strong>of</strong> deep relaxation, including listening to ocean<br />
waves or soothing music, sitting or walking meditation,<br />
bi<strong>of</strong>eedback, transcendental meditation, visualization,<br />
mindfulness meditation, neur<strong>of</strong>eedback, and progressive<br />
muscle relaxation. The restful alertness (relaxation response)<br />
induced by these interventions differs neurophysiologically<br />
from the sleep state; it enhances mental clarity<br />
and lowers body catechol levels, cortisol level, oxygen demand,<br />
blood pressure, and pulse.<br />
Although more than 90% <strong>of</strong> <strong>American</strong>s believe in God or a<br />
universal higher power, 22 spiritual self-care is not commonly<br />
recognized as one <strong>of</strong> the oncologist’s duties. Some<br />
individuals may believe that the stress management recommendations<br />
listed in this chapter adequately address their<br />
needs in the spiritual domain. Others, however, must engage<br />
in a regular spiritual practice (e.g., daily prayer or<br />
meditation or weekly attendance at religious services) to feel<br />
that they are attending well to their spiritual dimension.<br />
Finding meaning and higher purpose, beyond individual<br />
satisfaction and gain, in one’s pr<strong>of</strong>ession can support a sense<br />
<strong>of</strong> spiritual fulfillment. Self-care in this deeply personal<br />
domain requires self-knowledge and a certain level <strong>of</strong> intimacy<br />
with one’s self; cultivation <strong>of</strong> inner connectedness in<br />
whatever way the individual chooses can satisfy self-care<br />
needs and support well-being.<br />
Interpersonal. Like all people, oncologists live within a<br />
social context, a web <strong>of</strong> relationships. The extent to which<br />
the oncologist attends to his or her responsibilities to self<br />
therefore has an effect on others. Family commitments to<br />
significant others, spouses, and/or children require, above<br />
all else, time—a commodity in short supply in the oncology<br />
pr<strong>of</strong>ession. Inability or failure to meet commitments (im-<br />
plicit or explicit) to one’s family can induce considerable<br />
dissatisfaction and distress in all involved. A vital part <strong>of</strong><br />
oncologists’ self-care is thus attending to relationships.<br />
In spending time with others, both quality and quantity<br />
are important; no level <strong>of</strong> quality time, however exceptional<br />
in quality, can compensate for spending the bare minimum<br />
quantity <strong>of</strong> time with loved ones. Critical is the ability, in<br />
whatever time is available, to be truly present with others,<br />
rather than multitasking with mobile device in hand or<br />
being so exhausted as to barely engage in conversation.<br />
Strategies for ensuring that relationship responsibilities are<br />
met include family rituals, such as game night or Saturday<br />
breakfast out, and a regular date night with a spouse or<br />
partner. Making a commitment to a family dinner at least 5<br />
or 6 nights per week can be considered an evidence-based<br />
intervention; studies have shown that family dinners are<br />
associated with healthier eating patterns in older children<br />
and adolescents, 23 healthy adolescent development, and less<br />
engagement in risk behaviors. 24<br />
Social support is an important facilitator <strong>of</strong> well-being in<br />
busy individuals and those with plenty <strong>of</strong> time for socializing.<br />
Friends and acquaintances provide an emotional outlet<br />
and an opportunity to give and receive ideas, thoughts,<br />
feedback, concern, and affirmation. Although technology<br />
and social networking provide endless opportunities for<br />
connecting with others, there is no substitute for face-to-face<br />
encounters in conveying a true sense <strong>of</strong> support.<br />
Recognizing burnout in one’s self and colleagues and<br />
taking responsible action are important interpersonal duties<br />
<strong>of</strong> each oncologist. Signs and symptoms include the following:<br />
frequent irritability or anger, depression, disengagement<br />
and depersonalization, apathy, chronic fatigue and/or<br />
depressive symptoms, and substance abuse (e.g., alcohol,<br />
recreational drugs, and inappropriately used pharmaceutical<br />
drugs). These are neither normal nor acceptable attributes<br />
<strong>of</strong> an oncologist. On identifying these symptoms, seek<br />
a psychologist’s or counselor’s help for one’s self or, in the<br />
case <strong>of</strong> a colleague, either refer the individual to a practitioner<br />
who can provide direct help or report the concern to a<br />
senior supervisor or associate, as appropriate.<br />
In medicine, a system for accountability and attention to<br />
each other is <strong>of</strong>ten lacking. When a colleague is demonstrating<br />
unhealthy behaviors, one’s responsibility and course <strong>of</strong><br />
action are <strong>of</strong>ten unclear. Whom should we tell? How can we<br />
enlist help? Can we ensure appropriate follow-up? Although<br />
clear lines <strong>of</strong> authority and accountability exist in other<br />
disciplines, such as nursing, we in medicine are frequently<br />
unsure where to turn. Although lack <strong>of</strong> line management<br />
affords us a valued measure <strong>of</strong> autonomy and flexibility, it<br />
also leaves us unclear <strong>of</strong> how to maneuver when a colleague<br />
is troubled. In the absence <strong>of</strong> clear directives, oncologists<br />
should take seriously the responsibility to attend to each<br />
other and should discuss within their pr<strong>of</strong>essional teams a<br />
system for encouraging one another’s well-being and intervening<br />
when warning signs emerge. Until such systems<br />
become explicit aspects <strong>of</strong> the oncologists’ workplace, it will<br />
remain important to share a common understanding <strong>of</strong> this<br />
responsibility.<br />
Different Cultural Approaches to Duties to One’s Self<br />
An individual’s exploration into and definition <strong>of</strong> his or her<br />
duties to self will depend not only on personal factors, such<br />
as individual values, preferences, capacities, and circum-<br />
e13
stances (e.g., stage <strong>of</strong> career, family needs, and social configuration),<br />
but also on his or her cultural background. There<br />
is a paucity <strong>of</strong> literature examining cross-cultural differences<br />
in self-care between U.S. and foreign-born physicians<br />
and between members <strong>of</strong> racial/ethnic majority compared<br />
with minority populations. Studies have made cross-cultural<br />
comparisons <strong>of</strong> attitudes toward self-care between the<br />
United States and Germany and explored differences in<br />
U.S. and Swedish constructs such as self-concept and wellbeing.<br />
25 For current purposes, it is important to be aware<br />
that one’s own concept <strong>of</strong> self, definition <strong>of</strong> personal wellbeing<br />
in the domains discussed herein, and preferred selfcare<br />
strategies will be tempered by individual background—<br />
family, upbringing, culture, and life experiences. 26<br />
Conclusion<br />
Self-care is not purely a private matter; the extent to<br />
which oncologists attend to, or deny, the needs <strong>of</strong> their selves<br />
has repercussions in the personal, interpersonal, and pr<strong>of</strong>essional<br />
spheres. Burnout is a prevalent and serious phenomenon,<br />
directly related to inadequate self-care. This<br />
chapter presents a flexible framework to help the oncologist<br />
consider, and responsibly attend to, duties toward himself or<br />
Author’s Disclosures <strong>of</strong> Potential Conflicts <strong>of</strong> Interest<br />
Author<br />
Employment or<br />
Leadership<br />
Positions<br />
Consultant or<br />
Advisory Role<br />
Amy P. Abernethy Bristol-Myers<br />
Squibb; Helsinn<br />
1. Maslach C, Leiter MP. Early predictors <strong>of</strong> job burnout and engagement.<br />
J Appl Psychol. 2008;93:498-512.<br />
2. Maslach C. Job burnout: new directions in research and intervention.<br />
Curr Direct Psychol Sci. 2003;12:189-192.<br />
3. Maslach C, Schaufeli WB, Leiter MP. Job burnout. Annu Rev Psychol.<br />
2001;52:397-422.<br />
4. Spickard A Jr, Gabbe SG, Christensen JF. Mid-career burnout in<br />
generalist and specialist physicians. JAMA. 2002;288:1447-1450.<br />
5. Vachon ML. Staff stress in hospice/palliative care: a review. Palliat Med.<br />
1995;9:91-122.<br />
6. Maslach C, Jackson SE. The measurement <strong>of</strong> experienced burnout. J<br />
Occup Behav. 1981;2:99-113.<br />
7. Grunfeld E, Whelan TJ, Zitzelsberger L, Willan AR, Montesanto B,<br />
Evans WK. Cancer care workers in Ontario: prevalence <strong>of</strong> burnout, job stress<br />
and job satisfaction. CMAJ. 2000;163:166-169.<br />
8. Allegra CJ, Hall R, Yothers G. Prevalence <strong>of</strong> burnout in the U.S.<br />
<strong>Oncology</strong> community: results <strong>of</strong> a 2003 survey. J Oncol Pract. 2005;1:140-147.<br />
9. Chopra SS, Sotile WM, Sotile MO. Physician burnout. JAMA. 2004;291:<br />
633.<br />
10. Shanafelt TD, Bradley KA, Wipf JE, Back AL. Burnout and selfreported<br />
patient care in an internal medicine residency program. Ann Intern<br />
Med. 2002;136:358-367.<br />
11. Kearney MK, Weininger RB, Vachon ML, Harrison RL, Mount BM.<br />
Self-care <strong>of</strong> physicians caring for patients at the end <strong>of</strong> life: “Being connected<br />
...akeytomysurvival.” JAMA, 2009;301:1155-1164, E1.<br />
12. Leiter MP, Maslach C. A mediation model <strong>of</strong> job burnout. In: Antoniou<br />
CC, Cooper CL (eds). Research Companion to Organizational Health Psychology.<br />
Cheltenham, United Kingdom: Edward Elgar Publishing; 2005.<br />
13. Gastfriend DR. Physician substance abuse and recovery: what does it<br />
mean for physicians-and everyone else? JAMA. 2005;293:1513-1515.<br />
14. Hughes PH, Brandenburg N, Baldwin DC Jr., et al. Prevalence <strong>of</strong><br />
substance use among US physicians. JAMA. 1992;267:2333-2339.<br />
e14<br />
herself. Currently, metrics are not used to track oncologists’<br />
responsibility in caring for themselves; rather, performance<br />
measures typically evaluate patient care and other externally<br />
focused duties. With the understanding that we have a<br />
de facto honor system with respect to self-care, it is incumbent<br />
on every oncologist to adopt strategies for ensuring his<br />
or her health and well-being. Short-term and long-term<br />
benefits will be evident in multiple domains, including<br />
physical, mental, and psychological health, job satisfaction,<br />
and quality <strong>of</strong> life. In addition, our patients will benefit from<br />
having healthier, more present and available physicians.<br />
Acknowledgments<br />
Complete Funding Disclosure: As <strong>of</strong> February <strong>2012</strong>, Dr.<br />
Abernethy has research funding from the U.S. National Institutes<br />
<strong>of</strong> Health, U.S. Agency for Healthcare Research and<br />
Quality, Robert Wood Johnson Foundation, Pfizer, Eli Lilly,<br />
Bristol Meyers Squibb, Helsinn Therapeutics, Amgen, Kanglaite,<br />
Alexion, Biovex, DARA Therapeutics, Novartis, and<br />
Mi-Co; these funds are all distributed to Duke University<br />
Medical Center to support research. In the last 2 years she has<br />
had nominal consulting agreements (�$10,000) with Helsinn<br />
Therapeutics, Amgen, and Novartis.<br />
Stock<br />
Ownership Honoraria<br />
REFERENCES<br />
Research<br />
Funding<br />
Amgen; Novartis Amgen; BioVex;<br />
Bristol-Myers<br />
Squibb; DARA;<br />
Eisai; Helsinn;<br />
KangLaiTe; Lilly;<br />
Pfizer<br />
Expert<br />
Testimony<br />
AMY P. ABERNETHY<br />
Other<br />
Remuneration<br />
15. Institute <strong>of</strong> Medicine. To Err is Human: Building a Safer Health<br />
System. Washington, DC: National Academy <strong>of</strong> Sciences; 1999.<br />
16. West CP, Huschka MM, Novotny PJ, et al. Association <strong>of</strong> perceived<br />
medical errors with resident distress and empathy: a prospective longitudinal<br />
study. JAMA. 2006;296:1071-1078.<br />
17. Schernhammer ES, Colditz GA. Suicide rates among physicians: a<br />
quantitative and gender assessment (meta-analysis). Am J Psychiatry. 2004;<br />
161:2295-2302.<br />
18. Williams ES, Konrad TR, Scheckler WE, et al. Understanding physicians’<br />
intentions to withdraw from practice: the role <strong>of</strong> job satisfaction, job<br />
stress, mental and physical health. Health Care Manage Rev. 2001;26:7-19.<br />
19. Schwartz JS, Lewis CE, Clancy C, Kinosian MS, Radany MH, Koplan<br />
JP. Internists’ practices in health promotion and disease prevention. A<br />
survey. Ann Intern Med. 1991;114:46-53.<br />
20. Bazargan M, Makar M, Bazargan-Hejazi S, Ani C, Wolf KE. Preventive,<br />
lifestyle, and personal health behaviors among physicians. Acad Psychiatry.<br />
2009;33:289-295.<br />
21. World Health Organization. Health Promotion Glossary Update. Geneva,<br />
Switzerland: World Health Organization; 2000.<br />
22. Pew Charitable Trusts. US religious landscape survey. In: Pew Forum<br />
on Religion and Public Life. Philadelphia, PA: Pew Charitable Trusts; 2008.<br />
23. Gillman MW, Rifas-Shiman SL, Frazier AL, et al. Family dinner and<br />
diet quality among older children and adolescents. Arch Fam Med. 2000;9:<br />
235-240.<br />
24. Fulkerson JA, Story M, Mellin A, Leffert N, Neumark-Sztainer D,<br />
French SA. Family dinner meal frequency and adolescent development:<br />
relationships with developmental assets and high-risk behaviors. J Adolesc<br />
Health. 2006;39:337-345.<br />
25. Whetstone WR, Hansson AM. Perceptions <strong>of</strong> self-care in Sweden: a<br />
cross-cultural replication. J Adv Nursing. 1989;14:962-969.<br />
26. Whetstone WR. Perceptions <strong>of</strong> self-care in East Germany: a crosscultural<br />
empirical-investigation. J Adv Nursing. 1987;12:167-176.
Are Oncologists Accountable Only to<br />
Patients or Also to Their Families?<br />
An International Perspective<br />
By Antonella Surbone, MD, PhD, and Lea Baider, PhD<br />
Overview: In most societies, health pr<strong>of</strong>essionals traditionally<br />
carry responsibility only toward their patients. However,<br />
this is not the case in all cultures. In the contemporary<br />
practice <strong>of</strong> oncology in Western cultures, there is a shift<br />
toward assuming broader responsibility for patients with cancer’<br />
families during the illness course, the grieving stage, and<br />
in cancer prevention and genetic counseling.<br />
Traditional family, community, and religious values play a<br />
central role in determining people’s perceptions and attitudes<br />
toward life and death as well as toward caregiving for a sick<br />
relative. The meaning <strong>of</strong> cancer illness within the family<br />
culture is thus influenced not only by each individual’s values<br />
CANCER HAS become a global health emergency because<br />
<strong>of</strong> the aging <strong>of</strong> the world population that is<br />
occurring at unprecedented rates and portending an everincreasing<br />
demand for care. It is expected that by 2030, the<br />
US elderly population—conventionally defined as the population<br />
segment composed by people over age 65—will double<br />
and reach about 70 million people. 1 As the population ages,<br />
morbidity for chronic illness and disability increases, with<br />
consequent escalating demands for informal care delivered<br />
by family members or close friends. Oncologists are asked to<br />
recognize the key role <strong>of</strong> family in each patient’s experience<br />
<strong>of</strong> cancer and progressively assume more responsibilities<br />
toward families, during the entire course <strong>of</strong> the illness, from<br />
diagnosis to survivorship or end-<strong>of</strong>-life, as well as in the<br />
grieving stages. 2<br />
A family’s experience with cancer, including caregiving,<br />
always occurs in a particular cultural milieu: the traditional<br />
and religious values <strong>of</strong> each community play a central role in<br />
determining people’s perceptions and attitudes toward illness<br />
and death, as well as toward caregiving to a sick<br />
relative. The meaning <strong>of</strong> cancer is thus influenced not only<br />
by each individual’s values and beliefs but also by the<br />
family’s convergence and diversity <strong>of</strong> meaning, as well as<br />
their taboos and secrets. Consequently, in many cultures<br />
physicians have always included patients, families, and<br />
communities in their pr<strong>of</strong>essional responsibility. 3<br />
Although Western physicians’ pr<strong>of</strong>essional duty has traditionally<br />
being only toward their patients, today’s medicine<br />
increasingly acknowledges the need for health pr<strong>of</strong>essionals<br />
to involve the family in patient-centered care and to develop<br />
a functional system <strong>of</strong> collaboration and partnership with<br />
patients and their families. 2,4<br />
Patient- and Family-Centered Care in the United States<br />
The Institute <strong>of</strong> Medicine has identified the concept <strong>of</strong><br />
patient-centered care as one <strong>of</strong> the keys to improving health<br />
and quality <strong>of</strong> care in the United States, including efficiency<br />
and costs. 3 In patient-centered care, physicians clearly have<br />
responsibilities toward both the patient and the family. The<br />
<strong>American</strong> Academy <strong>of</strong> Family Physicians defines “family” as<br />
“a group <strong>of</strong> individuals with a continuing legal, genetic,<br />
and/or emotional relationship.” <strong>Society</strong>, the Academy states,<br />
“relies on the family group to provide for the economic and<br />
and beliefs but also by the family’s makeup and dynamics, as<br />
well as their taboos and secrets.<br />
Global cancer care should therefore be directed at the<br />
family as a unit, while respecting patient autonomy and<br />
privacy. This reappraisal <strong>of</strong> our traditional understanding <strong>of</strong><br />
physicians’ duty as solely directed at the patient is reflected<br />
in the recent US trend toward a patient- and family-centered<br />
care approach. An additional challenge for oncology pr<strong>of</strong>essionals<br />
is to integrate and tailor interventions toward the<br />
needs <strong>of</strong> both care recipients and caregivers and relate it to<br />
this dyad as the basic and enduring unit <strong>of</strong> care.<br />
protective needs <strong>of</strong> individuals, especially children and the<br />
elderly.” 5<br />
The <strong>American</strong> Academy <strong>of</strong> Pediatrics and the <strong>American</strong><br />
College <strong>of</strong> Emergency Physicians support the model <strong>of</strong><br />
patient- and family-centered care (PFCC). The latter defines<br />
the PFCC as “an approach to health care that recognizes the<br />
role <strong>of</strong> the family in providing medical care, encourages<br />
collaboration between the patient, family, and health care<br />
pr<strong>of</strong>essionals, and honors individual and family strengths,<br />
cultures, traditions, and expertise.” 6 This approach has<br />
proven effective in pediatric hospital and home care and<br />
emergency settings and is now being extended to adult<br />
care. 7<br />
According to The Institute for Patient- and Family-<br />
Centered Care, the patient- and family-centered approach<br />
focuses on the partnerships between patients, families, and<br />
providers at the clinical level. Patients and families are<br />
engaged as advisors and leaders in developing new policies<br />
on quality and safety. Practitioners recognize the vital role<br />
<strong>of</strong> family members <strong>of</strong> patients <strong>of</strong> all ages and acknowledge<br />
that emotional, social, and developmental support are integral<br />
components <strong>of</strong> health care. The definition <strong>of</strong> family, as<br />
well as the degree <strong>of</strong> the family’s involvement in health care,<br />
is determined by the competent patient, who maintains<br />
control <strong>of</strong> decisions concerning his or her own health care. 8,9<br />
In its expert report on cancer pain relief and palliative<br />
care, the World Heath Organization states the importance <strong>of</strong><br />
enhancing the quality <strong>of</strong> life <strong>of</strong> both patients and family<br />
members. 10 Teno et al developed a conceptual model <strong>of</strong><br />
patient-focused, family-centered medicine in end-<strong>of</strong>-life care,<br />
where the patient is at the center <strong>of</strong> care and the family,<br />
defined as “all people—whether or not they are blood relatives—with<br />
whom the patient has a very close or intimate<br />
From the Department <strong>of</strong> Medicine, New York University Medical School, New York, NY;<br />
and the Hebrew University Medical School, Sharett Institute <strong>of</strong> <strong>Oncology</strong>, Hadassah<br />
University Hospital, Jerusalem, Israel.<br />
Authors’ disclosures <strong>of</strong> potential conflicts <strong>of</strong> interest are found at the end <strong>of</strong> this article.<br />
Address reprint requests Antonella Surbone, MD, PhD, Department <strong>of</strong> Medicine, New<br />
York University Medical School, 550 First Ave., New York, NY 10016; email: antonella.<br />
surbone@gmail.com.<br />
© <strong>2012</strong> by <strong>American</strong> <strong>Society</strong> <strong>of</strong> <strong>Clinical</strong> <strong>Oncology</strong>.<br />
1092-9118/10/1-10<br />
e15
elationship” is also acknowledged. This approach is based<br />
on sharing information with family members on what to<br />
expect and the skills necessary to help care for the patient at<br />
home, providing desired physical comfort and emotional<br />
support, and promoting shared decision making focused on<br />
the patient, while also attending to the needs <strong>of</strong> family<br />
members.<br />
Role <strong>of</strong> the Family and Informal Caregiving in <strong>Oncology</strong><br />
Approximately 28.8 million adults in the United States<br />
are now family caregivers. 1,12 Sidebar 1 lists are some<br />
relevant data on the demographics <strong>of</strong> care in the United<br />
States, where unpaid caregiving is the largest source <strong>of</strong><br />
long-term care services (Sidebar 1).<br />
Caring is fundamental to human survival. 13 It is understood<br />
to imply a distinct way <strong>of</strong> being, believing, and acting<br />
that calls for commitment, emotional involvement, knowledge,<br />
and new coping skills. It motivates families and gives<br />
new meaning and structure to life. 14 At the same time,<br />
caregiving imposes a great burden on families at psychologic,<br />
social, and financial levels. As many as two-thirds <strong>of</strong><br />
caregivers report physical or mental health problems owing<br />
to caregiving, and nearly 50% have a chronic health condition<br />
that requires ongoing medical care. 1<br />
Because <strong>of</strong> the progressive shift from hospital to ambulatory<br />
or home cancer care, informal family caregivers are<br />
<strong>of</strong>ten required to replace skilled health care workers in the<br />
delivery <strong>of</strong> complex care to their ill family members. They<br />
KEY POINTS<br />
● Although Western health pr<strong>of</strong>essionals traditionally<br />
carry responsibility only toward their patients, the<br />
physicians’ duty in other cultures extends to families<br />
and communities.<br />
● Today’s Western medicine increasingly acknowledges<br />
the need for health pr<strong>of</strong>essionals to include the family<br />
in patient-centered care and to develop a functional<br />
system <strong>of</strong> collaboration and partnership with patients<br />
and their families.<br />
● The <strong>American</strong> Academy <strong>of</strong> Family Physicians defines<br />
“family” as “a group <strong>of</strong> individuals with a continuing<br />
legal, genetic, and/or emotional relationship” and<br />
acknowledges that society “relies on the family group<br />
to provide for the economic and protective needs <strong>of</strong><br />
individuals, especially children and the elderly.”<br />
● As informal caregiving by family members is rapidly<br />
increasing worldwide, oncologists are asked to assume<br />
broader responsibility toward their patients’<br />
families during the course <strong>of</strong> the patient’s illness and<br />
in the grieving stage.<br />
● Cancer is a “family illness” that can only be understood<br />
in terms <strong>of</strong> the social, cultural, religious, and<br />
family responses to the underlying disease, including<br />
the ways in which the patient and family conceptualize<br />
and elaborate it. For this reason, the family is<br />
increasingly recognized as the basic social and ethical<br />
unit <strong>of</strong> care.<br />
e16<br />
SURBONE AND BAIDER<br />
Sidebar 1. Demographics <strong>of</strong> Care in the United<br />
States<br />
● Fifty-two million informal/family caregivers provide<br />
care to an ill or disabled person 20 years and<br />
older.<br />
● Thirty-four million adults (16% <strong>of</strong> population)<br />
provide care to persons 50 years and older.<br />
● Between 6 to 7 million family caregivers provide<br />
care to adults 65 years and older who need<br />
assistance with everyday activities.<br />
● It is estimated there will be 37 million informal<br />
caregivers by 2050, an increase <strong>of</strong> 85% from 2000.<br />
are asked to perform multidimensional tasks, including<br />
treatment monitoring; treatment-related symptom management;<br />
emotional, financial, and spiritual support; and assistance<br />
with personal and instrumental care. 15,16<br />
Family members <strong>of</strong> patients with cancer strive to make<br />
sense <strong>of</strong> the unforeseen illness and its intrusion by enduring<br />
and reinterpreting the meaning <strong>of</strong> each other’s experience<br />
and finding a shared meaning within the private family<br />
culture. 17 Cultural expectations also regulate and influence<br />
basic meanings in the course <strong>of</strong> the socialization process and<br />
the roles that each member might assume. Even though<br />
caregiving can be a deeply rewarding spiritual experience <strong>of</strong><br />
connection with a loved one in need, caregivers <strong>of</strong> patients<br />
with cancer <strong>of</strong>ten perceive themselves as living a shrinking<br />
life, being forced to take responsibilities beyond what they<br />
feel comfortable doing, struggling to keep their home as a<br />
home, and no longer feeling the joy <strong>of</strong> being together with<br />
their family. 18,19 Sidebar 2 lists some <strong>of</strong> the positive and<br />
negative behavioral changes that can arise within a family<br />
when a member has cancer.<br />
To optimize the quality <strong>of</strong> caregivers’ lives and their<br />
ability to support and enhance the quality <strong>of</strong> care delivered<br />
to each patient with cancer, we must take into account the<br />
extent to which providing care results in physical, emotional,<br />
spiritual, and financial burdens for them and the<br />
ways in which these burdens can be addressed. Empirical<br />
studies show that the burden placed on family caregivers<br />
has negative effects on the quality <strong>of</strong> life <strong>of</strong> both the patients<br />
and their caregivers, particularly during advanced stages <strong>of</strong><br />
cancer. 20 Pr<strong>of</strong>essional interventions should therefore be directed<br />
at the family as a unit and not aimed solely at making<br />
caregiving easier. The challenge for oncology pr<strong>of</strong>essionals is<br />
to integrate and tailor interventions according to the specific<br />
needs <strong>of</strong> the care recipient and caregiver and relate to this<br />
Sidebar 2. Behavioral Changes in Response to<br />
Cancer in a Family Member<br />
● New coping styles and strengths<br />
● Discovery <strong>of</strong> meaning<br />
● Increased cohesiveness and spirituality<br />
● Stress and crises following recovery<br />
● Complaints, anger, and withdrawal
ONCOLOGISTS’ DUTIES TO PATIENT AND FAMILIES<br />
dyad as the basic and enduring unit <strong>of</strong> care, while respecting<br />
and fostering the patient’s autonomy. 19,21,22<br />
Oncologists’ Responsibilities to Their Patients and<br />
Families in Different Cultures<br />
Within each society, there are divergent groups with<br />
particular forms <strong>of</strong> family culture. Both the broad culture<br />
and private family culture frame people’s attitudes toward<br />
family roles, concepts <strong>of</strong> health, illness and suffering, and<br />
decisions about life and death. Systems <strong>of</strong> belief and cultural<br />
values are not static. Rather, cultures are dynamic and<br />
evolve over time under the influence <strong>of</strong> many demographic,<br />
social, and political factors. 23<br />
Culture has been powerfully described as a tapestry,<br />
where each single thread cannot be understood outside <strong>of</strong><br />
the entire fabric, and different attitudes and beliefs cannot<br />
be judged outside the entire context <strong>of</strong> any given culture. 22<br />
Applied to health care, this means that the same illness<br />
under different social and familial conditions could elicit diverse<br />
sets <strong>of</strong> emotional behaviors, suggesting the need for<br />
differential approaches to different patient–family units. 23,24<br />
Amani was a 34-year-old religious Muslim. She, her husband,<br />
and their eight children lived with the husband’s<br />
extended family. She was diagnosed in 2011 with stage IV<br />
metastatic breast cancer, along with brain and lung metastases.<br />
Shame and guilt prevented her from coming earlier for<br />
an examination. Her male physician communicated only<br />
with her family. During Amani’s hospitalization, the word<br />
“cancer” was never mentioned and she refused to have any<br />
direct discussions with physicians and nurses. Her preferences<br />
made perfect sense to her based on traditional values,<br />
but the Western medical team was frustrated and found it<br />
difficult to withhold the truth and allow the family to choose<br />
her place <strong>of</strong> death. Yet Amani’s family took her home, where<br />
she died a few days later, surrounded by the family’s prayers<br />
for Allah’s protection and hopes for an afterlife (Lea Baider,<br />
from a private diary).<br />
Disclosing information to family members, rather than<br />
patients themselves, and allowing families to make decisions,<br />
are among the best known examples <strong>of</strong> cultural<br />
differences in physicians’ responsibility toward patients<br />
and/or their families. They stem from a cultural belief in<br />
protecting patients from bad news and painful decisions.<br />
In contrast to the patient autonomy approach, considered<br />
potentially harmful to patients with cancer, the prevalent<br />
approach among most cultural groups is a “guided medical<br />
paternalism” that is family-oriented. Attitudes and practices<br />
regarding the most appropriate behavior toward seriously ill<br />
patients influences the physician’s duty in a manner that<br />
privileges family ties and family involvement in decision<br />
making over patient autonomy. 25,26 The view that physicians<br />
have a duty toward their patients’ families and that,<br />
together with them, they should protect patients from bad<br />
news is still dominant in many cultures. Terminal illness is<br />
usually disclosed to the family and not to the patient, who is<br />
generally being cared for at home rather than in a hospital<br />
or medical institution.<br />
A 2011 review related to disclosure <strong>of</strong> terminal illness in<br />
codes <strong>of</strong> medical ethics from 14 Islamic countries showed<br />
that approximately one-third <strong>of</strong> the physicians do not give<br />
any information about cancer prognosis and there is a<br />
general tendency to favor a religiously paternalistic/utilitarian,<br />
family-centered approach. 27 The Islamic religion not<br />
only shapes the social and personal lives <strong>of</strong> Muslims but also<br />
plays a significant role in guiding values and beliefs in every<br />
life event or circumstance, such as illness and death. 28<br />
Muslim physicians have a code <strong>of</strong> medical ethics consistent<br />
with cultural and religious norms, which is incorporated into<br />
their daily practice.<br />
A study <strong>of</strong> medical care in Lebanon reported that it is the<br />
physician’s duty to withhold the true nature <strong>of</strong> the diagnosis<br />
and prognosis from the patient, even though no false hope<br />
should be given to the family in the case <strong>of</strong> serious or<br />
terminal illness. 25 In Jordan, physicians and other health<br />
care providers attending to women with breast cancer are<br />
required to consider the influence <strong>of</strong> culture, religion,<br />
and the financial and personal characteristics <strong>of</strong> each<br />
patient before <strong>of</strong>fering information and support, and they<br />
<strong>of</strong>ten withhold them when asked to do so by the patient’s<br />
family. 30<br />
In Nepal, physicians are duty bound not to discuss a<br />
diagnosis <strong>of</strong> malignancy with their patients: communication<br />
should be only with family members, who then may filter the<br />
information to the patient. 36 In Pakistan, although breast<br />
cancer is the most common form <strong>of</strong> cancer among women,<br />
the experiences <strong>of</strong> patients and the suffering they endure<br />
routinely go unreported. 32 The social stigma attached to a<br />
diagnosis <strong>of</strong> breast cancer <strong>of</strong>ten prevents women from informing<br />
relatives and physicians <strong>of</strong> their illness as breast<br />
cancer is viewed as a taboo that diminishes the opportunities<br />
for single women to get married and there is also a social<br />
stigma to having a mastectomy. When discovering a lump<br />
in their breast, many women seek advice from local homeopaths<br />
on medicines or search spiritual support rather than<br />
medical advice. Often, they are afraid to seek medical help<br />
because <strong>of</strong> their mistrust <strong>of</strong> Western physicians and breast<br />
cancer is <strong>of</strong>ten diagnosed at advanced stages. 32<br />
In the Islamic Republic <strong>of</strong> Iran, breast cancer ranks first<br />
(21.4%) among all malignancies diagnosed in women. One in<br />
four women with breast cancer report for screening less<br />
frequently than recommended. The screening process itself<br />
is characterized by misinformation, lack <strong>of</strong> guidance and<br />
referral, and concealment on the part <strong>of</strong> the physician<br />
concerning the severity <strong>of</strong> the diagnosis and prognosis. 33 In<br />
a recent research study, several women who visited either a<br />
private physician or one working in the public health sector<br />
described how they felt when they learned that the physician<br />
had concealed the diagnosis <strong>of</strong> breast cancer. One<br />
woman described the physician who had diagnosed her case<br />
and referred her to the public hospital: “He did not tell me<br />
that I have breast cancer for month and a half. ...Allthat<br />
time, he did not spill out the word cancer.” Several women<br />
recruited from the public hospital interpreted this concealment<br />
as denial: “He—the physician—did not want to terrify<br />
us by saying, ‘you have cancer.’ ” Notwithstanding the physician’s<br />
intention, many women expressed the need to know<br />
the truth about their medical condition.<br />
These and other data show that, while <strong>of</strong>ten consonant<br />
with cultural and religious values and norms, patients with<br />
cancer are no longer satisfied with nondisclosure and wish to<br />
be more respected in their autonomy and right to take an<br />
active role in decision making about their care. Hence, it<br />
appears that, in family- and community-centered cultures,<br />
the duty <strong>of</strong> the oncologist is to involve his or her patients,<br />
rather than the family only. 34<br />
e17
Conclusion<br />
In more individualistic cultures, such as Western ones,<br />
family roles also change when a member has cancer and the<br />
psychologic, social, and financial burdens <strong>of</strong> family caregiving<br />
increase. Consequently, the patient–doctor relationship<br />
is shifting from its traditional dual partnership to a<br />
more extended collaboration among pr<strong>of</strong>essionals, patients,<br />
and their families. This involves a parallel shift in the<br />
culture <strong>of</strong> medicine in which each patient’s family is included<br />
in the team as part <strong>of</strong> a functional system. 2,35 A<br />
reappraisal <strong>of</strong> the notion <strong>of</strong> individual autonomy is necessary<br />
and there is a growing trend toward speaking <strong>of</strong><br />
relational autonomy, based on a renewed sense <strong>of</strong> connectedness<br />
among family members. “Relational autonomy” refers<br />
to an understanding <strong>of</strong> autonomy in light <strong>of</strong> family and<br />
community ties and <strong>of</strong> individual, cultural, and socioeconomic<br />
factors that affect our decision-making processes.<br />
The autonomy <strong>of</strong> each individual is no longer seen as an<br />
abstract but rather as a contextual concept. 34,36<br />
As oncologists, we consider part <strong>of</strong> our pr<strong>of</strong>essional responsibility<br />
to acknowledge our patients’ needs based on<br />
Authors’ Disclosures <strong>of</strong> Potential Conflicts <strong>of</strong> Interest<br />
Author<br />
Antonella Surbone*<br />
Lea Baider*<br />
*No relevant relationships to disclose.<br />
Employment or<br />
Leadership<br />
Positions<br />
Consultant or<br />
Advisory Role<br />
1. Spillman B, Black K. Staying the Course: Trends in Family Caregiving.<br />
Washington, DC: AARP; 2005.<br />
2. Baider L, Cooper CL, De-Nour K, eds. Cancer and the Family: Second<br />
Edition. Sussex, England: J Wiley & Sons, Ltd; 2000.<br />
3. Baider L, Surbone A. Cancer and the family: The silent words <strong>of</strong> truth.<br />
J Clin Oncol. 2010;28:1269-1272.<br />
4. Institute <strong>of</strong> Medicine, National Academy Press. “Envisioning the National<br />
Healthcare Quality Report (2001).” http://books.nap.edu/openbook.<br />
php?record_id�10073&page�41. Accessed Matrch 12, <strong>2012</strong>.<br />
5. <strong>American</strong> Academy <strong>of</strong> Family Physicians. “Family, Definition <strong>of</strong>.” http:<br />
//www.aafp.org/online/en/home/policy/policies/f/familydefinition<strong>of</strong>.html. Accessed<br />
March 12, <strong>2012</strong>.<br />
6. <strong>American</strong> Academy <strong>of</strong> Pediatrics (AAP) and <strong>American</strong> College <strong>of</strong> Emergency<br />
Physicians (ACEP). Patient- and family-centered care and the role <strong>of</strong><br />
the emergency physician providing care to a child in the emergency department.<br />
Policy Statement. Pediatrics. 2006;118:2242-2244.<br />
7. Committee on Hospital Care and <strong>American</strong> Academy <strong>of</strong> Pediatrics.<br />
Family-centered care and the pediatrician’s role. Pediatrics. 2003;112:691-<br />
696.<br />
8. Institute for Family-Centered Care. Advancing the Practice <strong>of</strong> Patientand<br />
Family-Centered Care: How to Get Started.” http://www.ipfcc.org/pdf/<br />
getting_started.pdf. Accessed March 12, <strong>2012</strong>.<br />
9. Institute for Family-Centered Care. Hospitals and Communities Moving<br />
Forward with Patient- and Family-Centered Care: Enhancing Quality and<br />
Safety for Patients and Their Families http://www.familycenteredcare.org/<br />
resources/multimedia/index.html. Accessed March 12, <strong>2012</strong>.<br />
10. World Health Organization. Cancer Pain Relief and Palliative Care:<br />
Report <strong>of</strong> a WHO Expert Committee Technical Report Series No. 804. Geneva,<br />
Switzerland: World Health Organization; 1990.<br />
11. Teno JM, Casey VA, Welch LC, et al. Patient-focused, family-centered<br />
end-<strong>of</strong>-life medical care. J Pain Sympt Manag. 2001;22:738-751.<br />
12. National Alliance for Caregiving and AARP. Caregiving in the US.<br />
Washington, DC: AARP and NAC; 2005.<br />
13. Soothill K, Morris SM, Thomas C, et al. The universal, situational, and<br />
personal needs <strong>of</strong> cancer patients and their main carers. Eur J Oncol Nurs.<br />
2003;7:5-13.<br />
e18<br />
religious beliefs, family traditions, and family roles and to<br />
involve family members, especially caregivers, in the processes<br />
<strong>of</strong> information giving and decision making <strong>of</strong>fering<br />
them support and counseling in full awareness and acceptance<br />
<strong>of</strong> differences.<br />
Family adaptation to cancer diagnosis is continuous<br />
across different stages—from hope to rebellion to despair—<br />
and family resilience, recovery, and growth may emerge<br />
from adversity. 37,38 Before encountering a major disease, a<br />
family may have never had to negotiate a situation perceived<br />
as threatening not only to the normal life <strong>of</strong> its<br />
members but also to the integrity <strong>of</strong> the family system. As a<br />
consequence <strong>of</strong> such complexities, part <strong>of</strong> the intimacy<br />
between the oncologist and his or her patient may be, or<br />
appear, lost. We still have to learn how to best keep our<br />
direct connection with cancer patients, while including<br />
among our responsibilities also those to their families and<br />
caregivers. This additional challenge for all oncologists is<br />
part <strong>of</strong> the novel paradigm <strong>of</strong> care where pr<strong>of</strong>essional responsibilities<br />
are no longer directed only at patients but also<br />
at their families.<br />
Stock<br />
Ownership Honoraria<br />
REFERENCES<br />
Research<br />
Funding<br />
Expert<br />
Testimony<br />
SURBONE AND BAIDER<br />
Other<br />
Remuneration<br />
14. Frank A. Relations <strong>of</strong> caring: Demoralization and remoralization in the<br />
clinic. Int J Human Caring. 2002;6:13-19.<br />
15. Given B, Given CW, Kozachik S. Family support in advanced cancer.<br />
CA Can J Clin. 2001;51:213-231.<br />
16. Yabr<strong>of</strong>f KR, Kim Y. Time costs associated with informal caregiving for<br />
cancer survivors. Cancer. 2009;115:362-4373.<br />
17. Fegg MJ, Brandstätter M, Kramer M, Kögler M, et al. Meaning in life<br />
in palliative care patients. J Pain Symptom Manag. 2010;40:502-509.<br />
18. Surbone A, Baider L, Balducci L. Caregiving to elderly patients:<br />
clinical, ethical, and psychosocial challenges. In Govindan R (ed). ASCO<br />
Educational Book. Alexandria, VA: <strong>American</strong> <strong>Society</strong> <strong>of</strong> <strong>Clinical</strong> <strong>Oncology</strong>,<br />
2010.<br />
19. Baider L, Goldzweig G. Exploration <strong>of</strong> Family Care: A Multicultural<br />
Approach. In <strong>Clinical</strong> Psycho-<strong>Oncology</strong>: An International Perspective. L<br />
Grassi, M Riba (eds.) Chichester, West Sussex, UK: J Wiley & Sons Ltd.;<br />
<strong>2012</strong>;187-196.<br />
20. Grov EK, Fossa SD, Sørebø O, et al. Primary caregivers <strong>of</strong> cancer<br />
patients in the palliative phase: A path analysis <strong>of</strong> variables influencing their<br />
burden. Soc Sci Med. 2006;63:2429-2439.<br />
21. Baider L. Unsteady Balance: The Constraints <strong>of</strong> Informal Care. J Pediatr<br />
Hematol Oncol. 2011;33:S108-S111.<br />
22. Surbone A. The difficult task <strong>of</strong> family caregiving in oncology: Which<br />
roles do autonomy and gender play? Supp Care Cancer. 2003;11:617-619.<br />
23. Kagawa-Singer M, Valdez D, Yu M, et al. Cancer, culture and health<br />
disparities: Time to chart a new course? CA Can J Clin. 2010;60:12-39.<br />
24. Searight HR, Gafford J. Cultural diversity at the end <strong>of</strong> life. Am Fam<br />
Phys. 2007;72:515-522.<br />
25. Kamen BA. Cancer pain, suffering, and spirituality: A Middle Eastern<br />
Cancer Consortium meeting. J Pediatr Hematol Oncol. 2010;32:341.<br />
26. Mobeireek AF, Al-Kassimi F, Al-Zahrani K, et al. Information disclosure<br />
and decision-making: The Middle East versus the Far East. J Med<br />
Ethics. 2009;34:225-229.<br />
27. Abdulhameed HE, Hammami MM, Hameed Mohamed EA. Disclosure<br />
<strong>of</strong> terminal illness to patients and families: Diversity <strong>of</strong> governing codes in 14<br />
Islamic countries. J Med Ethics. 2011;37:472-475.<br />
28. Bülow HH, Sprung CL, Reinhart K, et al. The world’s major religions’
ONCOLOGISTS’ DUTIES TO PATIENT AND FAMILIES<br />
points <strong>of</strong> view on end-<strong>of</strong>-life decisions in the intensive care unit. Intensive Care<br />
Med. 2008;34:423-430.<br />
29. Adib SA, Hamadeh GN. Attitudes <strong>of</strong> the Lebanese public regarding<br />
disclosure <strong>of</strong> serious illness. J Med Ethics. 2003;25:399-403.<br />
30. Alqaissi NM, Dickerson SS. Exploring common meanings <strong>of</strong> social<br />
support as experienced by Jordanian women with breast cancer. Cancer Nurs.<br />
2010;33:353-361.<br />
31. Gongal R, Vaidya P, Jha R, et al. Informing patients about cancer in<br />
Nepal: What do people prefer? Palliat Med. 2006;20:471-476.<br />
32. Banning M, Hafeez H, Faisal S, et al. The impact <strong>of</strong> culture and<br />
sociological and psychological issues on Muslim patients with breast cancer in<br />
Pakistan. Cancer Nurs. 2009;32:317-324.<br />
33. Lamyian M, Hydarnia A, Ahmadi F, et al. Barriers to and factors<br />
facilitating breast cancer screening among Iranian women. East Mediterr<br />
Health J. 2007;13:1160-1169.<br />
34. Surbone A. Telling Truth to Patients with Cancer: What is the Truth?<br />
Lancet Oncol. 2006;7:944-950.<br />
35. Rolland JS. Cancer and the family: An integrative model. Cancer.<br />
2005;104:2584-2595.<br />
36. Sherwin S. “A Relational Approach to Autonomy in Health Care.” In S<br />
Sherwin, The Feminist Health Care Ethics Research Network, The Politics <strong>of</strong><br />
Women’s Health: Exploring Agency and Autonomy. Philadelphia: Temple<br />
University Press; 1988;19-44.<br />
37. Vilchinsky N, Dekel R, Leibowitz, et al. Dynamics <strong>of</strong> support perception<br />
among couples coping with cardiac illness. Health Psychol. 2011;30:411-<br />
419.<br />
38. Naaman S, Radwan K, Johnson S. Coping with early breast cancer:<br />
Couple adjustment process. Psychiatr. 2009;72:321-342.<br />
e19
The Oncologist’s Duty to Provide Hope:<br />
Fact or Fiction?<br />
Overview: There are many sources <strong>of</strong> conflict in oncology.<br />
Conflicts arise because there are numerous therapeutic options,<br />
each <strong>of</strong> which is imperfect, and these conflicts produce<br />
ethical dilemmas. A recent <strong>American</strong> Medical Association<br />
(AMA) publication outlined the principles <strong>of</strong> medical ethics for<br />
managing conflicts. Common conflicts in oncology include<br />
whether to resuscitate, to give more chemotherapy, and how<br />
much truth to tell. These conflicts are magnified because <strong>of</strong><br />
the life and death scenario <strong>of</strong> advanced cancer. Denial,<br />
avoidance, and hope are psychologic mechanisms that enable<br />
adaptation to the life-threatening circumstances. Hope is<br />
widely written about though poorly understood and defined.<br />
Ethical statements regarding its virtue and importance to<br />
preserve are frequently given. In an effort to progress the<br />
THE ONCOLOGIST in clinical practice is conflicted in<br />
several domains, including truth telling, resuscitation,<br />
when to stop palliative chemotherapy, and changing goals <strong>of</strong><br />
care. The conflicts are magnified because cancer is a lifethreatening<br />
illness.<br />
Conflicts underpin ethical dilemmas. The conflicts today<br />
are in part because <strong>of</strong> the success <strong>of</strong> modern medical oncology,<br />
which has provided many therapeutic options. There<br />
are different models <strong>of</strong> how to solve ethical dilemmas. The<br />
AMA recently published the <strong>American</strong> College <strong>of</strong> Physicians<br />
(ACP) Ethics Manual, which listed four principles <strong>of</strong> medical<br />
ethics: beneficence, nonmaleficence, respect for patient autonomy,<br />
and justice. 1 These principles assist in ethically<br />
resolving clinical conflicts.<br />
The ACP Ethics Manual also defines pr<strong>of</strong>essionalism. It<br />
declares that pr<strong>of</strong>essionalism is “a code <strong>of</strong> ethics and a duty<br />
<strong>of</strong> service that put patient care above self-interest.” 1 This<br />
duty and attendant humility is another principle that encourages<br />
ethical behavior. 2 It was disappointing not to see<br />
noted in the AMA document the importance <strong>of</strong> selfawareness<br />
(or self-knowledge) as a tool for physicians to<br />
reflect on their motives in decision making and as a further<br />
safeguard to ethical practice.<br />
Are We Treating Patients with Cancer<br />
Too Aggressively?<br />
A recent Annals <strong>of</strong> <strong>Oncology</strong> editorial asked, “Why are we<br />
not ceasing chemotherapy when it is useless, toxic, logistically<br />
complex, and expensive?” This was based on statistics<br />
that show up to 25% <strong>of</strong> patients with solid tumors receive<br />
chemotherapy in the last month <strong>of</strong> life. 3 Unfortunately,<br />
these statistics tell only part <strong>of</strong> the story. They cannot<br />
include those patients who did respond to chemotherapy and<br />
whose lives were prolonged and therefore did not die within<br />
four weeks <strong>of</strong> initiating therapy. Furthermore, since we<br />
cannot know with certainty how each individual will respond,<br />
it is likely that we will have to treat a certain<br />
percent—in retrospect, in futility—so others will benefit.<br />
This holds true for many therapeutic scenarios in medicine.<br />
Nevertheless, there are instances when we recognize in<br />
advance that chemotherapy (and parenteral nutrition, radiation<br />
therapy, and antibiotics) should not be given. It can be<br />
harrowingly difficult to balance respect for autonomy and<br />
e20<br />
By Simon Wein, MD<br />
understanding <strong>of</strong> hope, two critical features are defined:<br />
(1) hope as a thought process only exists in the future, and<br />
(2) hope is only ever associated with positive and good<br />
thoughts. The future is unknown and uncertain; therefore,<br />
hoping can be manipulated by presenting statistics in a way<br />
to boost hoping. Thus a dilemma and specific ethical responsibility<br />
falls on oncologists when discussing conflicts. Furthermore,<br />
since hope is a subjective assessment <strong>of</strong> a<br />
possibility that is considered “good” by the hoper, it cannot<br />
be perceived as “false.” “False hope” is an erroneous assessment.<br />
Finally, this article introduces the concept that there<br />
might be a role to stop hoping—since hope <strong>of</strong> the future is also<br />
filled with doubt and fear—and instead live in the present and<br />
try to find joy and meaning today.<br />
desire for beneficence against nonmaleficence. There is a<br />
gray zone between ethical intent and clinical outcome.<br />
Various factors influence the decision making <strong>of</strong> the oncologist<br />
and include fear <strong>of</strong> death, medico-legal, pity, philosophy,<br />
fear <strong>of</strong> failure, and burn-out. Sometimes the oncologist is<br />
aware <strong>of</strong> these influences and at other times not. The<br />
psychologic mechanisms that influence these factors include,<br />
among others, denial, avoidance, and hope. This<br />
article will focus on hope.<br />
What Is Hope?<br />
There is no agreed-on definition <strong>of</strong> what constitutes hope.<br />
There is not even agreement on whether hope is an emotion<br />
or a cognition. Some authors describe generalized and particularized<br />
hope, while others see hope built around faith,<br />
love, meaning, dignity, peace, and spirituality, even using<br />
the words interchangeably. 4,5 More specifically, some <strong>of</strong><br />
these definitions mistake the mechanism (or process) <strong>of</strong> hope<br />
for its content or goal. 6<br />
Hope is a factor in many <strong>of</strong> the conflicts in decision<br />
making, particularly at the end <strong>of</strong> life: parenteral nutrition,<br />
chemotherapy, resuscitation, and truth telling (e.g., the<br />
hope that further treatment will prolong or “save” a life, or<br />
the concern that “telling the truth” will destroy a patient’s<br />
hope and shorten life).<br />
Daneault noted that there is “a growing awareness that<br />
nurturing hope is one <strong>of</strong> the most important tasks in the<br />
oncology clinic.” 7 Helft went further saying, “all patients<br />
with cancer have a vital and inalienable right to maintain<br />
hope.” 8 In the absence <strong>of</strong> a clear definition or even a<br />
consensus on a description <strong>of</strong> hope, and with no research<br />
measuring hope in oncologic care, such “important tasks”<br />
and “inalienable . . . rights” are not evidence based. So how<br />
From the Pain and Palliative Care Service, David<strong>of</strong>f Cancer Center, Rabin Medical<br />
Center, Petach Tikvah, Israel.<br />
Author’s disclosures <strong>of</strong> potential conflicts <strong>of</strong> interest are found at the end <strong>of</strong> this article.<br />
Address reprint requests Simon Wein, MD, Pain and Palliative Care Service, Room 36,<br />
David<strong>of</strong>f Cancer Center, Rabin Medical Center, 39 Jabotinsky St., Petach Tikvah, Israel,<br />
49100; email: simonwe@clalit.org.il.<br />
© <strong>2012</strong> by <strong>American</strong> <strong>Society</strong> <strong>of</strong> <strong>Clinical</strong> <strong>Oncology</strong>.<br />
1092-9118/10/1-10
DOCTOR’S DUTY TO PROVIDE<br />
do we know when, how, and if to “nurture hope” in the<br />
oncology clinic?<br />
Definitions <strong>of</strong> Hope<br />
Dufault and Martocchio define hope as “a multidimensional<br />
dynamic life force characterized by a confident yet<br />
uncertain expectation <strong>of</strong> achieving a future good, which, to<br />
the hoping person, is realistically possible and personally<br />
significant.” 4 This definition is broad and contains the core<br />
features <strong>of</strong> hope (“the future” and “subjective good”). However,<br />
by including multiple dimensions and “dynamic life<br />
force” in the definition, researchers tend to apply attributes<br />
such as faith, relationships and dignity to the definition<br />
itself. These ‘life forces’ support hope but are not hope.<br />
Snyder defines hope based on empirical research and a<br />
validated (his own) hope scale. 9 His work has not been<br />
extensively applied to oncology or palliative care. Snyder<br />
firmly holds the view that hope is a cognitive and not an<br />
emotional process. He defines hope as “the perceived capability<br />
to produce workable routes to desired goals and the<br />
requisite motivation to use those routes.” 9 That is, there is a<br />
goal, a plan to get there, and the will to do it. Hope is a<br />
consciously planned process, with emotions as an epiphenomenon.<br />
Meaning, faith, and love are adjuvants or struts to<br />
hope and influence Snyder’s “motivation” and “will to” but<br />
are not themselves integral to the definition <strong>of</strong> hope. Although<br />
Snyder’s definition is logical and measurable, it is<br />
not easy to apply at the bedside.<br />
Is Hope Good or Evil?<br />
The Greek myth <strong>of</strong> Pandora involves a woman who had a<br />
jar and was told not to open it. Curiosity got the better <strong>of</strong><br />
her, and upon opening the jar, various evils escaped: plague<br />
and disease. She quickly shut the lid, but one thing remained<br />
trapped inside: hope. 10 The ancient Greeks interpreted<br />
hope as evil: “Precisely because <strong>of</strong> its ability to keep<br />
the unfortunate in continual suspense, the Greeks considered<br />
hope the evil <strong>of</strong> evils, the truly insidious evil.” 11 That is,<br />
a person trapped in an unwinnable situation who still hopes<br />
to escape, merely tortures himself in the futility <strong>of</strong> the task.<br />
The parallel to incurable cancer and another round <strong>of</strong><br />
chemotherapy given “in the hope” is compelling.<br />
However, later in history, Judeo-Christian and Western<br />
philosophy associated life with an ultimate purpose and<br />
goal. Hope became identified with good things. Menninger<br />
said, “Love, faith, hope—in that order. The Greeks were<br />
wrong. Of course hope is real, and <strong>of</strong> course it is not evil. It<br />
is the enemy <strong>of</strong> evil, and an ally <strong>of</strong> love. Which is goodness.” 5<br />
A Practical Description <strong>of</strong> Hope<br />
The critical characteristic <strong>of</strong> hope is that it exists only in<br />
the future. We cannot hope in the past or the present. The<br />
KEY POINTS<br />
● <strong>Clinical</strong> conflicts create ethical dilemmas.<br />
● Hope exists only in the future.<br />
● Hope may cause more harm than good.<br />
● False hope is a self-contradiction.<br />
● There is benefit in living without hope.<br />
outcome <strong>of</strong> hoping is subjectively assessed as a probability<br />
but neither a certainty nor an impossibility. It is this<br />
characteristic that is important to bedside decision making.<br />
The second immutable characteristic is that hope is egosyntonic.<br />
That is, whenever we hope, it makes us feel good<br />
and optimistic, not sad, distressed, or anxious. Even if we<br />
hope something bad for ourselves (e.g., death), we note that<br />
its primary motivation is good.<br />
The key feature about the future is that it is unpredictable.<br />
Since hope exists only in the future, it too is unpredictable<br />
and unknowable. This is what Boris meant when he<br />
said that “we create ambiguity in order to preserve hope.” 12<br />
That is, by creating uncertain situations, we also create an<br />
opportunity to hope. Into uncertainty we can hope. This is<br />
the principle <strong>of</strong> gambling. An uncertainty is deliberately<br />
created and people visit casinos not for the fun but for the<br />
possibility to hope and the good feeling it creates.<br />
Similarly, when we <strong>of</strong>fer chemotherapy and the outcome is<br />
uncertain (as it must be), this ambiguity or uncertainty <strong>of</strong> a<br />
future outcome allows the patient to hope. A physician may<br />
be deliberately vague about the possible benefit <strong>of</strong> a new<br />
treatment or about the result <strong>of</strong> a scan, thus manipulating<br />
hope. The physician does not create hope, nor can the<br />
physician prescribe hope as such. More precisely, the physician<br />
amplifies an ambiguity about the future to which the<br />
patient may or may not respond.<br />
Husebo, however, states that “hope is not just something<br />
which is limited to an expectation <strong>of</strong> the future or afterlife.<br />
Hope is closely connected to a person’s life story.” 13 Of course<br />
hope is connected to a person’s values and history. Snyder<br />
noted this is what determines the content <strong>of</strong> the goals <strong>of</strong><br />
hope and influences the determination to succeed. However,<br />
hope—the verb and the noun—exists only in the future and<br />
in the imagination. Hope is a figment <strong>of</strong> the imagination.<br />
How can we understand hope in terms <strong>of</strong> good and evil?<br />
The Greeks thought hope was evil, whereas Western society<br />
sees hope as good. The appreciation that hope itself is “good”<br />
exists, in part, because the goals <strong>of</strong> hope are always perceived<br />
as subjectively good. The ancient Greeks were concerned<br />
about this characteristic <strong>of</strong> “goodness” being<br />
manipulated and abused.<br />
A more accurate description was furbished by Kertesz in<br />
his acceptance speech for the 2002 Nobel Prize for Literature.<br />
He described how after invading Hungary, the Russians<br />
controlled the people by manipulating hope and<br />
similarly the Germans manipulated the inmates <strong>of</strong> the<br />
concentration camps. Kertesz stated, “I understood that<br />
hope is an instrument <strong>of</strong> evil.” 14 Hope itself is not evil, since<br />
hope is a path and goal chosen by each individual in the hope<br />
<strong>of</strong> a good outcome. Neither, however, is hope necessarily<br />
beneficial to the hoper.<br />
False Hope<br />
Strictly speaking, hope can never be false since it is a<br />
subjective thought process whose primary function is to plan<br />
and achieve goals. The future is unknown, therefore, the<br />
subjective evaluation <strong>of</strong> a possibility can be contracted or<br />
expanded depending on psychologic needs. Thus, a 5%<br />
chance <strong>of</strong> response to chemotherapy may generate hope for<br />
some but appear futile to others. Similarly, after being told<br />
there is 0% chance <strong>of</strong> being cured, people who believe in<br />
miracles will still be able to hope. A miracle may be “false” to<br />
e21
an external observer, but to the person hoping for a miracle,<br />
it is true in all respects and “realistically possible.” 4<br />
A common understanding <strong>of</strong> false hope is a patient who<br />
mis-states the probability <strong>of</strong> a goal and hence harbors “false”<br />
hopes. It is common for patients to misunderstand statistics.<br />
This process is better labelled as denial or on occasions as a<br />
delusion. It is not uncommon for oncologists to be a little<br />
vague and ambiguous in communicating prognosis and statistics<br />
in order to encourage hope, as Boris taught us. 9<br />
Furthermore since hope exists in the unknowable future it<br />
cannot be disproven and therefore it cannot be shown to be<br />
false. For hope to function, it does not matter whether the<br />
statistics are true or false, or whether the patient misconstrued<br />
them. All that it matters is for the patient to believe<br />
them. Hope will then thrive, for better or for worse.<br />
Denial as an Elixir <strong>of</strong> Hope<br />
Daneault observed that “false hope, when it denies the<br />
reality <strong>of</strong> imminent death, may lead to a neglect <strong>of</strong> final<br />
arrangements and added burden for families,” while Von<br />
Roenn categorically stated that “false hope leads to disappointment<br />
and disillusionment.” 7,15 It might be more pragmatic<br />
to approach such hoping as we would denial and not<br />
to judge whether hope is “false” or “true.” If denial is not<br />
causing harm and is functioning as a psychological defense,<br />
then there is no need to intervene. Some people hope until<br />
the very end and thereby preserve their emotional wellbeing.<br />
Others, by hoping against hope, create problems—<br />
“evil” as the Greeks would have it—that cause harm or<br />
damage. It would be appropriate to consider counseling this<br />
latter group, similar to those where denial is causing problems.<br />
Denial is the gap between a subjective and objective<br />
assessment <strong>of</strong> a medical situation. Into this gap the patient<br />
can develop new goals and by reducing anxiety, can improve<br />
motivation. Clayton et al performed a systematic review <strong>of</strong><br />
hope in terminally ill patients and showed that some patients<br />
and caregivers “found hope in avoiding the facts, that<br />
is, denial. 16 The practical question to ask is: is hope functioning<br />
in a beneficial or maladaptive way?<br />
Life Without Hope<br />
Boris observed, “Indeed, there is no torment quite like<br />
hopelessness. But hopelessness marks the presence <strong>of</strong><br />
thwarted hope, not hope’s absence.” This fine distinction<br />
points the way to a therapeutic paradigm. A hope that has<br />
failed (i.e., hopelessness) causes frustration and pain. However,<br />
if one does not hope for anything, then one will not<br />
experience the anguish <strong>of</strong> frustrated hope. Not engaging in<br />
hoping means one cannot lose or, indeed, benefit from hope.<br />
Fear is also an expectation <strong>of</strong> the future. Fear in contrast<br />
to hope creates negative feelings and anxiety. It is in some<br />
ways an opposite to hope. If one were to disregard the future<br />
and just live in the here and now, then there would be no<br />
hope or fear as Gravlee observed, “The one with no hope also<br />
does not fear. Fear about the uncertain future remains, so<br />
long as hope about the future remains.” 17 We are anxious<br />
that our hope—our gamble in the future—will fail. Huxley<br />
said the same thing, though used drugs (“soma”) to stop<br />
worrying about what might happen: “Was and will make me<br />
ill, I can take a gramme and only am.” 18 Without memories<br />
there will be no anxiety about something going wrong in<br />
e22<br />
SIMON WEIN<br />
the future. Without a future there will be no hope or fear<br />
that the hope might fail. This fear-hope dyad is reflected in<br />
the ancient Greek’s jaundiced view <strong>of</strong> hope. By not hoping<br />
(clearly in some cases this is a form <strong>of</strong> denial), there will no<br />
longer be the merry-go-round <strong>of</strong> hope and despair.<br />
Hope in <strong>Oncology</strong><br />
In applying hope to clinical oncology a number <strong>of</strong> scenarios<br />
arise:<br />
• Should we give antitumor treatments to boost hope?<br />
How ethically bound are we to prescribe chemotherapy<br />
because a patient hopes it will work?<br />
• How prepared are we to say “no” to further treatment at<br />
the risk <strong>of</strong> the patient going elsewhere?<br />
• Is there a role for placebo to maintain hope?<br />
• Does telling the truth diminish hope?<br />
• Does loss <strong>of</strong> hope shorten life?<br />
• How do oncologists stop their own hopes (and fears) from<br />
influencing the decision?<br />
Three common strategies can be used when hope has<br />
failed in advanced cancer: to live today, to exchange one<br />
hope for another, and to reflect about the life lived. It is<br />
sometimes useful to discuss the nature <strong>of</strong> hope with the<br />
patient and how it influences decision making, especially if<br />
hope is driving the patient to unrealistic decisions or is<br />
causing psychologic symptoms <strong>of</strong> distress. A common approach<br />
is to suggest to the patient to live day to day and to<br />
enjoy the moment. Leave tomorrow to the future. Of course<br />
what we are doing is replacing the future (hope and fear)<br />
with the present. If the process <strong>of</strong> hoping cannot be relinquished—if<br />
the perceived subjective good is too comforting to<br />
give up—yet is starting to cause harm then Snyder’s hope<br />
model allows lost goals to be mourned and replaced with new<br />
ones. Instead <strong>of</strong> hoping for a cure, the goal might be switched<br />
to hoping to see a daughter married. Thus hoping in the<br />
future comforts the living in the present and pushes acceptance<br />
aside.<br />
Finally, a simple and effective strategy, “the life narrative,”<br />
is to talk to the patient about their lives and loves. This<br />
starts the process <strong>of</strong> replacing hope (in the future) with<br />
acceptance (in the present). Roy valued the physician “who<br />
know(s) how to drop the pr<strong>of</strong>essional role mask and relate to<br />
others simply and richly as a human being.” 19<br />
Conclusion<br />
Our pr<strong>of</strong>essional duty demands not only an adherence to<br />
the AMA ethical guidelines but also an effort to be aware <strong>of</strong><br />
the psychologic processes involved in decision making—both<br />
the patients’ and ours. Denial, avoidance, and hope are<br />
psychologic mechanisms used to avoid conflict. Hope has two<br />
core characteristics: it is a thought process that can exist<br />
only in the future and it is “hardwired” to make us feel good.<br />
Hope is neither inherently good nor bad. The outcome <strong>of</strong><br />
hope is neither necessarily good nor bad.<br />
“False hope” does not exist in the sense that each time we<br />
hope, we do so subjectively for our own psychologic benefit. If<br />
an external observer dismisses the likelihood <strong>of</strong> my hope<br />
materializing, this does not make my hope any less realistic<br />
or significant for me. Nor does it matter if the patient<br />
him/herself “misunderstands” the prognosis and build their<br />
hopes on a false premise. This is a form <strong>of</strong> denial and if it<br />
helps the patient overcome the stress <strong>of</strong> the illness without<br />
causing harm, then it is not “false” hope.
DOCTOR’S DUTY TO PROVIDE<br />
The value <strong>of</strong> studying hope in oncology and palliative care<br />
is to understand the role it plays in psychologic health, in<br />
making decisions about treatment, and in establishing goals<br />
<strong>of</strong> care. 20<br />
Talking about death and hope from the day <strong>of</strong> diagnosis<br />
means that when the disease has run its course, the discussion<br />
will be the continuation <strong>of</strong> a conversation, not a spinechilling<br />
revelation. The responsibility for presenting medical<br />
information in a way that enables the patient to balance<br />
acceptance against hope lies with the treating oncologist.<br />
Author’s Disclosures <strong>of</strong> Potential Conflicts <strong>of</strong> Interest<br />
Author<br />
Simon Wein*<br />
*No relevant relationships to disclose.<br />
Employment or<br />
Leadership<br />
Positions<br />
Consultant or<br />
Advisory Role<br />
1. Snyder L. <strong>American</strong> College <strong>of</strong> Physicians Ethics Manual, 6 th ed. Ann<br />
Intern Med. <strong>2012</strong>;156:73-104.<br />
2. Pellegrino ED. Pr<strong>of</strong>essionalism, Pr<strong>of</strong>ession and the Virtues <strong>of</strong> the Good<br />
Physician. Mt Sinai J Med. 2002;69:378-384.<br />
3. Braga S. Why do our patients get chemotherapy until the end <strong>of</strong> life?<br />
Annal Oncol. 2011;22:2345-2348.<br />
4. Dufault K, Martocchio BC. Hope: Its spheres and dimensions. Nurs Clin<br />
North Am. 1985;20:379-391.<br />
5. Menninger K. The Academic Lecture: Hope. Am J Psychiatry. 1959;116:<br />
481-491.<br />
6. Nekolaichuk CL, Bruera E. On the nature <strong>of</strong> hope in palliative care.<br />
J Palliat Care. 1998;14:36-42.<br />
7. Daneault S, Dion D, Sicotte C, et al. Hope and noncurative chemotherapies:<br />
Which affects the other? J Clin Oncol. 2010;28:2310-2313.<br />
8. Helft PR. Necessary collusion: Prognostic communication with advanced<br />
cancer patients. J Clin Oncol. 2005;23:3146-3150.<br />
9. Gum A, Snyder CR. Coping with terminal illness: The role <strong>of</strong> hopeful<br />
thinking. J Palliat Med. 2002;5:883-894.<br />
10. Wikipedia. Pandora. http://en.wikipedia.org/wiki/Pandora. Accessed<br />
January 31, <strong>2012</strong>.<br />
However, the oncologist is not obligated to prescribe hope<br />
and the oncologist can do better than collude or be disingenuous<br />
to reinforce hope.<br />
There are other values in life—family, work, hobbies,<br />
dignity, achievements—that the oncologist can explore a<br />
little at each visit. Discussion about these values may (or<br />
may not) enable acceptance <strong>of</strong> death. However, as Breitbart<br />
noted, “The paradox <strong>of</strong> the end <strong>of</strong> life dynamic is that<br />
through acceptance <strong>of</strong> the life one has lived comes acceptance<br />
<strong>of</strong> death.” 21<br />
Stock<br />
Ownership Honoraria<br />
REFERENCES<br />
Research<br />
Funding<br />
Expert<br />
Testimony<br />
Other<br />
Remuneration<br />
11. Hollingdale RJ (ed). The Nietzsche Reader. London: Penguin Classics,<br />
1977;277.<br />
12. Boris HN. On hope: Its nature and psychotherapy. Int R Psycho-Anal.<br />
1976;3:139-150.<br />
13. Husebo S. Is there hope, doctor? J Palliat Care. 1988;14:43-48.<br />
14. Kertesz I. Nobel Lecture, Nobel Prize in Literature 2002. http://<br />
www.nobelprize.org/nobel_prizes/literature/laureates/2002/kertesz-lecture-e.<br />
html. Accessed January 12, <strong>2012</strong>.<br />
15. Von Roenn JH, von Gunten CF. Setting goals to maintain hope. J Clin<br />
Oncol. 2003;21:570-574.<br />
16. Clayton J, Hancock K, Parker S, et al. Sustaining hope when communicating<br />
with terminally ill patients and their families: A systematic review.<br />
Psychooncology. 2008;17:641-659.<br />
17. Gravlee GS. Aristotle on Hope. J Hist Philos. 2000;38:461-477.<br />
18. Huxley A. Brave New World. New York: Perennial Classics, 1998.<br />
19. Roy D. Dying with dignity. J Palliat Care. 1986;2:3-4.<br />
20. Wein S. Hope: Concerning structure and function. Palliat Support<br />
Care. 2004;2:229-230.<br />
21. Breitbart W. Thoughts on the goals <strong>of</strong> psychosocial palliative care.<br />
Palliat Support Care. 2008;6:211-212.<br />
e23
MEDICAL ERRORS IN CANCER CARE: PREVENTION,<br />
DISCLOSURE, AND PATIENT AND<br />
FAMILY MEMBER RESPONSES<br />
CHAIR<br />
Antonella Surbone, MD, PhD<br />
New York University<br />
New York, NY<br />
SPEAKERS<br />
Itzhak Brook, MD, MSc<br />
Georgetown University<br />
Washington, DC<br />
Michael Rowe, PhD<br />
Yale School <strong>of</strong> Medicine, Program Recovery and Community Health<br />
New Haven, CT
Onclogists’ Difficulties in Facing and<br />
Disclosing Medical Errors: Suggestions for<br />
the Clinic<br />
Overview: Along with improved safety measures and changes<br />
in the culture <strong>of</strong> medicine, communication is key to reducing<br />
the effect <strong>of</strong> medical errors and to easing the medical,<br />
psychologic, and existential burdens they impose on all parties.<br />
Disclosure demonstrates respect for patients’ autonomy<br />
and promotes patient’s involvement in informed decision<br />
making about ways to correct or alleviate the effects <strong>of</strong> the<br />
error. It also enhances oncologists’ integrity and helps restore<br />
trust in the patient-doctor relationship.<br />
Because <strong>of</strong> the complexity <strong>of</strong> cancer treatments and the<br />
uncertainty regarding outcomes in oncology, oncologists may<br />
rationalize nondisclosure as a way to avoid adding to the<br />
physical and existential suffering <strong>of</strong> their patients. Although<br />
there is broad agreement among pr<strong>of</strong>essional and regulatory<br />
bodies, as well as medical ethicists, that physicians should<br />
disclose errors to patients—and physicians largely support<br />
IN RECENT years, research on the incidence and causes<br />
<strong>of</strong> error in medicine has led to calls for change in policy<br />
and practice. 1-2 Increasing attention has been paid to the<br />
role that appropriate communication plays in preventing<br />
and managing errors, not only at the medical level, but also<br />
with regard to the emotional and psychologic aftermath <strong>of</strong><br />
error for patients, family members and physicians. The<br />
limited research that has been conducted on the psychologic<br />
consequences <strong>of</strong> medical errors on physicians shows that<br />
they experience a wide range <strong>of</strong> emotions and thoughts—<br />
from guilt, feelings <strong>of</strong> inadequacy, anguish, shame (even to<br />
the point <strong>of</strong> leaving the pr<strong>of</strong>ession) to excessive caution<br />
toward other patients following the medical error. 3 Most <strong>of</strong><br />
what we know on this subject, however, comes from narrative<br />
accounts <strong>of</strong> physicians who have been brave enough to<br />
share with readers their experiences <strong>of</strong> committing or witnessing<br />
a medical error—the first starting with Dr. David<br />
Hilfiker in The New England Journal <strong>of</strong> Medicine in 1984. 4<br />
Communication about medical errors is key to easing the<br />
burden <strong>of</strong> errors on all parties with respect to all medical,<br />
psychologic, and existential dimensions. Given the lifethreatening<br />
nature <strong>of</strong> their illnesses and their dependence<br />
on their oncologists, patients with cancer can be especially<br />
hurt both physically and emotionally when errors or suspected<br />
errors occur. Oncologists who carry the emotional<br />
and psychologic burdens <strong>of</strong> caring for seriously ill patients<br />
over long periods <strong>of</strong> time can also be deeply affected by their<br />
own mistakes. 5 Proper disclosure, including a sincere apology,<br />
is the first step to restore trust in the patient-doctor<br />
relationship. It should be part <strong>of</strong> the immediate management<br />
<strong>of</strong> errors and <strong>of</strong> dealing with their long-term aftermaths.<br />
Disclosure <strong>of</strong> Medical Errors and the<br />
Patient-Doctor Relationship<br />
There is broad agreement among pr<strong>of</strong>essional and regulatory<br />
bodies and medical ethicists that physicians should<br />
disclose errors to their patients. 2,6,7 Empirical studies show<br />
that most patients want to be informed in detail about<br />
e24<br />
By Antonella Surbone, MD, PhD<br />
disclosure <strong>of</strong> error to patients—studies show discrepancy<br />
between physicians’ responses to hypothetical clinical scenarios<br />
<strong>of</strong> truth telling about medical errors and actual practices<br />
<strong>of</strong> withholding or tempering the truth. Among common<br />
reasons for avoiding disclosure are risk <strong>of</strong> malpractice lawsuits,<br />
fear <strong>of</strong> being exposed as incompetent, and feeling<br />
shame before patients and colleagues.<br />
Proper disclosure, however, including a sincere apology,<br />
should be part <strong>of</strong> the management <strong>of</strong> errors and <strong>of</strong> their<br />
long-term aftermaths. In disclosing medical errors, it is essential<br />
for oncologists to pay equal attention to the medical<br />
and the emotional aspects <strong>of</strong> the information they are giving<br />
and the reaction that it elicits in patients and families. Specific<br />
communication skills regarding disclosure <strong>of</strong> medical errors<br />
can be learned.<br />
medical errors that occur during their care, even when they<br />
do not lead to a negative outcome. 8 Disclosure demonstrates<br />
respect for patients’ autonomy and promotes their informed<br />
decision making about ways to correct or alleviate the effects<br />
<strong>of</strong> the error. 5,7<br />
Physicians largely support disclosure <strong>of</strong> error to patients,<br />
but there is discrepancy between responses to hypothetical<br />
clinical scenarios <strong>of</strong> medical errors, in which doctors support<br />
disclosure, and practice, in which they <strong>of</strong>ten do not provide<br />
full disclosure. 9 Among common reasons for avoiding<br />
disclosure are fear <strong>of</strong> malpractice lawsuits or <strong>of</strong> being<br />
exposed as incompetent, as well as feelings <strong>of</strong> shame before<br />
colleagues. 9-13 A pivotal study comparing 1,404 surgeons<br />
and general physicians practicing in Canada to 1,233 practicing<br />
in the United States showed that, even in a different<br />
medical-malpractice environment, a misleading culture <strong>of</strong><br />
medicine that privileges technical perfectionism and success<br />
rates over the humanist aspect <strong>of</strong> care in medical school,<br />
rather than fear <strong>of</strong> being sued, is the prime deterrent to full<br />
disclosure. 14,15 These attitudes toward the medical pr<strong>of</strong>ession<br />
are mostly transmitted through hidden curricula, which<br />
push doctors away from talking about their mistakes for fear<br />
<strong>of</strong> peer judgment or <strong>of</strong> punishment by senior staff.<br />
Retrospective studies suggest that patients and family<br />
members <strong>of</strong>ten sue because <strong>of</strong> doctors’ silence following<br />
adverse events and the accompanying sense that they have<br />
no other recourse for gaining respect, acknowledgment, and<br />
satisfaction. 16 By contrast, patients’ and families’ reasons<br />
not to sue after medical errors are far less known.<br />
From the Department <strong>of</strong> Medicine, New York University Medical School, New York, NY.<br />
Authors’ disclosures <strong>of</strong> potential conflicts <strong>of</strong> interest are found at the end <strong>of</strong> this article.<br />
Address reprint requests Antonella Surbone, MD, PhD, New York University Medical<br />
School, 5550 First Avenue, RBCD516, New York, NY 10016; e-mail: antonella.surbone@<br />
gmail.com.<br />
© <strong>2012</strong> by <strong>American</strong> <strong>Society</strong> <strong>of</strong> <strong>Clinical</strong> <strong>Oncology</strong>.<br />
1092-9118/10/1-10
DISCLOSING MEDICAL ERRORS<br />
Disclosure in oncology, as in other fields <strong>of</strong> medicine,<br />
follows directly from the fiduciary nature <strong>of</strong> the relationship.<br />
5,11,17,18 The relationship between patients with cancer<br />
and their oncologists occurs in a particular context <strong>of</strong> uncertainty<br />
about the evolution <strong>of</strong> the illness and prospects for<br />
treatment effectiveness that affects both parties and their<br />
reciprocal communication. 18,19 This uncertainty also contributes<br />
to burnout among oncologists. 20 In addition, the<br />
close relationships they may develop with their patients can<br />
make committing and disclosing error more difficult for<br />
them to bear. At the same time, severe illness and imminent<br />
mortality among their patients may tempt them to withhold<br />
information about error or adverse events. 5<br />
Even if barriers to disclosure could be reduced, the use <strong>of</strong><br />
experimental anticancer protocols and multiple medications—<strong>of</strong>ten<br />
carrying a high toxicity—along with the multidisciplinary<br />
nature <strong>of</strong> most cancer treatments, can render<br />
more difficult the oncologist’s assessment <strong>of</strong> whether an<br />
adverse event is a side effect <strong>of</strong> treatment or the result <strong>of</strong><br />
medical error. Given all these factors, oncologists may be<br />
tempted to rationalize away some errors and question the<br />
need to disclose them to patients. 5<br />
Questions <strong>of</strong> error, disclosure, and living with error in<br />
oncology should be viewed in the context <strong>of</strong> truth telling to<br />
patients with cancer. On one hand, because <strong>of</strong> the seriousness<br />
<strong>of</strong> their illnesses, patients with cancer may be particularly<br />
vulnerable to the inherent power asymmetry between<br />
them and their oncologists. On the other hand, oncologists<br />
<strong>of</strong>ten must deliver devastating news to their patients, and<br />
the manner and content <strong>of</strong> their delivery can mean the<br />
difference between mere fact giving and truth telling. 18<br />
They may hide behind a torrent <strong>of</strong> medical information,<br />
which, although factually correct, overwhelms and confuses<br />
their patients. On the other hand, however, oncologists can<br />
provide accurate and complete information to their patients,<br />
while still supporting their sense <strong>of</strong> hope through assurance<br />
that they will “be there” for them through the difficult paths<br />
toward cure or palliative care.<br />
KEY POINTS<br />
● The two most effective ways to reduce the effect <strong>of</strong><br />
medical errors are system management and improved<br />
communication.<br />
● Oncologists may experience psychologic and emotional<br />
difficulties, such us anguish, guilt, and shame,<br />
associated with acknowledging and disclosing a<br />
harmful medical error.<br />
● Errors should be communicated openly and clearly,<br />
and a sincere apology should be <strong>of</strong>fered to patients<br />
and their loved ones.<br />
● An apology is essential to restoring trust in the<br />
patient-doctor relationship and initiating a healing<br />
process for all parties in the aftermath <strong>of</strong> a medical<br />
error.<br />
● Making a shift toward increased disclosure requires<br />
humility, accountability and openness among oncologists,<br />
as well as communication skills that can be<br />
learned.<br />
The reality <strong>of</strong> each patient’s illness evolves in unique ways<br />
under the influence <strong>of</strong> many interrelated, contextual variables,<br />
including the patient-doctor relationship. 18 In disclosing<br />
to patients the seriousness <strong>of</strong> their cancer diagnosis and<br />
formulating a prognosis, oncologists can draw on the creative<br />
power <strong>of</strong> the reciprocal trust involved in their relationship,<br />
as much as focusing on the facts <strong>of</strong> the case. Patients<br />
with cancer and their families can thus know that honest<br />
dialogue and true cooperation will take place throughout the<br />
course <strong>of</strong> the illness and feel reassured that they will not be<br />
left alone to cope with the hardest truths, including those<br />
about medical errors. 21<br />
Disclosure <strong>of</strong> Medical Errors in <strong>Oncology</strong><br />
The two most effective ways to reduce the effect <strong>of</strong> medical<br />
errors are system management and improved communication.<br />
22 Telling the truth to patients about medical errors<br />
involves psychologic-emotional elements, in addition to pr<strong>of</strong>essional<br />
and legal ones. For patients, the physical effect<br />
<strong>of</strong> error and the shock <strong>of</strong> learning about it may be compounded<br />
by grief, loss, and a sense <strong>of</strong> isolation. Furthermore,<br />
patients may be forced to live with the possible long-term<br />
physical effects <strong>of</strong> error and the stigma <strong>of</strong> reduced family,<br />
social, and occupational roles sometimes associated with<br />
illness. 5 Finally, the effect on grieving families <strong>of</strong> knowing or<br />
suspecting that a loved one with an incurable illness has<br />
suffered or died from a medical error is potentially devastating,<br />
and thus requires special consideration in regard to<br />
disclosure. 21<br />
As in all clinical communication, content, setting, and<br />
emotional support are important. 23 The late Dr. Robert<br />
Buchman, a major expert in communication with patients<br />
with cancer summarized in the acronym CONES the essential<br />
elements <strong>of</strong> a discussion involving a medical error: C for<br />
choosing the appropriate “context” for carrying a difficult<br />
conversation; O for the “opening” statement that alerts the<br />
patient to a difficult conversation; N for a proper “narrative”<br />
to describe the error, starting with plain words such as “I<br />
found out that”; E for acknowledging and addressing “emotions”<br />
in a direct empathic way by stating upfront, “ it is very<br />
upsetting for you and it is awful for me too”; and S for<br />
“strategy and summary.” 24<br />
In disclosing medical errors it is essential for oncologists<br />
to attend to both medical and emotional aspects <strong>of</strong> the<br />
information provided and the reactions they may elicit from<br />
patients and families. The physician’s response must be<br />
clear at the clinical level and also address the emotional<br />
needs <strong>of</strong> patients and family members. In addition, in<br />
disclosing medical errors and discussing their causes and<br />
actual or potential consequences, oncologists should also<br />
share their own emotions. Four different levels <strong>of</strong> response—<br />
direct, escalationary, exploratory and empathic—have been<br />
identified. Although the first three predominantly address<br />
the factual aspects <strong>of</strong> the situation, only an empathic response<br />
has the potential to result in effective communication<br />
about something so painful and upsetting as a medical<br />
error. 24<br />
Culled from practical guidelines developed by institutional<br />
bodies, published research, and clinical experience,<br />
Table 1 lists suggestions for oncologists facing the difficult<br />
task <strong>of</strong> revealing to patients or family members that a<br />
medical error has occurred. 25-27<br />
e25
Table 1. Key Steps in Disclosing Medical Errors to Patients<br />
and Relatives<br />
● Initiate the conversation in a proper setting.<br />
● Start by saying that the conversation will be difficult and then give an explicit<br />
statement about what happened.<br />
● Talk and respond with humility and sorrow, not with defensiveness.<br />
● Offer a sincere clear apology, including admission <strong>of</strong> the error and remorse.<br />
● If you are still in training, ask for the presence <strong>of</strong> a senior staff member.<br />
● Refrain from pointing to someone else or to the system.<br />
● Disclose relevant data and take responsibility.<br />
● Promise that a thorough investigation <strong>of</strong> the error will be conducted.<br />
● Make sure that precise and timely answers are provided.<br />
● Guarantee constant monitoring <strong>of</strong> the patient as your priority.<br />
● While keeping an open attitude about possible good outcomes, avoid making<br />
promises you can’t keep, yet state that you hope for the best outcome.<br />
● Set up a follow-up meeting with family members, including anyone else they wish<br />
to involve.<br />
● Offer the help <strong>of</strong> a patient advocate and legal consultant to assist patient and<br />
relatives.<br />
Disclosure <strong>of</strong> Medical Errors: Individual and<br />
Cultural Aspects<br />
According to the Institute <strong>of</strong> Medicine (IOM), the standard<br />
definition <strong>of</strong> medical errors is “the failure <strong>of</strong> a planned action<br />
to be completed as intended or the use <strong>of</strong> a wrong plan to<br />
achieve an aim.” 1 Most patients, however, have a broader<br />
perspective on medical errors than physicians or researchers.<br />
5 For example, they tend to include among medical<br />
errors their physician’s failure to communicate effectively<br />
with them before and after an error has occurred. Individual<br />
physician or institutional arrogance has been described by<br />
some patients with cancer as contributing to their perception<br />
<strong>of</strong> medical errors and to the possibility <strong>of</strong> repairing a<br />
breach in trust. 28,29 Furthermore, the absence <strong>of</strong> physician<br />
empathy and honesty in the aftermath <strong>of</strong> harmful medical<br />
error not only exacerbates patients’ and family members’<br />
suffering but may be seen as part <strong>of</strong> the error itself.<br />
Disclosure <strong>of</strong> medical errors is now an ethical requirement<br />
<strong>of</strong> the medical pr<strong>of</strong>ession in the Western world. It cannot be<br />
ignored, however, that disclosing an error is inevitably<br />
related to the different attitudes and practices <strong>of</strong> providing<br />
information to patients with cancer in different countries<br />
and cultural groups. Clearly, it is considered more difficult,<br />
and it is less common to disclose medical errors in contexts<br />
where the truth about their overall condition and prognosis<br />
is withheld from patients with cancer.<br />
A recent study <strong>of</strong> surgeon-patient disclosure practices in<br />
Southwestern Nigeria asked 102 surgeons the question: “To<br />
whom should surgical errors be disclosed?” Only 15.7%<br />
replied that the information should be conveyed to the<br />
patient, whereas 30% indicated that it should be delivered to<br />
a family member. Others suggested that errors should be<br />
referred to the hospital ethics committee (22.5%), to hospital<br />
management staff (21.5%), to a colleague (21.5%), or to other<br />
staff members during a clinical meeting (2%). 30<br />
Overall, the authors <strong>of</strong> the study reported that surgeons<br />
were unsure about whether they should disclose medical<br />
errors. 30 Notably though, participating surgeons valued the<br />
potential benefits <strong>of</strong> disclosure in a similar way to that <strong>of</strong><br />
Western physicians, listing among those benefits: learning<br />
e26<br />
from errors; prevention <strong>of</strong> further complications; promotion<br />
<strong>of</strong> honesty, openness, trust and confidence; reduction <strong>of</strong> risk<br />
<strong>of</strong> litigation; relieving the doctor’s conscience; and overall<br />
practice improvement. Among the potential risks <strong>of</strong> disclosing<br />
medical errors they noted litigation and prosecution,<br />
negative effect on practice, loss <strong>of</strong> trust and confidence, loss<br />
<strong>of</strong> medical license, undue patient anxiety, and even risk <strong>of</strong><br />
physical assault. 30<br />
Conclusion<br />
ANTONELLA SURBONE<br />
Oncologists’ failure to disclose medical errors to their<br />
patients betrays the fiduciary nature <strong>of</strong> the patient-doctor<br />
relationship and diminishes the integrity <strong>of</strong> our pr<strong>of</strong>ession<br />
by silencing patients and family members through neglect <strong>of</strong><br />
their stories and experiences, as well as by obfuscating the<br />
effect that errors have on ourselves. 5 Silence may increase<br />
the risk <strong>of</strong> malpractice suits and is always morally wrong,<br />
for it inevitably adds to the pain <strong>of</strong> patients and their loved<br />
ones. 21<br />
Further research on the incidence and types <strong>of</strong> error, near<br />
errors, and perceived errors in oncology, using both qualitative<br />
and quantitative methods, and development <strong>of</strong> clear<br />
policies and practices regarding disclosure and other posterror<br />
interactions with patients and family members, are<br />
needed. Understanding that, like truth-telling about cancer<br />
diagnosis and prognosis, attitudes toward and practices <strong>of</strong><br />
disclosure <strong>of</strong> medical errors are subject to individual and<br />
cultural variability and should not prevent us from learning<br />
about sensitive communication about medical errors and<br />
from meeting adequate standards <strong>of</strong> ethical clinical practice<br />
worldwide. Analysis <strong>of</strong> institutional standards, policies, procedures,<br />
and training regarding disclosure <strong>of</strong> medical errors<br />
with a focus on how oncologists and other team members<br />
understand and apply these specific communication elements<br />
will guide researchers and practitioners in efforts to<br />
build on current practice.<br />
Mortality and Morbidity Conferences are <strong>of</strong>ten a proper<br />
setting for physicians to disclose and discuss their errors,<br />
but they do not provide the setting for expressing emotions<br />
and finding support. Didactic and experiential education<br />
and training regarding responsibility and accountability for<br />
optimal ongoing communication throughout the course <strong>of</strong><br />
the patient’s illness, including disclosure <strong>of</strong> medical errors,<br />
should be enhanced at all levels, from medical school<br />
through internship, to residency and oncology fellowship.<br />
Emphasis should be placed on the redemptive value <strong>of</strong><br />
apology and forgiveness for patients and their families and<br />
for oncologists, and on provision <strong>of</strong> emotional support for all<br />
parties affected by medical errors.<br />
Many compelling physicians’ narratives reporting the experience<br />
<strong>of</strong> medical errors are available to teach us that<br />
communication <strong>of</strong> medical errors is not only about disclosure<br />
to patients and families. Rather, it is also about creating a<br />
culture <strong>of</strong> medicine in which the patient’s interest and<br />
well-being always has priority over physician or institutional<br />
self-interest. This principle and goal can be achieved<br />
through establishing a serene atmosphere <strong>of</strong> collaboration,<br />
rather than competition, and open communication among all<br />
members <strong>of</strong> the staff, one in which a medical error is not a<br />
reason to fear punishment or loss <strong>of</strong> esteem from peers, but<br />
an incentive for all to share, understand, and improve.
DISCLOSING MEDICAL ERRORS<br />
Author’s Disclosures <strong>of</strong> Potential Conflicts <strong>of</strong> Interest<br />
Author<br />
Antonella Surbone*<br />
*No relevant relationships to disclose.<br />
Employment or<br />
Leadership<br />
Positions<br />
Consultant or<br />
Advisory Role<br />
1. Kohn LT, Corrigan J, Donaldson MS. To err is human: building a safer<br />
health system. Institute <strong>of</strong> Medicine (IOM) Washington, D.C.: National<br />
Academy Press; 2000.<br />
2. Ethics manual, <strong>American</strong> College <strong>of</strong> Physicians. 6th Edition. http://<br />
www.acponline.org/running_practice/ethics/manual.htm. Accessed February<br />
15, <strong>2012</strong>.<br />
3. Rowe M. Doctors’ responses to medical errors. Critical Reviews in<br />
<strong>Oncology</strong>/Hematology. 2004;52:147-163.<br />
4. Hilfiker D. Facing our mistakes. N Engl J Med. 1984;310:318-322.<br />
5. Surbone A, Rowe M, Gallagher TH. Confronting medical errors in<br />
oncology and disclosing them to cancer patients. J Clin Oncol. 2007;25:1463-<br />
1467.<br />
6. Finkelstein D, Wu A, Holtzman N, et al. When a physician harms a<br />
patient by a medical error: ethical, legal, and risk-management considerations.<br />
J Clin Ethics. 1997;8:330-335.<br />
7. Gallagher TH, Suddert D, Levinson L. Disclosing harmful errors to<br />
patients. N Engl J Med. 2007;356:2713-2719.<br />
8. Duclos CW, Eichler M, Taylor L, et al. Patient perspectives <strong>of</strong> patientprovider<br />
communication after adverse events. Intl J Qual Health Care.<br />
2005;17:479-486.<br />
9. Gallagher TH, Waterman AD, Ebers AG, et al. Patients’ and physicians’<br />
attitudes regarding the disclosure <strong>of</strong> medical errors. JAMA. 2003;289:1001-<br />
1007.<br />
10. Vincent C, Stanhope N, Crowley-Murphy M. Reasons for not reporting<br />
adverse events: an empirical study. J Eval Clin Pract. 1999;5:13-21.<br />
11. Beckman HB, Markakis KM, Suchman AL, et al. The doctor-patient<br />
relationship and malpractice. Lessons from plaintiffs’ depositions. Arch Intern<br />
Med. 1994;154:1365-1370.<br />
12. Hickson GB, Federspiel CF, Pichert JW, et al. Patient complaints and<br />
malpractice risk. JAMA. 2002;287:2951-2957.<br />
13. Mazor KM, Simon SR, Gurwitz JH. Communicating with patients<br />
about medical errors: a review <strong>of</strong> the literature. Arch Intern Med 2004;164:<br />
1690-1697.<br />
14. Gallagher TH, Garbutt JM, Waterman AD, et al. Choosing your words<br />
carefully: how physicians would disclose harmful medical errors to patients.<br />
Arch Intern Med. 2006;166:1585-1593.<br />
Stock<br />
Ownership Honoraria<br />
REFERENCES<br />
Research<br />
Funding<br />
Expert<br />
Testimony<br />
Other<br />
Remuneration<br />
15. Mahhod SC. Medical education. beware the hidden curriculum. Can<br />
Fam Phys. 2011;57:983-985.<br />
16. Vidmar N. Medical malpractice lawsuits: an essay on patient interests,<br />
the contingency fee system, juries, and social policy. Loy L. Law Rev. 2005; 38;<br />
1217-1266.<br />
17. Pellegrino ED, Thomasma DC. For the Patient’s Good. The Restoration<br />
<strong>of</strong> Beneficence in Health Care. London: Oxford University Press; 1988.<br />
18. Surbone A. Telling the truth to patients with cancer: what is the truth?<br />
Lancet Oncol. 2006;7:944-950.<br />
19. Schapira L. Shared uncertainty. J Support Oncol. 2004;2:14-18.<br />
20. Penson RT, Dignan FL, Canellos GP, et al. Burnout: caring for the<br />
caregivers. Oncologist. 2000;5:425-434.<br />
21. Rowe M. The rest is silence: hospitals and doctors should beware <strong>of</strong><br />
what can fill the space <strong>of</strong> their silence after a loved one’s death. Health Aff.<br />
2002;21:232-236.<br />
22. Penson RT, Svendsen SS, Chabenr BA, et al. Medical mistakes: a<br />
workshop on personal perspectives. Oncologist. 2001;6:92-99.<br />
23. Parker PA, Baile WF, de Moor C, et al. Breaking bad news about<br />
cancer: patients’ preferences for communication. J Clin Oncol. 2001;19:2049-<br />
2056.<br />
24. Buchman R. Interpersonal Communication And Relationship Enhancement:<br />
I*CARE. http://www.mdanderson.org/education-and-research/<br />
resources-for-pr<strong>of</strong>essionals/pr<strong>of</strong>essional-educational-resources/i-care/index.<br />
html. Accessed February 15, <strong>2012</strong>.<br />
25. Fallowfield L. Communication with patients after errors. J Heath Serv<br />
Res. 2010;15:56-59.<br />
26. Evans SB,Decker R. Disclosing medical errors: a practical guide and<br />
discussion <strong>of</strong> radiation oncology-specific controversies. Int J Radiat Oncol Biol<br />
Phys. 2011;5:1285-1288.<br />
27. O’Connell D, Kemp White M, Platt FW. Disclosing Unanticipated<br />
Outcomes and Medical Errors. JCOM. 2003;10:25-29.<br />
28. Ingelfinker FJ: Arrogance. N Engl J Med. 1980;303:1507-1511.<br />
29. Brook I. A physician’s experience as a cancer <strong>of</strong> the neck patient. Surg<br />
Oncol. 2010;19:188-192.<br />
30. Ogundiran TO, Adebamowo CA. Surgeon-Patient information disclosure<br />
practices in Southwestern Nigeria. Med Princ Pract. Epub 2011 Nov 23.<br />
e27
Preventing Errors in <strong>Oncology</strong>: Perspective<br />
<strong>of</strong> a Physician Who Is Also a Cancer Patient<br />
Overview: This article presents my personal experiences as a<br />
physician who underwent laryngectomy for hypopharyngeal<br />
squamous cell carcinoma. I describe the numerous medical<br />
and surgical errors that occurred during my hospitalization at<br />
ERRORS IN patient care are very common, occurring in<br />
up to 40% <strong>of</strong> patients who undergo surgery. These<br />
errors can cause complications in up to 18% <strong>of</strong> these patients.<br />
1,2 These errors <strong>of</strong>ten lead to malpractice lawsuits,<br />
increase expense <strong>of</strong> medical care, extend hospital stays, and<br />
can lead to greater morbidity and mortality. 3<br />
I have been a physician specializing in infectious diseases<br />
in the hospital setting for 40 years. I was not aware <strong>of</strong> how<br />
frequent medical errors were until I became a patient with<br />
cancer myself following the diagnosis <strong>of</strong> throat malignancy<br />
(hypopharyngeal carcinoma). For the first time, I had to deal<br />
with medical and surgical errors in my care as a patient and<br />
not as a physician.<br />
My primary small hypopharyngeal cancer (T1, N0, M0)<br />
was surgically removed, and I received local radiation;<br />
however, after 20 months, a local recurrence at a different<br />
location (pyriform sinus) was found (T2, N0, M0). Even<br />
though they tried three times, the surgeons failed to completely<br />
remove the cancer using laser. I then underwent<br />
complete pharynolaryngectomy with flap reconstruction at a<br />
different medical center that had greater experience with<br />
this type <strong>of</strong> tumor. The cancer was completely removed, and<br />
no local or systemic recurrence was noted.<br />
I was treated at three large academic medical centers (two<br />
military and one civilian) located at a major metropolitan<br />
center. Even though the overall medical care I received at all<br />
<strong>of</strong> these hospitals was adequate, I very quickly learned that<br />
medical errors occurred at all levels <strong>of</strong> my care. Despite<br />
these errors, I am grateful to the medical staff, including<br />
physicians, nurses, and other health care providers who<br />
cared for me.<br />
This article describes the medical errors I experienced<br />
during my hospitalizations at three hospitals and their<br />
impact on me. My inability to speak following my laryngectomy<br />
made it difficult for me to prevent many <strong>of</strong> these errors.<br />
Fortunately, I was able to prevent many <strong>of</strong> these mistakes,<br />
but not all <strong>of</strong> them.<br />
Delay in Diagnosing Cancer Recurrence<br />
My surgeons did not discover the recurrence <strong>of</strong> my tumor<br />
in a timely fashion, although they endoscopically examined<br />
me on a monthly basis following my initial surgery. This<br />
happened despite the fact that I had been complaining <strong>of</strong><br />
sharp and persistent pain in the right side <strong>of</strong> my throat for<br />
7 months. My surgeons reassured me that the pain was most<br />
likely caused by irritation because <strong>of</strong> the reflux <strong>of</strong> stomach<br />
acid. My physician increased the acid-reducing medication I<br />
was taking, but the pain did not disappear.<br />
The return <strong>of</strong> cancer was finally discovered by an astute<br />
resident who was the first examiner to perform an endoscopic<br />
examination using a Valsalva maneuver (exhaling<br />
and holding the air). This technique allows visualization <strong>of</strong><br />
e28<br />
By Itzhak Brook, MD, MSc<br />
three different medical centers. It is my hope that my presentation<br />
will contribute to the reduction <strong>of</strong> such errors and lead<br />
to a safer hospital environment for patients.<br />
the pyriform sinus where the tumor was found. In retrospect,<br />
I was wonder why the experienced head and neck<br />
surgeons who always examined me failed to perform such a<br />
basic procedure before. If they had done it earlier, the cancer<br />
recurrence (which was 4 by 2 cm) would have most likely<br />
been seen and removed at an earlier time, thus reducing the<br />
chance for the spread and need for laryngectomy.<br />
Only three weeks earlier, I was also examined endoscopically<br />
by a radiation oncologist who had seen no abnormality.<br />
This physician confessed to me later that he did not look<br />
down into the area where the new cancer was located<br />
because his instrument broke during the procedure.<br />
Failure to Excise the Recurrent Tumor Using Laser<br />
The first mistake during my admission for surgery was<br />
when my surgeons, using laser technology, mistakenly removed<br />
scar tissue instead <strong>of</strong> the cancerous lesion. The tumor<br />
was located further down in the pyriform sinus. It took a<br />
week before they realized that an error was made when the<br />
pathological studies showed that the excised material was<br />
scar tissue and not the lesion. This mistake could have been<br />
prevented if frozen sections <strong>of</strong> the suspected lesion, rather<br />
than just the margins <strong>of</strong> the pyriform sinus, had been<br />
studied in the operating room. As a result <strong>of</strong> this mistake, I<br />
had to undergo an additional surgical procedure with laser<br />
10 days later to remove the cancerous lesion.<br />
Following the second surgery, my surgeons informed me<br />
they were able to remove the cancer in its entirety by using<br />
the laser, and all the margins <strong>of</strong> the removed area were clear<br />
<strong>of</strong> cancer. I was therefore spared from a more extensive<br />
surgery, which would have included removal <strong>of</strong> tissues in<br />
the right side <strong>of</strong> my neck, requiring their replacement by<br />
tissue transplanted from my thighs or shoulder areas. I felt<br />
great relief when I heard the news and felt very fortunate.<br />
Even though the bulk <strong>of</strong> the tumor was finally removed, the<br />
surgeons could not get all <strong>of</strong> it out. I had to undergo two<br />
additional attempts to remove the cancer.<br />
I later realized that the failed attempt to remove the<br />
cancer on the first surgery made any repeated attempt more<br />
difficult because each surgery induces extensive local swelling<br />
and inflammation. This was especially true in my case<br />
because my cancer was located at a very narrow and inaccessible<br />
location. This made any follow-up interventions<br />
From the Department <strong>of</strong> Pediatrics, Georgetown University School <strong>of</strong> Medicine, Washington,<br />
DC.<br />
Author’s disclosures <strong>of</strong> potential conflicts <strong>of</strong> interest are found at the end <strong>of</strong> this article.<br />
Address reprint requests to Itzhak Brook, MD, MSc, 4431 Albemarle St. NW, Washington<br />
DC 20016; email: ib6@georgetown.edu.<br />
© <strong>2012</strong> by <strong>American</strong> <strong>Society</strong> <strong>of</strong> <strong>Clinical</strong> <strong>Oncology</strong>.<br />
1092-9118/10/1-10
PREVENTING MEDICAL ERRORS<br />
more difficult because insertion <strong>of</strong> an endoscope and visualization<br />
<strong>of</strong> the area were more difficult.<br />
I learned the hard way that experience is <strong>of</strong> primary<br />
importance with this kind <strong>of</strong> surgery. Since the frequency <strong>of</strong><br />
throat cancer has declined in the United States, there are<br />
fewer patients with this cancer and surgeons have less<br />
experience removing it. With fewer patients, the expertise in<br />
its removal and care is concentrated in fewer institutions.<br />
Even though my surgeons had very little experience using<br />
laser to remove my type <strong>of</strong> cancer, they still <strong>of</strong>fered to<br />
remove it using this technique. One <strong>of</strong> my surgeons summarized<br />
his philosophy about learning new procedures as: “See<br />
one, do one, teach one.” I believe the best approach should<br />
be: “See a hundred, do 200, teach one.”<br />
Although the error made by my surgeons was very regrettable,<br />
they admitted their responsibility for making it. This<br />
made it easier for me to accept the error and allow them to<br />
try to remove the cancer again, even though they suggested<br />
that I could seek care in another center.<br />
Failure <strong>of</strong> the Surgical Intensive Care Unit Staff to<br />
Respond to Breathing Difficulties<br />
I had experienced multiple unsafe situations because <strong>of</strong><br />
errors made by nurses. A serious situation occurred one day<br />
after my laryngectomy while I was still in the Surgical<br />
Intensive Care Unit (SICU). I suddenly felt an obstruction <strong>of</strong><br />
my airway and reached for the call button. I could not find it<br />
because it had fallen to the floor. The nurse who cared for me<br />
also cared for two more patients and was absent when I<br />
needed her help. I was unable to move because I was<br />
connected to multiple tubes and lines. I attempted to get the<br />
attention <strong>of</strong> the staff by disconnecting my heart- and oxygensaturation<br />
monitors, and even though I was a few feet away<br />
from the nursing station, I was ignored until my wife arrived<br />
about 10 minutes later and called for help. I was helpless in<br />
getting assistance without a voice and was desperately in<br />
need <strong>of</strong> air while medical personnel passed me by.<br />
When my wife complained about what happened to the<br />
nursing supervisor and SICU attending physician, she was<br />
rudely rebuffed. When I informed my surgeon about the<br />
incident, he just shrugged his shoulders and told me that he<br />
had little influence on what happens in the SICU, but he<br />
assured me that things would be much better for me when I<br />
was moved to the otolaryngology floor. The lack <strong>of</strong> willingness<br />
<strong>of</strong> my surgeon to act upon my complaint was very<br />
disappointing. Instead <strong>of</strong> addressing the problem in the<br />
SICU that cares for his patients when they are critically ill,<br />
KEY POINTS:<br />
● Medical, surgical, and nursing errors are common in<br />
patient care.<br />
● Some medical errors may be life threatening and can<br />
lead to increased morbidity and mortality.<br />
● Preventing such errors is <strong>of</strong> utmost importance.<br />
● A dedicated patient advocate, such as a family member<br />
or a friend, is very much needed for all hospitalized<br />
patients.<br />
● Open discussion <strong>of</strong> such errors may decrease them<br />
and eventually lead to better care.<br />
he comforted me by promising better care when I will be less<br />
needy for it.<br />
Failure to Respond to Breathing Difficulties in the<br />
Otolaryngology Ward<br />
A similar incident also occurred in the otolaryngology<br />
ward a week later when a nurse failed to respond to my<br />
urgent call to suction my trachea. It happened after I<br />
suddenly experienced difficulty in breathing because mucus<br />
was blocking my airway. Even though I pressed the call<br />
button, no one came to my assistance. I was finally able to<br />
get the attention <strong>of</strong> a nurse assistant who informed me that<br />
my nurse was on a break. The nurse assistant was not<br />
trained in suctioning airways, but she promised to look for a<br />
nurse. It took my nurse 15 minutes to come to suction my<br />
trachea. I later learned that she was busy ordering supplies<br />
on the phone during that time.<br />
I was very distressed as I was struggling to breathe in.<br />
Present in the otolaryngology ward during this time were<br />
two resident physicians and several nurse assistants, yet no<br />
one came to my help. It is clear that even on a ward<br />
dedicated to helping people with breathing difficulties, there<br />
were distractions that prevented physicians and nurses<br />
from paying attention to their patient’s urgent needs.<br />
Even though I brought the incident to the attention <strong>of</strong> the<br />
nurse supervisor and the head surgeon, I never received any<br />
feedback from them about what was to be done to prevent<br />
such incidents in the future. The lack <strong>of</strong> response was<br />
inappropriate and contributed to my frustration and anxiety.<br />
Prematurely Feeding by Mouth<br />
One <strong>of</strong> the most serious errors in my care was feeding me<br />
by mouth with food a week too early. Early mouth-feeding<br />
following laryngectomy with free-flap reconstruction can<br />
lead to its failure to integrate. The oral feeding continued for<br />
over 16 hours. The premature feeding was stopped only<br />
because I continued questioning this practice and brought it<br />
to the attention <strong>of</strong> the attending surgeon. I wonder what<br />
would have occurred if I would not have not persistently<br />
questioned the feeding. Would the error been eventually<br />
discovered?<br />
I repeatedly requested an explanation for this error, but<br />
the staff avoided responding to my inquiries. I finally<br />
learned only by looking in my medical records that this<br />
mistake occurred because a verbal order to start oral feeding<br />
intended for another patient was erroneously written in my<br />
chart.<br />
The handling <strong>of</strong> this incident was an example <strong>of</strong> a lack <strong>of</strong><br />
communication by the physicians and me. They failed to<br />
explain and apologize for the mistake. Accepting responsibility<br />
for the mistake and outlining what steps should be<br />
taken to prevent such mishaps in the future would have<br />
been the most appropriate way <strong>of</strong> handling the incident.<br />
Errors in Nursing Care<br />
Some <strong>of</strong> the errors by nursing and other staff included the<br />
following: not cleaning or washing their hands, not using<br />
gloves when indicated, taking an oral temperature without<br />
placing the thermometer in a plastic sheath, using an<br />
inappropriately sized blood pressure cuff (thus getting<br />
alarming readings), attempting to administer medications<br />
e29
y mouth that were intended to be given by nasogatric tube,<br />
dissolving pills in hot water and administering them<br />
through the feeding tube (which caused burning in the<br />
esophagus), delivering an incorrect dose <strong>of</strong> a medication,<br />
connecting a suction machine directly to the suction port in<br />
the wall without a bottle <strong>of</strong> water, forgetting to rinse away<br />
the hydrogen peroxide used for cleaning the tracheal breathing<br />
tube (thus causing severe irritation), forgetting to connect<br />
the call button when I was bedridden and unable to<br />
speak, and forgetting to write down verbal orders.<br />
I always informed the nurse supervisors and whenever I<br />
could the resident and/or attending physicians about the<br />
mistakes that were made; however, I was never informed<br />
what action was taken to prevent similar errors in the<br />
future.<br />
Conclusion<br />
One <strong>of</strong> the key recommendations to prevent medical errors<br />
is that patients should choose, whenever possible, experienced<br />
places that deal with their kind <strong>of</strong> disease regularly.<br />
Such familiarity and experience can reduce the occurrence <strong>of</strong><br />
errors and increase the safety <strong>of</strong> the patient. The way a<br />
patient can contribute to the prevention <strong>of</strong> medical errors is<br />
to be proactive and take these steps: be informed and not<br />
hesitate to challenge and ask for explanations; become<br />
knowledgeable about their medical condition; have family or<br />
friends serve as an advocate in the hospital; get a second<br />
opinion when making an important decision, such as deciding<br />
on the course <strong>of</strong> treatment; and educate the medical<br />
caregivers about their condition and needs (prior to and<br />
after surgery).<br />
All <strong>of</strong> the errors made in my care make me wonder what<br />
happens to individuals without medical background who<br />
cannot recognize and prevent many errors. Fortunately,<br />
despite the errors made in my care, I did not suffer any<br />
long-term consequences. However, to prevent medical errors,<br />
I had to be continuously on guard and vigilant, which<br />
Author’s Disclosure <strong>of</strong> Potential Conflicts <strong>of</strong> Interest<br />
Author<br />
Itzhak Brook*<br />
*No relevant relationships to disclose.<br />
Employment or<br />
Leadership<br />
Positions<br />
Consultant or<br />
Advisory Role<br />
1. Tezak B, Anderson C, Down A, et al. Looking ahead: the use <strong>of</strong><br />
prospective analysis to improve the quality and safety <strong>of</strong> care. Healthc Q.<br />
2009;12:80-84.<br />
2. Griffen FD, Turnage RH. Reviews <strong>of</strong> liability claims against surgeons:<br />
what have they revealed? Adv Surg. 2009;43:199-209.<br />
3. Studdert DM, Mello MM, Gawande AA, et al. Claims, errors, and<br />
e30<br />
was a very exhausting chore, especially during the difficult<br />
recovery period.<br />
My family members were instrumental in preventing<br />
many errors, highlighting the value <strong>of</strong> a dedicated patient<br />
advocate. My experiences taught me that it is essential that<br />
medical personnel openly discuss with their patients the<br />
mistakes that were made in their care. Since errors in<br />
patient care weaken patients’ trust in their caregivers,<br />
admission and acceptance <strong>of</strong> responsibility by the care<br />
providers can rebuild trust and re-establish the lost confidence.<br />
The establishment <strong>of</strong> a dialogue among patients,<br />
physicians, and other staff facilitates the discovery <strong>of</strong> the<br />
circumstances leading to the mistake, which assists in<br />
preventing similar ones in the future. Open discussion can<br />
reassure the patient that their care givers are taking the<br />
matter seriously and taking steps to make hospital stays<br />
safer for patients.<br />
Avoiding the discussion <strong>of</strong> the errors with the patient and<br />
their family only increases anxiety, frustration, and anger.<br />
This can interfere with the patient’s recovery and contribute<br />
to malpractice law suits.<br />
Medical mistakes should be prevented as much as humanly<br />
possible. Ignoring them can only lead to their repetition.<br />
Medical errors usually involve problems in caring for<br />
and treating patients and are improved if the processes are<br />
improved and a blame-free reporting mechanism is instituted.<br />
The recent development <strong>of</strong> a mandatory bedside<br />
checklist is a simple, cost-effective method to prevent many<br />
medical errors. 4<br />
I am sharing my personal experiences in the hope that<br />
they will encourage better medical training, contribute to<br />
greater diligence in care, and increase supervision and<br />
communication between health care providers and their<br />
patients. It is my hope that this presentation and article will<br />
contribute to the reduction <strong>of</strong> medical errors and create a<br />
safer environment in the hospital setting. It is also my hope<br />
that if mistakes do happen, medical caregivers will openly<br />
discuss them with their patients.<br />
Stock<br />
Ownership Honoraria<br />
REFERENCES<br />
Research<br />
Funding<br />
Expert<br />
Testimony<br />
ITZHAK BROOK<br />
Other<br />
Remuneration<br />
compensation payments in medical malpractice litigation. N Engl J Med.<br />
2006;354:2024-2033.<br />
4. Byrnes MC, Schuerer DJ, Schallom ME, et al. Implementation <strong>of</strong> a<br />
mandatory checklist <strong>of</strong> protocols and objectives improves compliance with a<br />
wide range <strong>of</strong> evidence-based intensive care unit practices. Crit Care Med.<br />
2009;37:2775-2781.
“PERSONALIZED” ONCOLOGY FOR COLORECTAL<br />
CANCER: READY FOR PRIME TIME OR<br />
STOP THE TRAIN?<br />
CHAIR<br />
Alan P. Venook, MD<br />
University <strong>of</strong> California, San Francisco<br />
San Francisco, CA<br />
SPEAKERS<br />
Johanna C. Bendell, MD<br />
The Sarah Cannon Research Institute<br />
Nashville, TN<br />
Robert S. Warren, MD<br />
University <strong>of</strong> California, San Francisco<br />
San Francisco, CA
Is There Currently an Established Role for<br />
the Use <strong>of</strong> Predictive or Prognostic Molecular<br />
Markers in the Management <strong>of</strong> Colorectal<br />
Cancer? A Point/Counterpoint<br />
By Alan P. Venook, MD, Johanna C. Bendell, MD, and Robert S. Warren, MD<br />
Overview: The term “personalized oncology” means different<br />
things to the oncologist than to the patient. But fundamentally,<br />
the phrase creates the expectation that decisions can be<br />
informed by the unique features <strong>of</strong> the patient and patient’s<br />
cancer. Much like determining antibiotic sensitivities in urinary<br />
tract infections, the oncologist is expected to choose the<br />
right treatment(s), for each individual patient.<br />
Numerous methods can be used to “personalize” management<br />
decisions, although truly useful biomarkers continue to<br />
escape our grasp. Positron Emission Tomography in patients<br />
with GI stromal tumors or genotyping <strong>of</strong> c-kit in chronic<br />
myelogenous leukemia cells can guide the use <strong>of</strong> imatinib,<br />
these scenarios represent a minority <strong>of</strong> patients. The promise<br />
<strong>of</strong> individualized therapy, however, has led to the commercialization<br />
<strong>of</strong> numerous assays to probe patient’s genetic<br />
make-up and that <strong>of</strong> the tumor. Breast cancer management<br />
THE TERM “personalized oncology” means different<br />
things to the oncologist than to the patient. But fundamentally,<br />
the phrase creates the expectation that decision<br />
making can be informed by the unique features <strong>of</strong> the<br />
patient and <strong>of</strong> that patient’s cancer. Much like culturing the<br />
urine in a urinary tract infection and determining antibiotic<br />
sensitivities, it implies that oncologists can pick just the<br />
right treatment, or combination <strong>of</strong> treatments, for each<br />
specific patient.<br />
There are numerous tools that can be used to “personalize”<br />
management decisions, although for much <strong>of</strong> the last<br />
decade the truly useful biomarkers were few and far between.<br />
Although positron emission tomography imaging in a<br />
patient with a GI stromal tumor patient or mutational<br />
analysis <strong>of</strong> chronic myelogenous leukemia cells could direct<br />
the proper decision about whether to use imatinib, these<br />
scenarios represented a distinct minority <strong>of</strong> patients. The<br />
promise, however, led to the commercialization <strong>of</strong> numerous<br />
assays to probe a patient’s genetic make-up and/or the<br />
cancer. We now know that gene recurrence scores can<br />
“personalize” breast cancer management in some patients,<br />
and more recently we have capitalized on mutations in<br />
non-small cell lung cancers and melanomas to subdivide<br />
these patients into subsets <strong>of</strong> diseases. We have shown<br />
recently that some <strong>of</strong> these infrequent mutations can be<br />
targeted successfully by small-molecule inhibitors.<br />
Despite the high incidence <strong>of</strong> colorectal cancer and our<br />
relatively long-standing grasp <strong>of</strong> the molecular pathways in<br />
colorectal cancer—the polyp-to-malignancy transformation—the<br />
management <strong>of</strong> disease in patients with colorectal<br />
cancer for years is mostly empiric and movement toward<br />
“personalization” has been incremental. Now, however, the<br />
availability <strong>of</strong> newer imaging modalities and commercial<br />
assays for genetic mutational analysis <strong>of</strong> tumor specimens<br />
has put the oncologist and oncologic surgeon in the crossfire<br />
with patients and families who believe the era <strong>of</strong> “personalization”<br />
is here.<br />
Now the debate: The medical oncologist (Johanna Bendell)<br />
has benefitted from the analysis <strong>of</strong> gene recurrence scores.<br />
More recently the analysis <strong>of</strong> germline or tumor-associated<br />
mutations in non-small cell lung cancer and melanoma has led<br />
to clinically meaningful molecular subsets <strong>of</strong> these diseases,<br />
guiding the successful targeting <strong>of</strong> such cancers with smallmolecule<br />
inhibitors.<br />
Despite the high incidence <strong>of</strong> colorectal cancer and our<br />
relatively long-standing grasp <strong>of</strong> the molecular pathways in<br />
colorectal carcinogenesis, the management <strong>of</strong> these patients<br />
remains mostly empiric and movement toward “personalization”<br />
has been slow and incremental. Now, however, molecular<br />
imaging and commercial assays for genetic makeup <strong>of</strong><br />
tumor specimens has put the oncologist and oncologic surgeon<br />
in the crossfire with patients and families who believe<br />
the era <strong>of</strong> “personalization” is here.<br />
wants the surgeon (Robert Warren) to carefully collect<br />
tumor tissue <strong>of</strong> the patient undergoing surgery. Always a<br />
willing collaborator, the surgeon—mindful <strong>of</strong> cost and the<br />
pathologist’s time—asks what makes the medical oncologist<br />
think these tests will be helpful. The patient expects a<br />
compromise that will help “personalize” care.<br />
Johanna Bendell, MD: Ready for Prime Time<br />
As part <strong>of</strong> the referral to the surgeon, the medical oncologist<br />
lists a number <strong>of</strong> specific tissue acquisition needs and<br />
handling instructions to the surgeon. The message: these<br />
are the biomarkers we will use to decide how to treat this<br />
patient and here are the data to support our decision.<br />
KRAS<br />
Many <strong>of</strong> the biomarkers currently in use or under active<br />
investigation are based around research done with epidermal<br />
growth factor receptor (EGFR) pathway inhibitors.<br />
These include KRAS, BRAF, and rash. The data from the<br />
CRYSTAL trial, where patients were randomly assigned in<br />
the first-line setting to receive fluorouracil (FU), leucovorin<br />
(LV), and irinotecan (IFL; FOLFIRI) with or without cetuximab,<br />
is well known. 1 This trial showed that the addition <strong>of</strong><br />
cetuximab improved response rate (49% vs. 39%; p � 0.0038)<br />
and progression-free survival (8.9 vs. 8.0 months; p � 0.05).<br />
For the patients with wild-type KRAS status, who comprised<br />
37% <strong>of</strong> the patients, the benefit <strong>of</strong> cetuximab over no cetuximab<br />
was greater for response rate (57.3% vs. 39.7%; p �<br />
From the Department <strong>of</strong> Medicine (Hematology/<strong>Oncology</strong>); Department <strong>of</strong> Surgical<br />
<strong>Oncology</strong>, University <strong>of</strong> California San Francisco Helen Diller Comprehensive Cancer<br />
Center, San Francisco, CA; Gastrointestinal <strong>Oncology</strong> Research, Sarah Cannon Research<br />
Institute, Nashville, TN.<br />
Authors’ disclosures <strong>of</strong> potential conflicts <strong>of</strong> interest are found at the end <strong>of</strong> this article.<br />
Address reprint requests to Robert S. Warren, MD, Pr<strong>of</strong>essor <strong>of</strong> Surgery, Chief, Surgical<br />
<strong>Oncology</strong>, University <strong>of</strong> California, San Francisco, 1600 Divisadero St., Room A-723, San<br />
Francisco, CA 94115; email: robert.warren@ucsfmedctr.org.<br />
© <strong>2012</strong> by <strong>American</strong> <strong>Society</strong> <strong>of</strong> <strong>Clinical</strong> <strong>Oncology</strong>.<br />
1092-9118/10/1-10<br />
193
0.001), progression-free survival (9.9 vs. 8.4 months; p �<br />
0.0012), and overall survival (OS; 23.5 vs. 20.0 months; p �<br />
0.0093). Patients who had KRAS mutations derived no<br />
benefit from the addition <strong>of</strong> cetuximab treatment. Multiple<br />
other large trials, including OPUS (infusional FU, LV, and<br />
oxaliplatin [FOLFOX] with or without cetuximab), COIN<br />
(FOLFOX/capecitabine plus oxaliplatin with or without cetuximab),<br />
NORDIC VII (infusional FU/folinic acid/oxaliplatin<br />
with or without cetuximab), PRIME (FOLFOX with or<br />
without panitumumab), PACCE (FOLFOX-bevacizumab<br />
with or without panitumumab), 181 (FOLFIRI with or<br />
without panitumumab), NCIC CO.17 (cetuximab vs. best<br />
supportive care), and panitumumab with or without best<br />
supportive care, have all consistently shown that patients<br />
with KRAS mutations do not have any improvement in<br />
outcomes with EGFR-based treatments.<br />
Determination <strong>of</strong> KRAS status is now standard <strong>of</strong> care<br />
and required before initiation <strong>of</strong> anti-EGFR therapy for<br />
patients with colorectal cancer. Although KRAS status does<br />
not define which patients will benefit, it can define patients<br />
KEY POINTS<br />
● Current staging methods for early colorectal cancer<br />
are inadequate to accurately define the prognosis <strong>of</strong><br />
stage 2 or stage 3 disease.<br />
● Similarly, there are very few biomarkers that have<br />
been proven to guide the choice adjuvant chemotherapy<br />
or biological therapy, with some notable<br />
exceptions.<br />
● A large body <strong>of</strong> evidence exists that argues for the use<br />
<strong>of</strong> genotyping the ras and raf oncogenes when the<br />
application <strong>of</strong> anti–epidermal growth factor receptor<br />
(EGFR) antibodies is being considered, or the determination<br />
<strong>of</strong> microsatellite stability when considering<br />
adjuvant treatment with fluorouracil and leucovorin.<br />
● At least three gene expression arrays have been<br />
commercialized for determining the prognosis <strong>of</strong> either<br />
stage 2 or stage 3 colon cancer. Each signature<br />
differs from the others in terms <strong>of</strong> genes that are<br />
assayed, and none has been found to predict response<br />
to any specific colon cancer therapy. And none <strong>of</strong> the<br />
available arrays has been approved for routine use<br />
under the National Comprehensive Cancer Network<br />
(NCCN) guidelines for colorectal cancer.<br />
● The road forward will require quality control over<br />
biospecimens; developing reference labs with state<strong>of</strong>-the-art<br />
analytical performance for DNA, RNA, and<br />
protein assays; using the best trial design and informatics<br />
that represent the consensus <strong>of</strong> leaders in the<br />
field; and working with the U.S. Food and Drug<br />
Administration (FDA) on the complex regulatory issues<br />
attendant to biomarker studies.<br />
● There is a surplus <strong>of</strong> possible prognostic and predictive<br />
biomarkers for colon cancer, but how best to<br />
reliably develop these to the point that they benefit<br />
our patients in individualized cancer care remains a<br />
huge challenge.<br />
194<br />
VENOOK, BENDELL, AND WARREN<br />
who will not benefit and spare them the toxicities, loss <strong>of</strong><br />
time, and loss <strong>of</strong> opportunity from receiving an ineffective<br />
therapy.<br />
KRAS-Specific Mutations: It Is Not As Simple As We Once<br />
Thought It Was<br />
What was first thought to be black or white—KRAS wild<br />
type or mutant—turns out to be shades <strong>of</strong> gray as we are<br />
learning that not all mutations are the same. Most mutations<br />
<strong>of</strong> KRAS are located in codon 12, but approximately<br />
20% <strong>of</strong> patients with KRAS mutations have a mutation in<br />
codon 13, the G13D mutation. 2 In a retrospective analysis,<br />
patients with tumors harboring G13D mutations had<br />
equivalent benefit from treatment with cetuximab therapy<br />
as did KRAS wild type patients. This was confirmed in a<br />
pooled analysis <strong>of</strong> patients in the CRYSTAL and OPUS<br />
studies where the progression-free survival hazard ratios<br />
(HRs) were similar for patients with wild-type (0.66; 95% CI,<br />
0.55 to 0.80) and G13D mutant (0.60; 95% CI, 0.32 to 1.12)<br />
KRAS. 3<br />
Although this observation needs to be validated in a<br />
prospective randomized study, it suggests that the specific<br />
KRAS mutation is an important detail and that tumor<br />
behavior may vary over codons 12 and 13 as well as other<br />
mutations that are lumped together.<br />
BRAF<br />
BRAF sits downstream <strong>of</strong> KRAS in the cell-signaling<br />
pathway, so it had been assumed that mutations in BRAF<br />
would have the same ramifications as KRAS mutations. The<br />
inactivity <strong>of</strong> EGFR-targeted monotherapy—either cetuximab<br />
or panitumumab—in patients who had BRAF mutations<br />
was confirmed in a retrospective report by<br />
DeNicolantonio. 4 However, when data from two first-line<br />
chemotherapy plus cetuximab trials (CRYSTAL and OPUS)<br />
were pooled and analyzed, it seemed that patients with<br />
BRAF mutations had much poorer prognoses than the larger<br />
number <strong>of</strong> patients without BRAF mutations, but that there<br />
was still some benefit to the EGFR antibody (OS BRAF<br />
mutant, 14.1 months with cetuximab vs. 10.3 months without<br />
cetuximab). 1<br />
This prognostic value makes knowledge <strong>of</strong> BRAF mutational<br />
status an important factor in management. Although<br />
it is a rare mutation (10% <strong>of</strong> colorectal cancer), patients with<br />
BRAF mutations have average OS <strong>of</strong> approximately 1 year,<br />
which is approximately half that over the overall metastatic<br />
colorectal cancer population. For these patients, that knowledge<br />
may be important when oncologist and patient consider<br />
issues related to treatment holidays given the aggressiveness<br />
<strong>of</strong> their disease. It also may lead to consideration <strong>of</strong><br />
directing these patients toward specific clinical trials. There<br />
are a number <strong>of</strong> inhibitors <strong>of</strong> the Ras/Raf/MEK pathway<br />
currently in development. A recent phase II study <strong>of</strong> the<br />
BRAF inhibitor PLX4032 (vemurafenib) in 21 patients with<br />
BRAF mutant colorectal cancer showed a response rate <strong>of</strong><br />
5%, with a median progression-free survival <strong>of</strong> 3.7 months. 5<br />
Though not as impressive as the data in patients with BRAF<br />
mutant melanoma, this agent does appear to warrant further<br />
evaluation. Trials are also ongoing looking at MEK<br />
inhibitors both as monotherapy and in combination (MEK is<br />
a downstream target) and combinations <strong>of</strong> BRAF and KRAS<br />
inhibitors. These agents already show promising data in<br />
BRAF mutated melanoma. 6
BIOMARKERS AND COLORECTAL CANCER<br />
Determination <strong>of</strong> BRAF status is important at the outset<br />
for guidance in terms <strong>of</strong> prognosis. As <strong>of</strong> today, do not order<br />
BRAF until the tumor is known to be KRAS wild type; by<br />
convention, KRAS mutant cancers do not have BRAF mutations.<br />
And I reserve the right to change the above recommendation.<br />
The BRAF story teaches us that we should not rush to<br />
judgment on the utility <strong>of</strong> biomarkers until the correct,<br />
adequately powered trials have been performed. To determine<br />
whether a biomarker is predictive, where the biomarker<br />
status can be used to determine whether a patient<br />
will or will not benefit from treatment, a four-arm trial must<br />
be done. The arms <strong>of</strong> the trial will include patients who do<br />
and do not receive active therapy and patients who do and do<br />
not have the presence <strong>of</strong> the biomarker. Trials that include<br />
only the presence or absence <strong>of</strong> a biomarker can tell us only<br />
whether that biomarker is prognostic, when the biomarker<br />
is associated with better or worse outcomes, no matter what<br />
the treatment.<br />
Rash: Acneiform Skin Change to EGFR Inhibitors<br />
Biomarkers not only include tests on tumor and blood, but<br />
also the reaction an individual patient may have to therapy.<br />
Skin rash reaction to EGFR inhibitors is an example. It has<br />
been well established that worsened skin reaction to EGFR<br />
inhibitors is associated with improved outcome to treatment.<br />
Data from the PRIME study, in which patients were<br />
randomly assigned to receive FOLFOX with or without<br />
panitumumab found that wild-type KRAS patients receiving<br />
panitumumab who had a grade 2 to 4 skin reaction compared<br />
with patients who had a grade 0 to 1 skin reaction had<br />
improved progression-free survival (11.1 vs. 6.0 months; p �<br />
0.001) and OS (28.3 vs. 11.5 months; p � 0.0001). 7 Given the<br />
importance <strong>of</strong> skin rash reaction to outcomes, the EVEREST<br />
study attempted to increase the dose <strong>of</strong> cetuximab therapy<br />
on the basis <strong>of</strong> rash. Patients received IFL and cetuximab,<br />
and patients who developed only grade 0 to 1 rash were<br />
randomly assigned to receive high-dose versus low-dose<br />
cetuximab. 8 Though the response rate was improved in the<br />
high dose patients (30% vs. 16%), there was no difference in<br />
OS. At the moment, we can say that the individual patient<br />
biomarker <strong>of</strong> skin reaction to EGFR inhibitors can help us<br />
predict improved outcomes with higher-grade rash.<br />
When you see our patient in follow-up, take note <strong>of</strong> the<br />
extent <strong>of</strong> acne. If it is quite severe, make sure the patient<br />
knows that is a good sign.<br />
Node-Negative Colon Cancer: Mismatch Repair<br />
The use <strong>of</strong> adjuvant chemotherapy for patients with stage<br />
II colon cancer has been the subject <strong>of</strong> much debate, and the<br />
ability to identify patients who would be more or less likely<br />
to benefit for adjuvant chemotherapy is needed. The status<br />
<strong>of</strong> DNA mismatch repair (MMR) seems to be a biomarker to<br />
guide the choice to use adjuvant chemotherapy for these<br />
patients. It is known that patients with tumors that are<br />
deficient in MMR (dMMR) have a better overall prognosis<br />
than those who have pr<strong>of</strong>icient MMR (pMMR). A pooled<br />
analysis <strong>of</strong> 457 patients and then 1,027 patients with stage<br />
II or III colon cancer found that patients with pMMR status<br />
had a benefit from adjuvant chemotherapy with a diseasefree<br />
survival HR <strong>of</strong> 0.67 (95% CI, 0.48 to 0.93), whereas<br />
patients with dMMR status derived no benefit from FU-<br />
based chemotherapy (HR � 1.10; 95% CI, 0.42 to 2.91). 9<br />
Further, for the stage II patients with dMMR there seemed<br />
to be a worse OS with the use <strong>of</strong> FU adjuvant chemotherapy<br />
(HR � 2.95; 95% CI, 1.02 to 8.54).<br />
When you refer the patient to the medical oncologist for<br />
discussion <strong>of</strong> adjuvant therapy for stage II disease, please<br />
also ask the pathologist to determine MMR status. If the<br />
tumor is dMMR, the use <strong>of</strong> adjuvant chemotherapy will be<br />
discouraged.<br />
Recurrence Score<br />
Although there are numerous gene signatures in development,<br />
the assay with the most supporting data is a 12-gene<br />
recurrence score. The recurrence score was derived from<br />
tumor samples from 1,851 stage II or III patients who<br />
received FU adjuvant chemotherapy or surgery alone, identifying<br />
a set <strong>of</strong> 12 genes that appeared to be associated with<br />
risk <strong>of</strong> disease recurrence. 10 This recurrence score was then<br />
analyzed in 1,436 tumor samples from patients with stage II<br />
cancer treated with or without FU chemotherapy on the<br />
QUASAR trial. 11 Patients with a recurrence score greater<br />
than 40 had a 25% rate <strong>of</strong> 3-year disease recurrence compared<br />
with patients with a recurrence score less than 30 who<br />
had a 3-year disease recurrence rate <strong>of</strong> 8%. However, although<br />
the recurrence score seems to identify patients with<br />
an increased risk <strong>of</strong> disease recurrence, when the recurrence<br />
score was applied to look for patients who would then have<br />
benefit from adjuvant chemotherapy, no conclusion could be<br />
drawn. Larger prospective trials are needed.<br />
I expect the recurrence score to be helpful in some patients,<br />
but let me discuss the ramifications <strong>of</strong> the score with<br />
the patient before we send the specimen, since the importance<br />
<strong>of</strong> the score is a function <strong>of</strong> the philosophy <strong>of</strong> the<br />
patient and oncologist. In general, though, we should get as<br />
much information as we can.<br />
Up-and-Coming Biomarkers<br />
EGFR Ligands<br />
Amphiregulin and epiregulin are ligands <strong>of</strong> EGFR. A<br />
study <strong>of</strong> examining DNA, RNA, and protein expression<br />
pr<strong>of</strong>ile in 110 patients who received cetuximab found that<br />
patients with increased expression <strong>of</strong> amphiregulin and<br />
epiregulin had an increased rate <strong>of</strong> disease control. 12 Further,<br />
a retrospective analysis <strong>of</strong> IFL-refractory patients<br />
treated with cetuximab and IFL for metastatic colorectal<br />
cancer found that patients with high epiregulin levels plus<br />
KRAS wild-type status showed a response rate <strong>of</strong> 58%<br />
compared with 40% for KRAS wild type alone, and a median<br />
progression-free survival <strong>of</strong> 36 weeks compared with 24<br />
weeks. 13 Another analysis for patients treated on the NCIC<br />
CO.17 study <strong>of</strong> cetuximab versus best supportive care<br />
showed that patients who had KRAS wild-type and high<br />
epiregulin had a much improved HR for OS benefit compared<br />
with the KRAS wild-type only or total population<br />
(HR � 0.46 vs. 0.55 vs. 0.70, respectively). 14 The combination<br />
<strong>of</strong> KRAS and EGFR ligand status appears promising to<br />
select patients who will have the most benefit from anti-<br />
EGFR therapy.<br />
Phosphoinositide 3-Kinase and Phosphatase and Tensin Homolog<br />
The phosphoinositide 3-kinase (PI3K) pathway is likely<br />
the most commonly activated pathway in cancers. Approxi-<br />
195
mately 40% <strong>of</strong> patients with colorectal cancer have alterations<br />
in the PI3K pathway including mutations and loss <strong>of</strong><br />
phosphatase and tensin homolog (PTEN) activity. 15,16 Multiple<br />
agents that affect this pathway are in development,<br />
including PI3K inhibitors, mammalian target <strong>of</strong> rapamycin<br />
(mTOR) inhibitors, and AKT inhibitors. For patients with<br />
colorectal cancer, studies have been reported looking at<br />
mTOR inhibitors 17 and AKT inhibitors. 18 It remains to be<br />
seen whether there is any association with outcomes with<br />
these agents in patients who have documented abnormalities<br />
in this pathway, but biomarker studies are underway.<br />
c-MET<br />
Overexpression <strong>of</strong> the c-met receptor in colorectal cancers<br />
is associated with a poor prognosis. 19 c-MET activation is<br />
associated with colon cancer tumorigenesis and metastasis,<br />
and in preclinical studies inhibition <strong>of</strong> c-met decreases colon<br />
tumor spread. 20 c-MET is also involved in signal transduction<br />
when growth factor receptors such as EGFR and vascular<br />
EGFR (VEGFR) are activated. 21 The combination <strong>of</strong><br />
c-MET and EGFR inhibition has shown in a randomized<br />
phase II study improved time to progression and OS in<br />
patients with non-small cell lung cancer with c-met overexpression.<br />
22 c-MET inhibitors are currently under investigation<br />
for patients with colorectal cancer. A phase I study <strong>of</strong><br />
IFL, cetuximab, and c-MET inhibitor ARQ-197 showed very<br />
promising results in pretreated disease. 23 First- and secondline<br />
studies are ongoing with inhibitors <strong>of</strong> the c-MET pathway.<br />
We await further data on the correlation in patients<br />
with colorectal cancer between c-MET expression and outcomes<br />
with treatment with c-MET pathway inhibitors.<br />
Ready for Prime Time: Conclusion<br />
As you can see, the era <strong>of</strong> personalized medicine has<br />
arrived for patients with colorectal cancer. We are already<br />
using markers to help guide clinical decisions. The markers<br />
we have will only continue to grow from here. We need to<br />
continue to find new markers, which can be done only if<br />
correlative biomarker studies are included in clinical trials.<br />
We are at only the beginning <strong>of</strong> fully realizing the potential<br />
<strong>of</strong> personalized biomarker status to deliver the best treatment<br />
to our patients.<br />
Robert S. Warren, MD: Personalized <strong>Oncology</strong> for<br />
Colorectal Cancer: Stop the Train<br />
The application <strong>of</strong> molecular biomarkers to prognosis, and<br />
markers that are predictive <strong>of</strong> benefit from specific chemotherapeutic<br />
regimens, is receiving tremendous interest on<br />
the part <strong>of</strong> patients, clinicians, and the pharmaceutical<br />
industry. Patients’ wishes, on the one hand, successful<br />
therapeutics, and on the other hand, avoiding potentially<br />
toxic therapy if the chance <strong>of</strong> benefit is remote; clinicians<br />
search for guidance in the care <strong>of</strong> their patients who have<br />
exhausted most standard therapy options, and the keen<br />
interest <strong>of</strong> the pharmaceutical and biotechnology companies<br />
in developing drugs that are effective, even if only in a<br />
subset <strong>of</strong> patients with colorectal cancer, are all driving the<br />
interest in the application <strong>of</strong> prognostic and predictive<br />
biomarkers. Identifying biomarkers that would permit personalized<br />
therapies and more successfully targeted drug<br />
development may save lives and require many fewer millions<br />
<strong>of</strong> dollars in drug development and in standard clinical<br />
196<br />
VENOOK, BENDELL, AND WARREN<br />
trials. If fact, a review <strong>of</strong> PubMed shows papers describing<br />
more than 60 new biomarkers for cancer just in 2011. Most<br />
are retrospective studies, small and inadequately powered<br />
to perform multivariable analysis in combination with more<br />
thoroughly examined potential markers.<br />
As a measure <strong>of</strong> the increasing interest in this pursuit, it<br />
may be noteworthy that in PubMed, searching under “colon<br />
cancer biomarkers,” 560 papers were published in 2008,<br />
whereas 1,228 were published in 2011. Under “personalized<br />
oncology,” 43 papers were published in 2008, whereas 161<br />
papers were published in 2011.<br />
The questions that Dr. Bendell so nicely discussed focus<br />
on a few <strong>of</strong> these: KRAS/BRAF mutations and the activity <strong>of</strong><br />
EGFR antibodies (MoAby), the utility <strong>of</strong> gene expression<br />
arrays in determining prognosis (generally in advanced<br />
colorectal cancer), and the impact <strong>of</strong> microsatellite instability<br />
on prognosis and response to chemotherapy. Despite a<br />
plethora <strong>of</strong> research publications, the U.S. Food and Drug<br />
Administration (FDA), the National Comprehensive Cancer<br />
Network (NCCN), and ASCO 24,25,26 have recommended the<br />
standard use <strong>of</strong> very few markers. The important question<br />
here is why? Are there guidelines for the development,<br />
verification, and validation <strong>of</strong> markers to hasten their application<br />
to the clinic? Following is a summary <strong>of</strong> definitions<br />
and a process to permit moving forward, part <strong>of</strong> a paper that<br />
gives a good perspective on the challenges we face in this<br />
pursuit 26 : biomarker—a characteristic that is objectively<br />
measured and evaluated as an indicator <strong>of</strong> normal biologic<br />
processes, pathogenic processes, or pharmacologic responses<br />
to a therapeutic intervention; diagnostic biomarkers—early<br />
detection biomarkers and disease classification; predictive<br />
biomarkers—predict patients likely to have an adverse<br />
event to a specific agent and predict patients likely to<br />
respond to a specific agent; outcome biomarkers—forecast<br />
response, progression, and recurrence; assay validation—<br />
the process <strong>of</strong> assessing the assay and its performance<br />
characteristics and determining the optimal conditions that<br />
will generate a reliable, reproducible, and accurate biomarker<br />
assay for the intended application; clinical qualification—linking<br />
a biomarker (using data obtained by a<br />
biomarker assay) with meaningful biologic or clinical outcomes;<br />
high-quality biospecimens (collection methods, quality<br />
assessment), accurate detection <strong>of</strong> markers (reference<br />
standards, analytic performance and methods), useful patient<br />
annotation, design <strong>of</strong> new bioinformatics approaches to<br />
the study <strong>of</strong> biomarkers, and the application <strong>of</strong> optimized<br />
retrospective and prospective study designs (adaptive clinical<br />
trials).<br />
The goal <strong>of</strong> this discovery process is to “elucidate the<br />
physiologic, toxicological, and pharmacologic, or clinical significance<br />
<strong>of</strong> the test results.” 26 The tools that we need to<br />
query our patients’ genomes and their tumors have developed<br />
rapidly. Analysis <strong>of</strong> DNA copy number by comparative<br />
genomic hybridization (CGH), mutation detection, epigenetic<br />
pr<strong>of</strong>iling, gene expression pr<strong>of</strong>iling, determining splice<br />
variants <strong>of</strong> significant genes and functional proteomics and<br />
metabolomics are becoming routine on frozen or paraffinfixed<br />
tumors. High-density single nucleotide polymorphism<br />
(SNP) arrays (containing 500,000 to 1,000,000 individual<br />
SNPs) permit an in-depth analysis to the host genome.<br />
Clearly, these methods are not rate limiting in terms <strong>of</strong><br />
progress toward individualized cancer care. Even the most<br />
robust <strong>of</strong> assays associated with therapeutic response or
BIOMARKERS AND COLORECTAL CANCER<br />
Fig. 1. Signalling pathways that are activated upon<br />
ligation <strong>of</strong> epidermal growth factor with its receptor<br />
(EGFR).<br />
Abbreviations: CRC, colorectal cancer; EGFR, epidermal<br />
growth factor receptor; PI3K, phosphoinositide 3-kinase;<br />
PTEN, phosphatase and tensin homolog.<br />
prognosis have various weak links. A few examples <strong>of</strong> this<br />
are discussed in the following sections.<br />
EGFR Antibodies and Mutations in KRAS and BRAF<br />
Dr. Bendell accurately points out that the majority <strong>of</strong><br />
retrospective clinical trials support the notion that patients<br />
whose tumors harbor mutations in codon 12 <strong>of</strong> KRAS show<br />
very little or no response to EGFR MoAby, whereas tumors<br />
with mutations in codon 13 seem to respond as well as Ras<br />
wild-type tumors. 27 But Dr. Bendell has omitted the observation<br />
from the COIN, NORDIC, PACCE, and CAIRO-2<br />
suggest that there is a detrimental effect <strong>of</strong> adding cetuximab<br />
to chemotherapy in patients with tumors containing<br />
mutant KRAS compared with chemotherapy alone. 28-31 The<br />
Fig. 2. Prevalence <strong>of</strong> epidermal growth factor (EGFR)<br />
pathway deregulations and response to monoclonal antibodies<br />
targeting EGFR in chemotherapy-refractory advanced<br />
colorectal cancer.<br />
Abbreviations: PI3K, phosphoinositide 3-kinase; PTEN,<br />
phosphatase and tensin homolog.<br />
assumption that approximately 40% <strong>of</strong> colorectal cancers harbor<br />
ras mutation is misleading. The focus most likely should<br />
be on the estimated 75% to 80% KRAS codon 12 mutations.<br />
Further, the story <strong>of</strong> Braf is more complicated. Dr. Bendell’s<br />
point is well taken that tumors with BRAF mutations<br />
(10% <strong>of</strong> colon cancer) have a poorer prognosis than wild-type<br />
tumors, but they may respond equally well to EGFR inhibition<br />
as do wild-type tumors. But, there are other abnormalities<br />
that can mimic (phenocopy) KRAS mutation vis à vis<br />
EGFR inhibition that merit more emphasis than discussed<br />
by Dr. Bendell. PTEN blocks signaling downstream <strong>of</strong><br />
growth factor receptors; PTEN especially manifests this<br />
effect by impairing signaling emanating from PI3K step 27<br />
(Fig. 1). Although the assay methodology for PTEN has not<br />
197
Table 1. Tumor Marker Utility Grading System Levels <strong>of</strong><br />
Evidence<br />
Level Definition<br />
I Prospective, marker primary objective<br />
Well-powered or meta-analysis<br />
II Prospective, marker the secondary objective<br />
III Retrospective, outcomes, multivariate analysis (most currently published<br />
marker studies are level <strong>of</strong> evidence III)<br />
IV Retrospective, outcomes, univariate analysis<br />
V Retrospective, correlation with other marker, no outcomes<br />
Adapted with permission from Hayes DF, Bast RC, Desch CE, et al. Tumor<br />
marker utility grading system: a framework to evaluate clinical utility <strong>of</strong> tumor<br />
markers. J Natl Cancer Inst. 1996;88:1464.<br />
been well established, the 20% to 40% <strong>of</strong> colon cancers<br />
lacking PTEN expression as a result <strong>of</strong> promoter methylation<br />
would be expected not to respond to EGFR inhibition.<br />
The same applies to tumors with activated PI3K, which<br />
again is downstream <strong>of</strong> ligand signaling through the EGFR<br />
and could explain why so few patients with wild-type KRAS<br />
and BRAF fail to benefit from EGFR inhibition. Figure 2<br />
(from Dienstmann, Vilar, and Tabernero 27 ) describes the<br />
possible abnormalities that can account for resistance to<br />
EGFR antibodies.<br />
Gene Expression Arrays<br />
There are numerous papers describing the correlation<br />
between gene expression and prognosis in colorectal cancer.<br />
Although Oncotype Dx (Genomic Health, Inc., Redwood<br />
City, CA) and Coloprint (Agendia, Irvine, CA) have been<br />
commercialized and are available for purchase, a number <strong>of</strong><br />
other investigators have reported, tested, and subsequently<br />
validated mRNA arrays as prognostic tools in colon cancer.<br />
In 2004, Wang and colleagues 32 reported a 23-gene signature<br />
based on the Affymetrix (Santa Clara, CA) Gene Chip<br />
that accurately predicted relapse in stage II colon cancers.<br />
Eschrich and colleagues as well as Arango 33,34 reported in<br />
2005 that a 43-gene signature predicted recurrence in stage<br />
II and III colon cancer. Barrier and colleagues in 2006<br />
identified a 30-gene signature using mRNA microarrays<br />
that predicted outcome in stage II colon cancer 35 In 2007,<br />
Gray and colleagues 36 reported the first experience using a<br />
set <strong>of</strong> 761 genes assayed by reverse transcriptase polymerase<br />
chain reaction (RT-PCR) to develop a signature for<br />
colon cancer recurrence, and this evolved in Oncotype Dx<br />
into a validation study in 2011. 37,38 The diagnostics company<br />
Almac (Souderton, PA) used a 634-probe set signature<br />
(Col Dx) 39 to identify patients with colon cancer with a high<br />
risk <strong>of</strong> recurrence in stage II colon cancers that is undergoing<br />
multiple independent validation studies. And in 2011,<br />
Salazar and colleagues demonstrated and validated the<br />
utility <strong>of</strong> an 18-gene signature, originally described in 2007,<br />
and is now commercially available 40,41 and sold as Coloprint<br />
for colon cancer prognosis. Febbo and colleagues identified<br />
levels <strong>of</strong> evidence for marker studies and NCCN categories<br />
<strong>of</strong> evidence. 24 Level 1 is used to characterize markers that<br />
were evaluated prospectively where the marker was the<br />
primary objective <strong>of</strong> the study. Most <strong>of</strong> the markers in the<br />
literature fall under Level IV, in which evaluation was<br />
retrospective and outcomes were determined in univariate<br />
analysis. Most <strong>of</strong> the array-based and RT-PCR–based<br />
genomic signatures fall under this category, including Colo-<br />
Print and Oncotype Dx. Interestingly, none <strong>of</strong> these arraybased<br />
signatures was informative in terms <strong>of</strong> response to<br />
chemotherapy. Also, there is very little overlap in the set <strong>of</strong><br />
genes chosen to create a prognostic signature in colon<br />
cancer, even though it seems that risk assessment is comparable<br />
regardless <strong>of</strong> the platform used. Choosing the best<br />
platform involves consideration <strong>of</strong> cost, rapidity <strong>of</strong> turnover,<br />
reproducibility, and whether any <strong>of</strong> these gene expression<br />
signatures have been assessed in multivariate analysis with<br />
other potential markers <strong>of</strong> outcome (p53 mutational status,<br />
p27 loss, thymidylate synthase expression, microsatellite<br />
instability, and others). Consequently, although available to<br />
the clinician and the patient, the exact utility <strong>of</strong> gene<br />
expression signatures remains a subject <strong>of</strong> debate. Neither<br />
Oncotype Dx nor ColoPrint is included in the NCCN guidelines<br />
for colon cancer.<br />
p53: When Genotype Interacts with Gender and<br />
Chemotherapeutic Regimen<br />
Table 2. Pair-Wise Comparisons <strong>of</strong> Survival by p53 Mutational Status<br />
The p53 tumor suppressor is frequently mutated in colon<br />
cancer, but the influence <strong>of</strong> such mutations on survival is<br />
still controversial 42 p53 mutations have been inferred by<br />
positive staining using IHC since the wild-type protein is<br />
degraded very rapidly, and is not demonstrable in the<br />
nucleus by immunohistochemistry, although the gold standard<br />
for p53 mutation analysis is Sanger sequencing. We<br />
investigated whether DNA-binding domain-specific mutations<br />
in p53 are predictive <strong>of</strong> survival in stage III colon<br />
cancer. p53 was evaluated in an intergroup trial, CALGB<br />
89803, 43 in patients with stage III colon cancer who were<br />
randomly assigned to receive adjuvant FU/LV or FU/LV<br />
with IFL, and various molecular markers were correlated<br />
with outcomes. p53 was genotyped in 607 patient tumors:<br />
p53 mutations were identified in 274 tumors, divided<br />
equally between zinc-binding and non–zinc-binding regions<br />
<strong>of</strong> the DNA-binding domain. Overall, p53 status was not<br />
predictive <strong>of</strong> benefit from either adjuvant regimen. Unexpectedly,<br />
however, the 5-year OS <strong>of</strong> women with tumors<br />
harboring non–zinc-binding mutations treated with FU/LV<br />
was 97% compared with OS <strong>of</strong> 72% for women with p53<br />
wild-type tumors (p � 0.004). Adding IFL to FU/LV negated<br />
this survival benefit (5-year OS <strong>of</strong> 81% vs. 72%). Conversely,<br />
5-year OS <strong>of</strong> women harboring tumors with zinc-binding<br />
Wild-Type vs. Zinc Binding Wild-Type vs. Non–Zinc Binding Zinc Binding vs. Non–Zinc Binding<br />
Patient Subset P OS DFS P OS DFS P OS DFS<br />
Men 0.58 0.28 0.19 0.18 0.48 0.7<br />
Women, FU/LV 0.04 0.24 0.004 0.002 �0.001 �0.001<br />
Women, IFL 0.71 0.79 0.18 0.53 0.49 0.83<br />
Men, FU/LV 0.48 0.66 0.80 0.48 0.35 0.29<br />
Men, IFL 0.12 0.05 0.10 0.24 0.85 0.63<br />
Abbreviations: DFS, disease-free survival; FU, fluorouracil; IFL, irinotecan; LV, leucovorin; OS, overall survival.<br />
198<br />
VENOOK, BENDELL, AND WARREN
BIOMARKERS AND COLORECTAL CANCER<br />
mutations who received FU/LV reversed the poor survival <strong>of</strong><br />
women with tumors harboring zinc-binding mutations and<br />
improved 5-year OS (50% vs. 73%; p � 0.1). No difference in<br />
OS was observed for men in either treatment arm or when<br />
genotype was considered. We conclude that CALGB 89803<br />
demonstrated a lack <strong>of</strong> survival benefit for patients with<br />
stage III colon cancer when IFL was added to FU/LV. We<br />
now show that in the setting <strong>of</strong> a large clinical trial, refined<br />
stratification <strong>of</strong> women, based on domain-specific mutations<br />
<strong>of</strong> p53 identifies subsets <strong>of</strong> patients likely to benefit from, or<br />
respond poorly to, adjuvant FU/LV. The interaction <strong>of</strong> p53<br />
genotype, gender, and adjuvant therapy regimen has the<br />
potential to be paradigm changing in the treatment <strong>of</strong> colon<br />
cancer, and possibly other malignancies (Table 2).<br />
These data, if validated, suggest that evaluation <strong>of</strong> p53<br />
genotype and gender may guide clinicians to make rational<br />
choices <strong>of</strong> adjuvant therapy. Clearly, prospective clinical<br />
trials must be sufficiently large and adequately powered to<br />
ferret out previously undescribed interactions between oncogenes,<br />
tumor suppressor genes, and gender.<br />
In summary, although many studies purporting to demonstrate<br />
either prognostic or predictive value in a variety <strong>of</strong><br />
markers (DNA based, protein based, and metabolite based)<br />
have been published, only a few have acquired sufficient<br />
evidence to warrant possible inclusion in clinical decision<br />
making in colon cancer. 46-48 Further, a variety <strong>of</strong> regulatory<br />
hurdles must be addressed during biomarker development<br />
in the future. 47<br />
The Road Ahead: Companion Diagnostics in<br />
<strong>Clinical</strong> Trials<br />
As use and interest in laboratory-developed tests continues<br />
to grow, it will be important to work with the FDA to<br />
define the oversight framework that takes into account<br />
higher-risk tests that require more complex validation,<br />
equipment, and s<strong>of</strong>tware. Regulatory clarity and predictability<br />
will be important to ensure that high-quality tests<br />
reach the market expeditiously.<br />
Authors’ Disclosures <strong>of</strong> Potential Conflicts <strong>of</strong> Interest<br />
Author<br />
Employment or<br />
Leadership<br />
Positions<br />
Consultant or<br />
Advisory Role<br />
Alan P. Venook Abbott<br />
Laboratories (U);<br />
Bristol-Myers<br />
Squibb (U);<br />
Chugai Pharma<br />
(U)<br />
Johanna C. Bendell*<br />
Robert S. Warren*<br />
*No relevant relationships to disclose.<br />
1. Van Cutsem E, Köhne CH, Láng I, et al. Cetuximab plus irinotecan,<br />
fluorouracil, and leucovorin as first-line treatment for metastatic colorectal<br />
cancer: updated analysis <strong>of</strong> overall survival according to tumor KRAS and<br />
BRAF mutation status. J Clin Oncol. 2011;29:2011-2019.<br />
2. De Roock W, Jonker DJ, Di Nicolantonio F, et al. Association <strong>of</strong> KRAS<br />
p.G13D mutation with outcome in patients with chemotherapy-refractory<br />
metastatic colorectal cancer treated with cetuximab. JAMA. 2010;304:1812-<br />
1820.<br />
3. Tejpar S, Bokemeyer C, Celik I, et al. The role <strong>of</strong> the KRAS G13D<br />
mutation in patients with metastatic colorectal cancer (mCRC) treated with<br />
first-line chemotherapy plus cetuximab. J Clin Oncol. 2011;29 (suppl; abstr<br />
630).<br />
Laboratory tests have been critical to recent advances in<br />
oncology because they can be developed rapidly and targeted<br />
locally. Currently, no controls have been placed on the<br />
marketing claims <strong>of</strong> theses tests, and limited control <strong>of</strong> test<br />
development exists. Likewise, there are limited premarket<br />
independent review or postmarket reporting requirements.<br />
These requirements are dependent on state regulations<br />
and/or the laboratory accreditation process. Currently, some<br />
<strong>of</strong> these functions for laboratories performing molecular<br />
testing, such as postmarket performance, are conducted<br />
voluntarily.<br />
Some sectors are concerned that additional regulation <strong>of</strong><br />
molecular testing may harm innovation because associated<br />
regulatory barriers may hinder the pace <strong>of</strong> progress.<br />
The product lifecycles <strong>of</strong> molecular testing assays <strong>of</strong>ten<br />
differ from those <strong>of</strong> drugs. This could present issues related<br />
to innovation with new therapies and their companion<br />
diagnostics.<br />
Most molecular testing is performed in laboratories that<br />
meet or exceed the laboratory quality standards currently<br />
set by the Centers for Medicare & Medicaid Services (CMS,<br />
Rockville, MD) under the <strong>Clinical</strong> Laboratory Improvement<br />
Amendments (CLIA). However, CLIA regulations do not<br />
measure the clinical validity or clinical utility <strong>of</strong> individual<br />
molecular tests. The FDA does have the authority to ensure<br />
the safety and effectiveness <strong>of</strong> molecular tests, which includes<br />
an evaluation <strong>of</strong> clinical validity.<br />
New therapies that are found to be effective in patients<br />
with a specific biomarker pr<strong>of</strong>ile may require a companion<br />
diagnostic test to be approved by the FDA if these tests are<br />
considered essential for the safe and effective use <strong>of</strong> a<br />
therapy. Label changes to already approved treatments can<br />
also be required if a companion diagnostic is shown to<br />
improve safety. It will be important to understand how<br />
clinical trials for drugs and their companion diagnostics<br />
should be designed and carried out, validated, and brought<br />
to the clinic. 49,50<br />
Stock<br />
Ownership Honoraria<br />
REFERENCES<br />
Research<br />
Funding<br />
Amgen; Bayer;<br />
Genentech;<br />
ImClone<br />
Systems;<br />
Novartis; Pfizer<br />
Expert<br />
Testimony<br />
Other<br />
Remuneration<br />
4. Di Nicolantonio F, Martini M, Molinari F, et al. Wild-type BRAF is<br />
required for response to panitumumab or cetuximab in metastatic colorectal<br />
cancer. J Clin Oncol. 2008;26:5705-5712.<br />
5. Kopetz, S, Desai J, Chan E, et al. PLX4032 in metastatic colorectal<br />
cancer patients with mutant BRAF tumors. J Clin Oncol. 2010;28 (suppl;<br />
abstr 3534).<br />
6. Infante JR, Falchook GS, Lawrence DP, et al. Phase I/II study to assess<br />
safety, pharmacokinetics, and efficacy <strong>of</strong> the oral MEK 1/2 inhibitor<br />
GSK1120212 (GSK212) dosed in combination with the oral BRAF inhibitor<br />
GSK2118436 (GSK436). J Clin Oncol. 2011;29 (suppl; abstr 630).<br />
7. Douillard JY, Cassidy J, Jassem F, et al. Randomized, open-label, phase<br />
3 study <strong>of</strong> panitumumab With FOLFOX4 vs FOLFOX4 alone as 1st-line<br />
199
treatment (tx) for metastatic colorectal cancer (mCRC): efficacy by skin<br />
toxicity. J Clin Oncol. 2010;28 (suppl; abstr 3528).<br />
8. Tejpar, S, et al. Phase I/II study <strong>of</strong> cetuximab dose-escalation in patients<br />
with metastatic colorectal cancer (mCRC) with no or slight skin reactions on<br />
cetuximab standard dose treatment (EVEREST): Pharmacokinetic (PK),<br />
Pharmacodynamic (PD) and efficacy data. J Clin Oncol. 2007;25 (suppl; abstr<br />
4037).<br />
9. Sargent DJ, Marsoni S, Monges G, et al. Defective mismatch repair as a<br />
predictive marker for lack <strong>of</strong> efficacy <strong>of</strong> fluorouracil-based adjuvant therapy in<br />
colon cancer. J Clin Oncol. 2010;28:3219-3226.<br />
10. O’Connell MJ, Lavery I, Yothers G, et al. Relationship between tumor<br />
gene expression and recurrence in four independent studies <strong>of</strong> patients with<br />
stage II/III colon cancer treated with surgery alone or surgery plus adjuvant<br />
fluorouracil plus leucovorin. J Clin Oncol. 2010;28:3937-3944.<br />
11. Gray RG, Quirke P, Handley K, et al. Validation study <strong>of</strong> a quantitative<br />
multigene reverse transcriptase-polymerase chain reaction assay for assessment<br />
<strong>of</strong> recurrence risk in patients with stage II colon cancer. J Clin Oncol.<br />
2011;29:4611-4619.<br />
12. Khambata-Ford S, Garrett CR, Meropol NJ, et al. Expression <strong>of</strong><br />
epiregulin and amphiregulin and K-ras mutation status predict disease<br />
control in metastatic colorectal cancer patients treated with cetuximab. J Clin<br />
Oncol. 2007;25:3230-3237.<br />
13. Tejpar S, De Roock W, Biesmans B, et al. High amphiregulin and<br />
epiregulin expression in KRAS wild-type colorectal primaries predicts response<br />
and survival benefit after treatment with cetuximab and irinotecan for<br />
metastatic disease. Presented at: 2008 Gastrointestinal Cancers Symposium;<br />
January 2010; Orlando, FL.<br />
14. Jonker DH, Karapetis C, Harbison C, et al. High epiregulin (EREG)<br />
gene expression plus K-ras wild-type (WT) status aspredictors <strong>of</strong> cetuximab<br />
benefit in the treatment <strong>of</strong> advanced colorectal cancer (ACRC): results from<br />
NCIC CTG CO. 17—a phase III trial <strong>of</strong> cetuximab versus best supportive care<br />
(BSC). J Clin Oncol. 2009;27 (suppl; abstr 4016).<br />
15. Brugge J, Hung MC, Mills GB. A new mutational AKTivation in the<br />
PI3K pathway. Cancer Cell. 2007;12:104-107.<br />
16. Baldus SE, Schaefer KL, Engers R, et al. Prevalence and heterogeneity<br />
<strong>of</strong> KRAS, BRAF, and PIK3CA mutations in primary colorectal adenocarcinomas<br />
and their corresponding metastases. Clin Cancer Res. 2010;16:790-799.<br />
17. Altomare I, Russell KB, Uronis HE, et al. Phase II trial <strong>of</strong> bevacizumab<br />
(B) plus everolimus (E) for refractory metastatic colorectal cancer (mCRC).<br />
J Clin Oncol. 2010;28 (suppl; abstr 3535).<br />
18. Bendell JC, Nemunaitis J, Vukelja SJ, et al. Randomized placebocontrolled<br />
phase II trial <strong>of</strong> perifosine plus capecitabine as second- or third-line<br />
therapy in patients with metastatic colorectal cancer. J Clin Oncol. 2011;29:<br />
4394-4400.<br />
19. Zeng ZS, Weiser MR, Kuntz E, et al. c-Met gene amplification is<br />
associated with advanced stage colorectal cancer and liver metastases. Cancer<br />
Lett. 2008;265:258-269.<br />
20. Takeuchi H, Kim J, Fujimoto A, et al. HGF activation in colorectal<br />
cancer via c-met receptor regulates IAP proteins. Presented at: 98th Annual<br />
Meeting <strong>of</strong> the <strong>American</strong> Association for Cancer Research; April 2007; Los<br />
Angeles, CA.<br />
21. Ding S, Merkulova-Rainon T, Han ZC, et al. HGF receptor upregulation<br />
contributes to the angiogenic phenotype <strong>of</strong> human endothelial cells<br />
and promotes angiogenesis in vitro. Blood. 2003;101:4816-4822.<br />
22. Spigel DR, Ervin TJ, Ramlau R, et al. Final efficacy results from<br />
OAM4558g, a randomized phase II study evaluating MetMAb or placebo in<br />
combination with erlotinib in advanced NSCLC. J Clin Oncol. 2011;29 (suppl;<br />
abstr 7505).<br />
23. Eng C, Eng C, Bendell JC, et al. Phase I results <strong>of</strong> the randomized,<br />
placebo controlled, phase I/II study <strong>of</strong> the novel oral c-Met inhibitor, ARQ 197,<br />
irinotecan (CPT-11), and cetuximab (C) in patients (pts) with wild-type (WT)<br />
KRAS metastatic colorectal cancer (mCRC) who have received front-line<br />
systemic therapy. Presented at: Gastrointestinal Cancers Symposium; January<br />
2011; San Francisco, CA.<br />
24. Febbo PG, Ladanyi M, Aldape KD, et al. NCCN Task Force report:<br />
evaluating the clinical utility <strong>of</strong> tumor markers in oncology. J Natl Compr<br />
Canc Netw. 2011;5:S1-S32.<br />
25. Tejpar S, Bertagnolli M, Bosman F, et al. Prognostic and predictive<br />
biomarkers in resected colon cancer: Current status and future perspectives<br />
for integrating genomics into biomarker discovery. Oncologist. 2010;15:390-<br />
404.<br />
26. Khleif, SN, Doroshow JH, Hait WN, et al. AACR-FDA-NCI Cancer<br />
Biomarkers Collaborative consensus report: advancing the use <strong>of</strong> biomarkers<br />
in cancer drug development. <strong>Clinical</strong> Cancer Res. 2010;16:3299-3318.<br />
200<br />
VENOOK, BENDELL, AND WARREN<br />
27. Dienstmann R, Vilar E, Tabernero J. Molecular predictors <strong>of</strong> response<br />
to chemotherapy in colorectal cancer. Cancer J. 2011;17:114-126.<br />
28. Maughan TS, Adams R, Smith CG, et al. Identification <strong>of</strong> potentially<br />
responsive subsets when cetuximab is added to oxaliplatin-fluoropyrimidine<br />
chemotherapy (CT) in first-line advanced colorectal cancer (aCRC): mature<br />
results <strong>of</strong> the MRC COIN trial. J Clin Oncol. 2010;28 (suppl; abstr 3502).<br />
29. Tveit K, Guren B, Glimelius P, et al. Randomized phase III study <strong>of</strong><br />
5-fluorouracil/folinate/oxaliplatin given continuously or intermittently with<br />
or without cetuximab, as first-line treatment <strong>of</strong> metastatic colorectal cancer:<br />
the Nordic VII study (NCT00145314), by the Nordic Colorectal Cancer<br />
Biomodulation Group. Ann. Oncol. 2010;21:viii9 (suppl; abstr LBA.20).<br />
30. Hecht JR, Mitchell E, Chidiac T, et al. A randomized phase IIIB trial <strong>of</strong><br />
chemotherapy, bevacizumab, and panitumumab compared with chemotherapy<br />
and bevacizumab alone for metastatic colorectal cancer. J Clin Oncol.<br />
2009;27:672-680.<br />
31. Tol J, Koopman. M, Cats M, et al. Chemotherapy, bevacizumab, and<br />
cetuximab in metastatic colorectal cancer. N Engl J Med. 2009;360:563-572.<br />
32. Wang Y, Jatkoe T, Zhang Y, et al. Gene expression pr<strong>of</strong>iles and<br />
molecular markers to predict recurrence <strong>of</strong> Dukes’ B colon cancer. J Clin<br />
Oncol. 2004;22:1564-1571.<br />
33. Eschrich S, Yang I, Bloom G, et al. Molecular staging for survival<br />
prediction <strong>of</strong> colorectal cancer patients. J Clin Oncol. 2005;23:3526-3535.<br />
34. Arango D, Laiho P, Kokko A, et al. Gene-expression pr<strong>of</strong>iling predicts<br />
recurrence in Dukes’ C colorectal cancer. Gastroenterology, 2005;129:874-<br />
884.<br />
35. Barrier A, Boelle PV, Roser R, et al. Stage II colon cancer prognosis<br />
prediction by tumor gene expression pr<strong>of</strong>iling. J Clin Oncol. 2006;24:4685-<br />
4691.<br />
36. Gray RG, Quirke P, Handley K, et al. Validation study <strong>of</strong> a quantitative<br />
multigene reverse transcriptase-polymerase chain reaction assay for assessment<br />
<strong>of</strong> recurrence risk in patients with stage II colon cancer. J Clin Oncol.<br />
2011;29:4611-4619.<br />
37. Clark-Langone KM, Sangli C, Krishnakumar J, et al. Translating<br />
tumor biology into personalized treatment planning: Analytical performance<br />
characteristics <strong>of</strong> the Oncotype DX Colon Cancer Assay. BMC Genomics.<br />
2007;8:279.<br />
38. O’Connell MJ, Lavery I, Yothers G, et al. Relationship between tumor<br />
gene expression and recurrence in four independent studies <strong>of</strong> patients with<br />
stage II/III colon cancer treated with surgery alone or surgery plus adjuvant<br />
fluorouracil plus leucovorin. J Clin Oncol. 2010;28:3937-3944.<br />
39. Kennedy RD, Bylesjo M, Kerr P, et al. Development and independent<br />
validation <strong>of</strong> a prognostic assay for stage II colon cancer using formalin-fixed<br />
paraffin-embedded tissue. J Clin Oncol. 2011;29:4620-4626.<br />
40. Salazar R, Roepman P, Capella G, et al. Gene expression signature to<br />
improve prognosis prediction <strong>of</strong> stage II and III colorectal cancer. J Clin<br />
Oncol. 2011;29:17-24.<br />
41. Tan IB, Tan P. Genetics: an 18-gene signature (ColoPrint) for colon<br />
cancer prognosis. Nat Rev Clin Oncol. 2011;8:131-133.<br />
42. Robles A, Harris CC. <strong>Clinical</strong> outcomes and correlates <strong>of</strong> TP53 mutations<br />
and cancer. Cold Spring Harb Perspect Biol. 2010;2:a001016.<br />
43. Saltz LB, Niedzwiecki D, Hollis D, et al. Irinotecan fluorouracil plus<br />
leucovorin is not superior to fluorouracil plus leucovorin alone as adjuvant<br />
treatment for stage III colon cancer: results <strong>of</strong> CALGB 89803. J Clin Oncol.<br />
2007;25:3456-3461.<br />
44. Van Schaeybroeck S, Allen W, Turkington, RC, et al. Implementing<br />
prognostic and predictive biomarkers in CRC clinical trials. Nat Rev Clin<br />
Oncol. 2011;8:222-232.<br />
45. Tabernero J, Baselga J. Multigene assays to improve assessment <strong>of</strong><br />
recurrence risk and benefit from chemotherapy in early-stage colon cancer:<br />
has the time finally arrived, or are we still stage locked? J Clin Oncol.<br />
2010;28:3904-3907.<br />
46. Arteaga CL, Baselga J. Impact <strong>of</strong> genomics on personalized cancer<br />
medicine. Clin Cancer Res. <strong>2012</strong>;18:612-618.<br />
47. Engstrom PF, Bloom MG, Demetri GD, et al. NCCN molecular testing<br />
white paper: effectiveness, efficiency, and reimbursement. J Natl Compr Canc<br />
Netw. 2011;6:S1-S16.<br />
48. Wistuba II, Gelovani JG, Jacoby JJ, et al. Methodological and practical<br />
challenges for personalized cancer therapies. Nat Rev Clin Oncol 2011;8:135-<br />
141.<br />
49. Wu W, Shi Q, Sargent DJ. Statistical considerations for the next<br />
generation <strong>of</strong> clinical trials. Semin Oncol 2011;38:598-604.<br />
50. McClellan M, Benner J, Schilsky R, et al. An accelerated pathway for<br />
targeted cancer therapies. Nat Rev Drug Discov. 2011;10:79-80.
CURATIVE-INTENT TREATMENT OF COLORECTAL<br />
CANCER METASTASES<br />
CHAIR<br />
Leonard Saltz, MD<br />
Memorial Sloan-Kettering Cancer Center<br />
New York, NY<br />
SPEAKERS<br />
Stephen B. Solomon, MD<br />
Memorial Sloan-Kettering Cancer Center<br />
New York, NY<br />
Steven A. Curley, MD<br />
University <strong>of</strong> Texas M. D. Anderson Cancer Center<br />
Houston, TX
The Interventional Radiologist Role<br />
in Treating Liver Metastases for<br />
Colorectal Cancer<br />
By Stephen B. Solomon, MD, and Constantinos T. S<strong>of</strong>ocleous, MD, PhD<br />
Overview: Interventional radiologists (IRs) have an expanding<br />
role in the treatment <strong>of</strong> liver metastases from colorectal<br />
cancer. Increasing data on the ability to treat liver metastases<br />
with locoregional therapies has solidified this position. Ablative<br />
approaches, such as radi<strong>of</strong>requency ablation and microwave<br />
ablation, have shown durable eradication <strong>of</strong> tumors.<br />
Catheter-directed therapies—such as transarterial chemoembolization<br />
(TACE), drug-eluting beads (DEB), Y90 radioembo-<br />
INTERVENTIONAL RADIOLOGISTS have an expanding<br />
role in the treatment <strong>of</strong> liver metastases from colorectal<br />
cancer. Understanding the various locoregional treatment<br />
options and their integration into the care <strong>of</strong> the metastatic<br />
patient is important for good oncologic care. This review will<br />
describe some <strong>of</strong> these treatment options and their existing<br />
supportive data.<br />
Ablative Therapies<br />
Patients undergoing liver resection <strong>of</strong> their colorectal metastases<br />
have prolonged survival. 1 However, only approximately<br />
20% <strong>of</strong> patients with liver metastases are surgical<br />
candidates. 2 Ablation technologies cause focal destruction <strong>of</strong><br />
tissue, and when coupled with imaging guidance, this targeted<br />
destruction can be aimed at a particular metastasis.<br />
Similar to surgical resection, the goal <strong>of</strong> ablation is to create<br />
a margin <strong>of</strong> destruction around the targeted tumor to prevent<br />
recurrence. There are a number <strong>of</strong> ablative tools available—including<br />
radi<strong>of</strong>requency ablation (RFA), microwave<br />
ablation, cryoablation, laser ablation, and focused ultrasound<br />
ablation—that rely on extreme temperature conditions<br />
<strong>of</strong> heat or cold to exact the tissue damage. 3 A new,<br />
nonthermal technique called irreversible electroporation<br />
uses electric fields to cause cell death without apparently<br />
harming tissue protein architecture that makes up structures<br />
such as bile ducts and vessels. This technique may<br />
open up new ablative opportunities near critical structures<br />
that were previously risky using thermal ablation tools,<br />
which could potentially damage these critical structures.<br />
Also, this nonthermal technique may be more effective than<br />
thermal ablation techniques near blood vessels where thermal<br />
techniques suffer because <strong>of</strong> a “heat sink effect” that<br />
limits how hot the tissue adjacent to a vessel can get. 4 More<br />
research is needed to better understand this modality. The<br />
technical differences among all <strong>of</strong> these techniques is beyond<br />
the scope <strong>of</strong> this review, suffice it to say, that RFA has been<br />
the most commonly used technique with the most extensive<br />
literature for treating liver metastases.<br />
Ablation techniques are less invasive and consequently<br />
less morbid than surgical resection. Patients can generally<br />
be treated as outpatients with a rapid recovery to normal<br />
activities. Percutaneous ablation procedures can be repeated<br />
if necessary and can be used to salvage recurrences after<br />
resection. 5 Although chemotherapy routines are frequently<br />
interrupted by surgical resection for 6 weeks, this same<br />
requirement is not present with percutaneous ablation techniques.<br />
6<br />
202<br />
lization, intra-arterial chemotherapy ports, and isolated<br />
hepatic perfusion (IHP)—are potential techniques for managing<br />
patients with unresectable liver metastases. Understanding<br />
the timing and role <strong>of</strong> these techniques in the<br />
multidisciplinary care <strong>of</strong> the patient is critical. Implementation<br />
<strong>of</strong> the IR clinic for consultation has enabled better integration<br />
<strong>of</strong> these therapies into the patient’s overall care and has<br />
facilitated improved opportunities for clinical studies.<br />
In a review <strong>of</strong> nine published articles for patients treated<br />
for unresectable colorectal liver metastases, the 5-year survival<br />
rate varied between 14% and 55% (median 30%). For<br />
the subgroup <strong>of</strong> patients with metastases smaller than or<br />
equal to 4 cm, this 5-year survival rate was better at 18% to<br />
56% (median 34%). 3 These are improved survival rates<br />
compared with chemotherapy alone and comparable to patients<br />
for whom metastases could be resected. 7 There has<br />
not been a randomized study for resectable liver metastases<br />
comparing resection with RFA. 8<br />
“Test <strong>of</strong> Time” Approach<br />
The “test <strong>of</strong> time” approach using percutaneous ablation<br />
was proposed by Livraghi et al in 2003 to allow the biology<br />
<strong>of</strong> the disease to express itself and to maintain quality <strong>of</strong> life<br />
in patients with liver metastases from colorectal cancer. 9<br />
The concept is that many patients may be able to avoid<br />
unnecessary surgery by undergoing ablation <strong>of</strong> liver metastases<br />
in the interval from the time <strong>of</strong> diagnosis <strong>of</strong> liver<br />
metastases to the time <strong>of</strong> hepatic metastatectomy. The<br />
theory is that RFA can completely treat many metastases<br />
with a small local recurrence rate. Those without recurrence<br />
will have avoided resection. In the delayed period from<br />
diagnosis to surgery, the patient who develops innumerable<br />
metastases would also be able to avoid unnecessary, unbeneficial<br />
surgery. For patients with local recurrence after<br />
ablation, there would still be an opportunity to repeat<br />
ablation or perform resection. In summary, this approach<br />
would allow the patient with a successful ablation and the<br />
patient for whom surgery would not have helped because <strong>of</strong><br />
an early explosion <strong>of</strong> metastases to benefit by avoiding<br />
resection. This theory has not been tested in a randomized,<br />
controlled study but, nonetheless, <strong>of</strong>fers an interesting management<br />
concept.<br />
Arterial Therapies<br />
Arterial therapies for colorectal liver cancer metastases<br />
can be performed to complement or salvage the effects <strong>of</strong><br />
systemic therapy. The application <strong>of</strong> a locoregional therapy<br />
From the Memorial Sloan-Kettering Cancer Center, New York, NY.<br />
Authors’ disclosures <strong>of</strong> potential conflicts <strong>of</strong> interest are found at the end <strong>of</strong> this article.<br />
Address reprint requests to Stephen B. Solomon, MD, Memorial Sloan-Kettering Cancer<br />
Center, 1275 York Ave., H-118, New York, NY; email: solomons@mskcc.org.<br />
© <strong>2012</strong> by <strong>American</strong> <strong>Society</strong> <strong>of</strong> <strong>Clinical</strong> <strong>Oncology</strong>.<br />
1092-9118/10/1-10
RADIOLOGY AND COLORECTAL LIVER METASTASES<br />
can occasionally downstage the patient with inoperable liver<br />
metastases to an operable status. As a whole, intra-arterial<br />
therapies rely on the fact that liver cancers derive their<br />
blood supply predominantly from hepatic arteries, whereas<br />
normal liver parenchyma has a predominantly portal vein<br />
source <strong>of</strong> blood supply. 10<br />
Intra-arterial Hepatic Chemotherapy<br />
Intra-arterial hepatic chemotherapy (IAHC) aims to increase<br />
the drug concentration in liver metastases and<br />
thereby improve response rates. 11 This approach can be best<br />
applied with drugs having a high first pass effect. Floxuridine<br />
with a first pass extraction rate <strong>of</strong> 95% can increase the<br />
liver dose by 100 to 300 times higher than the systemic<br />
perfusion. Historically, repeated or continuous IAHC has<br />
been delivered by a catheter and pump system requiring<br />
laparotomy. More recently, IAHC can be delivered through<br />
an interventional radiology approach with a subcutaneous<br />
port placed. 12 In one study <strong>of</strong> 36 patients with extensive<br />
nonresectable liver metastases (i.e., � 4 metastases in 86%<br />
and bilobar in 91%), IAHC was used with oxaliplatin (100<br />
mg/m 2 in 2 hours) plus intravenous 5-fluorouracil (5-FU)<br />
and leucovorin (leucovorin 400 mg/m 2 in 2 hours; 5-FU 400<br />
mg/m 2 bolus then 2,500 mg/m 2 in 46 hours) and cetuximab<br />
(400 mg/m 2 then 250 mg/m 2 /week or 500 mg/m 2 every 2<br />
weeks) as first-line treatment. Overall response rate (ORR)<br />
was 90% and disease control rate was 100%. Forty-eight<br />
percent <strong>of</strong> patients were downstaged enough to undergo an<br />
R0 resection and/or RFA. 13<br />
TACE<br />
There are several different regimens used to deliver<br />
TACE. One group using the regimen <strong>of</strong> cisplatin, doxorubicin,<br />
mitomycin C, ethiodol, and polyvinyl alcohol has shown<br />
an ORR <strong>of</strong> 43%. In this study, the median survival <strong>of</strong> 33<br />
months from initial diagnosis, 27 months from the time <strong>of</strong><br />
liver metastases, and 9 months from the start <strong>of</strong> chemoembolization<br />
suggests a possible improvement over reported<br />
survival time for systemic therapies alone. 14 Another group<br />
using mitomycin C alone (52.5%), mitomycin C with gemcitabine<br />
(33%), or mitomycin C and irinotecan (14.5%) has<br />
shown an ORR <strong>of</strong> 63%. 15<br />
KEY POINTS<br />
● Interventional radiologists play an important role in<br />
the multidisciplinary care <strong>of</strong> patients with colorectal<br />
cancer with liver metastases.<br />
● Ablation therapy can focally destroy liver metastases,<br />
providing long-term survival in select cases.<br />
● The “test <strong>of</strong> time” approach with ablation before<br />
hepatic metastatectomy can help select which patients<br />
will benefit from surgery and which patients’<br />
biology will lead to innumerable metastases making<br />
them ultimately not a good surgical candidate.<br />
● Intra-arterial therapies with chemoembolization, radioembolization,<br />
drug-eluting beads, intra-arterial<br />
chemotherapy, and isolated hepatic perfusion can<br />
play a role in treating unresectable liver metastases<br />
and liver-dominant disease.<br />
Recently, DEB have been developed that allow drug release<br />
after the bead has been embolized into the tumor<br />
microcirculation. The advantage <strong>of</strong> the beads is a reduced<br />
systemic delivery <strong>of</strong> chemotherapy. One <strong>of</strong> the drugs that<br />
has been loaded on these beads is irinotecan, which had a<br />
75% reduced systemic plasma level compared with intraarterial<br />
irinotecan alone. 16<br />
In a randomized study <strong>of</strong> two courses <strong>of</strong> DEB with irinotecan<br />
(36 patients) compared with eight courses <strong>of</strong> intravenous<br />
irinotecan, 5-FU, and leucovorin (FOLFIRI; 36<br />
patients) for 72 patients who failed at least two lines <strong>of</strong><br />
chemotherapy, the response rates were 70% for the DEB<br />
group compared with 30% for the systemic FOLFIRI<br />
group. 17 Similarly the 2-year overall survival (OS) was 38%<br />
compared with 18%, and the median OS was 690 days<br />
compared with 482 days. These both favored the DEB arm.<br />
Improvement in quality <strong>of</strong> life was 60% for the DEB group<br />
compared with 22% for the FOLFIRI group. Finally, overall<br />
cost was lower for the DEB treatment arm.<br />
In a multicenter, single-arm study <strong>of</strong> 55 patients who<br />
underwent DEB with irinotecan after failing systemic chemotherapy,<br />
response rates were 66% at 6 months and 75% at<br />
12 months with an OS <strong>of</strong> 19 months and a progression-free<br />
survival (PFS) <strong>of</strong> 11 months. 18<br />
Radioembolization<br />
External beam radiation therapy is challenged by the<br />
sensitivity <strong>of</strong> normal liver to radiation. The dose to treat a<br />
liver tumor is estimated at 70 Gy, while the liver tolerance<br />
dose is 35 Gy. Radioembolization refers to the targeted<br />
intra-arterial delivery <strong>of</strong> yttrium-90 (90Y) permanently<br />
bound to microspheres. Selective intra-arterial delivery enables<br />
doses over 120 Gy to target the tumor without reaching<br />
the liver toxicity threshold. An Italian multicenter,<br />
phase II study examined 50 patients with liver-only or<br />
liver-dominant colorectal metastases who failed at least<br />
three lines <strong>of</strong> systemic chemotherapy with at least one<br />
oxaliplatin and one irinotecan regimen and who underwent<br />
radioembolization. The ORR was 24%, the PFS was 3.7<br />
months with a median OS <strong>of</strong> 12.6 months, and the 1- and<br />
2-year survival rates were 50.4% and 19.6%, repectively. 19<br />
In a randomized study by Van Hazel et al comparing<br />
radioembolization plus systemic chemotherapy with chemotherapy<br />
alone as first-line therapy for colorectal liver metastases,<br />
the authors found an improved median survival <strong>of</strong><br />
29.4 months compared with 11.8 months in the<br />
chemotherapy-alone group. 20 In a Belgian multicenter<br />
phase III study, patients were randomly selected to receive<br />
Y90 microspheres in addition to intravenous 5-FU infusion<br />
compared with intravenous 5-FU alone. Median time to<br />
tumor progression was 4.5 compared with 2.1 months,<br />
respectively (hazard ratio � 0.51; 95% CI, 0.28 to 0.94; p �<br />
0.03). 21 There are at least 10 prospective clinical trials <strong>of</strong><br />
radioembolization that are open to accrual. 9<br />
IHP<br />
IHP allows high-dose chemotherapy to be delivered<br />
through the hepatic artery to the liver without reaching the<br />
systemic circulation. The venous outflow from the liver is<br />
circulated through extracorporeal filters to remove the drug<br />
before the blood is returned to the patient’s circulation. This<br />
technique opens the possibility <strong>of</strong> chemotherapy agents to<br />
those that do not necessarily have a high liver first pass<br />
203
effect. The most common drug that has been investigated<br />
has been melphalan at 200 mg doses. In a phase II study<br />
<strong>of</strong> 154 patients, the authors found a 50% ORR and a median<br />
PFS and OS <strong>of</strong> 7.4 and 24.8 months, respectively. 22 In<br />
another report comparing IHP with a comparable non-IHP<br />
cohort, the median OS for IHP was 25.0 months compared<br />
with 21.7 months for the control, which was not significantly<br />
different. Although the technique historically has<br />
been a surgical one, new angiographic tools have been<br />
developed, making interventional, image-guided implantation<br />
feasible.<br />
Conclusion<br />
IRs are playing an increasing role in the management <strong>of</strong><br />
patients with colorectal liver metastases. The increasing clinical<br />
role <strong>of</strong> the IR as a consultant—seeing patients in clinic—<br />
Authors’ Disclosures <strong>of</strong> Potential Conflicts <strong>of</strong> Interest<br />
Author<br />
Employment or<br />
Leadership<br />
Positions<br />
Consultant or<br />
Advisory Role<br />
Stephen B. Solomon Althera;<br />
AngioDynamics<br />
(U); Covidien (U);<br />
GE Healthcare;<br />
Johnson &<br />
Johnson<br />
Constantinos T. S<strong>of</strong>ocleous*<br />
*No relevant relationships to disclose.<br />
1. House MG, Ito H, Gonen M, et al. Survival after hepatic resection for<br />
metastatic colorectal cancer: Trends in outcomes for 1,600 patients during<br />
two decades at a single institution. J Am Coll Surg. 2010;210:744-752.<br />
2. Adam R, Vinet E. Regional treatment <strong>of</strong> metastasis: Surgery <strong>of</strong> colorectal<br />
liver metastases. Ann Oncol. 2004;15:103-106.<br />
3. Pathak S, Jones R, Tang JMF, et al. Ablative therapies for colorectal<br />
liver metastases: A systemic review. Colorectal Dis. 2011;13:252-265.<br />
4. Ben-David E, Appelbaum L, Sosna J, et al. Characterization <strong>of</strong> irreversible<br />
electroporation ablation in in vivo porcine liver. AJR Am J Roentgenol.<br />
<strong>2012</strong>;198:62-68.<br />
5. S<strong>of</strong>ocleous CT, Petre EN, Gonen M, et al. CT-guided radi<strong>of</strong>requency<br />
ablation as a salvage treatment <strong>of</strong> colorectal cancer hepatic metastases<br />
developing after hepatectomy. J Vasc Interv Radiol. 2011;22:755-761.<br />
6. Erinjeri JP, Fong AJ, Kemeny NE, et al. Timing <strong>of</strong> administration <strong>of</strong><br />
bevacizumab chemotherapy affects wound healing after chest wall port<br />
placement. Cancer. 2011;117:1296-1301.<br />
7. Machi J, Oishi AJ, Sumida K, et al. Long-term outcome <strong>of</strong> radi<strong>of</strong>requency<br />
ablation for unresectable liver metastases from colorectal cancer:<br />
evaluation <strong>of</strong> prognostic factors and effectiveness in first- and second-line<br />
management. Cancer J. 2006;12:318-326.<br />
8. Mulier S, Ni Y, Jamart J, et al. Radi<strong>of</strong>requency ablation versus resection<br />
for resectable colorectal liver metastases: time for a randomized trial? Ann<br />
Surg Oncol. 2008;15:144-157.<br />
9. Livraghi T, Solbiati L, Meloni F, et al. Percutaneous radi<strong>of</strong>requency<br />
ablation <strong>of</strong> liver metastases in potential candidate for resection: the “test-<strong>of</strong>time<br />
approach.” Cancer. 2003;97:3027-3035.<br />
10. de Baere T, Deschamps F. Arterial therapies <strong>of</strong> colorectal cancer<br />
metastases to the liver. Abdom Imaging. 2011;36:661-670.<br />
11. Kemeny NE, Melendez FD, Capanu M, et al. Conversion to resectability<br />
using hepatic artery infusion plus systemic chemotherapy for the treatment <strong>of</strong><br />
unresectable liver metastases from colorectal carcinoma. J Clin Oncol.<br />
2009;20:3465-3471.<br />
12. Ganeshan A, Upponi S, Hon L-Q, et al. Hepatic arterial infusion <strong>of</strong><br />
chemotherapy: the role <strong>of</strong> diagnostic and interventional radiology. Ann Oncol.<br />
2008;19:847-851.<br />
13. Malka D, Paris E, Caramella C, et al. Hepatic arterial infusion (HAI) <strong>of</strong><br />
204<br />
has increased their role as an important contributor to the<br />
multidisciplinary management and the customization <strong>of</strong> treatment<br />
for these patients. The IR armamentarium includes<br />
ablative tools that can focally destroy small numbers <strong>of</strong> liver<br />
metastases. The use <strong>of</strong> ablation in a “test <strong>of</strong> time” paradigm<br />
may limit unnecessary and morbid resections significantly<br />
contributing to the preservation <strong>of</strong> patient quality <strong>of</strong> life.<br />
Locoregional arterially directed therapies for liver-dominant<br />
metastases also allow the physician to manage the unresectable<br />
patient to extend disease-free periods and OS and, hopefully,<br />
convert him or her to a resection candidate for a potential<br />
cure. As all <strong>of</strong> these tools and techniques have become available<br />
and perfected over the past decade, it will become important<br />
for them to be investigated in clinical trials to best<br />
determine the appropriate use in the care <strong>of</strong> the patient with<br />
colorectal cancer with liver metastases.<br />
Stock<br />
Ownership Honoraria<br />
AngioDynamics;<br />
Johnson &<br />
Johnson<br />
REFERENCES<br />
Research<br />
Funding<br />
AngioDynamics;<br />
GE Healthcare<br />
SOLOMON AND SOFOCLEOUS<br />
Expert<br />
Testimony<br />
Other<br />
Remuneration<br />
oxaliplatin plus intravenous (iv) fluorouracil (FU), leucovorin (LV), and<br />
cetuximab for first-line treatment <strong>of</strong> unresectable colorectal liver metastases<br />
(CRLM) (CHOICE): A multicenter phase II study. J Clin Oncol. <strong>2012</strong>;28:15s<br />
(suppl; abstr 3558).<br />
14. Albert M, Kiefer MV, Sun W, et al. Chemoembolization <strong>of</strong> colorectal<br />
liver metastases with cisplatin, doxorubicin, mitomycin C, ethiodol, and<br />
polyvinyl alcohol. Cancer. 2011;117:343-352.<br />
15. Vogl TJ, Gruber T, Balzer JO, et al. Repeated transarterial chemoembolization<br />
in the treatement <strong>of</strong> liver metastases <strong>of</strong> colorectal cancer: Prospective<br />
study. Radiology. 2009;250:281-289.<br />
16. Taylor RR, Tang Y, Gonzalez MV, et al. Irinotecan drug eluting beads<br />
for use in chemoembolization: in vitro and in vivo evaluation <strong>of</strong> drug release<br />
properties. Eur J Pharm Sci. 2007;30:7-14.<br />
17. Fiorentini G, Aliberti C, Montagnani F, et al. Transarterial chemoembolization<br />
<strong>of</strong> metastatic colorectal carcinoma to the liver adopting polyvinyl<br />
alcohol microspheres loaded with irinotecan compared with FOLFIRI: Evaluation<br />
at two years <strong>of</strong> a phase III clinical trial. ESMO. 2010 (abstr 588).<br />
18. Martin RCG, Joshi J, Robbins K, et al. Hepatic intra-arterial injection<br />
<strong>of</strong> drug-eluting bead, irinotecan (DEBIRI) in unresectable colorectal liver<br />
metastases refractory to systemic chemotherapy: results <strong>of</strong> multiinstitutional<br />
study. Ann Surg Oncol. 2011;18:192-198.<br />
19. Cosmielli M, Golfieri R, Cagol PP, et al. Multi-centre phase II clinical<br />
trial <strong>of</strong> yttrium-90 resin microspheres alone in unresectable, chemotherapy<br />
refractory colorectal liver metastases. Br J Cancer. 2010;103:324-331.<br />
20. Van hazel G, Blackwell A, Anderson J, et al. Randomised phased 2 trial<br />
<strong>of</strong> SIR-Spheres plus fluorouracil/leucovorin chemotherapy versus fluorouracil/<br />
leucovorin chemotherapy alone in advanced colorectal cancer. J Surg Oncol.<br />
2004;88:78-85.<br />
21. Hendlisz A, Van Den Eynde M, Peeters M, et al. Phase III trial<br />
comparing protracted intravenous fluororacil infusion alone or with<br />
yttrium-90 resin microspheres radioembolization for liver-limited metastatic<br />
colorectal cancer refractory to standard chemotherapy. J Clin Oncol. 2010;<br />
28:3687-3694.<br />
22. Van Iersel LBJ, Gelderblom H, Vahrmeijer AL, et al. Isolated hepatic<br />
melphalan perfusion <strong>of</strong> colorectal liver metastases: outcome and prognostic<br />
factors in 154 patients. Ann Oncol. 2008;19:1127-1134.
Curative-Intent Treatment for Colorectal<br />
Liver Metastases: A Medical Oncologist’s<br />
Perspective<br />
Overview: Resection or ablation <strong>of</strong> CRC liver metastases can<br />
be <strong>of</strong>fered with curative intent in some, but not all patients<br />
in whom resection is technically possible. Chemotherapy<br />
can improve the potential for cure to some degree, either in<br />
the adjuvant or neoadjuvant setting, or, in relatively rare<br />
A SUBSTANTIAL<br />
NUMBER <strong>of</strong> colorectal cancer (CRC)<br />
patients with oligometastatic disease confined to the<br />
liver are potentially curable by surgical resection and/or<br />
ablation. The degree to which chemotherapy contributes to<br />
the curability <strong>of</strong> such patients has not been extensively<br />
studied. This article will discuss some <strong>of</strong> the relevant issues<br />
in patient selection and choice <strong>of</strong> management strategies<br />
and will review a number <strong>of</strong> studies that inform discussion<br />
on this topic. Ultimately, in the absence <strong>of</strong> specific, definitive<br />
data, conclusions must be drawn that are influenced by the<br />
available data but are ultimately based on subjective interpretation<br />
<strong>of</strong> those data. In this context, this article will<br />
present a number <strong>of</strong> opinions, all <strong>of</strong> which I believe to be well<br />
supported by data, but all <strong>of</strong> which are open to discussion<br />
and debate.<br />
Neoadjuvant Compared with Conversion<br />
to Resectability<br />
Chemotherapy in a curative-intent treatment strategy for<br />
CRC can be given in one <strong>of</strong> three modes; adjuvant, neoadjuvant,<br />
or conversion to resectability. Adjuvant chemotherapy<br />
is treatment given after an R0 resection in the hopes <strong>of</strong><br />
eradicating residual microscopic disease. There is an important<br />
difference between neoadjuvant and conversion chemotherapy,<br />
each <strong>of</strong> which is given before surgical intervention.<br />
Neoadjuvant chemotherapy refers to chemotherapy given<br />
before a planned resection <strong>of</strong> fully resectable disease. By<br />
definition, the disease that is treated with neoadjuvant<br />
chemotherapy could be resected with no chemotherapy at<br />
all, and no response is needed for the resection to be<br />
accomplished. If tumor shrinkage or regression is required<br />
for resection, then the chemotherapy is not correctly referred<br />
to as neoadjuvant; such chemotherapy is referred to<br />
as “conversion” chemotherapy, designed to convert an unresectable<br />
patient to a resectable one.<br />
Conversion to resectability is a complex and important<br />
concept, since some patients who might previously have<br />
been thought to be incurable might have a realistic chance<br />
through this multimodality approach. However, the numbers<br />
may be far smaller than are widely appreciated. For<br />
example, Adam et al reported in 2009 that <strong>of</strong> 184 patients<br />
with initially unresectable CRC metastases who were assessed<br />
as having become resectable and then were resected<br />
with an R0 resection, 24 were effectively cured (free <strong>of</strong><br />
disease <strong>of</strong>f therapy for 5 or more years). 1 So was the<br />
chemotherapy given with curative intent? If we take the<br />
denominator as the 184 patients who underwent an R0<br />
resection, then this gives a cure rate <strong>of</strong> 16%, as reported.<br />
However, a presumably larger cohort went to the operating<br />
By Leonard B. Saltz, MD<br />
circumstances, by converting truly unresectable disease into<br />
resectable. Careful and realistic patient selection, with an<br />
individualized and realistic assessment <strong>of</strong> curative potential,<br />
is key to providing each patient with the means to make<br />
realistic treatment choices.<br />
room, some <strong>of</strong> whom were found at operation to have tumors<br />
that were unresectable; this larger denominator would<br />
shrink the curative percentages somewhat. Consider further<br />
that the 184 patients were culled from the experiences <strong>of</strong><br />
14 years <strong>of</strong> treating patients with metastatic colorectal<br />
cancer at a busy tertiary referral center. Thus, the addition<br />
<strong>of</strong> chemotherapy to resection resulted in a cure <strong>of</strong> an otherwise<br />
incurable patient in 24 metastatic colorectal patients at<br />
this center over 14 years. It is a remarkable and laudable,<br />
albeit relatively rare, achievement. It is also notable that<br />
Adam’s group identified favorable risk factors within the<br />
16% <strong>of</strong> cured patients. Chances <strong>of</strong> being cured were higher in<br />
patients with three or fewer metastases, patients in whom<br />
the largest tumor before chemotherapy was no greater than<br />
3.0 cm, and in those patients who achieved a pathologic<br />
complete response. Thus, patients having complete resections<br />
but with bulky tumors, larger numbers <strong>of</strong> liver lesions,<br />
or absence <strong>of</strong> a complete pathologic response had a cure<br />
rate that was substantially smaller than 16%, a point that<br />
needs to be considered when discussing prognosis with such<br />
patients.<br />
It should be noted that most patients who are candidates<br />
for conversion to resectability are those who have disease<br />
confined to the liver, with lesions that are abutting critical<br />
structures, such as major vessels, whereby a resection is not<br />
possible, but in whom tumor shrinkage will create a plane to<br />
render resection possible. Since most CRC metastases that<br />
achieve a clinical complete response are not pathologic<br />
complete responses and are destined to grow back, the<br />
likelihood <strong>of</strong> resection <strong>of</strong> residual disease being curative is<br />
small if multiple areas <strong>of</strong> previous disease are left in. 2<br />
Adam’s group noted that resection <strong>of</strong> liver metastases in<br />
patients with resected celiac or retroperitoneal lymph nodes,<br />
even if those lymph nodes responded to chemotherapy, had<br />
a poor prognosis with no cures. 3<br />
Defining “Curative Intent”<br />
Terms such as “survival,“ “progression-free survival,” “response,“<br />
and, “clinical benefit,” while useful, are <strong>of</strong>ten mistaken<br />
to equate to more than they actually mean. The words<br />
“cure” and “curative” are pure; cure means the cancer is<br />
From the Department <strong>of</strong> Medicine, Memorial Sloan Kettering Cancer Center, New York,<br />
NY.<br />
Author’s disclosure <strong>of</strong> potential conflicts <strong>of</strong> interest are found at the end <strong>of</strong> this article.<br />
Address reprint requests to Leonard B. Saltz, MD, Memorial Sloan Kettering Cancer<br />
Center, 300 East 66th Street, Room 1049, New York, NY 10065; email: saltzl@mskcc.org.<br />
© <strong>2012</strong> by <strong>American</strong> <strong>Society</strong> <strong>of</strong> <strong>Clinical</strong> <strong>Oncology</strong>.<br />
1092-9118/10/1-10<br />
205
gone forever, no further therapy is needed. “Curative,” when<br />
referring to a treatment strategy—be it medical, surgical, or<br />
combined modality—means that the regimen will result, or<br />
has resulted, in a cure. However, the definition <strong>of</strong> the term<br />
“curative intent” is a bit more difficult to pin down. First, we<br />
must recognize that long-term survival is not the same as<br />
cure. A patient who is rendered free <strong>of</strong> known disease but<br />
who then recurs a year later and is alive 5 years after that<br />
has a good long-term survival but has not been cured. It is<br />
true that favorable long-term survival is a highly desirable<br />
outcome, but the degree to which surgery, ablation, and<br />
chemotherapy might or might not improve long-term survival<br />
is a discussion for another day; if the tumor is present,<br />
the patient has not been cured, and if a treatment strategy<br />
does not result in cure in some reasonable percentage <strong>of</strong><br />
patients, then we cannot realistically consider that we are<br />
using it with “curative intent.”<br />
There is no agreement on what degree <strong>of</strong> likelihood <strong>of</strong><br />
success constitutes curative intent. If an approach will<br />
result in a cure more than 90% <strong>of</strong> the time, is that a<br />
treatment given with curative intent? Most <strong>of</strong> us would say<br />
it is. What about approximately 75% <strong>of</strong> the time? Fifty<br />
percent? Twenty-five percent? What if a treatment strategy<br />
is curative only 10% or 5% <strong>of</strong> time? What about 3% or 1%? To<br />
provide some context for this discussion, consider the use <strong>of</strong><br />
systemic chemotherapy in patients with surgically unresectable<br />
disease. Dy et al reported the long-term outcome <strong>of</strong><br />
patients on intergroup protocol N9741. 4 A total <strong>of</strong> 1,508<br />
patients with metastatic CRC received one <strong>of</strong> three first-line<br />
regimens. Of these, 62 patients achieved a clinical complete<br />
response. However, with a median follow-up <strong>of</strong> 50 months,<br />
only 10 remained disease-free and could be considered<br />
cured. Thus, at first look, systemic chemotherapy would<br />
appear to have a cure rate <strong>of</strong> 10 out <strong>of</strong> 1,508, or 0.66%.<br />
However, if we take the group on N9741 who were treated<br />
with folinic acid, fluorouracil, and oxaliplatin (FOLFOX), 43<br />
<strong>of</strong> 679, or 6.3%, had a clinical complete response, and if we<br />
assume the same 16% cure rate for those achieving a clinical<br />
complete response, then the cure rate for FOLFOX is 1.0%.<br />
But the chance <strong>of</strong> achieving a complete response (a necessary<br />
first step toward a cure) was increased by having low<br />
tumor burden, a single site <strong>of</strong> disease, and measurable<br />
rather than evaluable disease, all <strong>of</strong> which are characteristics<br />
that would be present in the vast majority <strong>of</strong> patients<br />
who are candidates for resection or ablative therapy. Thus,<br />
for the population that might be considered for curative<br />
resection or ablation, greater than 1% <strong>of</strong> patients would be<br />
expected to be cured by FOLFOX alone. Clearly we would<br />
KEY POINTS<br />
● “Cure” is easy to define; “curative intent” is not.<br />
● Some, but not all, patients who can undergo resection<br />
or ablation have a realistic chance at cure.<br />
● Adjuvant or neoadjuvant FOLFOX can increase the<br />
likelihood <strong>of</strong> cure in FOLFOX-naive patients.<br />
● Therapies that are inactive in stage III adjuvant are<br />
unlikely to be active in stage IV adjuvant.<br />
● “Conversion to resectability” chemotherapy and “neoadjuvant”<br />
chemotherapy are not the same thing.<br />
206<br />
not regard FOLFOX as being given with curative intent to<br />
an unresectable patient. So the possibility <strong>of</strong> a rare cure<br />
is not sufficient to allow us to say a regimen has curative<br />
intent.<br />
We are probably more inclined to perceive a regimen as<br />
being given with curative intent if we think <strong>of</strong> it as curing<br />
10%, rather than if we consider that it fails to cure 90% <strong>of</strong><br />
patients. If more than four out <strong>of</strong> five or more than nine out<br />
<strong>of</strong> 10 patients treated by a strategy are destined to not<br />
achieve cure, can we say we are treating with curative<br />
intent? The answer to this question will vary from reader to<br />
reader, and this article is not proposing to set a definition;<br />
however, consideration <strong>of</strong> these points will be useful to the<br />
reader in making his or her own determination.<br />
Defining Curative Resection<br />
LEONARD B. SALTZ<br />
The reason that resection/ablation is an accepted standard<br />
practice without randomized data is that the cure rate with<br />
this approach appears to be substantially higher than what<br />
could realistically be expected from systemic therapy alone.<br />
It should be emphasized that it is the cure rate, and no other<br />
factor, that has led to the universal acceptance <strong>of</strong> this<br />
approach without a randomized trial. What is the cure rate,<br />
and to what degree does chemotherapy add to that? The<br />
European Organisation for Research and Treatment <strong>of</strong> Cancer<br />
(EORTC) has conducted one <strong>of</strong> the few large-scale<br />
randomized trials involving liver resection <strong>of</strong> hepatic metastases<br />
<strong>of</strong> CRC. 5 EORTC 40983 was limited to patients with no<br />
more than four liver-only metastases. As it turned out, more<br />
than one-half <strong>of</strong> the patients had only a solitary liver<br />
metastasis, 14% had three liver lesions, and only 7% had<br />
four liver lesions. Extrapolation to patients with more extensive<br />
disease must be done with caution. The three-year<br />
progression-free survival (effectively recurrence-free survival<br />
in this resected population) was the primary endpoint.<br />
Since the overwhelming majority <strong>of</strong> recurrences happen<br />
within the first 3 years, 3-year recurrence-free survival is a<br />
reasonable surrogate for cure. In this EORTC trial, <strong>of</strong> the<br />
patients actually undergoing resection, 3-year disease-free<br />
survival was 33% in patients treated with surgery only<br />
and 42% in patients treated with surgery plus perioperative<br />
FOLFOX (3 months preoperative and 3 months postoperative).<br />
Thus, <strong>of</strong> the patients with CRC with one to four<br />
(mostly one to two) liver metastases who were able to<br />
successfully undergo resection, approximately one-third<br />
were potentially cured. The administration <strong>of</strong> perioperative<br />
FOLFOX improved that potential cure rate by an absolute<br />
increase <strong>of</strong> 9%, or a relative increase <strong>of</strong> 27% (because 9%<br />
is 27% <strong>of</strong> the baseline <strong>of</strong> 33%). This 9% absolute increase<br />
was relatively disappointing, considering that FOLFOX increased<br />
disease-free survival by 7.2% in patients with stage<br />
III A and B disease and 11.5% in patients with stage IIIC<br />
disease in the MOSAIC trial. 6 The reason for this somewhat<br />
disappointing performance by FOLFOX has not been definitively<br />
identified. One possible contributing factor is that<br />
although no patient on the EORTC 40983 trial had received<br />
prior oxaliplatin, 42% had received prior fluoropyrimidinebased<br />
adjuvant chemotherapy for their primary cancer,<br />
thereby selecting for a relatively fluoropyrimidine-resistant<br />
population.<br />
The fact that the clinical trial was done with the FOLFOX<br />
split into 3 months before and 3 months after surgery should<br />
in no way be construed as creating an obligation to follow
TREATMENT FOR COLORECTAL LIVER METASTASES<br />
that same timing in clinical practice. Delivery <strong>of</strong> a full 6<br />
months <strong>of</strong> FOLFOX preoperatively, however, runs the risk<br />
<strong>of</strong> creating a higher-risk resection because <strong>of</strong> chemotherapyassociated<br />
steatohepatitis. For this reason, for clearly resectable<br />
patients, our group <strong>of</strong>ten favors up-front resection<br />
followed by postoperative FOLFOX for 6 months. No specific<br />
data exist, however, to settle this question.<br />
Resections with Four or More Liver Lesions<br />
The Memorial Sloan-Kettering Cancer Center hepatobiliary<br />
surgical service reviewed the outcome <strong>of</strong> all patients<br />
with four or more liver lesions resected over a 4-year period<br />
between 1998 and 2002. 7 It should be noted that oxaliplatin<br />
was largely unavailable during that time period, so the<br />
contribution <strong>of</strong> optimal therapies that are inactive in stage<br />
III adjuvant setting are unlikely to be active in stage IV<br />
adjuvant setting. Of 548 patients who underwent successful<br />
resection, 98 had four or more liver metastases. As is the<br />
case with many such surgical reports, the denominator <strong>of</strong><br />
how many patients were taken to the operating room with<br />
the intent <strong>of</strong> full resection but who would end up having<br />
an R1 or R2 resection is not available. Nevertheless, the<br />
long-term follow-up on these patients does provide useful<br />
insights, especially into the curative compared with noncurative<br />
outcomes and the important difference between overall<br />
survival and cure. Actuarial survival <strong>of</strong> these 98 patients<br />
was 33% at 5 years. This number was markedly different for<br />
those with four or five lesions (39% actuarial survival at 5<br />
years) compared with 19% 5-year actuarial survival for<br />
those with six or more lesions). These results should be<br />
interpreted within the context <strong>of</strong> modern imaging and modern<br />
chemotherapy, which make older historic comparisons<br />
moot. It is no longer correct or reasonable to say that the<br />
median survival with systemic therapy is 1 year and that<br />
most patients are dead within 2 years, as was the case 20<br />
years ago. In the N9741 trial <strong>of</strong> systemic therapy, the 5-year<br />
overall survival in patients treated with frontline FOLFOX<br />
was 9.8% (note that 10% <strong>of</strong> these 5-year survivors did<br />
ultimately undergo metastasectomy, while 90% did not). 8<br />
Five-year overall survival would be expected to be higher<br />
with systemic treatment for those patients with good performance<br />
status, relatively low volume <strong>of</strong> disease, and one site<br />
<strong>of</strong> metastases, so that the population that were candidates<br />
for liver resection would be expected to have a 5-year overall<br />
survival that would be somewhat higher than 10% with<br />
systemic therapy. Thus, the argument that surgery for four<br />
or more lesions improves long-term survival is compelling,<br />
although not airtight. When we look at the cure rate,<br />
however, a different picture emerges. The median diseasefree<br />
survival for these 98 patients was 12 months, with a<br />
range <strong>of</strong> 10 to 15 months, and the actuarial disease-free<br />
survival was 50% at 1 year, 12% at 3 years, and 0% at 5<br />
years. The intriguing and important question from a medical<br />
oncology perspective is whether or not active systemic<br />
oxaliplatin-containing adjuvant, neoadjuvant, or conversion<br />
chemotherapy could have had a substantial effect on the<br />
likelihood <strong>of</strong> cure in these patients.<br />
Role <strong>of</strong> Non-FOLFOX Regimens<br />
It is important to bear in mind that the role <strong>of</strong> adjuvant (or<br />
neoadjuvant) chemotherapy is to eradicate micrometastases.<br />
This is true for treatment <strong>of</strong> stage II, stage III, or fully<br />
resected stage IV. As such, there is no compelling argument<br />
for why a micrometastasis in a patient with stage IV disease<br />
should behave any differently from a micrometastasis in<br />
a patient with stage III disease. Numerous studies in the<br />
adjuvant setting <strong>of</strong> stage III and stage II colon cancer have<br />
demonstrated that irinotecan, bevacizumab, and cetuximab<br />
(and by reasonable extrapolation, panitumumab) are inactive<br />
in the adjuvant setting in stage III, with numerous<br />
large-scale trials showing failure to improve the 3-year or<br />
5-year recurrence-free survival rates or the overall survival<br />
rates. 9-14 Since it is well demonstrated that all <strong>of</strong> these<br />
agents have activity against macrometastatic disease, it is<br />
clear that we cannot extrapolate this activity to the micrometastatic<br />
setting. On the basis <strong>of</strong> these data, there is no<br />
compelling argument for why any agents that have been<br />
shown to be inactive in the adjuvant treatment <strong>of</strong> stage III<br />
colon cancer should be active in the adjuvant or neoadjuvant<br />
treatment <strong>of</strong> resected stage IV CRC. The only trial that<br />
directly addresses this question is the randomized ACCORD<br />
3 trial <strong>of</strong> 5-fluorouracil and leucorovin with or without<br />
irinotecan in the adjuvant treatment <strong>of</strong> resected liver metastases,<br />
which showed no benefit for irinotecan. 15 Note that<br />
this statement regarding irinotecan, bevacizumab, and antiepidermal<br />
growth factor receptor (EGFR) monoclonal antibodies<br />
does not apply to conversion chemotherapy; if the<br />
goal is to shrink a tumor you can see, then irinotecan,<br />
bevacizumab, and the anti-EGFR monoclonal antibodies<br />
are all reasonable considerations. However, if the goal is to<br />
eradicate micrometastatic disease and increase the cure<br />
rate, then in my opinion, they are not.<br />
Role <strong>of</strong> FOLFOX in Patients with Prior FOLFOX<br />
Adjuvant Treatment<br />
When metachronous liver metastases develop after adjuvant<br />
chemotherapy, it must be recalled that the cells that<br />
gave rise to those metastases were present during that<br />
adjuvant chemotherapy and were therefore, by definition,<br />
resistant to it. It is thus exceedingly unlikely that this same<br />
chemotherapy that failed to eradicate micrometastases in<br />
the first treatment will have activity on residual micrometastases<br />
this second time around after hepatic resection.<br />
Remember, you are not treating the resected metastases;<br />
you are treating residual micrometastases. Thus, adjuvant<br />
or neoadjuvant FOLFOX does not appear to be a reasonable<br />
maneuver in a patient who had previously received adjuvant<br />
FOLFOX and who, by virtue <strong>of</strong> the development <strong>of</strong> metastases,<br />
failed that adjuvant therapy. If a patient has received<br />
either no prior therapy or fluorpyrimidine adjuvant only,<br />
then adjuvant or neoadjuvant FOLFOX would be indicated.<br />
It is noteworthy that the EOTRC 40983 trial excluded<br />
patients with prior oxaliplatin adjuvant chemotherapy. In a<br />
patient who has had prior adjuvant FOLFOX (or capecitabine/oxaliplatin<br />
[Cape/Ox]), there unfortunately is not an<br />
active adjuvant or neoadjuvant systemic treatment to <strong>of</strong>fer,<br />
and therefore I do not believe that any adjuvant or neoadjuvant<br />
systemic chemotherapy is indicated in such patients.<br />
Note that FOLFOX and Cape/Ox can be used interchangeably.<br />
There is no evidence to support that one is superior to<br />
the other, that one can rescue the other, or that capecitabine<br />
has activity after failure <strong>of</strong> an infusional 5-fluorouracilcontaining<br />
regimen.) The fact that FOLFIRI (fluorouracil,<br />
leucovorin, irinotecan) with or without either bevacizumab<br />
or an anti EGFR monoclonal might be able to shrink a tumor<br />
207
is irrelevant if it is already resectable (and so neoadjuvant<br />
and not conversion chemotherapy) since we have no reason<br />
to believe that these agents or regimens will have activity<br />
against the residual micrometastases that we are targeting.<br />
Conclusion<br />
Resection or ablation <strong>of</strong> CRC liver metastases can be<br />
<strong>of</strong>fered with curative intent in some, but not all patients in<br />
whom resection is technically possible. Chemotherapy can<br />
improve the potential for cure to some degree, either in the<br />
Author’s Disclosure <strong>of</strong> Potential Conflicts <strong>of</strong> Interest<br />
Author<br />
Employment or<br />
Leadership<br />
Positions<br />
Consultant or<br />
Advisory Role<br />
Leonard B. Saltz Bristol-Myers<br />
Squibb (U);<br />
Genentech;<br />
ImClone<br />
Systems; Merck;<br />
Novartis; Pfizer;<br />
Roche<br />
1. Adam R, Wicherts DA, de Haas RJ, et al. Patients with initially<br />
unresectable colorectal liver metastases: Is there a possibility <strong>of</strong> cure? J Clin<br />
Oncol. 2009;27:1829-1835.<br />
2. Benoist S, Brouquet A, Penna C, et al. Complete response <strong>of</strong> colorectal<br />
liver metastases after chemotherapy: Does it mean cure? J Clin Oncol.<br />
2006;24:3939-3945.<br />
3. Adam R, de Haas RJ, Wicherts DA, et al. Is hepatic resection justified<br />
after chemotherapy in patients with colorectal liver metastases and lymph<br />
node involvement? J Clin Oncol. 2008;26:3672-3680.<br />
4. Dy GK, Krook JE, Green EM, et al. Impact <strong>of</strong> complete response to<br />
chemotherapy on overall survival in advanced colorectal cancer: Results from<br />
Intergroup N9741. J Clin Oncol. 2007;25:3469-3474.<br />
5. Nordlinger B, Sorbye H, Glimelius B, et al. Perioperative chemotherapy<br />
with FOLFOX4 and surgery versus surgery alone for resectable liver metastases<br />
from colorectal cancer (EORTC Intergroup trial 40983): a randomised<br />
controlled trial. Lancet. 2008;22:1007-1016.<br />
6. Andre T, Boni C, Navarro M, et al. Improved overall survival with<br />
oxaliplatin, fluorouracil, and leucovorin as adjuvant treatment in stage II or<br />
III colon cancer in the MOSAIC trial. J Clin Oncol. 2009;27:3109-3116.<br />
7. Kornprat P, Jarnagin WR, Gonen M, et al. Outcome after hepatectomy<br />
for multiple (four or more) colorectal metastases in the era <strong>of</strong> effective<br />
chemotherapy. Ann Surg Oncol. 2007;14:1151-1160.<br />
8. San<strong>of</strong>f HK, Sargent DJ, Campbell ME, et al. Five-year data and<br />
prognostic factor analysis <strong>of</strong> oxaliplatin and irinotecan combinations for<br />
advanced colorectal cancer: N9741. J Clin Oncol. 2008;26:5721-5727.<br />
9. Saltz LB, Niedzwiecki D, Hollis D, et al. Irinotecan fluorouracil plus<br />
208<br />
adjuvant or neoadjuvant setting, or, in relatively rare circumstances,<br />
by converting truly unresectable disease into<br />
resectable. Careful and realistic patient selection, with an<br />
individualized and realistic assessment <strong>of</strong> curative potential,<br />
is key to providing each patient with the means to make<br />
realistic treatment choices. Ultimately it is anticipated that<br />
molecular and immunologic assessments <strong>of</strong> individuals and<br />
their tumors will help guide rational selection <strong>of</strong> strategies<br />
to increase the curative potential <strong>of</strong> combined modality<br />
management <strong>of</strong> CRC liver metastases.<br />
Stock<br />
Ownership Honoraria<br />
REFERENCES<br />
Research<br />
Funding<br />
Amgen; Bayer;<br />
Bristol-Myers<br />
Squibb;<br />
Genentech;<br />
ImClone<br />
Systems; Merck;<br />
Pfizer; Roche;<br />
Taiho<br />
Pharmaceutical<br />
Expert<br />
Testimony<br />
LEONARD B. SALTZ<br />
Other<br />
Remuneration<br />
leucovorin is not superior to fluorouracil plus leucovorin alone as adjuvant<br />
treatment for stage III colon cancer: Results <strong>of</strong> CALGB 89803. J Clin Oncol.<br />
2007;25:3456-3461.<br />
10. Ychou M, Raoul J-L, Douillard J-Y, et al. A phase III randomised trial<br />
<strong>of</strong> LV5FU2 � irinotecan versus LV5FU2 alone in adjuvant high-risk colon<br />
cancer (FNCLCC Accord02/FFCD9802). Ann Oncol. 2009;20:674-680.<br />
11. Van Cutsem E, Labianca R, Bodoky G, et al. Randomized phase III trial<br />
comparing biweekly infusional fluorouracil/leucovorin alone or with irinotecan<br />
in the adjuvant treatment <strong>of</strong> stage III colon cancer: PETACC-3. J Clin<br />
Oncol. 2009;27:3117-3125.<br />
12. Allegra CJ, Yothers G, O’Connell MJ, et al. Phase III trial assessing<br />
bevacizumab in stages II and III carcinoma <strong>of</strong> the colon: Results <strong>of</strong> NSABP<br />
protocol C-08. J Clin Oncol. 2011;29:11-16.<br />
13. De Gramont A, Van Cutsem E, Tabernero J, et al. AVANT: Results<br />
from a randomized, three-arm multinational phase III study to investigate<br />
bevacizumab with either XELOX or FOLFOX4 versus FOLFOX4 alone as<br />
adjuvant treatment for colon cancer. J Clin Oncol. 2011;29 (suppl 4; abstr<br />
362).<br />
14. Alberts SR, Sargent DJ, Smyrk CJ, et al. Adjuvant mFOLFOX6 with or<br />
without cetuxiumab (Cmab) in KRAS wild-type (WT) patients (pts) with<br />
resected stage III colon cancer (CC): Results from NCCTG Intergroup Phase<br />
III Trial N0147. J Clin Oncol. 2010;28 (suppl; abstr CRA3507).<br />
15. Ychou M, Hohenberger W, Thezenas S, et al. A randomized phase III<br />
study comparing adjuvant 5-fluorouracil/folinic acid with FOLFIRI in patients<br />
following complete resection <strong>of</strong> liver metastases from colorectal cancer.<br />
Ann Oncol. 2009;20:1964-1970.
Surgical Treatment <strong>of</strong> Colorectal Cancer<br />
Liver Metastases<br />
Overview: Treatment strategies for patients with stage IV<br />
colorectal cancer have changed markedly in the last decade.<br />
Patients with colorectal cancer metastases to the liver have<br />
always been a fascinating group to consider biologically and<br />
for local-regional treatment strategies. In the late 1980s<br />
through the 1990s, resection was performed for a select<br />
subset <strong>of</strong> patients who had resectable disease. However, a<br />
high proportion <strong>of</strong> patients had bilobar unresectable disease<br />
and were treated with either 5-fluorouracil–based systemic<br />
THE LIVER, second only to lymph nodes as the most<br />
common site <strong>of</strong> metastasis from other solid tumors, is a<br />
common site <strong>of</strong> colorectal cancer (CRC) metastasis. A subset<br />
<strong>of</strong> patients with metastatic CRC have liver-only disease or<br />
liver and resectable extrahepatic disease and are candidates<br />
for surgical treatment. Resection <strong>of</strong> liver metastases is<br />
associated with minimal mortality rates in these patients,<br />
and resection can provide long-term survival benefit for a<br />
substantial proportion <strong>of</strong> patients. Beliefs regarding which<br />
patients with hepatic metastases are candidates for resection<br />
or other surgical treatment must be reevaluated<br />
through careful examination <strong>of</strong> extant data and also by<br />
considering the effect <strong>of</strong> improved diagnostic imaging techniques,<br />
new cytotoxic and targeted molecular therapies, and<br />
improved patient selection and treatment criteria.<br />
Multiple Liver Metastases<br />
Recent studies have demonstrated that resection for four<br />
or more hepatic metastases from CRC can yield long-term<br />
survival for some patients. For example, Minagawa and<br />
colleagues 1 reported that although survival was improved<br />
for patients with a solitary metastasis compared with those<br />
with multiple metastases, 10-year survival rates in excess <strong>of</strong><br />
20% were achieved for people who had resection <strong>of</strong> as many<br />
as 20 lesions. This study is the only report, to our knowledge,<br />
<strong>of</strong> a 10-year survival rate after resection <strong>of</strong> four or more CRC<br />
liver metastases. Our group has examined the prospective<br />
database at The University <strong>of</strong> Texas M. D. Anderson Cancer<br />
Center, which includes more than 1,600 patients who had<br />
resection <strong>of</strong> CRC liver metastases since 1995. Among 159<br />
patients with four to 12 liver metastases that were treated<br />
with surgery (resection or resection plus radi<strong>of</strong>requency<br />
ablation [RFA]), the 5-year disease-free survival rate was<br />
22% and the 5-year overall survival rate was 51%. 2 The<br />
median overall survival for these 159 patients was 62.1<br />
months. The favorable long-term survival data in our study<br />
relate to the fact that these patients were carefully selected.<br />
No patient had radiographic or intraoperative evidence <strong>of</strong><br />
extrahepatic disease at the time <strong>of</strong> surgical treatment <strong>of</strong> the<br />
hepatic metastases, most (89.9%) received neoadjuvant chemotherapy,<br />
most (72.7%) had reduction in tumor volume<br />
after preoperative chemotherapy, and all patients underwent<br />
thorough intraoperative ultrasonography to detect<br />
additional small lesions not evident on preoperative imaging<br />
studies. On multivariate analysis <strong>of</strong> predictive factors, only<br />
response to neoadjuvant chemotherapy remained an independent<br />
predictor <strong>of</strong> overall survival. Our data suggest that<br />
By Steven A. Curley, MD<br />
chemotherapy or implanted hepatic arterial infusion pumps.<br />
The advent <strong>of</strong> the new millennium was associated with the<br />
availability <strong>of</strong> several new cytotoxic and biologic agents active<br />
in colorectal cancer. These agents have completely changed<br />
the approach to the treatment <strong>of</strong> patients with colorectal<br />
cancer liver metastases and thus have increased the complexity<br />
<strong>of</strong> the decision-making process for treatment <strong>of</strong> these<br />
patients.<br />
tumor biology (response to cytotoxic therapy) rather than<br />
morphologic criteria (tumor number or size) determines<br />
long-term prognosis.<br />
Unresectable Liver Metastases or<br />
Extrahepatic Disease<br />
The presence <strong>of</strong> extrahepatic disease has clearly been<br />
demonstrated to be a negative prognostic factor in retrospective<br />
studies <strong>of</strong> CRC liver metastases treated with surgery<br />
alone. 3-6 The surgical dogma to not perform resection in<br />
these patients has been challenged in the modern era <strong>of</strong><br />
multimodality management for CRC metastatic to the liver.<br />
Adam and colleagues 7 reported a single-institution experience<br />
<strong>of</strong> 1,439 consecutive patients with CRC liver metastases<br />
at the time <strong>of</strong> diagnosis. At initial presentation, 1,104<br />
patients (77%) were deemed to have unresectable disease<br />
and were treated by chemotherapy, whereas 335 (23%) had<br />
resectable disease and underwent surgical treatment. There<br />
were two interesting findings from this study. First, among<br />
the 1,104 patients with nonresectable disease, 138 (12.5%)<br />
had a substantial reduction in the tumor volume with<br />
chemotherapy and became candidates for a potentially curative<br />
surgical procedure. These 138 patients were treated<br />
with either resection alone or resection combined with<br />
thermal ablation. The overall 5- and 10-year survival rates<br />
for these 138 patients were 33% and 23%, respectively, with<br />
disease-free survival rates at the same intervals <strong>of</strong> 22% and<br />
17%, respectively. The second important finding was that<br />
extrahepatic tumor was present at the time <strong>of</strong> surgical<br />
treatment in 52 (38%) <strong>of</strong> these 138 patients. All <strong>of</strong> these<br />
patients underwent surgical treatment <strong>of</strong> the liver metastases<br />
along with resection <strong>of</strong> lung, lymph node, peritoneal, or<br />
other metastatic disease. When it was possible to treat all<br />
metastatic disease surgically, there was no marked decrease<br />
in the long-term survival probability compared with that for<br />
patients who underwent treatment <strong>of</strong> liver metastases<br />
alone.<br />
From the Department <strong>of</strong> Surgical <strong>Oncology</strong>, University <strong>of</strong> Texas M. D. Anderson Cancer<br />
Center, Houston, TX.<br />
Author’s disclosure <strong>of</strong> potential conflicts <strong>of</strong> interest are found at the end <strong>of</strong> this article.<br />
Address reprint requests to: Steven A. Curley, MD, Division <strong>of</strong> Surgery, University <strong>of</strong><br />
Texas M. D. Anderson Cancer Center, 1400 Pressler St., Unit 1447, Houston, TX 77030;<br />
email: scurley@mdanderson.org.<br />
© <strong>2012</strong> by <strong>American</strong> <strong>Society</strong> <strong>of</strong> <strong>Clinical</strong> <strong>Oncology</strong>.<br />
1092-9118/10/1-10<br />
209
Resection and Thermal Ablation <strong>of</strong> Liver Metastases<br />
We have reported the results <strong>of</strong> a prospective study <strong>of</strong><br />
outcomes after surgical treatment <strong>of</strong> CRC liver metastases<br />
using hepatic resection only, RFA plus resection for multiple<br />
tumors, or RFA only when patients had tumor in an unresectable<br />
site in the liver. The study group consisted <strong>of</strong> 348<br />
consecutive patients with CRC liver metastases and no<br />
extrahepatic disease who were treated for cure with hepatic<br />
resection with or without RFA and 70 patients who were<br />
found at laparotomy to have liver-only disease (but were not<br />
candidates for curative treatment based on involvement <strong>of</strong><br />
too many liver segments) as outlined in Fig. 1. 8 An example<br />
<strong>of</strong> a case in which a combined approach <strong>of</strong> resection and<br />
ablation was used is illustrated in Fig. 2.<br />
The 5-year overall survival rate among the 348 patients<br />
treated with curative intent was 44%. Survival with resection<br />
alone (58% at 5 years) was significantly greater than<br />
that with either resection plus ablation (28%) or RFA<br />
alone (19%, p � 0.001), with no marked difference among<br />
the approaches, including RFA. The overall 4-year survival<br />
rates after resection, RFA plus resection, and RFA<br />
KEY POINTS<br />
● Management <strong>of</strong> colorectal cancer liver metastases<br />
requires a multidisciplinary team approach.<br />
● Patients should be considered candidates for resection<br />
whenever possible, regardless <strong>of</strong> the number <strong>of</strong><br />
metastatic lesions.<br />
● Modern systemic chemotherapy can produce response<br />
sufficient to convert unresectable liver metastases<br />
to resectable disease in some patients.<br />
● Chemotherapy can be hepatotoxic, so careful selection<br />
<strong>of</strong> neoadjuvant therapy agents and duration<br />
must be considered.<br />
● Resection provides better long-term survival probability<br />
compared with thermal tumor ablation<br />
techniques.<br />
210<br />
alone were 65%, 36%, and 22%, respectively (p � 0.0001);<br />
however, 4-year survival rates with resection plus RFA<br />
and RFA only were still substantial and significantly improved<br />
over that achieved with chemotherapy alone (5%) in<br />
the 70 patients not treated with resection or RFA (p �<br />
0.0017).<br />
Systemic Chemotherapy<br />
STEVEN A. CURLEY<br />
Fig. 1. Treatment <strong>of</strong> 418 consecutive<br />
patients with liver-only CRC metastases<br />
at M. D. Anderson Cancer Center. A total<br />
<strong>of</strong> 348 patients (83.3%) were treated<br />
for cure with hepatic resection only, RFA<br />
plus resection, or RFA only. Seventy patients<br />
were found to have disease too<br />
extensive for curative therapy and underwent<br />
chemotherapy (systemic, intraarterial<br />
chemotherapy via a hepatic<br />
artery infusion pump placed at the index<br />
laparotomy or intra-arterial plus<br />
systemic chemotherapy).<br />
Abbreviations: CRC, colorectal cancer;<br />
RFA, radi<strong>of</strong>requency ablation.<br />
Systemic chemotherapy can convert unresectable CRC<br />
liver metastases to resectable disease in some patients, and<br />
response to neoadjuvant chemotherapy in patients with<br />
resectable disease is an important prognostic indicator <strong>of</strong><br />
improved probability <strong>of</strong> long-term survival. 2,9,10 The use <strong>of</strong><br />
neoadjuvant chemotherapy in patients with resectable CRC<br />
liver metastases must be carefully considered in multidisciplinary<br />
treatment planning involving medical oncologists<br />
and hepatobiliary surgical oncologists. Neoadjuvant chemotherapy<br />
regimens that are based on oxaliplatin (Eloxatin;<br />
San<strong>of</strong>i Aventis) or irinotecan (Camptosar; Pfizer) can produce<br />
specific types <strong>of</strong> liver injury. Oxaliplatin-based chemotherapy<br />
was reported to cause sinusoidal obstruction, venoocclusive<br />
lesions in the microvasculature <strong>of</strong> nontumoral<br />
liver, and perisinusoidal fibrosis in more than half <strong>of</strong> the<br />
patients receiving neoadjuvant therapy. 11 Both irinotecan<br />
and oxaliplatin treatment have been found to produce nonalcoholic<br />
steatohepatitis (NASH), and the severity <strong>of</strong> NASH<br />
is greater in patients who are also obese. 12 Hepatic steatosis<br />
or steatohepatitis was identified as an independent variable<br />
that predicted increased perioperative morbidity and mortality<br />
rates in a retrospective study <strong>of</strong> 135 patients undergoing<br />
resection <strong>of</strong> CRC liver metastases. 13 Data suggest that<br />
steatosis impairs hepatic regeneration after resection, and<br />
the rates <strong>of</strong> liver insufficiency and hyperbilirubinemia are<br />
higher for patients with hepatic steatosis than for patients<br />
with normal livers.<br />
A multi-institution review <strong>of</strong> 406 patients who underwent<br />
resection <strong>of</strong> CRC liver metastases and had the severity <strong>of</strong><br />
NASH and nonalcoholic fatty liver disease scored on a<br />
standardized system was completed recently. 14,15 Preoperative<br />
chemotherapy was administered to 241 patients (61%),<br />
and 158 patients (39%) received no neoadjuvant treatment.
SURGICAL TREATMENT OF COLORECTAL LIVER METASTASES<br />
Fig. 2. The patient had a large tumor in the left<br />
lateral segment (image on the left); also visible is one<br />
<strong>of</strong> four smaller tumors in the right lobe. The large<br />
tumor was resected, as was one <strong>of</strong> the smaller tumors<br />
near the dome (image on the right). The other<br />
tumors were ablated, with ablation including a surrounding<br />
zone <strong>of</strong> the parenchyma to ensure a negative<br />
margin.<br />
Patients who received oxaliplatin were found to have sinusoidal<br />
dilation at a much higher frequency compared with<br />
patients who did not receive chemotherapy (p � 0.001). In<br />
contrast, the incidence <strong>of</strong> NASH was approximately 20%<br />
among patients who received irinotecan, which was significantly<br />
greater than the 4% rate among patients who received<br />
no chemotherapy (p � 0.001). The most striking<br />
finding <strong>of</strong> this study was that only NASH was associated<br />
with an increased 90-day mortality rate (14.7%) after liver<br />
resection, whereas other types <strong>of</strong> liver injury or normal liver<br />
were associated with a postoperative mortality rate <strong>of</strong> 1.6%<br />
(p � 0.001). Clearly, careful consideration must be given to<br />
the type and duration <strong>of</strong> neoadjuvant chemotherapy to be<br />
delivered to patients. In patients who have preoperative<br />
imaging or intraoperative evidence <strong>of</strong> either nonalcoholic<br />
fatty liver disease or NASH, consideration should be given to<br />
obtaining a core liver biopsy to assess the severity <strong>of</strong> steatohepatitis.<br />
The presence <strong>of</strong> severe NASH should lead the<br />
surgeon to question the safety <strong>of</strong> a major liver resection and<br />
lower the threshold to consider preoperative portal vein<br />
embolization in an attempt to reduce postoperative morbidity<br />
and liver failure rates.<br />
A subset <strong>of</strong> patients have borderline resectable or unresectable<br />
liver metastases that progress during first-line<br />
systemic chemotherapy. In addition, some patients who<br />
receive neoadjuvant chemotherapy before a planned liver<br />
resection have disease progression demonstrated on subsequent<br />
imaging examinations. Frequently, these patients<br />
will receive second-line chemotherapy. We recently evaluated<br />
60 patients who underwent resection <strong>of</strong> CRC liver<br />
metastases after receiving two or more different systemic<br />
chemotherapy regimens. 16 These patients tended to have<br />
more advanced CRC liver metastases as evidenced by a<br />
mean � SD number <strong>of</strong> tumors <strong>of</strong> 4 � 3.5 and a mean � SD<br />
maximum size <strong>of</strong> 5 � 3.2 cm. All the patients received<br />
irinotecan or oxaliplatin and frequently had been treated<br />
sequentially with regimens that contained both <strong>of</strong> these<br />
agents. Despite receiving a mean � SD <strong>of</strong> 17 � 8 cycles<br />
<strong>of</strong> chemotherapy, the postoperative morbidity rate was<br />
33%, and the 90-day mortality rate was 3%. The 1-, 3-, and<br />
5-year overall survival rates were 83%, 41%, and 22%,<br />
respectively. Thus, long-term survival after resection <strong>of</strong> CRC<br />
liver metastases in patients who have received second-line<br />
chemotherapy can produce a modest rate <strong>of</strong> long-term survival.<br />
Synchronous CRC Liver Metastases<br />
As many as 25% <strong>of</strong> patients with primary CRC will<br />
present with synchronous liver metastases. Historically,<br />
synchronous liver metastases from CRC has portended a<br />
poor prognosis. The traditional strategy for managing<br />
these patients was resection <strong>of</strong> the primary tumor followed<br />
by systemic chemotherapy and then consideration <strong>of</strong> liver<br />
resection if disease was considered to be resectable. This<br />
strategy arose in the 20th century before the advent <strong>of</strong><br />
effective modern cytotoxic and biologic agents for advanced<br />
CRC. Our group recently evaluated 156 consecutive patients<br />
with synchronous resectable liver metastases and<br />
intact primary tumor from CRC. 17 The patients in this<br />
series had nonobstructing primary tumors and received<br />
preoperative systemic chemotherapy. A subset <strong>of</strong> these<br />
patients had a notable antitumor response in the primary<br />
tumor, and the decision was made to perform a reverse<br />
strategy by resecting the liver metastases before surgical<br />
treatment <strong>of</strong> the primary tumor. When considering the<br />
entire group, this reverse approach was used in 19% <strong>of</strong><br />
the patients, whereas the remaining patients underwent<br />
either a classic approach <strong>of</strong> resection <strong>of</strong> the primary<br />
tumor followed by later resection <strong>of</strong> the CRC liver metastases<br />
or a combined resection <strong>of</strong> the primary tumor and<br />
liver metastases in one operation. The median number <strong>of</strong><br />
liver metastases was considerably higher in the reverse<br />
strategy group, but no substantial difference was found<br />
in postoperative morbidity or mortality rates when the<br />
reverse strategy was compared with either the classic or<br />
combined surgical approach groups. The 5-year survival<br />
rates for the classic, combined, and reverse strategy approach<br />
groups were 48%, 55%, and 39%, respectively. These<br />
data further underscore the importance <strong>of</strong> multidisciplinary<br />
team planning and management <strong>of</strong> patients with CRC liver<br />
metastases.<br />
Conclusion<br />
The criterion standard in the treatment <strong>of</strong> CRC liver<br />
metastases remains complete resection. Complete resection<br />
can be performed safely with low morbidity and mortality<br />
rates. Local hepatic tumor ablation techniques have a role in<br />
the treatment <strong>of</strong> patients with unresectable disease. However,<br />
it needs to be emphasized that no data currently exist<br />
to suggest that either RFA or RFA combined with resection<br />
provides survival comparable to that with complete resection<br />
and that long-term survival rates for patients treated<br />
with thermal ablation techniques are still being established.<br />
Aggressive surgical management combined with neoadjuvant<br />
and adjuvant chemotherapy regimens has yielded<br />
5-year survival rates that exceed 50%. All patients with CRC<br />
liver metastases should be evaluated by a multidisciplinary<br />
team consisting <strong>of</strong> hepatobiliary surgical oncologists, medical<br />
oncologists, pathologists, and diagnostic imaging specialists<br />
to optimize the management <strong>of</strong> patients with stage IV<br />
disease.<br />
211
Author’s Disclosures <strong>of</strong> Potential Conflicts <strong>of</strong> Interest<br />
Author<br />
Steven A. Curley*<br />
*No relevant relationships to disclose.<br />
Employment or<br />
Leadership<br />
Positions<br />
Consultant or<br />
Advisory Role<br />
1. Minagawa M, Makuuchi M, Torzilli G, et al. Extension <strong>of</strong> the frontiers <strong>of</strong><br />
surgical indications in the treatment <strong>of</strong> liver metastases from colorectal<br />
cancer: long-term results. Ann Surg. 2000;231:487-499.<br />
2. Pawlik TM, Abdalla EK, Ellis LM, et al. Debunking dogma: surgery for<br />
four or more colorectal liver metastases is justified. J Gastrointest Surg.<br />
2006;10:240-248.<br />
3. Hughes KS, Simon R, Songhorabodi S, et al. Resection <strong>of</strong> the liver for<br />
colorectal carcinoma metastases: a multi-institutional study <strong>of</strong> patterns <strong>of</strong><br />
recurrence. Surgery. 1986;100:278-284.<br />
4. Nordlinger B, Guiguet M, Vaillant JC, et al. Surgical resection <strong>of</strong><br />
colorectal carcinoma metastases to the liver. A prognostic scoring system to<br />
improve case selection, based on 1568 patients. Association Française de<br />
Chirurgie. Cancer. 1996;77:1254-1262.<br />
5. Fong Y, Fortner J, Sun RL, et al. <strong>Clinical</strong> score for predicting recurrence<br />
after hepatic resection for metastatic colorectal cancer: analysis <strong>of</strong> 1001<br />
consecutive cases. Ann Surg. 1999;230:309-318; discussion 318-321.<br />
6. Scheele J, Altendorf-H<strong>of</strong>mann A, Grube T, et al. Resection <strong>of</strong> colorectal<br />
liver metastases. What prognostic factors determine patient selection? [in<br />
German]. Chirurg. 2001;72:547-560.<br />
7. Adam R, Delvart V, Pascal G, et al. Rescue surgery for unresectable<br />
colorectal liver metastases downstaged by chemotherapy: a model to predict<br />
long-term survival. Ann Surg. 2004;240:644-657; discussion 657-658.<br />
8. Abdalla EK, Vauthey JN, Ellis LM, et al. Recurrence and outcomes<br />
following hepatic resection, radi<strong>of</strong>requency ablation, and combined resection/<br />
ablation for colorectal liver metastases. Ann Surg. 2004;239:818-825.<br />
9. Tanaka K, Adam R, Shimada H, et al. Role <strong>of</strong> neoadjuvant chemotherapy<br />
212<br />
Stock<br />
Ownership Honoraria<br />
REFERENCES<br />
Research<br />
Funding<br />
Expert<br />
Testimony<br />
STEVEN A. CURLEY<br />
Other<br />
Remuneration<br />
in the treatment <strong>of</strong> multiple colorectal metastases to the liver. Br J Surg.<br />
2003;90:963-969.<br />
10. Allen PJ, Kemeny N, Jarnagin W, et al. Importance <strong>of</strong> response to<br />
neoadjuvant chemotherapy in patients undergoing resection <strong>of</strong> synchronous<br />
colorectal liver metastases. J Gastrointest Surg. 2003;7:109-115<br />
11. Rubbia-Brandt L, Audard V, Sartoretti P, et al. Severe hepatic sinusoidal<br />
obstruction associated with oxaliplatin-based chemotherapy in patients<br />
with metastatic colorectal cancer. Ann Oncol. 2004;15:460-466.<br />
12. Fernandez FG, Ritter J, Goodwin JW, et al. Effect <strong>of</strong> steatohepatitis<br />
associated with irinotecan or oxaliplatin pretreatment on resectability <strong>of</strong><br />
hepatic colorectal metastases. J Am Coll Surg. 2005;200:845-853.<br />
13. Behrns KE, Tsiotos GG, DeSouza NF, et al. Hepatic steatosis as a<br />
potential risk factor for major hepatic resection. J Gastrointest Surg. 1998;2:<br />
292-298.<br />
14. Kleiner DE, Brunt EM, Van Natta M, et al. Design and validation <strong>of</strong> a<br />
histological scoring system for nonalcoholic fatty liver disease. Hepatology.<br />
2005;41:1313-1321.<br />
15. Vauthey JN, Pawlik TM, Ribero D, et al. Chemotherapy regimen<br />
predicts steatohepatitis and an increase in 90-day mortality after surgery for<br />
hepatic colorectal metastases. J Clin Oncol. 2006;24:2065-2072.<br />
16. Brouquet A, Overman MJ, Kopetz S. Is resection <strong>of</strong> colorectal liver<br />
metastases after a second-line chemotherapy regimen justified? Cancer.<br />
2011;117:4484-4492.<br />
17. Brouquet A, Mortenson MM, Vauthey J-N. Surgical strategies for<br />
synchronous colorectal liver metastases in 156 consecutive patients: classic,<br />
combined or reverse strategy? J Am Coll Surg. 2010;201:934-941.
RECTAL CANCER: NEW PARADIGMS BEYOND<br />
STANDARD CHEMOTHERAPY AND RADIATION<br />
CHAIR<br />
David P. Ryan, MD<br />
Massachusetts General Hospital Cancer Center<br />
Boston, MA<br />
SPEAKERS<br />
Martin R. Weiser, MD<br />
Memorial Sloan-Kettering Cancer Center<br />
New York, NY<br />
Karyn A. Goodman, MD<br />
Memorial Sloan-Kettering Cancer Center<br />
New York, NY
Minimally Invasive Surgery <strong>of</strong> Rectal Cancer:<br />
Current Evidence and Options<br />
By Atthaphorn Trakarnsanga, MD, and Martin R. Weiser, MD<br />
Overview: Minimally invasive surgery (MIS) <strong>of</strong> colorectal<br />
cancer has become more popular in the past two decades.<br />
Laparoscopic colectomy has been accepted as an alternative<br />
standard approach in colon cancer, with comparable oncologic<br />
outcomes and several better short-term outcomes com-<br />
MIS HAS become the standard procedure in some<br />
operations, such as cholecystectomy and appendectomy.<br />
Several short-term benefits are associated with MIS,<br />
including reduced postoperative pain and decreased need for<br />
analgesic drugs, more cosmetically pleasing incisions,<br />
shorter length <strong>of</strong> hospital stay, and earlier return to functionality.<br />
These are the major reasons for the increasing<br />
popularity <strong>of</strong> the minimally invasive surgical approach.<br />
Jacobs and colleagues 1 first reported a case series <strong>of</strong><br />
successful laparoscopic colectomies in 1990. In the beginning,<br />
its application was limited to treatment <strong>of</strong> benign<br />
lesions. However, the use <strong>of</strong> MIS in colon cancer raised<br />
concerns about the possibility <strong>of</strong> port site metastases and<br />
inadequate oncologic resection. Early reports indicated that<br />
the incidence <strong>of</strong> port site metastasis was approximately 4%<br />
in the laparoscopic group compared with 1% in the open<br />
group. 2,3 This contentious question was resolved by a large<br />
meta-analysis <strong>of</strong> several controlled studies, which concluded<br />
that the incidence <strong>of</strong> port site metastasis was 0.87% in the<br />
laparoscopic group compared with 0.34% in the open group<br />
(p � 0.16). 4 Correspondingly, concerns regarding oncologic<br />
outcome were dispelled by several well-designed large, randomized,<br />
controlled studies that demonstrated the equivalence<br />
<strong>of</strong> laparoscopic colectomy to the conventional open<br />
approach in terms <strong>of</strong> recurrence and survival. 5-8 Undoubtedly,<br />
laparoscopic colectomy is now considered an alternative<br />
standard approach in the treatment <strong>of</strong> colon cancer.<br />
Obviously, surgery <strong>of</strong> the rectum is more challenging for<br />
several reasons, mostly because dissection <strong>of</strong> the mesorectal<br />
plane is limited by the confines <strong>of</strong> the bony pelvis and the<br />
goal <strong>of</strong> preserving the autonomic nerves. Moreover, neoadjuvant<br />
treatment sequelae, especially those secondary to<br />
radiotherapy, may affect the surgical field. Thus, the role <strong>of</strong><br />
MIS for rectal cancer is still controversial and substantial<br />
evidence is lacking. We summarize the current status <strong>of</strong> MIS<br />
in rectal cancer, the findings <strong>of</strong> contemporary published<br />
studies, and the appropriate application <strong>of</strong> various existing<br />
techniques.<br />
Laparoscopic Rectal Cancer Surgery:<br />
Where Are We Now?<br />
At present, the <strong>American</strong> <strong>Society</strong> <strong>of</strong> Colon and Rectal<br />
Surgeons (ASCRS) has not endorsed laparoscopic proctectomy<br />
for rectal cancer because <strong>of</strong> concerns about achieving<br />
adequate mesorectal excision and clear surgical margins<br />
using this technique. 9 ASCRS has encouraged well-designed<br />
trials to examine the safety, efficacy, and benefits <strong>of</strong> MIS in<br />
rectal cancer surgery, especially in regard to long-term<br />
oncologic outcomes. A majority <strong>of</strong> these prospective randomized,<br />
controlled trials are ongoing, and the results have not<br />
yet been defined. Thus, MIS for rectal cancer has not become<br />
214<br />
pared to open surgery. Unlike the treatment for colon cancer,<br />
however, the minimally invasive approach in rectal cancer<br />
has not been established. In this article, we summarize the<br />
current status <strong>of</strong> MIS for rectal cancer and explore the various<br />
technical options.<br />
standard treatment in the United States. 10 Based on evidence<br />
presented in European trials, however, several countries<br />
in Europe and Asia have endorsed MIS as an<br />
alternative to standard open surgery in rectal cancer.<br />
United Kingdom MRC CLASICC Trial<br />
The United Kingdom Medical Research Council Trial <strong>of</strong><br />
Conventional vs. Laparoscopic-Assisted Surgery in Colorectal<br />
Cancer (United Kingdom MRC CLASICC) was a prospective<br />
randomized trial that included both patients with colon<br />
cancer and patients with rectal cancer. 7 Of the 794 enrolled<br />
patients, 381 had rectal cancer and the conversion rate was<br />
34%. Within the actual treatment group, 87 patients underwent<br />
open total mesorectal excision (TME) and 189 had<br />
laparoscopic-assisted TME. The primary endpoints were<br />
rate <strong>of</strong> positive-circumferential and longitudinally resected<br />
margins. No significant difference in either margin <strong>of</strong> resection<br />
was identified when comparing the two procedures.<br />
Among patients undergoing anterior resection, the rate <strong>of</strong><br />
positive circumferential resection (CRM) was slightly higher<br />
in the laparoscopic than in the open group (12% vs. 6%; p �<br />
0.80). In-hospital mortality rates between laparoscopic and<br />
open surgery were not significantly different (4% vs. 5%; p �<br />
0.57), but the mortality rate was higher in patients who<br />
were converted to open surgery.<br />
Regarding long-term outcomes, 5-year results <strong>of</strong> the<br />
United Kingdom MRC CLASICC have already been published.<br />
11 There was no significant difference in 5-year overall<br />
survival (OS) between the laparoscopic and open groups<br />
(60.3% vs. 52.9%; p � 0.132). Five-year disease-free survival<br />
did not differ either (53.2% vs. 52.1%, respectively; p �<br />
0.953). In converted patients, the overall survival rate was<br />
significantly worse, compared to the other patients who<br />
initially received randomized treatment (p � 0.05). There<br />
was no difference in rate <strong>of</strong> local recurrence between patients<br />
who had laparoscopic compared with open anterior<br />
resection, although as previously noted, the rate <strong>of</strong> positive<br />
CRM was slightly higher in the laparoscopic group (9.4% vs.<br />
7.6%; p � 0.74). No data was shown on local recurrence in<br />
patients who had abdominoperineal resection.<br />
From the Department <strong>of</strong> Surgery, Faculty <strong>of</strong> Medicine Siriraj Hospital, Mahidol University,<br />
Bangkok, Thailand; Department <strong>of</strong> Surgery, Memorial Sloan-Kettering Cancer Center,<br />
New York, NY.<br />
Authors’ disclosures <strong>of</strong> potential conflicts <strong>of</strong> interest are found at the end <strong>of</strong> this article.<br />
Address reprint requests to Martin R. Weiser, MD, Department <strong>of</strong> Surgery, Memorial<br />
Sloan-Kettering Cancer Center, 1275 York Ave., Room C-1075, New York, NY, 10065; email:<br />
weiser1@mskcc.org.<br />
© <strong>2012</strong> by <strong>American</strong> <strong>Society</strong> <strong>of</strong> <strong>Clinical</strong> <strong>Oncology</strong>.<br />
1092-9118/10/1-10
MINIMALLY INVASIVE SURGERY OF RECTAL CANCER<br />
An important finding <strong>of</strong> this study was the high rate <strong>of</strong><br />
conversion from laparoscopic to open surgery intraoperatively<br />
(34%), and the poorer outcomes in patients who were<br />
converted. The reasons for intraoperative conversion included<br />
tumor fixation, patient obesity, technical difficulty,<br />
practitioner uncertainty regarding adequate tumor clearance,<br />
and uncertainty regarding anatomic landmarks.<br />
COREAN Trial<br />
The Comparison <strong>of</strong> Open vs. laparoscopic surgery for mid<br />
and low REctal cancer After Neoadjuvant chemoradiotherapy<br />
(COREAN) trial is a multicenter randomized, controlled<br />
trial comparing open with laparoscopic surgery in Korean<br />
patients with mid- or low-rectal cancer (i.e., tumor within 9<br />
cm <strong>of</strong> the anal verge) following preoperative chemoradiotherapy.<br />
12 The primary endpoint was 3-year disease-free<br />
survival. Seven high-volume surgeons were qualified to<br />
enroll in the study. Three hundred and forty patients with<br />
T3/N0–2 rectal cancer were randomly assigned to laparoscopic<br />
and open groups in a ratio <strong>of</strong> 1:1. The conversion rate<br />
was 1.2% (2 in 170 patients), secondary to difficult dissection<br />
in the narrow pelvis and/or bleeding. The rates <strong>of</strong> coloanal<br />
anastomosis and abdominoperineal resection were equivalent<br />
between the laparoscopic and open surgery groups<br />
(19.4% vs. 19.4% and 11.2% vs. 14.1%, respectively; p �<br />
0.7085). Operative time was significantly longer in the<br />
laparoscopic group (244.9 min vs. 197 min, p � 0.006), but<br />
estimated blood loss was significantly less (217.5 mL vs. 200<br />
mL, p � 0.0001). The rate <strong>of</strong> positive CRM did not differ<br />
significantly between the two groups (2.9% laparoscopic vs.<br />
4.1% open; p � 0.770), nor did distal resection margin or<br />
pathologic stage. The short-term benefits <strong>of</strong> laparoscopic<br />
surgery were notable: speedier return <strong>of</strong> bowel function, less<br />
postoperative pain, and less need for/use <strong>of</strong> analgesic drugs.<br />
KEY POINTS<br />
● Minimally invasive colon surgery for cancer has become<br />
standard.<br />
● Rectal surgery is more challenging than colon surgery<br />
because <strong>of</strong> anatomic limitation <strong>of</strong> the pelvis and<br />
need to preserve autonomic nerves while maintaining<br />
a clear margin.<br />
● Minimally invasive rectal cancer surgery has proven<br />
feasible.<br />
● The UK MRC CLASICC trial demonstrated equivalent<br />
long-term recurrence and survival results between<br />
laparoscopic and open rectal surgery, and the<br />
results <strong>of</strong> several ongoing trials (the COREAN,<br />
COLOR II, JCOG 0404, and ACOSOG-Z6051 trials)<br />
are awaited.<br />
● Greater degree <strong>of</strong> rotation and instrument movement<br />
along with three-dimensional optics has led to the use<br />
<strong>of</strong> robotic technology in rectal cancer surgery. The UK<br />
MRC ROLARR trial is currently comparing laparoscopic<br />
to robotic resection.<br />
● The optimal learning curve and method to train<br />
surgeons in minimally invasive surgery is still in flux<br />
and include surgeon mentoring and didactic sessions.<br />
The COREAN trial is the first study to compare laparoscopic<br />
with open TME in patients who have undergone<br />
neoadjuvant chemoradiotherapy. The promising short-term<br />
outcomes <strong>of</strong> laparoscopic surgery for rectal cancer include a<br />
low rate <strong>of</strong> conversion and small number <strong>of</strong> positive CRMs.<br />
Both may be the result <strong>of</strong> neoadjuvant treatment, combined<br />
with optimal resection technique employed by experienced<br />
surgeons. Long-term follow-up data regarding recurrence<br />
and survival are awaited, and will be important factors in<br />
this large trial.<br />
COLOR II Trial<br />
The European Colon Cancer Laparoscopic or Open Resection<br />
(COLOR) II trial is an international randomized, multicenter<br />
study comparing the outcomes <strong>of</strong> curative<br />
laparoscopic and conventional open surgery for rectal cancer.<br />
13 The primary endpoint is 3-year locoregional recurrence.<br />
Secondary endpoints are recurrence-free and OS at 3,<br />
5, and 7 years; rate <strong>of</strong> port- and wound-site metastases;<br />
distant metastases; morbidity and mortality within 8 weeks<br />
following resection; macroscopic evaluation <strong>of</strong> the resected<br />
specimen; quality <strong>of</strong> life; and cost. Currently, the final<br />
results have been obtained but are not yet reported.<br />
JCOG 0404 Trial<br />
The Japanese <strong>Clinical</strong> <strong>Oncology</strong> Group Study (JCOG<br />
0404) is a multicenter randomized, controlled trial comparing<br />
laparoscopic and open surgery in Japanese patients with<br />
colon or rectal cancer. 14 This study was activated in October<br />
2004. The planned sample size is 818 cases; 409 cases per<br />
arm with 5 years <strong>of</strong> follow-up after 3 years <strong>of</strong> accrual. The<br />
primary endpoint is OS. Secondary endpoints are relapsefree<br />
survival, short-term clinical outcomes, adverse events,<br />
and rates <strong>of</strong> conversion.<br />
ACOSOG-Z6051 Trial<br />
The <strong>American</strong> College <strong>of</strong> Surgeons <strong>Oncology</strong> Group<br />
(ACOSOG)-Z6051 trial is a prospective, randomized trial<br />
aimed at evaluating the noninferiority <strong>of</strong> laparoscopic<br />
compared with open surgery for rectal cancer. 15 Primary<br />
measurements include completeness <strong>of</strong> TME and circumferential<br />
and distal resection margins. The secondary endpoints<br />
are short-term benefits; disease-free and overall<br />
survival; pelvic recurrence rates; and quality <strong>of</strong> life, including<br />
sexual and bowel function. The results are awaited.<br />
Meta-analysis<br />
The data on laparoscopic resection for mid- and low-rectal<br />
cancers are limited in predominantly retrospective series,<br />
prospective nonrandomized, controlled trials, and small randomized,<br />
controlled studies. As predicted, the conclusion<br />
from all meta-analyses is that patients in the laparoscopic<br />
group have better short-term outcomes than patients in the<br />
open surgery group. This is similar to the results <strong>of</strong> laparoscopic<br />
procedures in other diseases. 16-19 In terms <strong>of</strong> recurrence<br />
and survival, one meta-analysis (including a 5-year<br />
follow-up <strong>of</strong> the MRC CLASICC trial) evaluated 2,095 patients<br />
with rectal cancer (1,096 undergoing laparoscopic; 999<br />
undergoing open surgery) from 12 randomized, controlled<br />
studies. 20 Less blood loss, a more rapid return to oral diet,<br />
and briefer hospital stay were identified as short-term<br />
benefits <strong>of</strong> laparoscopic versus open surgery. Long-term<br />
215
outcomes; local, wound-site, and distant recurrences; 5-year<br />
overall and disease-free survival; and urinary and sexual<br />
function did not differ significantly between the two groups.<br />
Summary <strong>of</strong> Current Evidence<br />
The United Kingdom MRC CLASICC trial finding <strong>of</strong><br />
increased CRM positivity within the laparoscopic anterior<br />
resection group raised questions about the oncologic competence<br />
<strong>of</strong> this procedure. The question <strong>of</strong> adequacy is answered<br />
by the 5-year follow-up results. The COREAN trial<br />
concludes that laparoscopic surgery for rectal cancer, performed<br />
by skillful laparoscopic surgeons in high-volume<br />
centers, is feasible and has promising short-term results.<br />
As we await the long-term results <strong>of</strong> the aforementioned<br />
ongoing trials (the COREAN, COLOR II, JCOG 0404, and<br />
ACOSOG-Z6051 trials), one topic must be considered before<br />
applying the results to general practice: how to prepare and<br />
qualify operators so that patients receive the maximum<br />
benefits <strong>of</strong> MIS.<br />
Learning Curve for Laparoscopic Surgery for Rectal<br />
Cancer: Is There a Magic Number?<br />
As previously noted, rectal resection is more demanding<br />
than colectomy because <strong>of</strong> the anatomy <strong>of</strong> the pelvis, the<br />
desirability <strong>of</strong> preserving autonomic nerve function, and the<br />
sequelae <strong>of</strong> neoadjuvant treatment. In conventional open<br />
surgery, the retractor is a key factor in achieving adequate<br />
exposure, especially in male patients with a bulky mesorectal<br />
tumor and narrow pelvis. Appropriate TME should be<br />
performed by sharp (not blunt) dissection to achieve the best<br />
oncologic outcome. All <strong>of</strong> these considerations make laparoscopic<br />
rectal resection more challenging.<br />
One study from Japan analyzed the learning curve for<br />
laparoscopic low anterior resection. 21 Single surgeons performed<br />
250 operations, with patients divided into five<br />
groups. The learning curve analysis demonstrated that<br />
operative time stabilized after 50 cases. The conversion rate<br />
was significantly lower after 150 cases (p � 0.05), correlating<br />
with male sex and advanced T stage. Other studies have<br />
suggested that an adequate learning curve is passed at<br />
somewhere between 20 and 60 cases. 6,7,22,23 In laparoscopic<br />
TME, adequacy <strong>of</strong> exposure without use <strong>of</strong> the retractor is<br />
necessary. Therefore, not only the experience <strong>of</strong> the surgeon<br />
but the aptitude <strong>of</strong> the entire surgical team is important.<br />
The mirror image from the camera and the alignment <strong>of</strong> the<br />
first assistant’s working instruments in opposite positions<br />
may disorient the assistant operator. One study concluded<br />
that assisting in more than 30 to 40 cases is sufficient to<br />
overcome mirror-image movements. 24<br />
Existing published reports show no final agreement regarding<br />
the optimal number <strong>of</strong> cases necessary to achieve<br />
technical expertise. Extrapolated from the COST study,<br />
ASCRS and the <strong>Society</strong> <strong>of</strong> Gastrointestinal and Endoscopic<br />
Surgeons (SAGES) recommend that a practitioner complete<br />
at least 20 laparoscopic resections <strong>of</strong> benign colon lesions<br />
before being credentialed for colon cancer resection. 9 Undoubtedly,<br />
completing a large number <strong>of</strong> operations correlates<br />
with better outcomes. One good example is the<br />
correlation between high-volume surgeons (all ranked<br />
among the top one-third in operative experience in their<br />
respective cancer centers) and impressive short-term results<br />
216<br />
TRAKARNSANGA AND WEISER<br />
shown in the COREAN trial. 12 This begs the question: How<br />
can operators pass the learning curve without putting patients<br />
at risk? Simulation and animal and cadaveric training<br />
courses are currently available worldwide; however, it is not<br />
known if this kind <strong>of</strong> training sufficiently qualifies a surgeon<br />
to perform resection in human patients. From our perspective,<br />
the answer is still uncertain. Intraoperative supervision<br />
by an experienced surgeon as well as appropriate case<br />
selection should be major considerations in any training<br />
program, even at an advanced level. This is essential to safe<br />
clinical practice. 25<br />
Is There Any Benefit to Hand-Assisted Laparoscopy?<br />
It has been suggested that hand-assisted laparoscopic<br />
colectomy has potential benefits, including a shorter learning<br />
curve for the practitioner, less operative time, and lower<br />
conversion rate compared to conventional laparoscopic surgery.<br />
This may be especially true in complex procedures<br />
such as total proctocolectomy. 26-28 Unfortunately, several<br />
difficult issues are associated with hand-assisted laparoscopy.<br />
Placing a hand into the abdominal cavity may have<br />
iatrogenic effects including a traumatic affect on the immune<br />
system, which may eliminate the long-term benefits <strong>of</strong><br />
MIS (such as fewer postoperative adhesions). Interleukin 6<br />
(IL-6) and C-reactive protein (CRP) are sensitive markers <strong>of</strong><br />
the immune system’s acute inflammatory response, and<br />
exaggeration <strong>of</strong> these may actually have a deleterious effect<br />
on wound healing and increase the risk <strong>of</strong> postoperative<br />
infection. 29,30 Several human studies have shown significant<br />
lowering <strong>of</strong> IL-6 and CRP in laparoscopic compared with<br />
open surgery (p � 0.05, p � 0.007, and p � 0.001). 29 In two<br />
animal studies comparing hand-assisted laparoscopic surgery<br />
(HALS) with laparoscopic surgery, slightly higher levels<br />
<strong>of</strong> IL-6 and CRP were reported with HALS; however, the<br />
levels <strong>of</strong> these cytokines were significantly lower than the<br />
levels seen in open surgery (p � 0.04 and p � 0.05). 30,31 No<br />
comparative human study has been reported.<br />
Regarding long-term complications, one retrospective<br />
study compared 266 hand-assisted operations with 270<br />
laparoscopic operations for colorectal disease. After a 27month<br />
median follow-up, the incidence <strong>of</strong> small bowel obstruction<br />
was not significantly different between the handassisted<br />
versus laparoscopic groups (4.1% vs. 7.4%; p �<br />
0.10), nor was the rate <strong>of</strong> postoperative incisional hernia<br />
(6% vs. 4.8%; p � 0.54). 32 Based on these results, we may<br />
conclude that hand-assisted surgery does not appear to be<br />
inferior to laparoscopic surgery, and retains the potential<br />
benefits <strong>of</strong> MIS.<br />
HALS for rectal cancer is challenging because, as noted<br />
above, the pelvic space is restricted by its bony anatomic<br />
structure. On the positive side, the hand may be used as an<br />
effective retractor to achieve adequate exposure. Unfortunately,<br />
in a deep and narrow pelvis, the hand may not be a<br />
useful tool because <strong>of</strong> the limited anatomic space. A multicenter<br />
prospective, randomized study comparing handassisted<br />
with straight laparoscopic-assisted proctectomy for<br />
rectal cancer is currently in the recruitment phase. 33 The<br />
planned sample size is 128 cases; estimated accrual will end<br />
by December <strong>2012</strong>. The primary outcome is operative time.<br />
The secondary outcomes include adequacy <strong>of</strong> resection margins;<br />
in-hospital mortality and morbidity; and preoperative<br />
as well as 3- and 6-month postoperative follow-ups <strong>of</strong> uri-
MINIMALLY INVASIVE SURGERY OF RECTAL CANCER<br />
nary and sexual function. The investigators hypothesize<br />
that HALS results in shorter operative time while retaining<br />
the benefits associated with laparoscopic surgery.<br />
As we await the results <strong>of</strong> this trial, several publications<br />
have reported on the role <strong>of</strong> robotic-assisted laparoscopic<br />
TME. A prospective randomized, controlled trial comparing<br />
robotic-assisted and laparoscopic TME is also ongoing. Comparative<br />
analysis, based on the ultimate data comparing<br />
hand-assisted and robotic-assisted laparoscopic surgery for<br />
rectal cancer, will be challenging. Only one three-arm retrospective<br />
study comparing HALS, robotic-assisted, and laparoscopic<br />
TME has been reported (with 30 patients with<br />
rectal cancer per arm). 34 No significant differences in pathologic<br />
outcome or complications were identified. To ascertain<br />
the equivalence <strong>of</strong> HALS and robotic surgery, however, a<br />
prospective randomized, controlled trial is needed.<br />
Robotic-Assisted Laparoscopic Surgery for Rectal<br />
Cancer: the Ideal Solution?<br />
Robotic-assisted laparoscopic prostatectomy has become<br />
extremely popular. An excellent three-dimensional visual<br />
system using EndoWrist instruments eliminates many <strong>of</strong><br />
the limitations <strong>of</strong> operating with straight laparoscopic instruments<br />
in a restricted anatomic area. The extraperitoneal<br />
part <strong>of</strong> the rectum is situated in the pelvic cavity.<br />
Thus, the idea <strong>of</strong> robotic-assisted laparoscopic surgery for<br />
rectal cancer appeals to surgeons. Loss <strong>of</strong> free movement to<br />
several quadrants <strong>of</strong> the abdomen, as well as increases in<br />
cost without benefits when compared to laparoscopic surgery,<br />
limits the use <strong>of</strong> robotic-assisted laparoscopic colectomy<br />
for colon cancer. A retrospective study comparing 40<br />
robotic-assisted to 135 laparoscopic right hemicolectomies<br />
reported no significant difference in short-term benefits,<br />
including estimated blood loss, rate <strong>of</strong> conversion, complications,<br />
and hospital stay. Moreover, operative time and costs<br />
were significantly higher in the robotic versus the laparoscopic<br />
group (p � 0.001 vs. p � 0.003, respectively). 35 In<br />
concurrence with the findings <strong>of</strong> a meta-analysis <strong>of</strong> seven<br />
nonrandomized studies, the average operative time was 39<br />
minutes longer and $792 more expensive than in conventional<br />
laparoscopy, with no improvement in short-term benefits.<br />
36<br />
As noted, however, the rectum seems suited to robotic<br />
surgery and the learning curve is less steep. Data on 50<br />
robotic-assisted laparoscopic rectal surgery cases suggest<br />
that the learning curve is passed after 15 to 25 operations. 37<br />
Unfortunately, the utilization <strong>of</strong> this technology has been<br />
limited by cost. Most <strong>of</strong> the current evidence regarding the<br />
benefits <strong>of</strong> robotic-assisted laparoscopic rectal cancer surgery<br />
consists <strong>of</strong> case series. Case-match and nonrandomized<br />
studies conclude that robotic-assisted surgery is feasible and<br />
comparable to laparoscopic TME. 38-40 A large prospective,<br />
randomized, controlled trial is needed, however, to assess<br />
the equivalence <strong>of</strong> robotic surgery to conventional laparoscopic<br />
surgery.<br />
The United Kingdom MRC ROLARR trial<br />
The United Kingdom Medical Research Council Trial <strong>of</strong><br />
Robotic compared with Laparoscopic Resection for Rectal<br />
cancer (ROLARR) trial is a multicenter prospective, randomized,<br />
controlled trial <strong>of</strong> robotic-assisted compared with<br />
laparoscopic surgery in the curative treatment <strong>of</strong> rectal<br />
cancer. 41 Four hundred patients will be recruited and randomly<br />
assigned with a 1:1 ratio. This trial is currently in the<br />
randomization phase, which will end by mid-<strong>2012</strong>. The<br />
primary endpoint is conversion rates to open surgery. Secondary<br />
endpoints are intraoperative and postoperative complications;<br />
oncologic outcomes, including circumferential<br />
margin, 3-year overall and disease-free survival; and quality<br />
<strong>of</strong> life.<br />
Techniques: Total Robotic or Hybrid Approach?<br />
There is debate over port placements, docking techniques,<br />
and techniques for take-down <strong>of</strong> the splenic flexure in<br />
robotic-assisted laparoscopic TME. The two major techniques<br />
are 1) totally robotic surgery and 2) a hybrid approach,<br />
depending on which instrument is used in takedown.<br />
Some surgeons are practicing a hybrid approach<br />
consisting <strong>of</strong> a robotic surgical system for vessel control and<br />
pelvic dissection, and straight laparoscopy for splenic flexure<br />
mobilization. The idea is to decrease the prolonged<br />
operative time caused by multiple dockings <strong>of</strong> the robot.<br />
The technique <strong>of</strong> totally robotic surgery is challenging<br />
because <strong>of</strong> limitations in port placement and positioning <strong>of</strong><br />
the robot. Most <strong>of</strong> the time, more than one docking <strong>of</strong> the<br />
robot is needed to complete the operation. Some centers have<br />
proposed the feasibility <strong>of</strong> a single-stage technique. 42,43<br />
However, there are currently no published studies comparing<br />
docking techniques. Surgeon preference and experience<br />
are still the main factors in selecting an operation. From our<br />
perspective, there is no single ideal technique that will fit all<br />
patients. Knowledge <strong>of</strong> the benefits and limitations <strong>of</strong> each<br />
technique is crucial, and the surgeon must be prepared to<br />
tailor surgery to each individual patient.<br />
Conclusion<br />
Minimally invasive surgery for rectal cancer is challenging<br />
because <strong>of</strong> the anatomic restrictions <strong>of</strong> the bony pelvis<br />
and the necessity <strong>of</strong> autonomic nerve preservation. Various<br />
techniques have been proposed, including straight laparoscopic,<br />
hand-assisted, and robotic-assisted laparoscopic surgery.<br />
In laparoscopic rectal cancer surgery, the short-term<br />
benefits are similar to those associated with other minimally<br />
invasive techniques. Current data indicate that long-term<br />
oncologic outcomes are similar in terms <strong>of</strong> recurrence and<br />
survival. Several additional prospective, randomized, controlled<br />
trials are in progress. We believe that the results will<br />
demonstrate the noninferiority <strong>of</strong> laparoscopic surgery compared<br />
to open surgery.<br />
The data from small case series and one nonrandomized,<br />
controlled trial indicate that the emerging techniques <strong>of</strong><br />
hand-assisted and robotic-assisted laparoscopy are feasible<br />
and comparable to the results achieved with conventional<br />
laparoscopic surgery. Prospective randomized, controlled<br />
trials <strong>of</strong> both techniques are ongoing. At this time, it is not<br />
possible to determine which procedure is “best.” Surgeon<br />
preference and availability <strong>of</strong> instruments are crucial in<br />
choosing the right procedure for each individual patient.<br />
217
Authors’ Disclosures <strong>of</strong> Potential Conflicts <strong>of</strong> Interest<br />
Author<br />
Atthaphorn Trakarnsanga*<br />
Martin R. Weiser*<br />
*No relevant relationships to disclose.<br />
Employment or<br />
Leadership<br />
Positions<br />
Consultant or<br />
Advisory Role<br />
1. Jacobs M, Verdeja JC, Goldstein HS. Minimally invasive colon resection<br />
(laparoscopic colectomy). Surg Laparosc Endosc. 1991;1:144-150.<br />
2. Paolucci V, Schaeff B, Schneider M, et al. Tumor seeding following<br />
laparoscopy: international survey. World J Surg. 1999;23:989-997.<br />
3. Wexner SD, Cohen SM. Port site metastases after laparoscopic colorectal<br />
surgery for malignancy. Br J Surg. 1995;82:295-298.<br />
4. Kuhry E, Schwenk WF, Gaupset R, et al. Long-term results <strong>of</strong> laparoscopic<br />
colorectal cancer resection. Cochrane Database <strong>of</strong> Syst Rev. 2008;16:<br />
CD003432.<br />
5. Lacy AM, García-Valdecasas JC, Delgado S, et al. Laparoscopy assisted<br />
colectomy versus open colectomy for treatment <strong>of</strong> non-metastatic colon<br />
cancer: a randomised trial. Lancet. 2002;359:2224-2229.<br />
6. <strong>Clinical</strong> Outcomes <strong>of</strong> Surgical Therapy Study Group. A comparison <strong>of</strong><br />
laparoscopically assisted and open colectomy for colon cancer. N Engl J Med.<br />
2004;350:2050-2059.<br />
7. Jayne DG, Guillou PJ, Thorpe H, et al. Randomized trial <strong>of</strong> laparoscopicassisted<br />
resection <strong>of</strong> colorectal carcinoma: 3-year results <strong>of</strong> the UK MRC<br />
CLASSIC Trial Group. J Clin Oncol. 2007;25:3061-3068.<br />
8. Bonjer HJ, Hop WC, Nelson H, et al. Laparoscopically assisted vs. open<br />
colectomy for colon cancer: a meta-analysis. Arch Surg. 2007;142:298-303.<br />
9. Tjandra JJ, Kilkenny JW, Buie WD, et al. Practice parameters for the<br />
management <strong>of</strong> rectal cancer (revised). Dis Colon Rectum. 2005;48:411-423.<br />
10. Row D, Weiser MR. An update on laparoscopic resection for rectal<br />
cancer. Cancer Control. 2010;17:16-24.<br />
11. Jayne DG, Thorpe HC, Copeland J, et al. Five-year follow-up <strong>of</strong> the<br />
Medical Research Council CLASICC trial <strong>of</strong> laparoscopically assisted versus<br />
open surgery for colorectal cancer. Br J Surg. 2010;97:1638-1645.<br />
12. Kang SB, Park JW, Jeong SY, et al. Open versus laparoscopic surgery<br />
for mid or low rectal cancer after neoadjuvant chemoradiotherapy (COREAN<br />
trial): short-term outcomes <strong>of</strong> an open-label randomised controlled trial.<br />
Lancet Oncol. 2010;11:637-645.<br />
13. COLOR II: Laparoscopic Versus Open Rectal Cancer Removal. http://<br />
clinicaltrials.gov/ct2/show/NCT00297791. Accessed September 2, 2011.<br />
14. Kitano S, Inomata M, Sato A, et al. Randomized controlled trial to<br />
evaluate laparoscopic surgery for colorectal cancer: Japan <strong>Clinical</strong> <strong>Oncology</strong><br />
Group Study JCOG 0404. Jpn J Clin Oncol. 2005;35:475-477.<br />
15. <strong>American</strong> College <strong>of</strong> Surgeons. Laparoscopic-Assisted Resection or<br />
Open Resection in Treating Patients With Stage IIA, Stage IIIA, or Stage IIIB<br />
Rectal Cancer. http://clinicaltrials.gov/ct2/show/NCT00726622. Accessed September<br />
2, 2011.<br />
16. Aziz O, Constantinides V, Tekkis PP, et al. Laparoscopic versus open<br />
surgery for rectal cancer: a meta-analysis. Ann Surg Oncol. 2006;13:413-424.<br />
17. Breukink S, Pierie J, Wiggers T. Laparoscopic versus open total<br />
mesorectal excision for rectal cancer. Cochrane Database Systemat Rev.<br />
2006;CD005200.<br />
18. Anderson C, Uman G, Pigazzi A. Oncologic outcomes <strong>of</strong> laparoscopic<br />
surgery for rectal cancer: a systemic review and meta-analysis <strong>of</strong> the literature.<br />
Eur J Surg Oncol. 2008;34:1135-1142.<br />
19. Huang MJ, Liang JL, Wang LH, et al. Laparoscopic-assisted versus<br />
open surgery for rectal cancer: a meta-analysis <strong>of</strong> randomized controlled trials<br />
on oncologic outcomes. Int J Colorectal Dis. 2011;26:415-421.<br />
20. Ohtani H, Tamamori Y, Azuma T, et al. A meta-analysis <strong>of</strong> the shortand<br />
long-term results <strong>of</strong> randomized controlled trials that compared<br />
laparoscopy-assisted and conventional open surgery for rectal cancer. J<br />
Gastrointest Surg. 2011;15:1375-1385.<br />
21. Kayano H, Okuda J, Tanaka K, et al. Evaluation <strong>of</strong> the learning curve<br />
in laparoscopic low anterior resection for rectal cancer. Surg Endosc. 2011;<br />
25:2972-2979.<br />
22. Park IJ, Choi GS, Lim KH, et al. Multidimensional analysis <strong>of</strong> the<br />
218<br />
Stock<br />
Ownership Honoraria<br />
REFERENCES<br />
Research<br />
Funding<br />
TRAKARNSANGA AND WEISER<br />
Expert<br />
Testimony<br />
Other<br />
Remuneration<br />
learning curve for laparoscopic colorectal surgery: Lesson from 1,000 cases <strong>of</strong><br />
laparoscopic colorectal surgery. Surg Endosc. 2009;23:839-846.<br />
23. Li JC, Hon SS, Ng SS, et al. The learning curve for laparoscopic<br />
colectomy: Experience <strong>of</strong> surgical fellow in an university colorectal unit. Surg<br />
Endosc. 2009;23:1603-1608.<br />
24. Hwang MR, Seo GJ, Yoo SB, et al. Learning curve <strong>of</strong> assistants in<br />
laparoscopic colorectal surgery: overcoming mirror imaging. Surg Endosc.<br />
2010;24:2575-2580.<br />
25. Miskovic D, Wyles SM, Ni M, et al. Systematic review on mentoring and<br />
simulation in laparoscopic colorectal surgery. Ann Surg. 2010;252:943-951.<br />
26. Meshikhes AW. Controversy <strong>of</strong> hand-assisted laparoscopic colorectal<br />
surgery. World J Gastroenterol. 2010;16:5662-5668.<br />
27. Aalbers AG, Biere SS, van Berge Henegouwen MI, et al. Hand-assisted<br />
or laparoscopic-assisted approach in colorectal surgery: a systematic review<br />
and meta-analysis. Surg Endosc. 2008;22:1769-1780.<br />
28. Cima RR, Pendlimari R, Hobular SD, et al. Utility and short-term<br />
outcomes <strong>of</strong> hand-assisted laparoscopic colorectal surgery: a single-institution<br />
experience in 1103 patients. Dis Colon Rectum. 2011;54:1076-1081.<br />
29. Novitsky YW, Litwin DE, Callery MP. The net immunologic advantage<br />
<strong>of</strong> laparoscopic surgery. Surg Endosc. 2004;18:1411-1419.<br />
30. Orenstein SB, Kaban GK, Litwin DE, et al. Evaluation <strong>of</strong> serum<br />
cytokine release in response to hand-assisted, laparoscopic, and open surgery<br />
in a porcine model. Am J Surg. 2011;202:97-102.<br />
31. Matsumoto ED, Marulis V, Tunc L, et al. Cytokine response to surgical<br />
stress: comparison <strong>of</strong> pure laparoscopic, hand-assisted laparoscopic, and open<br />
nephrectomy. J Endourol. 2005;19:1140-1145.<br />
32. Sonoda T, Pandey S, Trencheva K, et al. Longterm complications <strong>of</strong><br />
hand-assisted versus laparoscopic colectomy. J Am Coll Surg. 2009;208:62-66.<br />
33. Hand-assisted versus “Pure” Laparoscopic Assisted Proctectomy for<br />
Rectal Cancer. http://clinicaltrials.gov/ct2/show/study/NCT00651677?term�<br />
hand&show_desc�Y. Accessed September 2, 2011.<br />
34. Patel CB, Ragupathi M, Ramos-Valadez DI, et al. A three-arm (laparoscopic,<br />
hand-assisted and robot) matched-case analysis <strong>of</strong> intraoperative<br />
and postoperative outcomes in minimally invasive colorectal surgery. Dis<br />
Colon Rectum. 2011;54:144-150.<br />
35. deSouza AL, Prasad LM, Park JJ, et al. Robotic assistance in right<br />
hemicolectomy: is there a role? Dis Colon Rectum. 2010;53:1000-1006.<br />
36. Maeso S, Reza M, Mayol JA, et al. Efficacy <strong>of</strong> the Da Vinci surgical<br />
system in abdominal surgery compared with that <strong>of</strong> laparoscopy: A systematic<br />
review and meta-analysis. Ann Surg. 2010;252:254-262.<br />
37. Bokhari MB, Patel CB, Ramoz-Valadez DI, et al. Learning curve for<br />
robotic-assisted laparoscopic colorectal surgery. Surg Endosc. 2011;25:855-<br />
860.<br />
38. Pigazzi A, Luca F, Patriti A, et al. Multicentric study on robot<br />
tumor-specific mesorectal excision for the treatment <strong>of</strong> rectal cancer. Ann<br />
Surg Oncol. 2010;17:1614-1620.<br />
39. Baik SH, Kwon HY, Kim JS, et al. Robotic versus laparoscopic low<br />
anterior resection <strong>of</strong> rectal cancer: Short-term outcome <strong>of</strong> a prospective<br />
comparative study. Ann Surg Oncol. 2009;16:1480-1487.<br />
40. Antoniou SA, Antoniou GA, Koch OO, et al. Robot-assisted laparoscopic<br />
surgery <strong>of</strong> the colon and rectum. Surg Endosc. <strong>2012</strong>;26:1-11.<br />
41. Robotic versus Laparoscopic Resection for Rectal Cancer. http://ctru.<br />
leeds.ac.uk/rolarr. Accessed September 2, 2011.<br />
42. Choi DJ, Kim SH, Lee RJ, et al. Single-stage totally robotic dissection<br />
for rectal cancer surgery: Technique and short-term outcome in 50 consecutive<br />
patients. Dis Colon Rectum. 2009;52:1824-1830.<br />
43. Park YA, Kim JM, Kim SA, et al. Totally robotic surgery for rectal<br />
cancer: From splenic flexure to pelvic floor in one setup. Surg Endosc.<br />
2010;24:715-720.
Radiation in Rectal Cancer: What Are the<br />
Options and If/When Can It Be Avoided?<br />
Overview: Treatment <strong>of</strong> rectal cancer has improved greatly<br />
over recent decades. This review looks at the pivotal trials in<br />
the development <strong>of</strong> the current standard <strong>of</strong> therapy as well<br />
as new directions for more individualized therapy for rectal<br />
cancer. Rates <strong>of</strong> local recurrence and overall survival (OS) for<br />
surgery alone have improved with the use <strong>of</strong> (neo)adjuvant<br />
5-fluorouracil (5-FU)-based chemoradiation. New surgical<br />
techniques have improved outcomes, but preoperative radiotherapy<br />
still confers an additional benefit. Despite benefits in<br />
the metastatic setting, the addition <strong>of</strong> oxaliplatin to 5-FU and<br />
RECTAL CANCER treatment has made great strides<br />
over the last 50 years from the time <strong>of</strong> surgery alone,<br />
employing a blunt dissection technique that was associated<br />
with local recurrence rates <strong>of</strong> 30% to 60% for locally advanced<br />
disease. Adjuvant 5-FU-based chemoradiation reduced<br />
local failure rates to 10% to 12% and improved OS<br />
by 10% to 15%. 1-3 These results led to the National Cancer<br />
Institute consensus statement in 1990 recommending adjuvant<br />
chemoradiation for all patients with T3 to T4 or<br />
node-positive rectal cancer. 4 This was the first attempt to<br />
risk stratify patients with rectal cancer.<br />
Subsequently, the introduction <strong>of</strong> the total mesorectal<br />
excision (TME) and vastly improved surgical techniques<br />
have reduced the risk <strong>of</strong> positive margins and node disease<br />
left behind. As a result, local recurrence rates have dropped<br />
to 10% to 15% with surgery alone. Despite better surgical<br />
technique, the Dutch TME study (CKVO 95–04) still showed<br />
a benefit in local control for preoperative radiotherapy<br />
compared with surgery alone. 5 These investigators randomly<br />
selected 1,805 eligible patients to receive either<br />
surgery alone or short-course radiation therapy (5 � 5 Gy)<br />
followed by surgery and concluded that the addition <strong>of</strong><br />
radiation significantly decreases the rate <strong>of</strong> local recurrence<br />
at 2 years even with high-quality surgery (p � 0.001). In<br />
addition, the MRC07 trial randomly selected 1,350 patients<br />
with rectal cancer to receive either neoadjuvant short-course<br />
radiation therapy (5 � 5 Gy) or selective postoperative<br />
concurrent chemoradiotherapy for patients with a positive<br />
circumferential resection margin. Preoperative radiotherapy<br />
was associated with improved 3-year local control compared<br />
with selecting patients to receive adjuvant long-course<br />
radiation therapy. 6 Thus, the use <strong>of</strong> radiotherapy even in<br />
the setting <strong>of</strong> a high-quality TME appears to improve local<br />
control rates.<br />
To determine the optimal sequence for surgery and chemoradiotherapy,<br />
several randomized trials compared preoperative<br />
with postoperative 5-FU-based chemoradiotherapy.<br />
The German Rectal Study demonstrated an improvement in<br />
local relapse rates and reduced acute and overall toxicity<br />
using preoperative chemoradiation compared with postoperative<br />
chemoradiation. 7 This phase III randomized trial has<br />
established the “standard <strong>of</strong> care” for T3/4 or node-positive<br />
rectal cancer as neoadjuvant infusional 5-FU with conventionally<br />
fractionated radiotherapy to 5,040 cGy. With this<br />
approach, investigators reported local recurrence rates <strong>of</strong><br />
6% and distant recurrence rates <strong>of</strong> 36%; therefore, bringing<br />
By Karyn A. Goodman, MD<br />
radiotherapy has not shown improved outcomes, although it<br />
increased toxicity. Preoperative therapy also allows for the<br />
identification <strong>of</strong> predictive and prognostic markers for response<br />
to treatment, which has great potential to individualize<br />
treatment. Currently, the search for validated biomarkers and<br />
the refinement <strong>of</strong> risk stratification are the focus <strong>of</strong> ongoing<br />
study. Tailored therapy may include modification <strong>of</strong> the pelvic<br />
radiotherapy field, nonoperative therapy, or avoidance <strong>of</strong><br />
radiotherapy.<br />
rates <strong>of</strong> local recurrence well below rates <strong>of</strong> distant recurrence.<br />
The subsequent publications <strong>of</strong> the EORTC 22921<br />
and FFCD 9203 further confirmed the benefit <strong>of</strong> combining<br />
5-FU-based chemotherapy with preoperative radiation by<br />
improving the rates <strong>of</strong> pathologic complete response (pCR)<br />
and local control with acceptable increase in toxicity. 8,9<br />
Intensifying Neoadjuvant Therapy: Is More Better?<br />
The next step in the evolution <strong>of</strong> therapy for rectal cancer<br />
has been the attempt to improve our current standard <strong>of</strong><br />
care with more intensive therapy. Given the multitude <strong>of</strong><br />
new chemotherapy combinations and molecularly targeted<br />
systemic therapies being investigated in metastatic colorectal<br />
cancer, it was natural to consider incorporating these<br />
regimens into the adjuvant/neoadjuvant setting. 10<br />
The endpoints for evaluating the benefit <strong>of</strong> combination<br />
chemotherapy agents for neoadjuvant therapy have been<br />
tailored to allow for earlier assessment <strong>of</strong> efficacy such as<br />
pCR. For example, pCR following preoperative chemoradiation<br />
is predictive <strong>of</strong> OS and should therefore be an appropriate<br />
surrogate for evaluating the effect <strong>of</strong> neoadjuvant<br />
chemoradiation regimens for rectal cancer. 11-16 Other clinically<br />
relevant endpoints may include sphincter preservation<br />
and toxicity.<br />
The wide array <strong>of</strong> new cytotoxic and targeted agents has<br />
propagated numerous small phase II studies evaluating<br />
pCR rates. Based on promising results <strong>of</strong> these phase II<br />
studies, there have now been four recently reported phase<br />
III randomized trials that have added oxaliplatin to 5-FUbased<br />
preoperative chemoradiation with the intent <strong>of</strong> improving<br />
pCR and disease-free survival. Both the ACCORD<br />
trial and the STAR trial found no difference in pCR rate with<br />
the addition <strong>of</strong> oxaliplatin to standard 5-FU infusional<br />
chemotherapy and 50.4 Gy. 17,18 The toxicity was significantly<br />
greater (ACCORD: 25% v. 11%, p � .001; STAR:<br />
24% v. 8%, p � 0.001) in the oxaliplatin arm as well. The<br />
preliminary reports <strong>of</strong> the NSABP R-04 and the German<br />
From the Department <strong>of</strong> Radiation <strong>Oncology</strong>, Memorial Sloan-Kettering Cancer Center,<br />
New York, NY.<br />
Author’s disclosure <strong>of</strong> potential conflicts <strong>of</strong> interest are found at the end <strong>of</strong> this article.<br />
Address reprint requests to Karyn A. Goodman, MD, Department <strong>of</strong> Radiation <strong>Oncology</strong>,<br />
Memorial Sloan-Kettering Cancer Center, 1275 York Ave., New York, NY, 10065; email:<br />
goodmank@mskcc.org.<br />
© <strong>2012</strong> by <strong>American</strong> <strong>Society</strong> <strong>of</strong> <strong>Clinical</strong> <strong>Oncology</strong>.<br />
1092-9118/10/1-10<br />
219
CAO/ARO 04 study—which added oxaliplatin to continuous<br />
intravenous infusion 5-FU or capecitabine and 50.4 Gy <strong>of</strong><br />
radiotherapy—demonstrated that oxaliplatin did not result<br />
in an overwhelming improvement in pCR rates, although<br />
it was statistically significant (13% v. 17%, p � 0.033) in the<br />
German study. 19,20 The NSABP study showed significantly<br />
more grade 3 and 4 diarrhea with oxaliplatin (15% v. 7%, p �<br />
0.0001), similar to the STAR and ACCORD trials. Interestingly,<br />
the toxicity for the oxaliplatin arm was not higher in<br />
the German study, likely because <strong>of</strong> a lower cumulative dose<br />
<strong>of</strong> oxaliplatin and a split course treatment with the 5-FU.<br />
Oxaliplatin did not improve sphincter preservation—one <strong>of</strong><br />
the goals <strong>of</strong> preoperative therapy—in any <strong>of</strong> the studies.<br />
Clearly, there is no role for oxaliplatin in combination with<br />
neoadjuvant 5-FU-based chemoradiation. It appears to only<br />
increase toxicity without improving pCR rates, sphincter<br />
preservation, or local control.<br />
Individualization <strong>of</strong> Neoadjuvant Therapy<br />
The neoadjuvant approach is particularly appealing for<br />
assessing the biology <strong>of</strong> the tumor and identifying active<br />
regimens, which can then be used for the patient in the<br />
postoperative setting. Preoperative therapy also allows us<br />
to identify predictive and prognostic markers for treatment<br />
response. This leads to the potential for individualizing<br />
treatment regimens using information about the tumor and<br />
the patient. Basic prognostic information, such as tumor<br />
stage, histologic grade, and location, are being reinforced by<br />
the use <strong>of</strong> biomarkers, which may help us identify patients<br />
most likely to benefit from certain drugs. Although there are<br />
many markers that have been evaluated, none have been<br />
validated for clinical use. This is an area <strong>of</strong> greatly needed<br />
research. Future studies that add new agents may have<br />
the best chance <strong>of</strong> showing effect if we can use predictive<br />
KEY POINTS<br />
● Preoperative chemoradiation improved local control<br />
rates—even after total mesorectal excision—for patients<br />
with locally advanced rectal cancer.<br />
● Oxaliplatin combined with 5-fluorouracil–based<br />
chemoradiation does not improve pathologic response<br />
rates, sphincter preservation, or disease-free survival<br />
and increases the toxicity <strong>of</strong> neoadjuvant chemoradiation.<br />
● Pooled analyses <strong>of</strong> patients who received adjuvant<br />
treatment for rectal cancer have demonstrated intermediate<br />
and high risk disease, potentially identifying<br />
subgroups <strong>of</strong> patients who may not benefit from<br />
adjuvant radiotherapy.<br />
● Nonoperative approaches for patients with clinical<br />
complete responses to chemoradiation may allow for<br />
a more individualized approach for patients with<br />
distal rectal cancers.<br />
● The PROSPECT trial is investigating the selective<br />
use <strong>of</strong> chemoradiation in patients treated with neoadjuvant<br />
folinic acid, 5-FU, and oxaliplatin<br />
(FOLFOX) chemotherapy for locally advanced rectal<br />
cancers.<br />
220<br />
KARYN A. GOODMAN<br />
biomarkers as eligibility criteria to enrich the population <strong>of</strong><br />
patients who will potentially respond.<br />
Until validated biomarkers are available, further refinements<br />
in risk stratification based on clinical factors should<br />
be pursued to allow for better identification <strong>of</strong> patients<br />
who may or may not benefit from more aggressive multimodality<br />
therapy. In an attempt to stratify patients with locally<br />
advanced rectal cancer, investigators analyzed pooled data<br />
from 2,551 patients enrolled in three North <strong>American</strong> rectal<br />
trials. Patients with intermediate-risk rectal cancer—defined<br />
as T1 to T2N1 or T3N0—had only approximately 6%<br />
to 8% risk <strong>of</strong> local recurrence and may not derive any benefit<br />
from adjuvant radiation. 21 In addition, validated nomograms<br />
based on clinical and pathologic features are now<br />
available to assess risks for local or distant recurrence and<br />
neoadjuvant chemoradiation. 22 In Northern Europe, features<br />
on rectal magnetic resonance imaging (MRI) have been<br />
used to stratify patients into low-, intermediate-, and highrisk<br />
groups. Based on MRI findings <strong>of</strong> extramural extent<br />
<strong>of</strong> the tumor, the relation to the mesorectal fascia, and the<br />
presence <strong>of</strong> multiple suspicious lymph nodes, low-risk patients<br />
are treated with surgery alone, patients with<br />
intermediate-risk rectal tumors with “short-course” 5 Gy �<br />
5 preoperative protocol, and patients with high-risk tumors<br />
based on MRI findings with “long-course” preoperative<br />
chemoradiation. 23<br />
Patient-specific risk factors such as age, performance<br />
status, and other medical comorbidities must be considered<br />
when adding chemotherapy or targeted agents to neoadjuvant<br />
therapy. The benefits <strong>of</strong> bevacizumab are likely to be<br />
outweighed by cardiovascular or thrombotic risks in the<br />
patient with a history <strong>of</strong> cardiac disease. For young, premenopausal<br />
women who are trying to preserve fertility, the<br />
risks and benefits <strong>of</strong> pelvic radiation must be addressed. For<br />
these young, female patients, the argument to avoid radiotherapy<br />
could be made.<br />
Selective Use <strong>of</strong> Therapies for Rectal Cancer<br />
Tailoring therapy can also be achieved by modifying the<br />
pelvic radiotherapy field. With better imaging techniques,<br />
it may not be necessary to treat the whole pelvis for all<br />
patients. There may be patients whose tumors have a lower<br />
risk <strong>of</strong> lateral pelvic sidewall involvement and, therefore,<br />
can be spared treating internal iliac nodes. Also, after the<br />
use <strong>of</strong> induction chemotherapy—an area <strong>of</strong> growing interest—it<br />
may be possible to direct radiation at the postchemotherapy<br />
volume, potentially making the treatment<br />
fields smaller. The use <strong>of</strong> intensity-modulated radiotherapy<br />
(IMRT) can minimize dose to radiosensitive structures in<br />
the pelvis and potentially minimize toxicity but requires<br />
more in-depth knowledge <strong>of</strong> pelvic anatomy and nodal drainage<br />
patterns. IMRT may also allow dose escalation to higher<br />
risk areas and to achieve better pCR rates in patients not<br />
receiving surgery.<br />
The option <strong>of</strong> nonoperative therapy is also emerging as<br />
an option for patients who have clinical complete responses<br />
to neoadjuvant therapy. This approach was initially reported<br />
by investigators from Brazil. They reported on 361<br />
patients with rectal cancer, 99 (27%) <strong>of</strong> whom demonstrate<br />
a clinical complete response. These patients were observed<br />
without radical resection. The local recurrence rate was<br />
remarkably low (5%), and all five <strong>of</strong> these patients have been
salvaged with either surgery or additional radiation and<br />
have no evidence as yet <strong>of</strong> subsequent recurrence. 24 However,<br />
the time to recurrence in this group is quite long. Most<br />
data from patients who receive surgery demonstrate the<br />
majority <strong>of</strong> local recurrences within 2 to 3 years <strong>of</strong> treatment.<br />
Yet in this group, the mean interval is 52 months.<br />
This highlights the need for continued close surveillance <strong>of</strong><br />
these patients, many years out from treatment. There is<br />
interest in studying nonoperative management in Europe<br />
and the United States, particularly in the low-lying tumors<br />
that would require an abdominoperineal resection and permanent<br />
colostomy.<br />
Finally, tailored therapy may come in the way <strong>of</strong> selective<br />
Author’s Disclosure <strong>of</strong> Potential Conflicts <strong>of</strong> Interest<br />
Author<br />
Karyn A. Goodman*<br />
*No relevant relationships to disclose.<br />
Employment or<br />
Leadership<br />
Positions<br />
Consultant or<br />
Advisory Role<br />
1. Douglass HO Jr, Moertel CG, Mayer RJ, et al. Survival after postoperative<br />
combination treatment <strong>of</strong> rectal cancer. N Engl J Med. 1986;315:1294-<br />
1295.<br />
2. Krook JE, Moertel CG, Gunderson LL, et al. Effective surgical adjuvant<br />
therapy for high-risk rectal carcinoma. N Engl J Med. 1991;324:709-715.<br />
3. O’Connell MJ, Martenson JA, Weiand HS, et al. Improving adjuvant<br />
therapy for rectal cancer by combining protracted-infusion fluorouracil with<br />
radiation therapy after curative surgery. N Engl J Med. 1994;331:502-507.<br />
4. NIH consensus conference. Adjuvant therapy for patients with colon and<br />
rectal cancer. JAMA. 1990;264:1444-1450.<br />
5. Kapiteijn E, Marijnen CA, Nagtegaal ID, et al. Preoperative radiotherapy<br />
combined with total mesorectal excision for resectable rectal cancer.<br />
N Engl J Med. 2001;345:638-646.<br />
6. Sebag-Montefiore D, Stephens RJ, Steele R, et al. Preoperative radiotherapy<br />
versus selective postoperative chemoradiotherapy in patients with<br />
rectal cancer (MRC CR07 and NCIC-CTG C016): a multicentre, randomised<br />
trial. Lancet. 2009;373:811-820.<br />
7. Stahl M, Stuschke M, Lehmann N, et al. Chemoradiation with and<br />
without surgery in patients with locally advanced squamous cell carcinoma <strong>of</strong><br />
the esophagus. J Clin Oncol. 2005;23:2310-2317.<br />
8. Bosset JF, Calais G, Mineur L, et al. Enhanced tumorocidal effect <strong>of</strong><br />
chemotherapy with preoperative radiotherapy for rectal cancer: preliminary<br />
results-EORTC 22921. J Clin Oncol. 2005;23:5620-5627.<br />
9. Gérard JP, Conroy T, Bonnetain F, et al. Preoperative radiotherapy with<br />
or without concurrent fluorouracil and leucovorin in T3-4 rectal cancers:<br />
results <strong>of</strong> FFCD 9203. J Clin Oncol. 2006;24:4620-4625.<br />
10. André T, Boni C, Mounedji-Boudiaf L, et al. Oxaliplatin, fluorouracil,<br />
and leucovorin as adjuvant treatment for colon cancer. N Engl J Med.<br />
2004;350:2343-2351.<br />
11. Rödel C, Martus P, Papadoupolos T, et al. Prognostic significance <strong>of</strong><br />
tumor regression after preoperative chemoradiotherapy for rectal cancer.<br />
J Clin Oncol. 2005;23:8688-8696.<br />
12. Janjan NA, Crane C, Feig BW, et al. Improved overall survival among<br />
responders to preoperative chemoradiation for locally advanced rectal cancer.<br />
Am J Clin Oncol. 2001;24:107-112.<br />
13. Garcia-Aguilar J, Hernandez de Anda E, Sirivongs P, et al. A pathologic<br />
complete response to preoperative chemoradiation is associated with lower<br />
local recurrence and improved survival in rectal cancer patients treated by<br />
mesorectal excision. Dis Colon Rectum. 2003;46:298-304.<br />
14. Ruo L, Tickoo S, Klimstra DS, et al. Long-term prognostic significance<br />
therapy for patients who have a good response to induction<br />
chemotherapy. In the near future, we will see the opening <strong>of</strong><br />
the PROSPECT trial: a large phase III randomized trial <strong>of</strong><br />
no radiotherapy for patients with good response to induction<br />
chemotherapy. This study will accrue 1,000 patients to<br />
determine if preoperative folinic acid, 5-FU, and oxaliplatin<br />
(FOLFOX) is equivalent to standard preoperative chemoradiation.<br />
This study is based on excellent outcomes in a pilot<br />
study <strong>of</strong> 30 patients who received preoperative FOLFOX and<br />
bevacizumab and had a 27% pCR rate. 25 The PROSPECT<br />
trial is large enough to allow for a better understanding<br />
<strong>of</strong> prognostic and predictive factors to help identify which<br />
patients benefit from different therapeutic approaches.<br />
Stock<br />
Ownership Honoraria<br />
REFERENCES<br />
Research<br />
Funding<br />
Expert<br />
Testimony<br />
Other<br />
Remuneration<br />
<strong>of</strong> extent <strong>of</strong> rectal cancer response to preoperative radiation and chemotherapy.<br />
Ann Surg. 2002;236:75-81.<br />
15. Vecchio FM, Valentini V, Minsky B D, et al. The relationship <strong>of</strong><br />
pathologic tumor regression grade (TRG) and outcomes after preoperative<br />
therapy in rectal cancer. Int J Radiat Oncol Biol Phys. 2005;62:752-760.<br />
16. Guillem JG, Chessin DB, Cohen AM, et al. Long-term oncologic outcome<br />
following preoperative combined modality therapy and total mesorectal<br />
excision <strong>of</strong> locally advanced rectal cancer. Ann Surg. 2005;241:829-836.<br />
17. Gerard JP, Azria D, Gourgou-Bourgade S, et al. Comparison <strong>of</strong> two<br />
neoadjuvant chemoradiotherapy regimens for locally advanced rectal cancer:<br />
results <strong>of</strong> the phase III trial ACCORD 12/0405-Prodige 2. J Clin Oncol.<br />
2010;28:1638-1644.<br />
18. Aschele C, Cionini L, Lonardi S, et al. Primary tumor response to<br />
preoperative chemoradiation with or without oxaliplatin in locally advanced<br />
rectal cancer: pathologic results <strong>of</strong> the STAR-01 randomized phase III trial.<br />
J Clin Oncol. 2011;29:2773-2780.<br />
19. Roh MS, Yothers GA, O’Connell MJ, et al. The impact <strong>of</strong> capecitabine<br />
and oxaliplatin in the preoperative multimodality treatment in patients with<br />
carcinoma <strong>of</strong> the rectum: NSABP R-04. J Clin Oncol. 2011;29 (suppl 15; abstr<br />
3503).<br />
20. Roedel C, Becker H, Fietkau R, et al. Preoperative chemoradiotherapy<br />
and postoperative chemotherapy with 5-fluorouracil and oxaliplatin versus<br />
5-fluorouracil alone in locally advanced rectal cancer: First results <strong>of</strong> the<br />
German CAO/ARO/AIO-04 randomized phase III trial. J Clin Oncol. 2011;29<br />
(suppl 15; abstr 3505).<br />
21. Gunderson LL, Sargent DJ, Tepper JE, et al. Impact <strong>of</strong> T and N stage<br />
and treatment on survival and relapse in adjuvant rectal cancer: a pooled<br />
analysis. J Clin Oncol. 2004;22:1785-1796.<br />
22. Valentini V, van Stiphout RG, Lammering G, et al. Nomograms for<br />
predicting local recurrence, distant metastases, and overall survival for<br />
patients with locally advanced rectal cancer on the basis <strong>of</strong> European<br />
randomized clinical trials. J Clin Oncol. 2011;29:3163-3172.<br />
23. Smith N, Brown G. Preoperative staging <strong>of</strong> rectal cancer. Acta Oncol.<br />
2008;47:20-31.<br />
24. Habr-Gama A, Perez RO, Proscurshim I, et al. Patterns <strong>of</strong> failure and<br />
survival for nonoperative treatment <strong>of</strong> stage c0 distal rectal cancer following<br />
neoadjuvant chemoradiation therapy. J Gastrointest Surg. 2006;10:1319-<br />
1328.<br />
25. Schrag D, Weiser MR, Goodman KA, et al. Neoadjuvant FOLFOX-bev,<br />
without radiation, for locally advanced rectal cancer. J Clin Oncol. 2010;28<br />
(suppl 15; abstr 3511).<br />
221
STAGE III COLON CANCER: WHAT WORKS, WHAT<br />
DOESN’T AND WHY, AND WHAT’S NEXT (eQ&A)<br />
CHAIR<br />
Thierry André, MD<br />
Pitie-Salpetriere Hospital<br />
Paris, France<br />
SPEAKERS<br />
Bert H. O’Neil, MD<br />
University <strong>of</strong> North Carolina at Chapel Hill<br />
Chapel Hill, NC<br />
Jeffrey A. Meyerhardt, MD, MPH<br />
Dana-Farber Cancer Institute<br />
Boston, MA
Stage III Colon Cancer: What Works, What<br />
Doesn’t and Why, and What’s Next<br />
By Thierry André, MD, Bert H. O’Neil, MD, and Jeffrey A. Meyerhardt, MD, MPH<br />
Overview: Adjuvant treatment for patients with stage III<br />
colon cancer, one <strong>of</strong> the most common malignancies, is an<br />
important issue in oncology. The use <strong>of</strong> adjuvant chemotherapy<br />
in this setting has undoubtedly improved prognosis. This<br />
article describes the development <strong>of</strong> adjuvant therapy and<br />
COLORECTAL CANCER is the third most common<br />
cancer in both men and women, and the fourth leading<br />
cause <strong>of</strong> cancer deaths in the world. 1 The estimated worldwide<br />
incidence <strong>of</strong> colorectal cancer is 1.2 million per year. 1<br />
In Western countries, the median age at diagnosis is 71, and<br />
nearly 25% <strong>of</strong> all colon cancers are diagnosed at stage III. 2<br />
The main prognostic factor for estimating survival or relapse<br />
after surgery for localized disease is the tumor, node, and<br />
metastasis staging system. 3 The 3-year disease-free survival<br />
(DFS) for patients with stage III colon cancer without<br />
any postoperative chemotherapy ranges between 44% and<br />
52%. 4,5<br />
This review focuses on adjuvant therapy for patients with<br />
stage III colon cancer. The goal is to illustrate how adjuvant<br />
therapy has improved survival 3-fold in patients with stage<br />
III colon cancer and to acknowledge those who contributed<br />
to this achievement. We will also review efforts that have<br />
not succeeded and discuss next steps.<br />
Early Chemotherapy and 5-Fluorouracil/Levamisole<br />
The first drugs evaluated in colon cancer trials were the<br />
alkylating agent thiotepa and the antimetabolites fluorodeoxyuridine<br />
and 5-fluorouracil (5-FU), which was discovered<br />
by Charles Heidelberger in 1957. 6 The first adjuvant trials<br />
using various routes <strong>of</strong> administration were performed in<br />
the late 1960s and early 1970s. However, all proved unsuccessful<br />
except for trials using 5-FU. According to the criteria<br />
<strong>of</strong> that time, positive results were reported with intraluminal<br />
or intravenous administration <strong>of</strong> 5-FU, but the small<br />
number <strong>of</strong> undersized randomized studies failed to show any<br />
statistical significance. 7 In 1988, Buyse and colleagues carried<br />
out a meta-analysis <strong>of</strong> 17 trials involving 6,791 patients<br />
with colorectal cancer (mean <strong>of</strong> 400 patients per trial), and<br />
found a small benefit in overall survival (OS) with the<br />
5-FU–based regimens (hazard ratio [HR] � 0.83). 8 They also<br />
concluded that much larger trials were needed.<br />
Adjuvant Therapy for Patients with Colon Cancer: the<br />
First Step (1990)<br />
The first study demonstrating the value <strong>of</strong> adjuvant chemotherapy<br />
in patients with stage III colon cancer (TX, N1 or<br />
N2, M0) was published by Moertel in 1990. 4 This study<br />
showed an increase in OS and DFS in patients receiving<br />
5-FU/levamisole chemotherapy during 1 year compared with<br />
patients treated with levamisole alone or not receiving<br />
any treatment. After a mean follow-up <strong>of</strong> 6.5 years, patients<br />
treated with 5-FU/levamisole showed a 40% reduction<br />
in their recurrence rate and an estimated 33% reduction<br />
in overall death rate. 4 The 3-year DFS and 5-year<br />
OS estimated from the survival curves were 64% and 63%,<br />
progress in the past decade as well as failures in multiple<br />
agents that have demonstrated efficacy in the metastatic<br />
setting. Finally, the current clinical trials will be reviewed, as<br />
well as complementary therapies including diet and exercise<br />
for survivors <strong>of</strong> colorectal cancer.<br />
respectively. 4,5 These results constituted an impressive<br />
therapeutic advance, and 5-FU/levamisole adjuvant chemotherapy<br />
became the standard <strong>of</strong> care for patients with stage<br />
III colon cancer.<br />
Two other studies demonstrated the benefit <strong>of</strong> adjuvant<br />
chemotherapy with 5-FU and leucovorin (LV) compared<br />
with no treatment for patients with colon cancer. 9,10 The<br />
International Multicenter Pooled Analysis <strong>of</strong> colorectal cancer<br />
Trials (IMPACT) showed a 22% relative risk reduction in<br />
mortality with 5-FU/LV. 9<br />
Modulation <strong>of</strong> 5-FU with Levamisole or Leucovorin<br />
After the publication <strong>of</strong> NSABP C-04 study results in<br />
1999, 11 which showed a small DFS advantage <strong>of</strong> 5-FU/LV<br />
compared with FU/levamisole in patients with high-risk<br />
stage II and III colon cancer, the addition <strong>of</strong> levamisole to<br />
5-FU was withdrawn. The combination <strong>of</strong> 5-FU (bolus or<br />
short infusion) and LV, administered either daily for 5 days<br />
per month (according to the Mayo Clinic regimen) or weekly<br />
for 6 months (according to the Roswell Park regimen),<br />
became the new standard treatments for patients with stage<br />
III colon cancer. 11,12 In fact, the Intergroup study 0089<br />
(INT-0089) showed that the Roswell Park regimen, the<br />
Mayo Clinic regimen, 5-FU/LV/levamisole, and 5-FU/levamisole<br />
(control arm) treatments were equivalent. In this<br />
four-arm study, the authors concluded that 5-FU/LV (either<br />
the Roswell Park or the Mayo Clinic regimens) could replace<br />
5-FU/levamisole. 12 Interestingly, the biweekly LV5FU2 regimen<br />
(LV followed by both a bolus and a 22-hour infusion <strong>of</strong><br />
5-FU for 2 consecutive days) was also compared with<br />
monthly 5-FU/LV (a modified Mayo Clinic regimen) in the<br />
GERCOR C96-1 trial, which included patients with stage II<br />
and III colon cancer. 13 There was no significant improvement<br />
found in DFS, but the LV5FU2 regimen became<br />
another accepted standard because <strong>of</strong> its improved safety<br />
pr<strong>of</strong>ile. Finally, the ACCENT (Adjuvant Colon Cancer<br />
ENdpoinTs) meta-analysis, which included individual data<br />
from more than 20,000 patients followed for more than 8<br />
years, confirmed a 10% absolute improvement in OS for<br />
patients with stage III colon cancer who received adjuvant<br />
From the Service d’Oncologie Médicale, Hôpital Saint-Antoine, Assistance Publique des<br />
Hôpitaux de Paris, Paris, France and Université Pierre et Marie Curie (Paris 6); University<br />
<strong>of</strong> North Carolina, Chapel Hill, NC; Dana-Farber Cancer Institute, Harvard Medical<br />
School, Boston, MA.<br />
Authors’ disclosures <strong>of</strong> potential conflicts <strong>of</strong> interest are found at the end <strong>of</strong> this article.<br />
Address reprint requests to Jeffrey A. Meyerhardt, MD, MPH, Dana-Farber Cancer<br />
Institute, 44 Binney Street, Boston, MA 02215; email: Jeffrey_Meyerhardt@dfci.harvard.<br />
edu.<br />
© <strong>2012</strong> by <strong>American</strong> <strong>Society</strong> <strong>of</strong> <strong>Clinical</strong> <strong>Oncology</strong>.<br />
1092-9118/10/1-10<br />
223
Table 1. Comparison <strong>of</strong> Fluoropyrimidines Alone (5-FUl/LV) and Fluoropyrimidines (5-FU/LV)/Oxaliplatin Combination in Patients with<br />
Stage III Colon Cancer: Survival in 3 Phase III Randomized Controlled Trials<br />
ENDPOINTS MOSAIC 61 Stage III MOSAIC 18 Stage III NSABP 07 20 Stage III NO1698 19 Stage III<br />
Disease-Free Survival<br />
Follow-up duration, y 3 5 5 3<br />
DFS without vs. with oxaliplatine, % 65.3 vs. 72.2 58.9 vs. 66.4 57.8 vs. 64.4 66.5 vs. 70.9<br />
Absolute difference in DFS, % � 6.3 � 7.5 � 6.6 � 4.4<br />
HR (95% CI) 0.76 (0.62–0.92) 0.78 (0.65–0.93) 0.78 (0.68–0.90) 0.80 (0.69–0.93)<br />
p value � .005 � .005 � .0007 � .0045<br />
Overall Survival<br />
Follow-up duration, y 3 6 5 4.75<br />
OS without vs. with oxaliplatine, % ND 68.7 vs. 72.9 73.8 vs. 76.5 ND<br />
Absolute difference in OS, % – � 4.2 � 2.7 � 3.4<br />
HR (95% CI) ND 0.80 (0.65–0.97) 0.85 (0.72–1.00) 0.87 (0.72–1.05)<br />
p value – � .023 � .052 � .1486<br />
Abbreviations: 5-FU, 5-fluorouracil; CI, confidence interval; DFS, disease-free survival; HR, hazard ratio; LV, leucovorin; ND, not done; OS: overall survival.<br />
chemotherapy with fluoropyrimidines compared with surgery<br />
alone. 14<br />
Oral Fluoropyrimidines, Capecitabine, and<br />
Uracil/Tegafur (UFT) Are Equivalent to 5-FU/LV<br />
Oral drugs are more convenient than intravenous chemotherapy,<br />
at least when used as single agents. The X-ACT<br />
trial compared bolus 5-FU/LV (the Mayo Clinic regimen)<br />
with oral capecitabine for 6 months in patients with stage III<br />
colon cancer. DFS was at least equivalent (HR � 0.87,<br />
p � 0.001 for noninferiority). 15 In the NSABP C-06 study,<br />
oral uracil and tegafur (UFT) plus LV was compared to the<br />
Roswell Park regimen and the results showed similar DFS<br />
and OS. 16<br />
KEY POINTS<br />
● Adjuvant chemotherapy for patients with stage III<br />
colon cancer has substantially evolved in the past 2<br />
decades, increasing disease-free and overall survival.<br />
● Until the early 1990s, research focused on modulation<br />
and duration <strong>of</strong> 5-fluorouracil and standard <strong>of</strong><br />
care became 5-fluorouracil and leucovorin in 1990.<br />
● The addition <strong>of</strong> oxaliplatin to 5-fluorouracil/leucovorin<br />
further improved disease-free and overall survival<br />
for patients with stage III disease; however,<br />
additional toxicities during therapy and cumulative<br />
neuropathy require active monitoring and adjustments<br />
in caring for patients with stage III disease.<br />
● Three drugs that have shown benefit in metastatic<br />
colorectal cancer (irinotecan, bevacizumab, and cetuximab)<br />
failed to provide significant additional benefit<br />
in the management <strong>of</strong> stage III colon cancer, and<br />
the reasons for the discordance between metastatic<br />
disease and adjuvant therapy remains unclear.<br />
● Adjunctive therapies to standard treatment are actively<br />
being considered in colon cancer survivorship,<br />
including physical activity, diet, obesity, vitamin D,<br />
and nonsteroidal anti-inflammatory drugs, and data<br />
are forthcoming to further understand the role these<br />
may play in the outcomes <strong>of</strong> patients with stage III<br />
colon cancer.<br />
224<br />
ANDRÉ, O’NEIL, AND MEYERHARDT<br />
A 3-Year DFS Is the New Endpoint in Adjuvant Trials<br />
The traditional endpoint for clinical trials <strong>of</strong> patients<br />
with adjuvant colon cancer treatment was 5-year OS. The<br />
ACCENT meta-analysis <strong>of</strong> adjuvant studies demonstrated<br />
that 3-year DFS was a predictor <strong>of</strong> 5-year OS results and<br />
could be an appropriate primary endpoint for adjuvant<br />
studies in colon cancer. The data led to the approval <strong>of</strong><br />
3-year DFS as the primary endpoint in adjuvant therapy<br />
studies for patients with stage III colon cancer by the United<br />
States Food and Drug Administration (FDA). 17 Using 3-year<br />
DFS as the main endpoint reduces trial duration, drug<br />
development time, and cost, and as a result, new and better<br />
treatments can be made available to patients more rapidly.<br />
Oxaliplatin Plus Fluoropyrimidines Is Better than<br />
Fluoropyrimidines Alone: The Second Step (2004)<br />
Combined fluoropyrimidines and oxaliplatin led to significant<br />
improvements in survival in three phase III<br />
randomized controlled trials (Table 1). 18-20 The MOSAIC<br />
(Multicenter International Study <strong>of</strong> Oxaliplatin/5-FU/LV in<br />
the Adjuvant Treatment <strong>of</strong> Colon Cancer) trial compared the<br />
efficacy <strong>of</strong> the LV5FU2 regimen and the same regimen plus<br />
oxaliplatin (FOLFOX4) in patients with stage II and III<br />
colon cancer. Patients were randomly assigned to receive<br />
12 biweekly cycles <strong>of</strong> LV5FU2 or FOLFOX4. 18 For patients<br />
with stage III disease, FOLFOX4 improved DFS by 24%<br />
(HR � 0.76, p � 0.005). On the basis <strong>of</strong> these results,<br />
FOLFOX4 was approved as adjuvant therapy after surgery<br />
for patients with stage III colon cancer. Updated results<br />
showed 5-year DFS rates <strong>of</strong> 58.9% and 66.4% for LV5FU2<br />
and FOLFOX4, respectively (HR � 0.78, p � 0.005). A<br />
benefit in OS for patients treated with FOLFOX4 was also<br />
observed: the 6-year OS rates comparing LV5FU2 and<br />
FOLFOX4 were 68.7% and 72.9%, respectively (HR � 0.80,<br />
p � 0.023). 18 The National Surgical Adjuvant Breast and<br />
Bowel Project (NSABP) trial C-07 evaluated the FLOX<br />
regimen (i.e., oxaliplatin added to a weekly bolus <strong>of</strong> 5-FU/<br />
LV) in patients with stage II and III colon cancer. 20 For both<br />
stages, the benefit provided by oxaliplatin on 3-year DFS<br />
was similar to that reported in the MOSAIC study (HR 0.80,<br />
p � 0.004). A longer follow-up indicated a 6.6% increase in<br />
DFS at 5 years (57.8% vs. 64.4%, p � 0.0007) for patients<br />
with stage III disease, with a nonsignificant benefit in OS:<br />
5-year OS rates with 5-FU/LV and FLOX were 73.8% and<br />
76.5%, respectively (HR � 0.85, p � 0.052).
ADJUVANT THERAPY FOR STAGE III COLON CANCER<br />
For both studies, the combination <strong>of</strong> 5-FU/oxaliplatin,<br />
however, proved substantially more toxic than 5-FU alone,<br />
resulting in a greater risk <strong>of</strong> severe cytopenia, nausea,<br />
vomiting, diarrhea, and peripheral neuropathy. 18,20<br />
Finally, the superiority <strong>of</strong> XELOX (i.e., capecitabine/<br />
oxaliplatin) over bolus 5-FU/LV as adjuvant treatment for<br />
patients with stage III colon cancer was shown in the<br />
NO16968 trial. 19 Haller and colleagues reported a 3-year<br />
DFS <strong>of</strong> 66.5% for 5-FU/LV (Mayo Clinic or Roswell Park<br />
regimens) compared with 70.9% for XELOX (HR � 0.80, p �<br />
0.0045). As anticipated with oxaliplatin, neurosensory toxicity<br />
occurred in the majority <strong>of</strong> patients in these three<br />
studies, but the frequency <strong>of</strong> peripheral sensory neuropathy<br />
decreased during the follow-up period. 18 Preliminary<br />
patient-reported outcomes from the NSABP C-07 trial suggest<br />
that symptoms <strong>of</strong> oxaliplatin-associated neurotoxicity<br />
persisting for 6 years after treatment failed to reach clinical<br />
significance. 21 The total dose <strong>of</strong> oxaliplatin administered<br />
in NSABP C-07 was approximately 30% lower than in<br />
MOSAIC, yet provided comparable efficacy. The per-protocol<br />
planned oxaliplatin dose was 1,020 mg/m 2 (12 cycles) in<br />
MOSAIC and 765 mg/m 2 (nine cycles) in NSABP C-07,<br />
although the median dose <strong>of</strong> oxaliplatin received per patient<br />
was 810 mg/m 2 (9.5 cycles) in MOSAIC and 667 mg/m 2 (7.8<br />
cycles) in NSABP C-07. 18,20 Stopping oxaliplatin in cases <strong>of</strong><br />
neuropathy at grades 2 (e.g., paresthesia/dysesthesia interfering<br />
with function, but not with activities <strong>of</strong> daily living<br />
[i.e., permanent dysesthesia]) or 3 (e.g., paresthesia/dysesthesia<br />
with pain or functional impairment) must be done in<br />
clinical practice. If oxaliplatin has to be stopped because <strong>of</strong><br />
peripheral neuropathy, 5-FU/LV or capecitabine should be<br />
administered per protocol (i.e., for a total <strong>of</strong> 6 months).<br />
Could the mFOLFOX6 Regimen Replace FOLFOX4?<br />
The mFOLFOX6 regimen (85 mg/m 2 <strong>of</strong> intravenous oxaliplatin<br />
with 400 mg/m 2 <strong>of</strong> LV over 2 hours, followed by 400<br />
mg/m 2 bolus injection <strong>of</strong> 5-FU, then 2,400 mg/m 2 intravenous<br />
5-FU for 46 hours) is more convenient for patients and<br />
less expensive than the FOLFOX4 regimen (i.e., 1 day<br />
compared with 2 days in the outpatient unit). In addition, no<br />
evidence <strong>of</strong> further toxicity has resulted from the increased<br />
dose <strong>of</strong> 5-FU. To date, no study has compared FOLFOX4 to<br />
mFOLFOX6 in the adjuvant setting. However, in the<br />
NSABP C-08 study, 22 which evaluated the mFOLFOX6<br />
regimen with or without bevacizumab in patients with stage<br />
III disease, the 3-year DFS was nearly identical to that<br />
observed in patients treated with FLOX in the oxaliplatin<br />
arms <strong>of</strong> NSABP C-07 or those treated with FOLFOX4 in the<br />
MOSAIC trial. 18,20<br />
Can Older Patients Benefit from Adjuvant Therapy?<br />
The value and feasibility <strong>of</strong> 5-FU–based adjuvant chemotherapy<br />
for patients older than age 70 was demonstrated in<br />
a pooled analysis by Sargent and colleagues, 23 who found no<br />
evidence <strong>of</strong> interaction between age and the efficacy <strong>of</strong><br />
chemotherapy. Moreover, except for leucopenia in one study,<br />
the incidence <strong>of</strong> toxic effects <strong>of</strong> chemotherapy was not higher<br />
in older patients.<br />
Adjuvant treatment with FOLFOX4 was well tolerated in<br />
patients older than age 70, with higher rates <strong>of</strong> only neutropenia<br />
and thrombocytopenia compared with younger patients<br />
(49% vs. 43%, p � 0.04%, and 5% vs. 2%, p � 0.04,<br />
respectively). 24 A combined analysis <strong>of</strong> the two pivotal<br />
adjuvant trials comparing 5-FU/oxaliplatin with 5-FU alone<br />
failed to demonstrate a benefit in DFS and OS in older<br />
patients, despite a positive trend for time to recurrence. 25<br />
On the other hand, the NO1698 trial, 19 and the posthoc<br />
analysis <strong>of</strong> MOSAIC data 26 showed efficacy in both the<br />
young and the older adult populations. Therefore no formal<br />
recommendation can be <strong>of</strong>fered. From a clinical point <strong>of</strong><br />
view, the decision to add oxaliplatin to fluoropyrimidines as<br />
adjuvant treatment should be at least based on performance<br />
status and comorbidities. In addition to these cancer-specific<br />
prognostic factors, the use <strong>of</strong> a comprehensive geriatric<br />
assessment is strongly recommended to evaluate the appropriateness<br />
<strong>of</strong> chemotherapy.<br />
Lack <strong>of</strong> Benefit <strong>of</strong> Irinotecan in the Adjuvant Setting<br />
As discussed, 5-FU leads to a relatively large improvement<br />
in OS for patients with stage III colon cancer. Interestingly,<br />
this benefit occurs despite a relatively low<br />
radiographic response rate <strong>of</strong> 15% to 20%. 27 With the addition<br />
<strong>of</strong> oxaliplatin and irinotecan for patients with metastatic<br />
disease, response rates more than doubled, suggesting<br />
synergy in the cytotoxic effects on colorectal cancer cells. 27<br />
As such, it was anticipated that oxaliplatin would improve<br />
the chance <strong>of</strong> remaining disease-free longer after surgery<br />
and increase OS when added to 5-FU. 18-20 Perhaps a presaging<br />
event to the current situation, however, was the story<br />
<strong>of</strong> irinotecan in the adjuvant setting. When compared headto-head<br />
in the metastatic setting, irinotecan has proven<br />
itself to be essentially equivalent to oxaliplatin (i.e., as an<br />
addition to 5-FU). 27 Despite this, four randomized trials <strong>of</strong><br />
irinotecan as an adjuvant addition for patients with stages<br />
II and III colorectal cancer failed to demonstrate benefit,<br />
demonstrating that improvements in response and OS in the<br />
metastatic setting might not reliably translate to clear<br />
benefit in the adjuvant setting. 28-30 To this day, the failure<br />
<strong>of</strong> irinotecan remains a puzzle, but it is also important to<br />
understand that even a doubling <strong>of</strong> response rate with<br />
oxaliplatin led to a fairly modest improvement in DFS<br />
(HR � 0.78 to 0.8) and OS (HR � 0.8 to 0.85) in the MOSAIC<br />
and NSABP C-07 studies, respectively. 18,20 Perhaps, then,<br />
what we can really conclude about adjuvant therapy for<br />
patients with stage III colorectal cancer is that it is amazing<br />
how large the benefit <strong>of</strong> 5-FU has been given the very low<br />
response rate in stage IV disease.<br />
Lack <strong>of</strong> Benefit <strong>of</strong> Biologics in the Adjuvant Setting:<br />
Where Did We Go Wrong?<br />
The first biologic agent to fail in the adjuvant setting was<br />
bevacizumab. Mouse models <strong>of</strong> metastasis (tail-vein injection)<br />
performed in the 1990s suggested substantial inhibition<br />
<strong>of</strong> metastasis by the depletion <strong>of</strong> vascular endothelial<br />
growth factor (VEGF) that could potentially translate to<br />
benefit in the human adjuvant situation. 31 It is welldocumented<br />
that bevacizumab led to improvements in<br />
progression-free survival and OS in several trials <strong>of</strong> patients<br />
with metastatic colorectal cancer, with an impressive HR for<br />
death <strong>of</strong> 0.66 in the pivotal randomized study <strong>of</strong> bevacizumab.<br />
32 As such, hopes were high for benefit in the adjuvant<br />
setting.<br />
The first published study <strong>of</strong> adjuvant bevacizumab was<br />
NSABP C-08. 22 This study randomly assigned 2,710 pa-<br />
225
tients with stages II and III colorectal cancer after surgery<br />
to a standard 6-month course <strong>of</strong> FOLFOX or FOLFOX plus<br />
5 mg/kg <strong>of</strong> bevacizumab every other week followed by 6<br />
months <strong>of</strong> single-agent bevacizumab. The study accrued<br />
quickly, which reflected enthusiasm for the agent in this<br />
setting. Disappointingly, the primary endpoint <strong>of</strong> 3-year<br />
DFS did not improve significantly with bevacizumab (HR<br />
0.87, p � 0.08), although there was improvement in DFS at<br />
1 year, suggesting either a biologic effect during bevacizumab<br />
treatment that did not lead to cure or that relapse<br />
was observed later because <strong>of</strong> later surveillance scanning in<br />
patients on the bevacizumab arm (i.e., median time to first<br />
imaging was earlier in the control arm). Subsequent to this,<br />
the 3,451-patient European AVANT (AVastinAdjuvaNT)<br />
trial, which had a similar design to the NSABP C-08 trial<br />
but included a third capecitabine/oxaliplatin/bevacizumab<br />
arm, reported similarly negative results, including a transient<br />
benefit during the first year. 33 Unlike NSABP C-08,<br />
AVANT actually demonstrated a nonsignificant trend toward<br />
a decrement in survival for patients in both<br />
bevacizumab-containing arms compared with FOLFOX<br />
alone. In neither study did excess toxicity <strong>of</strong> bevacizumab<br />
appear to be a contributing issue.<br />
So what went wrong with bevacizumab? One question is<br />
whether the scientific rationale for (relatively) short-term<br />
VEGF inhibition was sound. Conceptually, we believe that<br />
adjuvant therapy works by killing tumor cells, something<br />
that bevacizumab is not known to do. If one were to consider<br />
the static effects <strong>of</strong> VEGF inhibition to be valid, then ideally<br />
bevacizumab would be continued for a long period, with the<br />
expectation <strong>of</strong> potential relapse whenever one chose to end<br />
therapy. We may never know if prolonged use <strong>of</strong> bevacizumab<br />
would be <strong>of</strong> benefit. In some experimental models,<br />
bevacizumab has been shown to work in metastatic tumors<br />
by normalizing aberrant tumor neovasculature, 34 a mechanism<br />
that would not be expected to be <strong>of</strong> benefit in tumor<br />
micrometastases. At this time, whether further study <strong>of</strong><br />
bevacizumab (or other VEGF pathway–targeting strategies)<br />
will be pursued remains unclear.<br />
Antibodies that prevent the binding <strong>of</strong> ligands to the<br />
epidermal growth factor receptor (EGFR) are the other class<br />
<strong>of</strong> agent that has recently been studied in the adjuvant<br />
setting. Cetuximab and panitumumab both have clear activity<br />
in subsets <strong>of</strong> patients with metastatic disease, 35 and both<br />
have been shown to increase response rates by roughly 10%<br />
when added to FOLFOX or FOLFIRI (5-FU/leucovorin/<br />
irinotecan) and when analysis is restricted to patients with<br />
KRAS wild-type tumors. This increase in response rate,<br />
along with single-agent responses that are similar to those<br />
seen with 5-FU, would have suggested that the EGFR<br />
antibodies might cause an increase in tumor-cell killing that<br />
would make them good adjuvant therapy agents. With this<br />
rationale in mind, study N0147 was designed by the North<br />
Central Cancer Treatment Group (NCCTG) and conducted<br />
by multiple U.S. and Canadian cooperative groups. N0147<br />
went through several design iterations before settling as a<br />
study <strong>of</strong> FOLFOX with or without cetuximab for patients<br />
with resected stage III colorectal cancer. 36,37 An interesting<br />
subgroup <strong>of</strong> patients randomly assigned to receive FOLFIRI<br />
with or without cetuximab was removed when it became<br />
clear that irinotecan was an ineffective addition to 5-FU in<br />
this setting. 38 Over the course <strong>of</strong> the N0147 study, the KRAS<br />
story emerged, and the study was amended to include only<br />
226<br />
patients with KRAS wild-type tumors. The final reported<br />
sample size included 1,864 patients with KRAS wild-type<br />
tumors. One issue that became clear was related to toxicity.<br />
A clear signal emerged in patients older than 70 with high<br />
rates <strong>of</strong> study discontinuation for toxicity, which resulted in<br />
yet another amendment. After all was said and done, the<br />
results <strong>of</strong> N0147 were again disappointing. Even patients<br />
with KRAS wild-type tumors treated with cetuximab did not<br />
benefit; in fact the trend was toward harm (DFS HR � 1.2,<br />
p � NS). 36 In addition, there was clear harm in patients<br />
with KRAS-mutant cancers. 37 In patients older than 70, the<br />
HR was 1.79 (p � 0.03), significantly favoring the FOLFOXonly<br />
arm. 36<br />
So again, what went wrong with cetuximab? In this case,<br />
unexpected toxicities that could have shortened chemotherapy<br />
contributed to the negative outcome. As with irinotecan,<br />
however, the ultimate cause <strong>of</strong> failure is unclear. One hint<br />
that this study might have been negative was the relatively<br />
small incremental improvements in outcome in most trials<br />
<strong>of</strong> patients with metastatic cancer (a disconnect from the<br />
relatively robust increase in radiographic response). Could it<br />
be that micrometastases have not yet become dependent on<br />
EGFR ligands for their growth? Could early-stage tumors<br />
be biologically different and less likely to depend on EGFR<br />
signaling?<br />
Curiously, data from the N0147 study now suggest that<br />
cetuximab might benefit patients when given with the<br />
discontinued irinotecan backbone, although interpreting<br />
these data has to be tempered by the small sample size. 38<br />
In a poster presented at the ASCO Gastrointestinal Cancers<br />
Symposium 2011, N0147 investigators demonstrated a<br />
fairly robust benefit trend in the FOLFIRI subgroup (3-year<br />
DFS HR � 0.4, p � 0.05; 3-year OS HR � 0.3, p � 0.08). 38<br />
Given the irinotecan data, whether this information will be<br />
followed up in larger trials remains to be seen.<br />
Current Adjuvant Therapy Trials<br />
ANDRÉ, O’NEIL, AND MEYERHARDT<br />
To date, adjuvant therapy trials for patients with colon<br />
cancer, as in most other cancers, develop because <strong>of</strong> promising<br />
data in metastatic disease. Whether this strategy is<br />
optimal is debatable since certain agents might have more<br />
efficacy in a micrometastatic setting, though this remains<br />
the current approach. Hampering the testing <strong>of</strong> new agents<br />
in adjuvant colon cancer, therefore, has been the recent lack<br />
<strong>of</strong> new agents with definitive efficacy in metastatic disease.<br />
The last drug to be approved by the FDA was panitumumab<br />
in 2006.<br />
Given the lack <strong>of</strong> agents to test in the adjuvant setting,<br />
questions have arisen related to the necessary duration <strong>of</strong><br />
therapy. In the late 1980s and early 1990s, INT-0089 demonstrated<br />
noninferiority between 6 months <strong>of</strong> adjuvant chemotherapy<br />
with 5-FU/LV and 12 months with 5-FU/<br />
levamisole. 12 In addition, the GERCOR trial showed that<br />
the LV5FU2 regimen administered for either 6 or 9 months<br />
achieved similar results. 13 Therefore, a 6-month duration <strong>of</strong><br />
treatment became the standard. Given the cumulative neuropathy<br />
from oxaliplatin that can remain with patients for<br />
years after completing adjuvant therapy, 18 the question <strong>of</strong><br />
needed duration <strong>of</strong> adjuvant therapy became pertinent.<br />
Further, one small trial by Chau and colleagues demonstrated<br />
noninferiority <strong>of</strong> a protracted venous infusion regimen<br />
<strong>of</strong> 5-FU for 3 months compared to the Mayo Clinic<br />
(5-FU/LV) regimen in adjuvant therapy for patients with
ADJUVANT THERAPY FOR STAGE III COLON CANCER<br />
Table 2. Ongoing Adjuvant Trials for Colon Cancer Testing Duration <strong>of</strong> FOLFOX or XELOX Therapy (3 versus 6 Months <strong>of</strong> Therapy)<br />
stage II and III colorectal cancer. 39 In response to these<br />
concerns, at least five randomized trials have opened and<br />
are accruing patients to test the noninferiority <strong>of</strong> 3 months<br />
compared with 6 months <strong>of</strong> a fluoropyrimidine/oxaliplatin<br />
combination therapy (Table 2). Investigators have agreed to<br />
pool data from all patients with stage III disease enrolled in<br />
the trials to test for noninferiority through a preplanned<br />
pooling project, the International Duration Evaluation <strong>of</strong><br />
Adjuvant Chemotherapy (IDEA) prospective pooled analysis.<br />
Noninferiority will be declared if the two-sided 95%<br />
confidence interval for the HR comparing 3 to 6 months <strong>of</strong><br />
therapy lies entirely below 1.10, which translates to an<br />
approximately 1.5% difference in DFS.<br />
Adjunctive Therapy Considerations<br />
Epidemiologic and scientific research indicates that diet<br />
and other lifestyle factors influence the risk <strong>of</strong> developing<br />
colorectal cancer. 40 Obesity, consumption <strong>of</strong> red meat, a<br />
Western-pattern diet, alcohol, and smoking influence one’s<br />
risk <strong>of</strong> developing colorectal cancer; physical activity, calcium,<br />
vitamin D, postmenopausal estrogen use, aspirin,<br />
nonsteroidal anti-inflammatory drugs (NSAIDs), and possibly<br />
folate decrease one’s risk. Until recently, whether these<br />
factors influence outcomes in patients already diagnosed<br />
with colorectal cancer was largely unknown. Data on factors<br />
that may influence disease recurrences and mortality for<br />
survivors <strong>of</strong> colon cancer are emerging.<br />
Energy-Balance Factors and Colon Cancer Outcomes<br />
Multiple prospective cohort studies have tested the influence<br />
<strong>of</strong> physical activity before and after diagnosis and<br />
changes in physical activity before and after diagnosis. 41-45<br />
Haydon and colleagues reported that physical activity before<br />
diagnosis was associated with a 51% improvement in DFS in<br />
patients with stage II and III colorectal cancer. 46 In contrast,<br />
physical activity before diagnosis did not influence<br />
colorectal cancer–specific mortality in women diagnosed<br />
with stages I to III colorectal cancer participating in the<br />
Nurse’s Health Study, but activity after diagnosis did improve<br />
outcomes. Compared to women engaged physical activity<br />
less than 3 metabolic equivalent task (MET)-hours<br />
per week (i.e., considered fairly inactive), those engaged in<br />
at least 18 MET-hours per week (i.e., the equivalent to 1<br />
hour <strong>of</strong> moderate walking daily at least 6 days per week)<br />
had an adjusted HR for colorectal cancer–specific mortality<br />
<strong>of</strong> 0.39 (95% CI, 0.18 to 0.82) and an adjusted HR for overall<br />
TOSCA SCOT CALGB/SWOG 80,702 GERCOR HORG<br />
Country Italy United Kingdom plus other<br />
European sites<br />
Principal Investigator (s) Alberto Sobrero Timothy Iveson and<br />
Jim Paul<br />
United States and Canada France Greece<br />
Jeffrey Meyerhardt and<br />
Anthony Shields<br />
Thierry Andre and<br />
Julien Taieb<br />
Accrual Goal 3,500 9,500 2,500 2,000 1,000<br />
Inclusion Criteria High risk stage II and III colon<br />
cancer<br />
Additional Features Previously included bevacizumab<br />
randomization but dropped<br />
after other bevacizumab trials<br />
reports<br />
Stage II and III colon and<br />
rectal cancer<br />
Stage III colon cancer Stage III colon<br />
cancer<br />
2 � 2 randomization - duration<br />
question and celecoxib or<br />
placebo<br />
Ioannis Souglakos<br />
High risk stage II<br />
and stage III<br />
colon cancer<br />
mortality <strong>of</strong> 0.43 (95% CI, 0.25 to 0.74). 41 Further, women<br />
who increased their activity after diagnosis had an HR <strong>of</strong><br />
0.48 (95% CI, 0.24 to 0.97) for death from colorectal cancer<br />
and an HR <strong>of</strong> 0.51 (95% CI, 0.30 to 0.85) for death from<br />
any cause, compared to those with no change in activity.<br />
Similar results were seen in a cohort <strong>of</strong> male health pr<strong>of</strong>essionals<br />
participating in the Health Pr<strong>of</strong>essionals Follow-up<br />
Study. 44 These results were further supported by data from<br />
a prospective cohort study within an adjuvant therapy trial<br />
sponsored by the National Cancer Institute (Cancer and<br />
Leukemia Group B [CALGB] trial 89803). 42 Among 832<br />
patients with stage III colon cancer who survived and were<br />
recurrence-free approximately 6 months after adjuvant<br />
chemotherapy, physical activity after diagnosis improved<br />
DFS by approximately 50% beyond surgery and adjuvant<br />
chemotherapy. In addition, a recent cohort from Australia<br />
<strong>of</strong> more than 1,800 survivors <strong>of</strong> stage I to III colorectal<br />
cancer demonstrated that physical activity after diagnosis<br />
improved overall mortality by approximately 25%. 45 These<br />
data led to a multinational trial in Canada and Australia,<br />
the Colon Health and Life-Long Exercise Change<br />
(CHALLENGE) trial, designed to determine the effects <strong>of</strong><br />
a 3-year structured and supervised physical activity intervention<br />
on disease outcomes in 962 high-risk survivors <strong>of</strong><br />
stage II and III colon cancer who completed adjuvant chemotherapy<br />
within the previous 2 to 6 months. 47 The trial is<br />
currently open to accrual.<br />
Although the initial assumption may be that physically<br />
active survivors <strong>of</strong> colorectal cancer have lower body mass<br />
indexes or that exercise lowers body weight, most studies<br />
examining obesity and colorectal cancer survival have<br />
shown only modest associations, if any. 48 Only one study has<br />
reported on the effect <strong>of</strong> change in weight on outcomes in<br />
survivors <strong>of</strong> colorectal cancer. Patients enrolled in CALGB<br />
89803 reported weight during chemotherapy and approximately<br />
6 months after completion <strong>of</strong> chemotherapy and<br />
change in weight or body mass index was not associated<br />
with either DFS or OS. 49<br />
Although various dietary factors have been associated<br />
with the development <strong>of</strong> colorectal cancer, only one large,<br />
prospective cohort study has tested whether diet is associated<br />
with outcomes in survivors <strong>of</strong> colon cancer. 50 In CALGB<br />
89803, participants completed a food frequency questionnaire<br />
during adjuvant therapy and 6 months after the<br />
completion <strong>of</strong> adjuvant therapy. Using these data, two major<br />
dietary patterns were identified: prudent pattern, character-<br />
227
ized by high fruit, vegetable, poultry, and fish intakes, and<br />
Western pattern, characterized by high meat, fat, refined<br />
grains, and dessert intakes. All patients were assigned a<br />
relative value along the spectrum <strong>of</strong> both dietary patterns<br />
and the two patterns were not correlated with each other.<br />
Compared with patients in the lowest quintile <strong>of</strong> the<br />
Western-pattern diet, those in the highest quintile experienced<br />
an adjusted HR for DFS <strong>of</strong> 3.25 (95% CI, 2.04 to 5.19;<br />
p for trend � 0.0001). In contrast, the prudent-pattern diet<br />
was not significantly associated with cancer recurrence or<br />
mortality. Studies on the components <strong>of</strong> the Westernpattern<br />
diet that influenced these findings are underway.<br />
NSAIDs, Vitamin D, and Colon Cancer Outcomes<br />
Outside <strong>of</strong> energy-balance factors, research is emerging on<br />
the effects <strong>of</strong> aspirin, NSAIDs, and vitamin D. Prospective<br />
cohort studies have demonstrated that aspirin and NSAIDs<br />
reduce the risk <strong>of</strong> developing colorectal cancer, and several<br />
randomized trials have shown that these agents reduce the<br />
number <strong>of</strong> subsequent polyps in patients with a history <strong>of</strong><br />
adenomatous polyps. 51-54 Initial studies have also shown a<br />
beneficial association <strong>of</strong> these agents for survivors <strong>of</strong> colorectal<br />
cancer. 55 In the CALGB 89803 cohort <strong>of</strong> patients with<br />
stage III colon cancer, regular use <strong>of</strong> aspirin (defined as<br />
consistent use both while receiving adjuvant therapy and for<br />
6 months after the completion <strong>of</strong> adjuvant therapy) had a<br />
significant 54% improvement in DFS compared with patients<br />
that did not regularly use aspirin (p � 0.001). 55<br />
Similarly, for patients with stages I to III colorectal cancer<br />
from both the Nurse’s Health Study and Health Pr<strong>of</strong>essional<br />
Follow-up Study, regular aspirin users had a more modest<br />
but still significant improvement in colorectal cancer mortality<br />
compared to nonusers (HR � 0.72; 95% CI, 0.54 to<br />
0.97, p � 0.03). 56<br />
The Vioxx in Colorectal Cancer Therapy: Definition <strong>of</strong><br />
Optimal Regime (VICTOR) trial was a randomized controlled<br />
trial in the United Kingdom evaluating the role <strong>of</strong><br />
r<strong>of</strong>ecoxib, a cyclooxygenase 2 inhibitor, for patients with<br />
stage II and III colon cancer who completed adjuvant therapy.<br />
57 The trial activated in April 2002 but was terminated<br />
in September 2004 after the withdrawal <strong>of</strong> r<strong>of</strong>ecoxib from<br />
the market. Before study closure, 2,434 patients were randomly<br />
assigned to placebo, 2 years <strong>of</strong> r<strong>of</strong>ecoxib, or 5 years <strong>of</strong><br />
r<strong>of</strong>ecoxib, with a median duration <strong>of</strong> time on trial <strong>of</strong> 7.4<br />
months. The HR for DFS with a median follow-up <strong>of</strong> 3 years<br />
was 0.90 (95% CI, 0.77 to 1.06), favoring r<strong>of</strong>ecoxib among<br />
patients with stage II and III disease. However the early<br />
closure <strong>of</strong> the trial and the limited time patients received the<br />
study medication in relationship to the intent <strong>of</strong> the trial<br />
limit conclusions from this trial. Given the observational<br />
Authors’ Disclosures <strong>of</strong> Potential Conflicts <strong>of</strong> Interest<br />
Author<br />
Employment or<br />
Leadership<br />
Positions<br />
Consultant or<br />
Advisory Role<br />
data and the potential benefit seen in VICTOR, the CALGB<br />
and Southwestern <strong>Oncology</strong> Group (SWOG) have developed<br />
a trial (CALGB/SWOG 80702, Table 2) that tests the duration<br />
<strong>of</strong> FOLFOX therapy, with a second randomization <strong>of</strong><br />
celecoxib compared with placebo initiated concurrent with<br />
the start <strong>of</strong> FOLFOX and continuing for 3 years. The trial<br />
will test for superiority in DFS <strong>of</strong> celecoxib compared with<br />
placebo.<br />
Increasing evidence suggests an association between vitamin<br />
D and development <strong>of</strong> colorectal cancer. A metaanalysis<br />
<strong>of</strong> five epidemiologic studies found a 51% decrease<br />
in the risk <strong>of</strong> colorectal cancer associated with plasma<br />
25(OH)D levels in the highest quintile compared to those in<br />
the lowest quintile (p � 0.0001). 58 Although studies on<br />
vitamin D in patients with colon cancer are limited, two<br />
recent studies by Ng and colleagues using the Nurse’s<br />
Health Study and the Health Pr<strong>of</strong>essional Follow-up Study<br />
cohorts suggest that either higher levels <strong>of</strong> serum vitamin D<br />
before diagnosis or greater predictive vitamin D serum<br />
levels are associated with improved colorectal cancer–specific<br />
mortality, overall mortality, or both. 59,60<br />
Conclusion<br />
Adjuvant chemotherapy has taken an important place in<br />
the treatment <strong>of</strong> patients with stage III colon cancer. The<br />
5-year survival <strong>of</strong> patients with stage III colon cancer has<br />
increased from less than 50% with surgery alone (before<br />
1990) to more than 70% with surgery followed by adjuvant<br />
fluoropyrimidines combined with oxaliplatin (since 2003).<br />
However, despite this progress, multiple disappointments<br />
have emerged including the realization that three <strong>of</strong> the four<br />
agent classes approved for metastatic colorectal cancer since<br />
1995 have failed to improve the chance <strong>of</strong> cure for patients<br />
with stages II and III colorectal cancer. We must learn from<br />
the experiences <strong>of</strong> these trials to avoid future expensive<br />
failures, and must also begin to consider that not all colon<br />
cancers are the same. The future, as is the case for most<br />
areas <strong>of</strong> cancer, must involve rational trial designs based<br />
on well-characterized and clinically validated biology. The<br />
study <strong>of</strong> agents with marginal survival benefit in unselected<br />
populations is discouraged based on the experiences outlined<br />
above. Although most <strong>of</strong> the current trials are not<br />
adding new agents to FOLFOX (with the exception <strong>of</strong> the<br />
CALGB/SWOG 80702 trial), the focus on the potential for<br />
less therapy and presumed less long-term neuropathy is<br />
important. Finally, beyond standard chemotherapies, continued<br />
research on other adjunctive therapies is critical to<br />
further understand the role <strong>of</strong> energy balance and other host<br />
factors in colon cancer survival.<br />
Stock<br />
Ownership Honoraria<br />
Research<br />
Funding<br />
Thierry André Roche; San<strong>of</strong>i Baxter<br />
International;<br />
Roche; Yacult<br />
Bert H. O’Neil Amgen;<br />
Bayer/Onyx<br />
Amgen<br />
Jeffrey A. Meyerhardt Bayer AstraZeneca;<br />
Bristol-Myers<br />
Squibb<br />
228<br />
ANDRÉ, O’NEIL, AND MEYERHARDT<br />
Expert<br />
Testimony<br />
Other<br />
Remuneration
ADJUVANT THERAPY FOR STAGE III COLON CANCER<br />
1. Ferlay J, Shin HR, Bray F, et al. Estimates <strong>of</strong> worldwide burden <strong>of</strong><br />
cancer in 2008: GLOBOCAN 2008. Int J Cancer. 2010;127:2893-2917.<br />
2. Chou JF, Row D, Gonen M, et al. <strong>Clinical</strong> and pathologic factors that<br />
predict lymph node yield from surgical specimens in colorectal cancer: a<br />
population-based study. Cancer. 2010;116:2560-2570.<br />
3. Edge SB, Compton CC. The <strong>American</strong> Joint Committee on Cancer: the<br />
7th edition <strong>of</strong> the AJCC Cancer Staging Manual and the future <strong>of</strong> TNM. Ann<br />
Surg Oncol. 2010;17:1471-1474.<br />
4. Moertel CG, Fleming TR, Macdonald JS, et al. Levamisole and fluorouracil<br />
for adjuvant therapy <strong>of</strong> resected colon carcinoma. N Engl J Med.<br />
1990;322:352-358.<br />
5. Moertel CG, Fleming TR, Macdonald JS, et al. Fluorouracil plus levamisole<br />
as effective adjuvant therapy after resection <strong>of</strong> stage III colon carcinoma:<br />
a final report. Ann Intern Med. 1995;122:321-326.<br />
6. Heidelberger C, Chaudhuri NK, Danneberg P, et al. Fluorinated pyrimidines,<br />
a new class <strong>of</strong> tumour-inhibitory compounds. Nature. 1957;179:663-<br />
666.<br />
7. Moertel CG. Chemotherapy <strong>of</strong> gastrointestinal cancer. N Engl J Med.<br />
1978;299:1049-1052.<br />
8. Buyse M, Zeleniuch-Jacquotte A, Chalmers TC. Adjuvant therapy <strong>of</strong><br />
colorectal cancer. Why we still don’t know. JAMA. 1988;259:3571-3578.<br />
9. Efficacy <strong>of</strong> adjuvant fluorouracil and folinic acid in colon cancer. International<br />
Multicentre Pooled Analysis <strong>of</strong> Colon Cancer Trials (IMPACT)<br />
investigators. Lancet. 1995;345:939-944.<br />
10. O’Connell MJ, Mailliard JA, Kahn MJ, et al. Controlled trial <strong>of</strong><br />
fluorouracil and low-dose leucovorin given for 6 months as postoperative<br />
adjuvant therapy for colon cancer. J Clin Oncol. 1997;15:246-250.<br />
11. Wolmark N, Rockette H, Mamounas E, et al. <strong>Clinical</strong> trial to assess the<br />
relative efficacy <strong>of</strong> fluorouracil and leucovorin, fluorouracil and levamisole,<br />
and fluorouracil, leucovorin, and levamisole in patients with Dukes’ B and C<br />
carcinoma <strong>of</strong> the colon: results from National Surgical Adjuvant Breast and<br />
Bowel Project C-04. J Clin Oncol. 1999;17:3553-3559.<br />
12. Haller DG, Catalano PJ, Macdonald JS, et al. Phase III study <strong>of</strong><br />
fluorouracil, leucovorin, and levamisole in high-risk stage II and III colon<br />
cancer: final report <strong>of</strong> Intergroup 0089. J Clin Oncol. 2005;23:8671-8678.<br />
13. Andre T, Colin P, Louvet C, et al. Semimonthly versus monthly<br />
regimen <strong>of</strong> fluorouracil and leucovorin administered for 24 or 36 weeks as<br />
adjuvant therapy in stage II and III colon cancer: results <strong>of</strong> a randomized<br />
trial. J Clin Oncol. 2003;21:2896-2903.<br />
14. Sargent D, Sobrero A, Grothey A, et al. Evidence for cure by adjuvant<br />
therapy in colon cancer: observations based on individual patient data from<br />
20,898 patients on 18 randomized trials. J Clin Oncol. 2009;27:872-877.<br />
15. Twelves C, Wong A, Nowacki MP, et al. Capecitabine as adjuvant<br />
treatment for stage III colon cancer. N Engl J Med. 2005;352:2696-2704.<br />
16. Lembersky BC, Wieand HS, Petrelli NJ, et al. Oral uracil and tegafur<br />
plus leucovorin compared with intravenous fluorouracil and leucovorin in<br />
stage II and III carcinoma <strong>of</strong> the colon: results from National Surgical<br />
Adjuvant Breast and Bowel Project Protocol C-06. J Clin Oncol. 2006;24:2059-<br />
2064.<br />
17. Sargent DJ, Wieand HS, Haller DG, et al. Disease-free survival versus<br />
overall survival as a primary end point for adjuvant colon cancer studies:<br />
individual patient data from 20,898 patients on 18 randomized trials. J Clin<br />
Oncol. 2005;23:8664-8670.<br />
18. André T, Boni C, Navarro M, et al. Improved overall survival with<br />
oxaliplatin, fluorouracil, and leucovorin as adjuvant treatment in stage II or<br />
III colon cancer in the MOSAIC trial. J Clin Oncol. 2009;27:3109-3116.<br />
19. Haller DG, Tabernero J, Maroun J, et al. Capecitabine plus oxaliplatin<br />
compared with fluorouracil and folinic acid as adjuvant therapy for stage III<br />
colon cancer. J Clin Oncol. 2011;29:1465-1471.<br />
20. Kuebler JP, Wieand HS, O’Connell MJ, et al. Oxaliplatin combined<br />
with weekly bolus fluorouracil and leucovorin as surgical adjuvant chemotherapy<br />
for stage II and III colon cancer: results from NSABP C-07. J Clin<br />
Oncol. 2007;25:2198-2204.<br />
21. Land SR, Kopec JA, Cecchini RS, et al. Neurotoxicity from oxaliplatin<br />
combined with weekly bolus fluorouracil and leucovorin as surgical adjuvant<br />
chemotherapy for stage II and III colon cancer: NSABP C-07. J Clin Oncol.<br />
2007;25:2205-2211.<br />
22. Allegra CJ, Yothers G, O’Connell MJ, et al. Phase III trial assessing<br />
bevacizumab in stages II and III carcinoma <strong>of</strong> the colon: results <strong>of</strong> NSABP<br />
protocol C-08. J Clin Oncol. 2011;29:11-16.<br />
23. Sargent DJ, Goldberg RM, Jacobson SD, et al. A pooled analysis <strong>of</strong><br />
adjuvant chemotherapy for resected colon cancer in elderly patients. N Engl<br />
J Med. 2001;345:1091-1097.<br />
24. Goldberg RM, Tabah-Fisch I, Bleiberg H, et al. Pooled analysis <strong>of</strong> safety<br />
and efficacy <strong>of</strong> oxaliplatin plus fluorouracil/leucovorin administered bi-<br />
REFERENCES<br />
monthly in elderly patients with colorectal cancer. J Clin Oncol. 2006;24:<br />
4085-4091.<br />
25. Jackson McCleary NA, Meyerhardt J, Green E, et al. Impact <strong>of</strong> older<br />
age on the efficacy <strong>of</strong> newer adjuvant therapies in �12,500 patients (pts) with<br />
stage II/III colon cancer: findings from the ACCENT database. J Clin Oncol.<br />
2009;27:15s (suppl; abstr 4010).<br />
26. Tournigand C, Andre T, Bachet J, et al. FOLFOX4 as adjuvant therapy<br />
in elderly patients (pts) with colon cancer (CC): subgroup analysis <strong>of</strong> the<br />
MOSAIC trial. J Clin Oncol. 2010;28:15s (suppl; abstr 3522).<br />
27. Lucas AS, O’Neil BH, Goldberg RM. A decade <strong>of</strong> advances in cytotoxic<br />
chemotherapy for metastatic colorectal cancer. Clin Colorectal Cancer. 2011;<br />
10:238-244.<br />
28. Papadimitriou CA, Papakostas P, Karina M, et al. A randomized phase<br />
III trial <strong>of</strong> adjuvant chemotherapy with irinotecan, leucovorin and fluorouracil<br />
versus leucovorin and fluorouracil for stage II and III colon cancer: a<br />
Hellenic Cooperative <strong>Oncology</strong> Group study. BMC Med. 2011;9:10.<br />
29. Saltz LB, Niedzwiecki D, Hollis D, et al. Irinotecan fluorouracil plus<br />
leucovorin is not superior to fluorouracil plus leucovorin alone as adjuvant<br />
treatment for stage III colon cancer: results <strong>of</strong> CALGB 89803. J Clin Oncol.<br />
2007;25:3456-3461.<br />
30. Van Cutsem E, Labianca R, Bodoky G, et al. Randomized phase III trial<br />
comparing biweekly infusional fluorouracil/leucovorin alone or with irinotecan<br />
in the adjuvant treatment <strong>of</strong> stage III colon cancer: PETACC-3. J Clin<br />
Oncol. 2009;27:3117-3125.<br />
31. Warren RS, Yuan H, Matli MR, et al. Regulation by vascular endothelial<br />
growth factor <strong>of</strong> human colon cancer tumorigenesis in a mouse model <strong>of</strong><br />
experimental liver metastasis. J Clin Invest. 1995;95:1789-1797.<br />
32. Hurwitz H, Fehrenbacher L, Novotny W, et al. Bevacizumab plus<br />
irinotecan, fluorouracil, and leucovorin for metastatic colorectal cancer.<br />
N Engl J Med. 2004;350:2335-2342.<br />
33. De Gramont A, Van Cutsem E, Tabernero J, et al. AVANT: Results<br />
from a randomized, three-arm multinational phase III study to investigate<br />
bevacizumab with either XELOX or FOLFOX4 versus FOLFOX4 alone as<br />
adjuvant treatment for colon cancer. J Clin Oncol. 2011; (suppl; abstr 362)<br />
34. Carmeliet P, Jain RK. Principles and mechanisms <strong>of</strong> vessel normalization<br />
for cancer and other angiogenic diseases. Nat Rev Drug Discov. 2011;10:<br />
417-427.<br />
35. Stein S, Cohen DJ, Hochster HS. Treatment paradigms with epidermal<br />
growth factor receptor-targeted therapies in colorectal cancer. Clin Colorectal<br />
Cancer. 2010; 9 Suppl 1:S44-S50.<br />
36. Alberts SR, Sargent DJ, Smyrk TC, et al. Adjuvant mFOLFOX6 with or<br />
without cetuxiumab (Cmab) in KRAS wild-type (WT) patients (pts) with<br />
resected stage III colon cancer (CC): results from NCCTG Intergroup Phase<br />
III Trial N0147. J Clin Oncol. 2010;28:18s (suppl; abstr CRA3507).<br />
37. Goldberg RM, Sargent DJ, Thibodeau SN, et al. Adjuvant mFOLFOX6<br />
plus or minus cetuximab (Cmab) in patients (pts) with KRAS mutant (m)<br />
resected stage III colon cancer (CC): NCCTG Intergroup Phase III Trial<br />
N0147. J Clin Oncol. 2010;28:15s (suppl; abstr 3508).<br />
38. Huang J, Sargent DJ, Mahoney MR, et al. Adjuvant FOLFIRI with or<br />
without cetuximab in patients with resected stage III colon cancer: NCCTG<br />
Intergroup phase III trial N0147. J Clin Oncol. 2011; (suppl; abstr 363).<br />
39. Chau I, Norman AR, Cunningham D, et al. A randomised comparison<br />
between 6 months <strong>of</strong> bolus fluorouracil/leucovorin and 12 weeks <strong>of</strong> protracted<br />
venous infusion fluorouracil as adjuvant treatment in colorectal cancer. Ann<br />
Oncol. 2005;16:549-557.<br />
40. Wei EK, Giovannucci E, Wu K, et al. Comparison <strong>of</strong> risk factors for<br />
colon and rectal cancer. Int J Cancer. 2004;108:433-442.<br />
41. Meyerhardt JA, Giovannucci EL, Holmes MD, et al. Physical activity<br />
and survival after colorectal cancer diagnosis. J Clin Oncol. 2006;24:3527-<br />
3534.<br />
42. Meyerhardt JA, Heseltine D, Niedzwiecki D, et al. Impact <strong>of</strong> physical<br />
activity on cancer recurrence and survival in patients with stage III colon<br />
cancer: findings from CALGB 89803. J Clin Oncol. 2006;24:3535-3541.<br />
43. Haydon AM, Macinnis R, English D, et al. Effect <strong>of</strong> physical activity<br />
and body size on survival after diagnosis with colorectal cancer. Gut. 2006;<br />
55:62-67.<br />
44. Meyerhardt JA, Giovannucci EL, Ogino S, et al. Physical activity and<br />
male colorectal cancer survival. Arch Intern Med. 2009;169:2102-2108.<br />
45. Baade PD, Meng X, Youl PH, et al. The impact <strong>of</strong> body mass index and<br />
physical activity on mortality among patients with colorectal cancer in<br />
Queensland, Australia. Cancer Epidemiol Biomarkers Prev. 2011;20:1410-1420.<br />
46. Haydon AM, Macinnis RJ, English DR, et al. Physical activity, insulinlike<br />
growth factor 1, insulin-like growth factor binding protein 3, and survival<br />
from colorectal cancer. Gut. 2006;55:689-694.<br />
47. Courneya KS, Booth CM, Gill S, et al. The Colon Health and Life-Long<br />
229
Exercise Change trial: a randomized trial <strong>of</strong> the National Cancer Institute <strong>of</strong><br />
Canada <strong>Clinical</strong> Trials Group. Curr Oncol. 2008;15:271-278.<br />
48. Meyerhardt JA, Ma J, Courneya KS. Energetics in colorectal and<br />
prostate cancer. J Clin Oncol. 2010;28:4066-4073.<br />
49. Meyerhardt JA, Niedzwiecki D, Hollis D, et al. Impact <strong>of</strong> body mass<br />
index and weight change after treatment on cancer recurrence and survival in<br />
patients with stage III colon cancer: findings from Cancer and Leukemia<br />
Group B 89803. J Clin Oncol. 2008;26:4109-4115.<br />
50. Meyerhardt JA, Niedzwiecki D, Hollis D, et al. Association <strong>of</strong> dietary<br />
patterns with cancer recurrence and survival in patients with stage III colon<br />
cancer. JAMA. 2007;298:754-764.<br />
51. Baron JA, Cole BF, Sandler RS, et al. A randomized trial <strong>of</strong> aspirin to<br />
prevent colorectal adenomas. N Engl J Med. 2003;348:891-899.<br />
52. Baron JA, Sandler RS, Bresalier RS, et al. A randomized trial <strong>of</strong><br />
r<strong>of</strong>ecoxib for the chemoprevention <strong>of</strong> colorectal adenomas. Gastroenterology.<br />
2006;131:1674-1682.<br />
53. Bertagnolli MM, Eagle CJ, Zauber AG, et al. Celecoxib for the prevention<br />
<strong>of</strong> sporadic colorectal adenomas. N Engl J Med. 2006;355:873-884.<br />
54. Chan AT, Giovannucci EL, Meyerhardt JA, et al. Long-term use <strong>of</strong><br />
aspirin and nonsteroidal anti-inflammatory drugs and risk <strong>of</strong> colorectal<br />
cancer. JAMA. 2005;294:914-923.<br />
230<br />
ANDRÉ, O’NEIL, AND MEYERHARDT<br />
55. Fuchs C, Meyerhardt JA, Heseltine DL, et al. Influence <strong>of</strong> regular<br />
aspirin use on survival for patients with stage III colon cancer: findings from<br />
Intergroup trial CALGB 89803. J Clin Oncol. 2005;23: (suppl abstr 3530).<br />
56. Chan AT, Ogino S, Fuchs CS. Aspirin use and survival after diagnosis<br />
<strong>of</strong> colorectal cancer. JAMA. 2009;302:649-659.<br />
57. Kerr DJ, Dunn JA, Langman MJ, et al. VICTOR: a phase III placebocontrolled<br />
trial <strong>of</strong> r<strong>of</strong>ecoxib in colorectal cancer patients following surgical<br />
resection. J Clin Oncol. 2008;26: (suppl abstr 4038).<br />
58. Gorham ED, Garland CF, Garland FC, et al. Optimal vitamin D status<br />
for colorectal cancer prevention: a quantitative meta analysis. Am J Prev<br />
Med. 2007;32:210-216.<br />
59. Ng K, Meyerhardt JA, Wu K, et al. Circulating 25-hydroxyvitamin d<br />
levels and survival in patients with colorectal cancer. J Clin Oncol. 2008;26:<br />
2984-2991.<br />
60. Ng K, Wolpin BM, Meyerhardt JA, et al. Prospective study <strong>of</strong> predictors<br />
<strong>of</strong> vitamin D status and survival in patients with colorectal cancer. Br J<br />
Cancer. 2009;101:916-923.<br />
61. André T, Boni C, Mounedji-Boudiaf L, et al. Oxaliplatin, fluorouracil,<br />
and leucovorin as adjuvant treatment for colon cancer. N Engl J Med.<br />
2004;350:2343-2351.
EVOLVING PARADIGMS IN THE MANAGEMENT OF<br />
UNRESECTABLE PANCREATIC CANCER<br />
CHAIR<br />
Hedy Lee Kindler, MD<br />
The University <strong>of</strong> Chicago Medical Center<br />
Chicago, IL<br />
SPEAKERS<br />
Theodore S. Hong, MD<br />
Massachusetts General Hospital<br />
Boston, MA<br />
Lauren A. Wiebe, MD<br />
Medical College <strong>of</strong> Wisconsin<br />
Milwaukee, WI
A New Direction for Pancreatic Cancer<br />
Treatment: FOLFIRINOX in Context<br />
Overview: Since 1996, the cornerstone <strong>of</strong> chemotherapy for<br />
advanced pancreatic cancer has been gemcitabine, which has<br />
a genuine, but modest effect on survival and quality <strong>of</strong> life. It<br />
has been remarkably difficult to improve on these outcomes.<br />
Many phase III studies <strong>of</strong> gemcitabine doublets have been<br />
uniformly negative, with the exception <strong>of</strong> a trial <strong>of</strong> gemctabine<br />
plus erlotinib, which provided only marginal benefit. In 2010,<br />
the FOLFIRINOX regimen (bolus and infusional 5-fluorouracil,<br />
irinotecan, and oxaliplatin) emerged as a major treatment<br />
advance for patients with metastatic pancreatic cancer. In a<br />
trial with 342 patients, FOLFIRINOX yielded a longer median<br />
overall survival (11.1 vs. 6.8 months, hazard ratio [HR] 0.57,<br />
p < 0.001), a superior progression-free survival (6.4 vs. 3.3<br />
months, HR 0.47, p < 0.001), a higher objective response rate<br />
(31.6% vs. 9.4%, p < 0.001), and a significant increase in time<br />
Background<br />
THERE HAS been a long-standing, well-deserved therapeutic<br />
nihilism regarding chemotherapy for advanced<br />
pancreatic cancer. In countless trials over the past few<br />
decades, many drugs and drug combinations have demonstrated<br />
minimal to no activity against this devastating<br />
disease.<br />
Since 1996, the cornerstone <strong>of</strong> therapy has been gemcitabine,<br />
an agent with a genuine, but modest impact. In the<br />
pivotal trial that compared gemcitabine to weekly bolus<br />
5-fluorouracil (5-FU), gemcitabine treatment yielded a response<br />
rate <strong>of</strong> 5% and a median overall survival <strong>of</strong> 5.7<br />
months. 1 Although these results do represent a real advance<br />
over 5-FU, gemcitabine is principally given because it provides<br />
a clinical benefit, a combination <strong>of</strong> an improvement <strong>of</strong><br />
pain and performance status, and a stabilization <strong>of</strong> weight.<br />
Despite these very modest outcomes, it has been difficult<br />
for any agent to displace gemcitabine for advanced pancreatic<br />
cancer. Most drugs simply do not work in this disease.<br />
Although innumerable phase II trials have reported “promising<br />
activity” for various gemcitabine-based cytotoxic and<br />
targeted doublets, phase III trials <strong>of</strong> these combinations<br />
have been uniformly disappointing, generally yielding<br />
greater toxicity for the multidrug regimen with no improvement<br />
in overall survival. 2-7<br />
This dismal outlook changed slightly in 2005, when the<br />
National Cancer Institute <strong>of</strong> Canada PA.3 trial demonstrated<br />
a small improvement in survival for patients treated<br />
with gemcitabine plus erlotinib, an epidermal growth factor<br />
receptor tyrosine kinase inhibitor. 8 Although the results<br />
were statistically significant, with a hazard ratio <strong>of</strong> 0.82, the<br />
absolute improvement in median overall survival, 5.91<br />
months with gemcitabine compared with 6.24 months for<br />
gemcitabine/erlotinib, was minimal. This combination also<br />
came with a substantial economic cost 9 and did not improve<br />
quality <strong>of</strong> life. One could easily question whether such an<br />
incremental improvement in overall survival was worth the<br />
expense and toxicity. 10 Over the next 5 years, several more<br />
negative phase III trials were reported, 11-16 and it certainly<br />
appeared that any major improvements were a long way <strong>of</strong>f.<br />
It was in this context that the results <strong>of</strong> the PRODIGE<br />
232<br />
By Hedy Lee Kindler, MD<br />
until definitive deterioration in quality <strong>of</strong> life, compared with<br />
gemcitabine. FOLFIRINOX is also more cost-effective than<br />
gemcitabine. Because <strong>of</strong> higher rates <strong>of</strong> grade 3 to 4 neutropenia<br />
(46% vs. 21%), febrile neutropenia (5% vs. 1%), and<br />
diarrhea (13% vs. 2%) with FOLFIRINOX, vigilant patient<br />
selection, education, and monitoring are essential. Retrospective<br />
single-institution series confirm the substantial activity <strong>of</strong><br />
FOLFIRINOX in metastatic, locally advanced, and previouslytreated<br />
patients; demonstrate its safety in individuals with<br />
biliary stents; and elucidate how physicians routinely modify<br />
drug doses without clear evidence or guidelines. Ongoing and<br />
planned studies will prospectively evaluate FOLFIRINOX in<br />
the adjuvant, locally advanced, and borderline resectable<br />
settings, will add targeted agents to FOLFIRINOX, and will<br />
evaluate how to adjust doses to ameliorate toxicity.<br />
4/ACCORD 11 trial, presented first at ASCO in 2010 17 and<br />
published in the New England Journal <strong>of</strong> Medicine in<br />
2011, 18 represent a substantial treatment advance for patients<br />
with pancreatic cancer. The multidrug combination<br />
FOLFIRINOX significantly improved median and<br />
progression-free survival, objective response rates, and quality<br />
<strong>of</strong> life, albeit with greater toxicity. Given all <strong>of</strong> the<br />
negative trials that preceded it, these data alone are a major<br />
achievement. What is truly unprecedented is the magnitude<br />
<strong>of</strong> the benefit achieved with this regimen.<br />
In this article, we will review the data behind this pivotal<br />
study, examine how oncologists are currently using this<br />
regimen, and assess the ways that FOLFIRINOX may be<br />
incorporated into pancreatic cancer treatment in the future.<br />
Development <strong>of</strong> the FOLFIRINOX Regimen: Phase I<br />
and II Studies<br />
It would seem logical to combine 5-FU, irinotecan, and<br />
oxaliplatin: these 3 drugs have activity in several gastrointestinal<br />
malignancies, including pancreatic cancer, without<br />
many overlapping toxicities. The doses for the FOLFIRI-<br />
NOX regimen were established in a phase I trial <strong>of</strong> 34<br />
evaluable patients with advanced solid tumors enrolled<br />
between 1998 and 2000. 19 Although the investigators may<br />
have been primarily interested in developing this combination<br />
for metastatic colon cancer, they also observed two<br />
responses (one complete and one partial) among the six<br />
patients with pancreatic cancer enrolled on the study.<br />
These encouraging data prompted the same investigators<br />
to evaluate this regimen in a single-arm phase II trial. 20<br />
Forty-six chemotherapy-naïve patients with pancreatic cancer<br />
with a World Health Organization (WHO) performance<br />
status <strong>of</strong> 0 or 1 enrolled at nine French centers from 2000 to<br />
2002. Most subjects (76%) had metastatic disease, a perfor-<br />
From the University <strong>of</strong> Chicago, Chicago, IL.<br />
Author’s disclosures <strong>of</strong> potential conflicts <strong>of</strong> interest are found at the end <strong>of</strong> this article.<br />
Address reprint requests to Hedy Lee Kindler, MD, Section <strong>of</strong> Hematology/<strong>Oncology</strong>,<br />
University <strong>of</strong> Chicago, 5841 S. Maryland Ave., MC 2115, Chicago, IL 60637; email:<br />
hkindler@medicine.bsd.uchicago.edu.<br />
© <strong>2012</strong> by <strong>American</strong> <strong>Society</strong> <strong>of</strong> <strong>Clinical</strong> <strong>Oncology</strong>.<br />
1092-9118/10/1-10
FOLFIRINOX FOR PANCREATIC CANCER<br />
mance status (PS) <strong>of</strong> 1 (74%), and liver metastases (61%).<br />
Responses, assessed using WHO criteria 21 by an external<br />
review committee, were observed in 26% <strong>of</strong> the patients<br />
(95% CI, 13% to 39%) and included 4% complete responses.<br />
Thirty-five percent <strong>of</strong> the subjects had stable disease, yielding<br />
a disease control rate <strong>of</strong> 61%. The median progressionfree<br />
survival was 8.2 months (95% CI, 5.3 to 11.6 months).<br />
The median overall survival was 10.2 months (95% CI, 8.1 to<br />
14.4 months); in patients with metastatic disease, the median<br />
overall survival was 9.9 months. Grades 3–4 neutropenia<br />
and diarrhea developed in 52% and 17% <strong>of</strong> patients,<br />
respectively. Quality <strong>of</strong> life, measured by the EORTC QLQ-<br />
C30, was improved in all functional scales except for cognitive<br />
functioning.<br />
The PRODIGE 4/ACCORD 11 Study<br />
The impressive activity observed with FOLFIRINOX in<br />
the phase I and II trials 19,20 prompted the development <strong>of</strong><br />
a randomized phase II/III study that compared this new<br />
regimen with the benchmark therapy, gemcitabine. 18 Key<br />
eligibility criteria included no prior chemotherapy, an<br />
ECOG Performance Status <strong>of</strong> 0 or 1, age less than 76 years,<br />
measurable metastatic disease, and a total bilirubin less<br />
than 1.5 times the upper limit <strong>of</strong> normal.<br />
Patients, stratified by center, performance status (0 vs. 1),<br />
and location <strong>of</strong> the primary tumor (head vs. body/tail)<br />
were randomized 1:1 to either FOLFIRINOX (oxaliplatin<br />
KEY POINTS<br />
● In a pivotal trial in 342 patients with metastatic<br />
pancreatic cancer and a good performance status,<br />
FOLFIRINOX (bolus and infusional 5-fluorouracil,<br />
irinotecan, and oxaliplatin) produced a longer median<br />
overall survival (11.1 vs. 6.8 months), a superior<br />
progression-free survival (6.4 vs. 3.3 months), a<br />
higher objective response rate (31.6% vs. 9.4%), and a<br />
significant increase in the time until definitive deterioration<br />
in quality <strong>of</strong> life compared with gemcitabine.<br />
● FOLFIRINOX is more cost-effective than gemcitabine.<br />
● Vigilant patient selection, education, and monitoring<br />
are essential with FOLFIRINOX treatment, because<br />
<strong>of</strong> higher rates <strong>of</strong> grade 3–4 neutropenia, febrile<br />
neutropenia, and diarrhea compared with gemcitabine.<br />
● Retrospective single-institution series confirm the<br />
substantial activity <strong>of</strong> FOLFIRINOX in metastatic,<br />
locally advanced, and previously-treated patients;<br />
demonstrate its safety in patients with biliary stents;<br />
and elucidate how physicians routinely modify drug<br />
doses without clear evidence or guidelines.<br />
● Ongoing and planned studies will prospectively evaluate<br />
FOLFIRINOX in the adjuvant, locally advanced,<br />
and borderline resectable settings, will add targeted<br />
agents to FOLFIRINOX, and will evaluate how to<br />
adjust doses to ameliorate toxicity.<br />
85 mg/m 2 , leucovorin 400 mg/m 2 , irinotecan 180 mg/m 2 ,<br />
bolus 5-fluorouracil 400 mg/m 2 followed by infusional<br />
5-fluorouracil 2,400 mg/m 2 given over 46 hours, every 14<br />
days) or gemcitabine (1,000 mg/m 2 over 30 minutes weekly<br />
for 7 <strong>of</strong> 8 weeks, then weekly for 3 <strong>of</strong> 4 weeks). Each cycle<br />
was defined as a 2-week interval for both regimens. Six<br />
months <strong>of</strong> treatment were recommended for responding<br />
patients. Filgrastim was not routinely administered for<br />
primary prophylaxis, though it was permitted for high-risk<br />
patients. Computed tomography (CT) scans were obtained<br />
every 2 months. RECIST criteria were employed for response<br />
assessment. 22 Quality <strong>of</strong> life was measured by<br />
EORTC QLQ-C30 questionnaires completed every 2 weeks.<br />
Response was the primary efficacy endpoint <strong>of</strong> the phase<br />
II portion <strong>of</strong> the study, which was planned to proceed to<br />
phase III if at least 12 objective responses occurred in the<br />
first 40 FOLFIRINOX-treated patients. Overall survival<br />
was the primary phase III endpoint. Phase II patients were<br />
included in the phase III analysis.<br />
The study was designed to have an 80% power to detect an<br />
increase in median overall survival from 7 to 10 months (HR<br />
0.70, � � 0.05). For the final analysis, 360 patients would be<br />
required to reach 250 events; an interim analysis was<br />
planned after 167 events occurred. 23 In September 2009, the<br />
Independent Data Management Committee recommended<br />
that accrual be stopped early, because a planned interim<br />
analysis determined that the primary endpoint was<br />
achieved with a p value <strong>of</strong> less than 0.001. 18,23<br />
Between 2005 and 2009, 342 patients enrolled at 48<br />
French centers. Patient characteristics were balanced between<br />
the arms for age, sex, performance status, tumor<br />
location, biliary stenting, metastatic sites, and baseline Ca<br />
19–9 level, except that a greater percentage <strong>of</strong> patients on<br />
the gemcitabine arm had measurable pulmonary metastases<br />
(29% vs. 19%). The median age was 61. Approximately 60%<br />
<strong>of</strong> the subjects in both arms had a PS <strong>of</strong> 1, and approximately<br />
87% had liver metastases. Only 38% <strong>of</strong> patients had<br />
tumors <strong>of</strong> the pancreatic head, and only 14% had a biliary<br />
stent.<br />
The median number <strong>of</strong> 2-week treatment cycles was 10 in<br />
the FOLFIRINOX arm and 6 in the gemcitabine arm (p �<br />
0.001). The median relative dose intensity was approximately<br />
80% for each <strong>of</strong> the component drugs in the FOL-<br />
FIRINOX regimen and 100% for gemcitabine.<br />
Independent radiologic review confirmed that 15 <strong>of</strong> the<br />
first 44 FOLFIRINOX-treated patients (34%) in the phase II<br />
portion <strong>of</strong> the trial had an objective response, meeting the<br />
criteria for the study to proceed to phase III. Patients<br />
treated with FOLFIRINOX achieved a much higher objective<br />
response rate (31.6%) than those who received singleagent<br />
gemcitabine (9.4%).<br />
Median overall survival was significantly longer in the<br />
patients treated with FOLFIRINOX (11.1 months vs. 6.8<br />
months, HR � 0.57, 95% CI, 0.45 to 0.73; p � 0.001). Overall<br />
survival rates at 6, 12, and 18 months were also superior for<br />
FOLFIRINOX-treated patients (76%, 48%, and 19% respectively),<br />
compared with 58%, 21%, and 6%, respectively for<br />
those who received gemcitabine. Patients on the multidrug<br />
regimen also achieved a superior progression-free survival<br />
(6.4 months vs. 3.3 months, HR � 0.47; 95% CI, 0.37 to<br />
0.59; p � 0.001). The beneficial effect <strong>of</strong> FOLFIRINOX was<br />
similar in all patient subgroups. These data are summarized<br />
in Table 1.<br />
233
Table 1. FOLFIRINOX versus Gemcitabine: Efficacy 18<br />
FOLFIRINOX Gemcitabine<br />
Patients who received FOLFIRINOX experienced significantly<br />
higher rates <strong>of</strong> grade 3 and 4 neutropenia (46% vs.<br />
21%), febrile neutropenia (5% vs. 1%), thrombocytopenia<br />
(9% vs. 4%), diarrhea (13% vs. 2%), and sensory neuropathy<br />
(9% vs. 0%) than those who received gemcitabine. The<br />
presence <strong>of</strong> a biliary stent did not increase the risk <strong>of</strong><br />
infection in either arm, and no cholangitis was reported.<br />
Filgrastim was given to 43% <strong>of</strong> FOLFIRINOX-treated patients.<br />
Toxicity data are summarized in Table 2.<br />
Significantly more patients on the gemcitabine arm had a<br />
definitive decrease in their scores on the Global Health<br />
Status and Quality <strong>of</strong> Life scale compared with those on the<br />
FOLFIRINOX arm (66% vs. 31%, HR � 0.47, p � 0.001). A<br />
significant increase in the time until definitive deterioration<br />
in quality <strong>of</strong> life was observed in the FOLFIRINOX-treated<br />
subjects for all functional and symptom scales.<br />
FOLFIRINOX Usage in <strong>Clinical</strong> Practice<br />
Hazard<br />
Ratio p value<br />
Patients 171 171<br />
Complete response<br />
Partial response (PR)<br />
0.6%<br />
31%<br />
0%<br />
9%<br />
0.0001<br />
Stable disease (SD) 39% 42%<br />
Disease control (PR � SD) 71% 51% 0.0003<br />
Median overall survival (months) 11.1 6.8 0.57 �0.0001<br />
1-yr survival 48% 21%<br />
18-mo survival 19% 6%<br />
Progression-free survival (months) 6.4 3.3 0.47 �0.0001<br />
Almost immediately after Conroy presented the FOLFIRI-<br />
NOX data at ASCO in 2010, Xcenda, LLC analyzed the<br />
prescribing plans <strong>of</strong> <strong>American</strong> oncologists using the NMCR<br />
Challenging Cases live research vehicle. 24 From July 31 to<br />
August 28, 2010, they assessed the prescribing plans <strong>of</strong> more<br />
than 370 U.S. oncologists for first-line therapy <strong>of</strong> patients<br />
with metastatic pancreatic cancer and PS 1 or 2. They<br />
observed that the FOLFIRINOX data produced an immediate<br />
change in the distribution <strong>of</strong> planned first-line prescribing.<br />
For the PS 1 scenario, FOLFIRINOX had an 18% share;<br />
in the previous year, those patients would mostly have<br />
received gemcitabine/erlotinib. As expected, plans for FOL-<br />
FIRINOX were negligible (3%) for PS 2 patients.<br />
In the phase I, II, and III trials described above, FOLFIRI-<br />
Table 2. FOLFIRINOX versus Gemcitabine: Selected Grade 3<br />
and 4 Toxicities 18<br />
FOLFIRINOX Gemcitabine<br />
Toxicity<br />
Hematologic<br />
(171 patients) (171 patients) p value<br />
Neutropenia 46% 21% �0.001<br />
Febrile neutropenia 5% 1% 0.03<br />
Thrombocytopenia 9% 4% 0.04<br />
Anemia<br />
Non-hematologic<br />
8% 6% NS<br />
Fatigue 24% 18% NS<br />
Vomiting 15% 8% NS<br />
Diarrhea 13% 2% �0.001<br />
Sensory neuropathy 9% 0% �0.001<br />
Increased alanine aminotransferase 7% 21% �0.001<br />
Thrombosis 7% 4% NS<br />
Abbreviation: NS, not significant.<br />
234<br />
HEDY LEE KINDLER<br />
NOX was tested in European patients who had predominantly<br />
body/tail tumors and a good performance status,<br />
which may not be fully representative <strong>of</strong> other populations <strong>of</strong><br />
patients with pancreatic cancer. In the absence <strong>of</strong> data from<br />
additional prospective clinical trials, retrospective singleinstitution<br />
series from highly selected academic centers<br />
provide some insight into the current state <strong>of</strong> FOLFIRINOX<br />
usage in the United States and elsewhere. 25-30 These data<br />
confirm the considerable activity <strong>of</strong> this regimen in the<br />
metastatic, locally advanced, and previously-treated settings,<br />
demonstrate the safety <strong>of</strong> this combination in patients<br />
with indwelling biliary stents, and elucidate how some<br />
physicians are routinely modifying drug doses without clear<br />
evidence or guidelines.<br />
Between July 2010 and April 2011, investigators at Massachusetts<br />
General Hospital treated 29 patients with pancreatic<br />
cancer with FOLFIRINOX; 59% had metastatic<br />
disease. 25 The median age was 60. Most (62%) were<br />
chemotherapy-naïve, 17% and 21%, respectively, had received<br />
one and two prior regimens. Almost half (48%) had a<br />
PS <strong>of</strong> 1; 55% had head or uncinate tumors, and 28% had<br />
biliary stents. A median <strong>of</strong> eight cycles were delivered. The<br />
objective response rate on formal radiologic review was 38%.<br />
In chemotherapy-naïve patients, the response rate was 56%;<br />
stable disease was reported in 39%, for a disease control rate<br />
<strong>of</strong> 94%. In previously-treated patients, the response rate was<br />
9%; the disease control rate was 73%. Response rates were<br />
similar in patients with locally advanced and metastatic<br />
disease (42% and 35%, respectively). Emergency room visits<br />
or hospitalizations were required in 41% <strong>of</strong> patients, most<br />
commonly for neutropenic fever or dehydration (14% each).<br />
Seven <strong>of</strong> the 10 patients who developed grades 3 or 4<br />
neutropenia had not received prophylactic growth factors<br />
from the start <strong>of</strong> FOLFIRINOX treatment.<br />
Washington University physicians used a registry to document<br />
the tolerance and efficacy <strong>of</strong> FOLFIRINOX. 26<br />
Twenty-nine patients, with a median age <strong>of</strong> 57, received at<br />
least one cycle <strong>of</strong> FOLFIRINOX. Most (93%) had an ECOG<br />
PS <strong>of</strong> 0 or 1, 38% had pancreatic head tumors, and 24% had<br />
biliary stents. The majority (52%) had metastatic disease;<br />
the rest were locally advanced. The regimen was empirically<br />
modified in the majority <strong>of</strong> patients because <strong>of</strong> concern for<br />
potential toxicities. The 5-FU bolus was deleted in 48%.<br />
Irinotecan was dose-reduced in four patients who had the<br />
UGT1A1*28/*28 genotype, and in nine patients who did not.<br />
Prophylactic growth factor support was administered to 48%<br />
<strong>of</strong> patients beginning with the first cycle; 10% initiated<br />
granulocyte-colony stimulating factor (G-CSF) in subsequent<br />
cycles. Grades 3 and 4 neutropenia developed in 14%,<br />
but only one patient experienced a neutropenic fever. Fourteen<br />
percent <strong>of</strong> the patients were hospitalized. Of the patients<br />
who began treatment on the full-dose regimen, only<br />
25% required any dose adjustment in future cycles, suggesting<br />
to these investigators that prophylactic dose adjustments<br />
may not have been necessary. On independent<br />
review, 26% achieved partial responses and 63% had stable<br />
disease. Thus, despite frequent dose reductions, the majority<br />
<strong>of</strong> patients achieved disease control with FOLFIRINOX.<br />
Investigators at Yale University observed that oncologists<br />
in community and academic practices were reluctant to use<br />
full-dose FOLFIRINOX because <strong>of</strong> its toxicity pr<strong>of</strong>ile. 27 To<br />
ascertain the potential effect <strong>of</strong> dose attenuation on toxicity<br />
and efficacy, they performed a retrospective review <strong>of</strong> pa-
FOLFIRINOX FOR PANCREATIC CANCER<br />
tients with pancreatic cancer who were treated with FOL-<br />
FIRINOX at their institution between June 2010 and June<br />
2011, and compared patient characteristics, toxicities, and<br />
response rates with those reported in the pivotal phase III<br />
trial. 18 Thirty-five patients, with a median age <strong>of</strong> 61, were<br />
treated with FOLFIRINOX; 68% had an ECOG PS <strong>of</strong> 0, 57%<br />
had pancreatic head tumors, 46% had locally advanced<br />
disease, and only 14% had received prior chemotherapy.<br />
Only 17% <strong>of</strong> patients received full-dose FOLFIRINOX<br />
with the first cycle. Irinotecan was reduced in 93% <strong>of</strong><br />
patients and omitted in 3%; oxaliplatin was reduced in 34%;<br />
bolus 5-FU was reduced in 31% and omitted in 24%; leucovorin<br />
was decreased in 37%; and the 5-FU continuous<br />
infusion was decreased in 10%. A median <strong>of</strong> 10 cycles was<br />
delivered. The median relative doses <strong>of</strong> oxaliplatin, irinotecan,<br />
bolus 5-FU, and infusional 5-FU were 90%, 68%, 68%,<br />
and 100%, respectively (in the phase III trial the median<br />
relative doses <strong>of</strong> oxaliplatin, irinotecan, and 5-FU [bolus and<br />
infusion] were 78%, 81%, and 82%, respectively). Yale patients<br />
experienced significantly less grade 3/4 fatigue (p �<br />
0.0089) and neutropenia (p � 0.0001) compared with patients<br />
in the phase III trial. Despite routine dose modifications,<br />
the response rate, progression-free survival, and<br />
overall survival were not significantly different from historic<br />
controls. The authors concluded that modest dose attenuations<br />
<strong>of</strong> FOLFIRINOX reduce toxicity but do not appear to<br />
compromise its efficacy.<br />
The activity <strong>of</strong> FOLFIRINOX in previously treated patients<br />
has been described in two retrospective series from<br />
France. 28,29 In a retrospective review <strong>of</strong> 27 patients who<br />
received second-line FOLFIRINOX from 2003 to 2009, a<br />
median <strong>of</strong> six cycles were delivered. Grades 3–4 neutropenia<br />
developed in 56%, and one patient experienced grade 5<br />
febrile neutropenia; 44% received growth factors as secondary<br />
prophylaxis. Partial responses were achieved in 19%;<br />
44% had stable disease. The median time to progression was<br />
5.4 months, and median overall survival was 8.5 months. 28<br />
In a second French series, 13 previously-treated patients,<br />
only 69% <strong>of</strong> whom had a PS <strong>of</strong> 0 or 1, were treated with<br />
FOLFIRINOX, mostly (85%) in the second-line setting.<br />
There were no objective responses, but 46% had stable<br />
disease. No grade 4 toxicities were reported. 29<br />
FOLFIRINOX Is Cost-effective<br />
Using a Markov model, Attard and colleagues assessed<br />
the cost-effectiveness <strong>of</strong> FOLFIRINOX compared with gemcitabine<br />
in Canadian patients undergoing first-line treatment<br />
for metastatic pancreatic cancer. 31 Since the initial<br />
treatment choice affects subsequent therapy, second-line<br />
treatment was also included in their analysis.<br />
Their first model, based on the ACCORD 11 trial data,<br />
assumed that in one arm patients received first-line FOL-<br />
FIRINOX and second-line gemcitabine, and in the other<br />
arm, they received first-line gemcitabine and second-line<br />
platinum-based chemotherapy; in both groups G-CSF usage<br />
was allowed.<br />
The second analysis, which mirrored current Ontario<br />
treatment patterns, assumed that first-line FOLFIRINOX<br />
was followed by second-line gemcitabine in one arm, and<br />
first-line gemcitabine was followed by best supportive care<br />
in the other arm; no G-CSF was permitted.<br />
In both scenarios, first-line FOLFIRINOX produced more<br />
life years and quality-adjusted life years (QALY) than firstline<br />
gemcitabine. The costs per QALYs for FOLFIRINOX<br />
were about $45,000 to $55,000. Thus, even though the<br />
component drugs <strong>of</strong> the FOLFIRINOX regimen cost more<br />
than single-agent gemcitabine, FOLFIRINOX is more costeffective<br />
than gemcitabine. FOLFIRINOX has therefore<br />
received a favorable funding recommendation in most Canadian<br />
provinces.<br />
Future Directions with FOLFIRINOX<br />
Given the impressive activity <strong>of</strong> FOLFIRINOX in the<br />
metastatic setting, plans are underway to prospectively<br />
evaluate this regimen in the adjuvant, locally advanced, and<br />
borderline resectable settings, studies are ongoing to add<br />
targeted agents to FOLFIRINOX, and trials are in development<br />
to determine how to adjust doses and ameliorate<br />
toxicity.<br />
PRODIGE 24-ACCORD 24/0610 will be the first trial to<br />
evaluate FOLFIRINOX in the adjuvant setting. This study<br />
will use a modified regimen, called mFOLFIRINOX, in<br />
which the bolus 5-FU has been omitted and all other drug<br />
doses remain unchanged. Eligible patients with resected<br />
head, body, or tail lesions, PS 0–1, age less than 80 years,<br />
total bilirubin less than1.5 X ULN, and Ca 19–9 less than<br />
180, will be stratified by center, node status, postoperative<br />
Ca 19–9 level (90 vs. 91–180), and surgical margin (R0 vs.<br />
R1), and randomized to 24 weeks <strong>of</strong> gemcitabine or mFOL-<br />
FIRINOX. The primary endpoint will be 3-year disease-free<br />
survival. A 30-patient lead-in safety analysis will soon be<br />
initiated to ascertain that the rate <strong>of</strong> grade 3–4 diarrhea is<br />
less than 5%. The study requires 490 patients to demonstrate<br />
a 10% increase in disease-free survival at 3 years,<br />
from 17% to 27%. It will be conducted in France and Canada.<br />
The safety <strong>of</strong> FOLFIRINOX in borderline resectable patients<br />
has been described in a retrospective series. 30<br />
A021101 is a 50-patient, single-arm, neoadjuvant phase II<br />
U.S. Intergroup study <strong>of</strong> mFOLFIRINOX followed by chemoradiation<br />
then surgery and postoperative gemcitabine for<br />
patients with borderline resectable pancreatic cancer. This<br />
benchmark trial will assess the feasibility <strong>of</strong> a multiinstitutional<br />
effort in this patient subgroup and provide a<br />
foundation for future trials. The primary endpoint is 1-year<br />
overall survival.<br />
FOLFIRINOX may also serve as a platform for the addition<br />
<strong>of</strong> targeted agents, though caution must be used because<br />
<strong>of</strong> the potential for overlapping toxicities. The Cancer<br />
and Leukemia Group B (CALGB) will be leading a phase<br />
IB/randomized phase II trial <strong>of</strong> mFOLFIRINOX plus placebo<br />
or ganitumumab (a monoclonal antibody to the insulin<br />
growth factor 1 receptor), in patients with previously untreated<br />
metastatic pancreatic cancer. This will be the first<br />
U.S. cooperative group trial to confirm the European experience<br />
with FOLFIRINOX.<br />
A phase IB dose-finding study <strong>of</strong> FOLFIRINOX plus<br />
IPI-926, a hedgehog pathway inhibitor, is currently ongoing.<br />
Once a phase II dose is determined, this combination will be<br />
incorporated in a randomized phase II study in development<br />
in CALGB and ECOG, in which patients with locally advanced<br />
diseaase are randomized to FOLFIRINOX plus IPI-<br />
926 or placebo, followed by chemoradiation. A phase IB<br />
study <strong>of</strong> FOLFIRINOX plus the hedgehog inhibitor LDE225<br />
is also accruing patients.<br />
235
The Southwest <strong>Oncology</strong> Group (SWOG) is in the process<br />
<strong>of</strong> designing a randomized trial that compares FOLFIRI-<br />
NOX with FOLFOX, to determine whether irinotecan is an<br />
essential component <strong>of</strong> the regimen. Another approach is to<br />
ascertain which patients would be most likely to develop<br />
toxicity from irinotecan and adjust doses accordingly.<br />
UGT1A1 is the enzyme responsible for clearing SN-38, the<br />
active metabolite <strong>of</strong> irinotecan; germline polymorphisms in<br />
the UGT1A1 gene are known to reduce enzymatic activity. 32<br />
Thus, patients treated with FOLFIRINOX who have different<br />
UGT1A1 genotypes may tolerate different doses <strong>of</strong><br />
irinotecan. Using genotype-guided dosing, investigators at<br />
the University <strong>of</strong> Chicago will soon open a phase I study to<br />
establish the optimal safe doses <strong>of</strong> irinotecan in the mFOL-<br />
FIRINOX regimen for each <strong>of</strong> three UGT1A1 genotype<br />
groups (*1*1, *1*28, and *28*28).<br />
Conclusion<br />
After so many negative randomized trials <strong>of</strong> gemcitabine<br />
doublets, the unprecedented outcomes achieved with the<br />
FOLFIRINOX regimen clearly represent a major treatment<br />
advance for those patients with pancreatic cancer who have<br />
a good performance status. No other randomized study has<br />
ever achieved a median survival <strong>of</strong> nearly a year. In no other<br />
Author’s Disclosures <strong>of</strong> Potential Conflicts <strong>of</strong> Interest<br />
Author<br />
Employment or<br />
Leadership<br />
Positions<br />
Consultant or<br />
Advisory Role<br />
Hedy Lee Kindler AstraZeneca;<br />
Bristol-Myers<br />
Squibb; Clovis<br />
<strong>Oncology</strong>;<br />
Genentech;<br />
GlaxoSmithKline;<br />
Merck<br />
1. Burris HA 3rd, Moore MJ, Andersen J, et al. Improvements in survival<br />
and clinical benefit with gemcitabine as first-line therapy for patients with<br />
advanced pancreas cancer: A randomized trial. J Clin Oncol. 1997;15:2403-<br />
2413.<br />
2. Berlin JD, Catalano P, Thomas JP, et al. Phase III study <strong>of</strong> gemcitabine<br />
in combination with fluorouracil versus gemcitabine alone in patients with<br />
advanced pancreatic carcinoma: Eastern Cooperative <strong>Oncology</strong> Group Trial<br />
E2297. J Clin Oncol. 2002;20:3270-3275.<br />
3. Rocha Lima CM, Green MR, Rotche R, et al. Irinotecan plus gemcitabine<br />
results in no survival advantage compared with gemcitabine monotherapy in<br />
patients with locally advanced or metastatic pancreatic cancer despite increased<br />
tumor response rate. J Clin Oncol 2004;22:3776-3783.<br />
4. Louvet C, Labianca R, Hammel P, et al. Gemcitabine in combination<br />
with oxaliplatin compared with gemcitabine alone in locally advanced or<br />
metastatic pancreatic cancer: Results <strong>of</strong> a GERCOR and GISCAD phase III<br />
trial. J Clin Oncol. 2005;23:3509-3516.<br />
5. Oettle H, Richards D, Ramanathan RK, et al. A phase III trial <strong>of</strong><br />
pemetrexed plus gemcitabine versus gemcitabine in patients with unresectable<br />
or metastatic pancreatic cancer. Ann Oncol. 2005;16:1639-1645.<br />
6. Heinemann V, Quietzsch D, Gieseler F, et al. Randomized phase III trial<br />
<strong>of</strong> gemcitabine plus cisplatin compared with gemcitabine alone in advanced<br />
pancreatic cancer. J Clin Oncol. 2006;24:3946-3952.<br />
7. Herrmann R, Bodoky G, Ruhstaller T, et al. Gemcitabine plus capecitabine<br />
compared with gemcitabine alone in advanced pancreatic cancer: A<br />
randomized, multicenter, phase III trial <strong>of</strong> the Swiss Group for <strong>Clinical</strong><br />
Cancer Res and the Central European Cooperative <strong>Oncology</strong> Group. J Clin<br />
Oncol. 2007;25:2212-2217.<br />
8. Moore MJ, Goldstein D, Hamm J, et al. Erlotinib plus gemcitabine<br />
compared with gemcitabine alone in patients with advanced pancreatic<br />
236<br />
trial have we even whispered about an 18-month survival in<br />
a proportion <strong>of</strong> patients with metastatic disease. No other<br />
phase III study has achieved such a high objective response<br />
rate. And despite substantial, though manageable toxicities,<br />
FOLFIRINOX also helps patients feel better for longer than<br />
if they received gemcitabine, a drug used principally for its<br />
effect on symptoms. Remarkably, this new drug combination<br />
is even cost-effective.<br />
The investigators in this study are to be commended not<br />
only for the decade they spent in developing the highly<br />
active FOLFIRINOX regimen. They are also to be applauded<br />
for a very well-designed pivotal study, which tested this<br />
therapy in a uniform population <strong>of</strong> patients (all metastatic),<br />
who, with their good performance status, were most likely to<br />
tolerate the rigors <strong>of</strong> the multidrug combination and were<br />
thus most able to benefit from it.<br />
Unanswered questions remain about the optimal way to<br />
dose the component drugs to minimize toxicity while preserving<br />
activity. Upcoming studies will address the potential<br />
role <strong>of</strong> this regimen in other disease settings and will use<br />
FOLFIRINOX as a platform on which to add new agents.<br />
It has been a long journey, but with the advent <strong>of</strong> FOL-<br />
FIRINOX, we are finally beginning to make progress against<br />
this dismal disease.<br />
Stock<br />
Ownership Honoraria<br />
REFERENCES<br />
Research<br />
Funding<br />
CanBas;<br />
Genentech;<br />
Infinity; Lilly;<br />
Merck;<br />
Morphotek<br />
Expert<br />
Testimony<br />
HEDY LEE KINDLER<br />
Other<br />
Remuneration<br />
cancer: a phase III trial <strong>of</strong> the National Cancer Institute <strong>of</strong> Canada <strong>Clinical</strong><br />
Trials Group. J Clin Oncol. 2007;25:1960-1966.<br />
9. Grubbs SS, Grusenmeyer PA, Petrelli NJ, Gralla RJ. Is it cost-effective<br />
to add erlotinib to gemcitabine in advanced pancreatic cancer? J Clin Oncol.<br />
2006; 24:18s(suppl; abstr 6048).<br />
10. Miksad RA, Schnipper L, Goldstein M. Does a statistically significant<br />
survival benefit <strong>of</strong> erlotinib plus gemcitabine for advanced pancreatic cancer<br />
translate into clinical significance and value? J Clin Oncol. 2007;25:4506-<br />
4507.<br />
11. Kindler HL, Niedzwiecki D, Hollis D, et al. Gemcitabine plus bevacizumab<br />
compared with gemcitabine plus placebo in patients with advanced<br />
pancreatic cancer: Phase III trial <strong>of</strong> the Cancer and Leukemia Group B<br />
(CALGB 80303). J Clin Oncol. 2010;28:3617-3622.<br />
12. Kindler HL, Ioka T, Richel DJ, et al. Axitinib plus gemcitabine versus<br />
placebo plus gemcitabine in patients with advanced pancreatic adenocarcinoma:<br />
A double-blind randomized phase 3 study. Lancet <strong>Oncology</strong>. 2011;12:<br />
256-262.<br />
13. Poplin E, Feng Y, Berlin J, et al. Phase III, randomized study <strong>of</strong><br />
gemcitabine and oxaliplatin versus gemcitabine (fixed-dose rate infusion)<br />
compared with gemcitabine (30-minute infusion) in patients with pancreatic<br />
carcinoma E6201: a trial <strong>of</strong> the Eastern Cooperative <strong>Oncology</strong> Group. J Clin<br />
Oncol. 2009;27:3778-3785.<br />
14. Philip PA, Benedetti J, Corless CL, et al. Phase III study comparing<br />
gemcitabine plus cetuximab versus gemcitabine in patients with advanced<br />
pancreatic adenocarcinoma: Southwest <strong>Oncology</strong> Group-directed intergroup<br />
trial S0205. J Clin Oncol. 2010;28:3605-3610.<br />
15. Cunningham D, Chau I, Stocken DD, et al. Phase III randomized<br />
comparison <strong>of</strong> gemcitabine versus gemcitabine plus capecitabine in patients<br />
with advanced pancreatic cancer. J Clin Oncol. 2009;27:5513-5518.<br />
16. Van Cutsem E, Vervenne WL, Bennouna J, et al. Phase III trial <strong>of</strong>
FOLFIRINOX FOR PANCREATIC CANCER<br />
bevacizumab in combination with gemcitabine and erlotinib in patients with<br />
metastatic pancreatic cancer. J Clin Oncol. 2009;27:2231-2237.<br />
17. Conroy T, Desseigne F, Ychou M, et al. Randomised phase III trial<br />
comparing FOLFIRINOX (F : 5FU/leucovorin [LV], irinotecan [I], oxaliplatin<br />
[O]) versus gemcitabine (G) as first-line treatment for metastatic pancreatic<br />
adenocarcinoma : Preplanned interim analysis results <strong>of</strong> the PRODIGE<br />
4/ACCORD 11 trial. J Clin Oncol. 2010;28: 15S (suppl; abstr 4010).<br />
18. Conroy T, Desseigne F, Ychou M, et al. FOLFIRINOX versus gemcitabine<br />
for metastatic pancreatic cancer. N Engl J Med. 2011;364:1817-1825.<br />
19. Ychou M, Conroy T, Seitz JF, et al. An open phase I study assessing<br />
the feasibility <strong>of</strong> the triple combination: oxaliplatin plus irinotecan plus<br />
leucovorin/5-fluorouracil every 2 weeks in patients with advanced cancers.<br />
Ann Oncol. 2003;14:481-489.<br />
20. Conroy T, Paillot B, Francois E, et al. Irinotecan plus oxaliplatin and<br />
leucovorin-modulated fluorouracil in advanced pancreatic cacner—A Groupe<br />
Tumeurs Digestives <strong>of</strong> the Federation Nationale des Centres de Lutte Contre<br />
le Cancer Study. J Clin Oncol. 2005;23:1228-1236.<br />
21. Miller AB, Hoogstraten B, Staquet M, et al. Reporting results <strong>of</strong> cancer<br />
treatment. Cancer. 1981;47:201-214.<br />
22. Therasse P, Arbuck SG, Eisenhauer EA, et al. New guidelines to<br />
evaluate the response to treatment in solid tumors. European Organization<br />
for Research and Treatment <strong>of</strong> Cancer, National Cancer Institute <strong>of</strong> the<br />
United States, National Cancer Institute <strong>of</strong> Canada. J Natl Cancer Inst.<br />
2000;92:205-216.<br />
23. Conroy T, Gavoille C, Adenis A. Metastatic pancreatic cancer: Old<br />
drugs, new paradigms. Curr Opin Oncol. 2011;23:390-395.<br />
24. Bendell JC, Britton S, Green MR, et al. Immediate impact <strong>of</strong> the<br />
FOLFIRINOX phase III data reported at the 2010 ASCO Annual Meeting on<br />
prescribing plans <strong>of</strong> <strong>American</strong> oncology physicians for patients with metastatic<br />
pancreas cancer. J Clin Oncol. 2011;29 (suppl, abstract 286).<br />
25. Faris JE, Hong TS, McDermott S, et al. FOLFIRINOX in locally<br />
advanced or metastatic pancreatic cancer. J Clin Oncol. <strong>2012</strong>;30 (suppl; abstr<br />
313).<br />
26. Peddi PF, Tan BR, Picus J, et al. Washington University experience<br />
with FOLFIRINOX in pancreatic cancer. J Clin Oncol. <strong>2012</strong>;30 (suppl; abstr<br />
354).<br />
27. Gunturu KS, Thumar JR, Hochster HS, et al. Single-institution experience<br />
with FOLFIRINOX in advanced pancreatic cancer. J Clin Oncol.<br />
<strong>2012</strong>;30 (suppl; abstr 330).<br />
28. Assaf E, Verlindhe-Carvalho M, Delbaldo C, et al. 5-Fluorouracil/<br />
leucovorin combined with irinotecan and oxaliplatin (FOLFIRINOX) as<br />
second-line chemotherapy in patients with metastatic pancreatic adenocarcinoma.<br />
<strong>Oncology</strong>. 2011;80:301-306.<br />
29. Breysacher G, Kaatz O, Lemarignier C, et al. Safety and clinical<br />
effectiveness <strong>of</strong> FOLFIRINOX in metastatic pancreas cancer after first-line<br />
therapy. J Clin Oncol. 2010;28 (suppl; abstr 269).<br />
30. Hosein PJ, Kawamura C, Macintyre J, et al. Pilot study <strong>of</strong> neoadjuvant<br />
FOLFIRINOX in unresectable locally advanced pancreatic carcinoma. J Clin<br />
Oncol. 2011;29 (suppl; abstr 324).<br />
31. Attard CL, Brown S, Alloul K, et al. Cost-effectiveness <strong>of</strong> FOLFIRINOX<br />
for first-line treatment <strong>of</strong> metastatic pancreatic cancer. J Clin Oncol. <strong>2012</strong>;30<br />
(suppl; abstr 199).<br />
32. Fujita K, Sparreboom. Pharmacogenetics <strong>of</strong> irinotecan disposition and<br />
toxicity: A review. Curr Clin Pharmacol. 2010;5:209-219.<br />
237
A Matter <strong>of</strong> Timing: Is There a Role for<br />
Radiation in Locally Advanced Pancreatic<br />
Cancer, and If So, When?<br />
By Theodore S. Hong, MD, Jennifer Y. Wo, MD, and Eunice L. Kwak, MD, PhD<br />
Overview: The role <strong>of</strong> radiation therapy in the management <strong>of</strong><br />
locally advanced pancreatic cancer is controversial. Despite<br />
its localized presentation, locally advanced pancreatic cancer<br />
is characterized by high rates <strong>of</strong> metastases. Historic data<br />
have been mixed, and newer studies have called into question<br />
the use <strong>of</strong> radiation therapy. However, it appears that patients<br />
more likely to benefit from chemoradiation can be identified<br />
PANCREATIC CANCER is the fourth leading cause <strong>of</strong><br />
cancer deaths in the United States, with 5-year survival<br />
rates <strong>of</strong> less than 5%. 1 Although surgical resection <strong>of</strong>fers<br />
the best chance at long-term survival, only 10% to 20% <strong>of</strong><br />
patients have resectable disease. Nearly half <strong>of</strong> all patients<br />
with pancreatic cancer have clinically evident metastatic<br />
disease, and the remaining 40% <strong>of</strong> patients present with<br />
localized but unresectable disease because <strong>of</strong> vascular involvement,<br />
which is considered locally advanced pancreatic<br />
cancer (LAPC). However, even patients with apparently<br />
localized disease have high rates <strong>of</strong> occult metastases, with<br />
staging laparoscopy studies demonstrating that at least<br />
30% <strong>of</strong> patients harbor metastatic peritoneal disease that is<br />
undetectable by imaging. 2<br />
With the greater appreciation <strong>of</strong> the systemic nature <strong>of</strong><br />
localized pancreatic cancer, there has been a renewed interest<br />
in evaluating the optimal timing <strong>of</strong> radiation therapy,<br />
including the early use <strong>of</strong> systemic chemotherapy both to<br />
address metastatic disease and to select patients most likely<br />
to benefit from local therapy.<br />
Historic Perspective<br />
Although the trend toward upfront or neoadjuvant chemotherapy<br />
has gained traction in recent years, the benefit <strong>of</strong><br />
chemoradiation therapy has always been controversial.<br />
Early randomized trials suggested mixed benefits from<br />
chemoradiation. Chemoradiation first came to the forefront<br />
<strong>of</strong> therapy for LAPC based on a randomized trial conducted<br />
by the Gastrointestinal Tumor Study Group (GITSG) in<br />
1974. 3 In this study, 228 patients with LAPC were randomly<br />
assigned to receive 60 Gy <strong>of</strong> radiation therapy alone compared<br />
with 40 Gy <strong>of</strong> radiation with concurrent 5-fluorouracil<br />
(5-FU) and 60 Gy <strong>of</strong> radiation with 5-FU. Patients assigned<br />
to the two chemoradiation arms also received maintenance<br />
5-FU. Both chemoradiation arms were associated with significantly<br />
improved survival compared with the radiation<br />
alone arm (p � 0.05 for 40 Gy arm, p � 0.001 for 60 Gy arm).<br />
No significant difference was reported between the 60 Gy<br />
and 40 Gy chemoradiation arms.<br />
In 1977, the Eastern Cooperative <strong>Oncology</strong> Group (ECOG)<br />
initiated a randomized trial comparing 5-FU alone to chemoradiation<br />
therapy. Ninety-one patients with LAPC were<br />
randomly assigned to groups receiving either weekly 5-FU<br />
alone or 40 Gy <strong>of</strong> radiation with bolus 5-FU followed by<br />
weekly 5-FU. 4 This study showed no difference in survival<br />
with or without radiation therapy (median survival [MS]<br />
8.3 months vs. 8.2 months, not significant [NS]). However,<br />
this study was criticized because <strong>of</strong> the poor survival in both<br />
238<br />
with an induction phase <strong>of</strong> chemotherapy. Data evaluating this<br />
approach suggest that approximately 30% <strong>of</strong> patients will<br />
develop metastatic disease within the first 3 to 4 months <strong>of</strong><br />
chemotherapy. Patients without progression who receive<br />
chemoradiation therapy may experience improved survival.<br />
Future directions include the validation <strong>of</strong> this strategy and<br />
the integration <strong>of</strong> biologic agents.<br />
groups. Because <strong>of</strong> these results, GITSG initiated another<br />
randomized trial comparing a three-drug regimen <strong>of</strong> streptozocin,<br />
mitomycin C, and 5-FU (SMF) or radiation therapy<br />
to 54 Gy with concurrent 5-FU followed by SMF chemotherapy.<br />
5 In this small 43-patient study, the 1-year survival was<br />
improved with the inclusion <strong>of</strong> chemoradiation compared to<br />
chemotherapy alone (41% vs. 19%, p � 0.02). Based on this<br />
study, chemoradiation therapy was felt to be superior to<br />
chemotherapy alone.<br />
Role <strong>of</strong> Radiation Questioned: the 2000–01<br />
FFCD/SFRO Trial<br />
Because <strong>of</strong> the limitations <strong>of</strong> the data defining the role <strong>of</strong><br />
radiation therapy and the lack <strong>of</strong> formal evaluation <strong>of</strong><br />
radiation therapy in the modern treatment era, the Federation<br />
Francophone de Cancerologie Digestive (FFCD) and<br />
the Societe Francophone de Radiotherapie Oncologique<br />
(SFRO) evaluated the role <strong>of</strong> chemoradiation therapy in<br />
patients with LAPC. 6 In this trial, patients were randomly<br />
assigned to receive either chemoradiation followed by maintenance<br />
gemcitabine or gemcitabine alone until progression.<br />
Radiation was delivered to 60 Gy in 2 Gy/fraction with an<br />
infusion <strong>of</strong> 5-FU (300 mg/m 2 /day) on days 1 to 5 for 6 weeks<br />
and cisplatin (20 mg/m 2 /day) on days 1 to 5 during weeks 1<br />
and 5. This phase III study was designed to demonstrate<br />
an improvement from 6 months with gemcitabine alone to<br />
12 months with chemoradiation. However, the study was<br />
stopped early because <strong>of</strong> slow accrual, and a review by an<br />
independent data monitoring committee followed. At that<br />
time, it was noted that the chemoradiation arm had a<br />
shorter overall survival (OS). Survival analysis showed<br />
worse median survival with chemoradiation (8.6 months)<br />
compared with gemcitabine alone (13.0 months) by intentto-treat<br />
analysis (p � 0.03). Because <strong>of</strong> concerns <strong>of</strong> a possibly<br />
confounding toxicity, a per-protocol analysis was performed<br />
restricting analysis to those who received at least 75% <strong>of</strong><br />
protocol treatment. By this analysis, patients receiving<br />
chemoradiation therapy still experienced a shorter median<br />
survival (9.5 months vs. 15.0 months, p � 0.006).<br />
From the Departments <strong>of</strong> Radiation <strong>Oncology</strong> and Medicine, Massachusetts General<br />
Hospital and Harvard Medical School, Boston, MA.<br />
Authors’ disclosures <strong>of</strong> potential conflicts <strong>of</strong> interest are found at the end <strong>of</strong> this article.<br />
Address reprint requests to Theodore S. Hong, MD, Department <strong>of</strong> Radiation <strong>Oncology</strong>,<br />
Massachusetts General Hospital, 100 Blossom Street, Cox 3, Boston, MA 02114; email:<br />
tshong1@partners.org.<br />
© <strong>2012</strong> by <strong>American</strong> <strong>Society</strong> <strong>of</strong> <strong>Clinical</strong> <strong>Oncology</strong>.<br />
1092-9118/10/1-10
RADIATION AND PANCREATIC CANCER<br />
Although these results suggest that chemoradiation<br />
should not be used, the study design limits the ability to<br />
interpret the outcome. The choices <strong>of</strong> chemotherapy agents<br />
and radiation dose were nonstandard and may have compromised<br />
tolerability in the chemoradiation arm. For instance,<br />
although highly emetogenic, cisplatin has not demonstrated<br />
significant activity in pancreatic cancer and the dose <strong>of</strong><br />
radiation used was higher than tested in most clinical trials.<br />
Furthermore, this nonstandard regimen was not tested in<br />
phase II trials before its use in this randomized trial. The<br />
toxicity <strong>of</strong> this regimen led to fewer than half <strong>of</strong> the patients<br />
receiving 75% <strong>of</strong> the chemoradiation dose and contributed to<br />
the poor tolerance <strong>of</strong> subsequent maintenance chemotherapy.<br />
The median total dose <strong>of</strong> gemcitabine received by<br />
patients in the chemoradiation arm was well below half the<br />
total dose received by patients in the chemotherapy-alone<br />
group.<br />
Despite the flaws, this trial underscores a key lesson: even<br />
when localized in presentation, pancreatic cancer is a<br />
largely systemic disease, and it is important for patients to<br />
receive chemotherapy early in the treatment course.<br />
Does Anyone Benefit from Radiation Therapy?<br />
The GERCOR Study<br />
In the late 1990s through the early 2000s, the Groupe<br />
Coopérateur Multidisciplinaire en Oncologie (GERCOR) ran<br />
a number <strong>of</strong> phase II and III trials evaluating different<br />
chemotherapy regimens. 7 Following the trend <strong>of</strong> including<br />
patients with locally advanced disease in chemotherapy<br />
trials but acknowledging the controversial role <strong>of</strong> radiation<br />
therapy, investigators recommended patients for these trials<br />
who had no progression after 3 months <strong>of</strong> receiving chemotherapy<br />
and who had ECOG performance status <strong>of</strong> 2 or<br />
better. However, investigators could also continue chemotherapy<br />
at their discretion. Radiation therapy was delivered<br />
to 45 Gy with a boost to a total dose <strong>of</strong> 55 Gy with continuous<br />
infusion 5-FU. Computed tomography (CT) planning was<br />
mandated.<br />
In a retrospective analysis <strong>of</strong> prospective trials, 167 <strong>of</strong> the<br />
497 patients enrolled on these trials had locally advanced<br />
disease. After 3 months <strong>of</strong> chemotherapy, 71% <strong>of</strong> patients<br />
with LAPC did not have progressive disease and had performance<br />
status scores eligible for chemoradiation. Of the<br />
remaining 29% <strong>of</strong> patients ineligible for chemoradiation, 45<br />
<strong>of</strong> the 53 patients were ineligible because they had progres-<br />
KEY POINTS<br />
● Most patients with locally advanced pancreatic cancer<br />
will succumb to metastatic disease.<br />
● A subset <strong>of</strong> patients with more locally aggressive<br />
disease may be identified by the use <strong>of</strong> upfront<br />
chemotherapy.<br />
● The use <strong>of</strong> radiation therapy may be beneficial in<br />
patients with nonprogressive disease.<br />
● DPC4 may be a biomarker that can predict biologic<br />
behavior.<br />
● Future directions include improving the evaluation <strong>of</strong><br />
new radiation techniques and integrating biologic<br />
therapies.<br />
sive disease. Of the 128 patients eligible for chemoradiation,<br />
72 (56%) patients received chemoradiation and 56 (44%)<br />
patients continued receiving chemotherapy alone. Patients<br />
in both groups were well matched by chemotherapy regimen,<br />
age, performance status, weight loss, and response to<br />
chemotherapy. The median progression-free survival (PFS)<br />
for the chemoradiation and chemotherapy groups were 10.8<br />
months and 7.4 months, respectively (p � 0.005). The<br />
median OS was 15 months for the chemoradiation group and<br />
11.7 months for the chemotherapy group (p � 0.0009).<br />
Because <strong>of</strong> its retrospective nature, this study was limited<br />
by potential patient selection bias; however, the two treatment<br />
groups were well balanced for performance status, age,<br />
and chemotherapy response. This study suggests that some<br />
patients might benefit from chemoradiation therapy. Although<br />
previous studies have recognized that some patients<br />
have rapidly metastatic disease, this study suggests that an<br />
effective strategy to evaluate which patients have aggressive<br />
systemic biology is to start with systemic chemotherapy and<br />
to move only those patients whose disease does not progress<br />
after several months <strong>of</strong> chemotherapy to receive chemoradiation.<br />
This study caused a substantial paradigm shift in<br />
the way radiation therapy was timed. Increasingly, institutions<br />
would initiate radiation therapy with chemotherapy,<br />
primarily to identify patients with rapid, early metastatic<br />
progression who would not benefit from aggressive local<br />
therapy.<br />
Further Evidence <strong>of</strong> the Benefit <strong>of</strong> Radiation Therapy:<br />
ECOG 4201<br />
During this period, investigators in the United States<br />
were also investigating the role <strong>of</strong> radiation therapy. Using<br />
a gemcitabine platform, ECOG 4201 directly compared<br />
gemcitabine alone with gemcitabine-based chemoradiation<br />
therapy followed by gemcitabine. 8 The chemotherapy-alone<br />
group received one 7-week induction cycle, followed by five<br />
additional 4-week cycles (3 weeks on, 1 week <strong>of</strong>f). The<br />
chemoradiation arm received 50.4 Gy <strong>of</strong> radiation as 1.8<br />
Gy/fraction with concurrent gemcitabine (600 mg/m 2 )<br />
weekly during radiation treatment followed by five additional<br />
cycles <strong>of</strong> gemcitabine. The intended study design<br />
included 316 patients but was closed after enrolling only 74<br />
patients because <strong>of</strong> slow accrual.<br />
As expected, the radiation arm had greater toxicity, with<br />
41% <strong>of</strong> patients experiencing grade 4 toxicity compared to<br />
9% in the chemotherapy-alone arm. Despite this, both arms<br />
received a median <strong>of</strong> three cycles <strong>of</strong> chemotherapy. Median<br />
survival was improved in the radiation arm (11.1 months vs.<br />
9.2 months, p � 0.017). There was no difference in PFS.<br />
This trial was more straightforward than the FFCD trial<br />
because it directly compared radiation therapy with gemcitabine<br />
to gemcitabine alone. The chemoradiation regimen<br />
demonstrated safety in phase I testing, which in contrast to<br />
the FFCD trial, is reflected in the fact that both arms<br />
received a similar amount <strong>of</strong> chemotherapy. However, the<br />
small number <strong>of</strong> patients enrolled in this study limited the<br />
ability to understand the role <strong>of</strong> radiation therapy. Additionally,<br />
because radiation therapy was used at the beginning<br />
<strong>of</strong> treatment, issues <strong>of</strong> timing were not addressed. Finally,<br />
this small study elicited an OS benefit, despite a lack <strong>of</strong><br />
difference in PFS. It is unclear why this would be the case,<br />
especially in light <strong>of</strong> the increased toxicity <strong>of</strong> the radiation<br />
arm. However, this study was the first study to demonstrate<br />
239
an OS benefit with radiation therapy against a backbone <strong>of</strong><br />
gemcitabine.<br />
Further Evidence <strong>of</strong> Delaying Radiation Therapy<br />
Investigators at the University <strong>of</strong> California, San Francisco<br />
(UCSF), seeking to optimize systemic control, performed<br />
a phase II study <strong>of</strong> fixed-dose rate gemcitabine in<br />
combination with cisplatin for six cycles, followed by chemoradiation<br />
therapy. 9 The gemcitabine was given at 1,000<br />
mg/m 2 infused at 10 mg/m 2 /minute followed by cisplatin (20<br />
mg/m 2 ) administered on days 1 and 15 <strong>of</strong> a 28-day cycle.<br />
Patients initiated chemoradiation therapy 2 to 6 weeks after<br />
completing six cycles <strong>of</strong> chemotherapy. Radiation was delivered<br />
to a standard 50.4 Gy with continuous infusion 5-FU.<br />
Of the 25 patients enrolled on the study, eight (28%)<br />
developed disease progression while receiving chemotherapy<br />
between two and 4.5 treatment cycles (median three cycles),<br />
consistent with the rate <strong>of</strong> metastatic progression seen in<br />
the GERCOR study. Thirteen patients (56%) proceeded to<br />
the chemoradiation phase. The median OS for the entire<br />
cohort was 13.5 months, with a median time to progression<br />
(TTP) <strong>of</strong> 10.5 months.<br />
The patterns <strong>of</strong> treatment failure were informative for the<br />
selection that occurs with upfront chemotherapy. Seven <strong>of</strong><br />
the eight patients who had progressive disease during chemotherapy<br />
had metastatic disease. In contrast, six <strong>of</strong> the 12<br />
patients who received all six cycles <strong>of</strong> chemotherapy and<br />
chemoradiation developed local progression, rather than<br />
metastatic progression. This study suggests that delayed<br />
radiation therapy enriches the group <strong>of</strong> patients receiving<br />
radiation for those with more localized biology.<br />
Investigators at Massachusetts General Hospital (MGH)<br />
also compared upfront chemoradiation therapy with delayed<br />
chemoradiation therapy. At MGH, patients with LAPC were<br />
historically treated with early chemoradiation therapy.<br />
With the publication <strong>of</strong> the GERCOR study in 2007, a<br />
gradual shift to delayed chemoradiation therapy occurred.<br />
In a retrospective analysis, investigators compared patients<br />
who received chemotherapy before chemoradiation therapy<br />
with those who started with chemoradiation therapy to<br />
determine the relative outcomes associated with patient<br />
selection based on the timing <strong>of</strong> chemoradiation treatment.<br />
10<br />
In this study, 70 consecutive patients with unresectable<br />
(46 patients) or borderline resectable (24 patients) LAPC<br />
were treated with chemoradiation from 2005 to 2009. Patients<br />
typically received 50.4 Gy <strong>of</strong> radiation in 28 fractions<br />
(91%) with concurrent 5-FU (84%) or capecitabine (14%).<br />
Forty patients received chemoradiation alone, and 30 patients<br />
received a median <strong>of</strong> 4 months <strong>of</strong> chemotherapy before<br />
chemoradiation, typically gemcitabine (93%). All patients<br />
without progression after chemotherapy were <strong>of</strong>fered<br />
chemoradiation.<br />
Fifty-three percent <strong>of</strong> the patients in the early chemoradiation<br />
group compared with 83% in the delayed chemoradiation<br />
group had categorically unresectable tumors at<br />
diagnosis. Median OS for the early and delayed chemoradiation<br />
groups was 12.4 months and 18.7 months, respectively<br />
(p � 0.02). Median PFS for early compared with delayed<br />
chemoradiation was 6.7 months and 11.4 months, respectively<br />
(p � 0.02). On multivariate analysis, administration<br />
<strong>of</strong> chemotherapy before chemoradiation (adjusted hazard<br />
240<br />
HONG, WO, AND KWAK<br />
ratio [AHR] � 0.49; 95% CI, 0.28 to 0.87; p � 0.02) and<br />
surgical resection (AHR � 0.38; 95% CI, 0.17 to 0.85; p �<br />
0.02) were associated with increased OS.<br />
These studies are representative <strong>of</strong> the paradigm shift<br />
that has occurred at many institutions over the past 5 years.<br />
The use <strong>of</strong> upfront chemotherapy represents a way to<br />
identify patients more likely to benefit from chemoradiation<br />
treatment.<br />
A Marker for Different Biologies: DPC4 Status<br />
Although clinical evidence has suggested that upfront<br />
chemotherapy could help identify patients more likely to<br />
benefit from chemoradiation, a biomarker predictive <strong>of</strong> clinical<br />
course had yet to be identified. To this end, investigators<br />
at Johns Hopkins University evaluated 76 patients who had<br />
consented to participate in their rapid autopsy program. 11<br />
Patients were grouped according to death with widely metastatic<br />
disease or death with locally progressive disease.<br />
Grouping was correlated with KRAS and TP53 mutations<br />
and DPC4 (which stands for Deleted in Pancreatic Cancer,<br />
locus 4) status.<br />
Of the 76 patients, 9 had no metastases at the time <strong>of</strong><br />
death, 13 had 10 metastases or fewer, 26 had 11 to 99<br />
metastases, and 26 patients had more than 100 metastases.<br />
Investigators found a striking correlation between DPC4<br />
status and patterns <strong>of</strong> treatment failure. Generally, patients<br />
with DPC4 loss had widespread disease, whereas patients<br />
with intact DPC4 were more likely to have locally destructive<br />
disease (p � 0.007).<br />
This study highlights two major points. First, almost 30%<br />
<strong>of</strong> patients died with either no metastases or fewer than 10<br />
metastases, highlighting that a substantial number <strong>of</strong> patients<br />
succumb to locally destructive disease rather than<br />
metastatic disease. Second, intact DPC4 presence as assayed<br />
by immunohistochemistry may identify a subgroup <strong>of</strong><br />
tumors that are destined to be locally destructive as opposed<br />
to widely metastatic. Because this finding may have implications<br />
for the relative roles <strong>of</strong> chemotherapy and radiation<br />
therapy, the Radiation Therapy <strong>Oncology</strong> Group (RTOG)<br />
has proposed a study in which systemic chemotherapy with<br />
a combination <strong>of</strong> 5-FU, oxaliplatin, irinotecan, and leucovorin<br />
(FOLFIRINOX) will be studied in patients with loss <strong>of</strong><br />
DPC4 and radiation dose escalation will be studied in<br />
patients with intact DPC4.<br />
Current Recommendations and Future Directions<br />
The evidence that LAPC represents a heterogeneous<br />
group has been mounting. Although the majority <strong>of</strong> patients<br />
will succumb to metastatic disease, there appears to be a<br />
group <strong>of</strong> patients that has more locally destructive disease.<br />
Currently, the use <strong>of</strong> chemotherapy before radiation therapy<br />
has the advantages <strong>of</strong> early treatment <strong>of</strong> micrometastatic<br />
disease and better selection <strong>of</strong> patients who may benefit<br />
from chemoradiation therapy. Efforts are currently ongoing<br />
to further clarify the timing <strong>of</strong> radiation for LAPC. At this<br />
writing, the LAP 07 study is evaluating the role <strong>of</strong> radiation<br />
therapy in this very manner (Fig. 1). Led by GERCOR, LAP<br />
07 is evaluating chemoradiation therapy after four cycles <strong>of</strong><br />
chemotherapy compared with two more cycles <strong>of</strong> chemotherapy.<br />
This trial involving 902 patients will hopefully provide<br />
a definitive answer to the role <strong>of</strong> delayed chemoradiation.<br />
In the future, the use <strong>of</strong> biomarkers, such as DPC4, may
RADIATION AND PANCREATIC CANCER<br />
Fig. 1. The schema <strong>of</strong> the ongoing LAP 07 randomized trial for locally advanced pancreatic cancer.<br />
provide this biologic information at the time <strong>of</strong> diagnosis.<br />
However, the DPC4 assay has not yet been standardized for<br />
routine clinical use. The RTOG is currently in the process <strong>of</strong><br />
validating the feasibility <strong>of</strong> this assay on fine needle aspiration<br />
samples for its proposed trial <strong>of</strong> patients with LAPC.<br />
Questions remain regarding the optimal duration <strong>of</strong> therapy.<br />
No standard recommendation exists about how much<br />
chemotherapy should be given before chemoradiation therapy.<br />
As discussed above, the GERCOR study used 3 months.<br />
The RTOG plans to use 2 months for the induction phase <strong>of</strong><br />
chemotherapy. At our institution, we have chosen to use 4<br />
months, in part because adjuvant trials like RTOG 9704<br />
have administered 4 months <strong>of</strong> chemotherapy when used<br />
with 5.5 weeks <strong>of</strong> chemoradiation therapy. Additionally,<br />
because the median TTP on gemcitabine is between 3 and 4<br />
months, 4 months allows adequate time for metastatic<br />
biology to declare itself, as demonstrated in the UCSF trial.<br />
The use <strong>of</strong> 2 to 4 months <strong>of</strong> chemotherapy seems reasonable.<br />
In addition, no consensus has been set on which chemotherapy<br />
should be used with standard fractionation radiation<br />
therapy. Classically, most studies have used 5-FU,<br />
shifting from bolus 5-FU to continuous infusion 5-FU, which<br />
follows the trend in rectal cancer. Gemcitabine-based<br />
chemoradiation therapy has been studied extensively, including<br />
in the randomized ECOG study. For example, RTOG<br />
and Alliance cooperative groups are using capecitabine as<br />
the standard chemoradiation backbone. Whether a particular<br />
platform has substantial advantages over another remains<br />
unclear.<br />
There has been increasing interest in the use <strong>of</strong> stereotactic<br />
body radiation therapy (SBRT), where advanced technology<br />
delivers high doses <strong>of</strong> radiation alone to the tumor. In<br />
2004, an initial phase I study out <strong>of</strong> Stanford University<br />
demonstrated the feasibility and safety <strong>of</strong> SBRT for LAPC.<br />
In this study, 15 patients with LAPC and an ECOG performance<br />
status <strong>of</strong> less than 2 were treated with 15 Gy, 20 Gy,<br />
or 25 Gy <strong>of</strong> radiation in a single fraction via CyberKnife®.<br />
One <strong>of</strong> the three patients treated at 15 Gy, two <strong>of</strong> the four<br />
patients treated at 20 Gy, and zero <strong>of</strong> the six patients<br />
treated with 25 Gy had local progression. 12 This phase I<br />
study was followed by a phase II study combining conven-<br />
tionally fractionated chemoradiation with a stereotactic radiosurgery<br />
(SRS) boost. 13 Sixteen patients were treated with<br />
45 Gy <strong>of</strong> radiation in 1.8 Gy/fraction to the tumor and<br />
regional lymphatics with concurrent 5-FU or capecitabine.<br />
Within 1 month <strong>of</strong> chemoradiation treatment, patients were<br />
given an SRS boost <strong>of</strong> 25 Gy <strong>of</strong> radiation alone to the tumor<br />
using CyberKnife. Sixteen patients were treated. Fifteen <strong>of</strong><br />
the16 patients were free from local progression until death.<br />
However, TTP (17.5 weeks) and median survival (33 weeks)<br />
were modest. In a retrospective study from the Beth Israel<br />
Deaconness Medical Center, 47 patients with LAPC were<br />
given two cycles <strong>of</strong> gemcitabine followed by restaging. 14<br />
Patients without metastatic disease were given a third cycle<br />
<strong>of</strong> gemcitabine while undergoing planning. Patients were<br />
then treated with 24 Gy to 36 Gy <strong>of</strong> radiation in three<br />
fractions followed by maintenance gemcitabine. Eight patients<br />
(17%) developed metastatic disease before undergoing<br />
SBRT. Patients undergoing SBRT had a median OS <strong>of</strong> 20<br />
months.<br />
Efforts are also ongoing to reduce the toxicity <strong>of</strong> chemoradiation<br />
treatment by using advanced radiation technology<br />
such as intensity modulated radiation therapy (IMRT).<br />
IMRT uses multiple beam angles and a computational<br />
process called inverse planning to create irregular dose<br />
distributions that can conform to irregular shapes, thereby<br />
affording the potential to decrease the toxicity associated<br />
with chemoradiation. In a retrospective review from the<br />
University <strong>of</strong> Maryland, 46 patients with pancreatic or<br />
ampullary cancers treated with IMRT were evaluated. 15<br />
Investigators noted a grade 3 or 4 nausea and vomiting rate<br />
<strong>of</strong> 0%, compared to the 11% seen with standard CT-based<br />
radiation therapy used in RTOG 9704. This study suggests<br />
that further improving radiation delivery can positively<br />
influence the toxicity pr<strong>of</strong>ile that was reported in the FFCD<br />
and ECOG studies.<br />
Based on the current research, for the treatment <strong>of</strong> patients<br />
with LAPC we recommend 2 to 4 months <strong>of</strong> chemotherapy<br />
followed chemoradiation therapy. Patients typically<br />
receive gemcitabine, although select patients are now also<br />
receiving FOLFIRINOX. Chemoradiation therapy is delivered<br />
with either continuous infusion 5-FU or capecitabine to<br />
241
a dose <strong>of</strong> 50.4 Gy to 59.4 Gy <strong>of</strong> radiation in 1.8 Gy/fraction.<br />
Future trials are now focused on the prospective evaluation<br />
Authors’ Disclosures <strong>of</strong> Potential Conflicts <strong>of</strong> Interest<br />
Author<br />
Theodore S. Hong*<br />
Jennifer Y. Wo*<br />
Eunice L. Kwak*<br />
*No relevant relationships to disclose.<br />
Employment or<br />
Leadership<br />
Positions<br />
Consultant or<br />
Advisory Role<br />
1. Jemal A, Siegel R, Xu J, et al. Cancer statistics, 2010. CA Cancer J Clin.<br />
2010;60:277-300.<br />
2. Jimenez RE, Warshaw AL, Fernandez-del Castillo C. Laparoscopy and<br />
peritoneal cytology in the staging <strong>of</strong> pancreatic cancer. J Hepatobiliar<br />
Pancreat Surg. 2000;7:15-20.<br />
3. Gastrointestinal Tumor Study Group. A multi-institutional comparative<br />
trial <strong>of</strong> radiation therapy alone and in combination with 5-fluorouracil for<br />
locally unresectable pancreatic carcinoma. Ann Surg. 1979;189:205-208.<br />
4. Klaassen DJ, MacIntyre JM, Catton GE, et al. Treatment <strong>of</strong> locally<br />
unresectable cancer <strong>of</strong> the stomach and pancreas: a randomized comparison<br />
<strong>of</strong> 5-fluorouracil alone with radiation plus concurrent and maintenance<br />
5-fluorouracil-an Eastern Cooperative <strong>Oncology</strong> Group study. J Clin Oncol.<br />
1985;3:373-378.<br />
5. Gastrointestinal Tumor Study Group. Treatment <strong>of</strong> locally unresectable<br />
carcinoma <strong>of</strong> the pancreas: comparison <strong>of</strong> combined-modality therapy (chemotherapy<br />
plus radiotherapy) to chemotherapy alone. J Natl Cancer Inst.<br />
1988;80:751-755.<br />
6. Chauffert B, Mornex F, Bonnetain F, et al. Phase III trial comparing<br />
intensive induction chemoradiotherapy (60 Gy, infusional 5-FU and intermittent<br />
cisplatin) followed by maintenance gemcitabine with gemcitabine alone<br />
for locally advanced unresectable pancreatic cancer. Definitive results <strong>of</strong> the<br />
2000-01 FFCD/SFRO study. Ann Oncol. 2008;19:1592-1599.<br />
7. Huguet F, André T, Hammel P, et al. Impact <strong>of</strong> chemoradiotherapy after<br />
disease control with chemotherapy in locally advanced pancreatic adenocarcinoma<br />
in GERCOR phase II and III studies. J Clin Oncol. 2007;25:326-331.<br />
8. Loehrer PJ Sr, Feng Y, Cardenes H, et al. Gemcitabine alone versus<br />
242<br />
<strong>of</strong> biomarkers such as DPC4 and the integration <strong>of</strong> targeted<br />
therapies into this platform.<br />
Stock<br />
Ownership Honoraria<br />
REFERENCES<br />
Research<br />
Funding<br />
HONG, WO, AND KWAK<br />
Expert<br />
Testimony<br />
Other<br />
Remuneration<br />
gemcitabine plus radiotherapy in patients with locally advanced pancreatic<br />
cancer: an Eastern Cooperative Group <strong>Oncology</strong> Trial. J Clin Oncol. 2011;29:<br />
4105-4112.<br />
9. Ko AH, Quivey JM, Venook AP, et al. A phase II study <strong>of</strong> fixed-dose rate<br />
gemcitabine plus low-dose cisplatin followed by consolidative chemoradiation<br />
for locally advanced pancreatic cancer. Int J Radiat Oncol Biol Phys. 2007;<br />
68:809-816.<br />
10. Arvold ND, Ryan DP, Niemierko A, et al. Long-term outcomes <strong>of</strong><br />
neoadjuvant chemotherapy before chemoradiation for locally advanced pancreatic<br />
cancer. Cancer. 2011 Oct 21.<br />
11. Iacobuzio-Donahue CA, Fu B, Yachida S, et al. DPC4 gene status <strong>of</strong> the<br />
primary carcinoma correlates with patterns <strong>of</strong> failure in patients with<br />
pancreatic cancer. J Clin Oncol. 2009;27:1806-1813.<br />
12. Koong AC, Le QT, Ho A, et al. Phase I study <strong>of</strong> stereotactic radiosurgery<br />
in patients with locally advanced pancreatic cancer. Int J Radiat Oncol Biol<br />
Phys. 2004;58:1017-21.<br />
13. Koong AC, Christ<strong>of</strong>ferson E, Le QT, et al. Phase II study to assess the<br />
efficacy <strong>of</strong> conventionally fractionated radiotherapy followed by a stereotactic<br />
radiosurgery boost in patients with locally advanced pancreatic cancer. Int J<br />
Radiat Oncol Biol Phys. 2005;63:320-3.<br />
14. Mahadevan A, Miksad R, Goldstein M, et al. Induction gemcitabine and<br />
stereotactic body radiotherapy for locally advanced nonmetastatic pancreas<br />
cancer. Int J Radiat Oncol Biol Phys. 2011;81:e615-22.<br />
15. Yovino S, Poppe M, Jabbour S, et al. Intensity-modulated radiation<br />
therapy significantly improves acute gastrointestinal toxicity in pancreatic<br />
and ampullary cancers. Int J Radiat Oncol Biol Phys. 2011;79:158-62.
A Myriad <strong>of</strong> Symptoms: New Approaches to<br />
Optimizing Palliative Care <strong>of</strong> Patients with<br />
Advanced Pancreatic Cancer<br />
Overview: Patients with advanced pancreatic cancer (APC)<br />
require early and frequent palliative interventions to achieve<br />
optimal quality <strong>of</strong> life for the duration <strong>of</strong> illness. Evidencebased<br />
supportive treatments exist to maximize quality <strong>of</strong> life<br />
for any patient, whether receiving chemotherapy or not. This<br />
article provides a comprehensive review <strong>of</strong> symptoms with<br />
current treatment recommendations and directions for future<br />
development. Celiac plexus neurolysis improves pain in the<br />
majority <strong>of</strong> patients with APC and should be moved earlier in<br />
the analgesic paradigm. Malignant bowel obstruction can be<br />
palliated quickly with optimal management via gastric decompression,<br />
octreotide, parenteral opioids, and standing antiemetics.<br />
Recommendations are provided for best treatment <strong>of</strong><br />
malignant gastroparesis, gastric outlet obstruction, and<br />
chemotherapy-induced nausea and vomiting in this popula-<br />
PATIENTS WITH advanced pancreatic cancer (APC)<br />
require early and frequent palliative interventions to<br />
achieve optimal quality <strong>of</strong> life for the duration <strong>of</strong> illness.<br />
Despite recent notable advances in multidisciplinary antineoplastic<br />
therapy, the majority <strong>of</strong> patients with APC ultimately<br />
die after facing numerous physical and emotional<br />
hurdles. 1,2 As a physician caring for these patients, there<br />
are opportunities to relieve suffering from predictable complications<br />
<strong>of</strong> pancreatic cancer. Evidence-based supportive<br />
treatments exist to maximize quality <strong>of</strong> life for any patient,<br />
whether receiving chemotherapy or not. This article provides<br />
a comprehensive review <strong>of</strong> symptoms facing patients<br />
wtih APC with current treatment recommendations and<br />
directions for future development.<br />
Pain<br />
Pain from cancer in the pancreas is <strong>of</strong>ten constant and<br />
severe, felt predominantly in the midback and epigastrium.<br />
Noxious sensory input from inflamed pancreatic tissue and<br />
direct neural invasion is transmitted via the celiac plexus as<br />
pain. 3 Opioids and adjuvant medications remain standard <strong>of</strong><br />
care for analgesia; however, the titration <strong>of</strong> opioids can be<br />
limited by systemic toxicities and may not adequately address<br />
the pain. Successful locoregional intervention minimizes<br />
systemic opioid requirements early in the disease. 3<br />
In the majority <strong>of</strong> patients, celiac plexus neurolysis (CPN)<br />
has been shown to provide effective pain relief simultaneous<br />
with reduction in systemic opioids. 3,4 An injection <strong>of</strong> either<br />
ethanol or phenol destroys afferent nerve fibers 3,4 and disrupts<br />
pain signals for an average <strong>of</strong> 3 months, though<br />
sometimes permanently. CPN can be performed with equivalent<br />
efficacy surgically, percutaneously under radiologic<br />
guidance, or endoscopically via ultrasound. 3 The most common<br />
risks include transient hypotension, constipation, or<br />
diarrhea; no serious adverse events were noted in a metaanalysis.<br />
3 More than 80% <strong>of</strong> patients note significantly<br />
improved analgesia after CPN in blinded or sham studies. 3,4<br />
To evaluate both effect and timing <strong>of</strong> CPN in pancreatic<br />
cancer, Wyse and colleagues performed a double-blind, randomized<br />
controlled trial <strong>of</strong> patients found to have inoperable<br />
By Lauren A. Wiebe, MD<br />
tion. Malignant ascites can be treated initially with diuretics<br />
and sodium-restriction in patients with an exudative process;<br />
however, an indwelling catheter is recommended for patients<br />
with recurrent ascites, particularly because <strong>of</strong> carcinomatosis<br />
or a refractory process. With exocrine insufficiency contributing<br />
to weight loss, pancreatic enzyme replacement is essential<br />
to improve nourishment in the majority <strong>of</strong> patients. Presently,<br />
megestrol acetate is the only U.S. Food and Drug Administration<br />
(FDA)-approved therapy for the anorexia-cachexia syndrome,<br />
although future developments are promising. Finally,<br />
patients with advanced pancreatic cancer should be screened<br />
and treated early for depression as a common comorbid<br />
diagnosis. Early palliative care consultation also helps address<br />
the existential and psychosocial concerns <strong>of</strong> patients<br />
facing death from pancreatic cancer in a holistic manner.<br />
pancreatic cancer at time <strong>of</strong> diagnostic endoscopic ultrasound<br />
(EUS). The 96 participants were randomized to either<br />
CPN or usual medical management at time <strong>of</strong> EUS diagnosis.<br />
Persistently increasing pain scores were noted in the<br />
control group during the study; however, patients who<br />
underwent neurolysis reported improvements in analgesia<br />
both 1 and 3 months later with a statistically significant<br />
decrease <strong>of</strong> 49% in mean pain score. 4<br />
The study from Wyse and colleagues adds to the body <strong>of</strong><br />
evidence supporting early CPN for patients with pancreatic<br />
cancer. 3 Optimal timing would be at moment <strong>of</strong> diagnosis if<br />
a patient reports abdominal pain attributable to inoperable<br />
pancreatic cancer.<br />
Because these investigators took a detailed pain history<br />
before diagnostic EUS/endoscopic retrograde cholangiopancreatography<br />
(ERCP), patients were able to benefit from<br />
CPN with early, lasting pain relief. 4 For patients undergoing<br />
surgical exploration, CPN should be considered intraoperatively<br />
for early, seamless analgesia once a diagnosis<br />
is secured. With recurrent pain, repeat CPN is indicated<br />
and effective, particularly if a patient had benefit from prior<br />
neurolysis. 3<br />
Nausea and Vomiting<br />
Nausea or vomiting in a patient with APC can arise from<br />
multiple etiologies. With new onset, a patient should be<br />
evaluated for potentially reversible causes, some <strong>of</strong> which<br />
are discussed later in this article. While receiving antineoplastic<br />
therapy, adequate support with antiemetics should<br />
be provided per the guidelines <strong>of</strong> the <strong>American</strong> <strong>Society</strong> <strong>of</strong><br />
<strong>Clinical</strong> <strong>Oncology</strong> (ASCO) or the National Cancer Care<br />
Network (NCCN), available online. As FOLFIRINOX is<br />
From the Department <strong>of</strong> Hematology and <strong>Oncology</strong>, The Medical College <strong>of</strong> Wisconsin,<br />
Froedtert Memorial Hospital.<br />
Author’s disclosures <strong>of</strong> potential conflicts <strong>of</strong> interest are found at the end <strong>of</strong> this article.<br />
Address reprint requests to Lauren Wiebe, MD, Department <strong>of</strong> Hematology and <strong>Oncology</strong>,<br />
The Medical College <strong>of</strong> Wisconsin, 9200 W. Wisconsin Ave., Milwaukee, WI 53226; email:<br />
laurenwiebe@gmail.com.<br />
© <strong>2012</strong> by <strong>American</strong> <strong>Society</strong> <strong>of</strong> <strong>Clinical</strong> <strong>Oncology</strong>.<br />
1092-9118/10/1-10<br />
243
more commonly given for APC, special attention should be<br />
paid to this aspect <strong>of</strong> patient care.<br />
ASCO provides evidence-based guidelines for medications<br />
based on likelihood <strong>of</strong> emesis by chemotherapy regimen. 5<br />
Published in 2011, the recommended support for highlyemetogenic<br />
chemotherapy consists <strong>of</strong> a three-drug regimen:<br />
dexamethasone, a 5HT3 inhibitor, and a neurokinin-1<br />
(NK-1) inhibitor. For moderately-emetogenic chemotherapy,<br />
data supports a single dose <strong>of</strong> palonestron followed by 3<br />
days <strong>of</strong> dexamethasone. The guidelines suggest dexamethasone<br />
alone as premedication for patients starting a lowemetogenic<br />
regimen such as gemcitabine, though the steroid<br />
dose may be decreased in APC patients with glucose intolerance<br />
or new diabetes. 5<br />
In the most recent edition, both the NCCN and ASCO<br />
guidelines acknowledge the antiemetic effect <strong>of</strong> olanzapine.<br />
5,6 The authors <strong>of</strong> ASCO guidelines suggest olanzapine<br />
for patients with refractory chemotherapy-induced nausea<br />
and vomiting (CINV) that occurs despite appropriate prophylaxis.<br />
The NCCN guidelines include olanzapine 2.5–5 mg<br />
twice daily for patients with breakthrough CINV. 6 However,<br />
subsequent to publication <strong>of</strong> these guidelines, additional<br />
data has emerged showing superior control <strong>of</strong> delayed nausea<br />
with an olanzapine-based regimen over the standard<br />
recommendations for patients undergoing highly emetogenic<br />
chemotherapy, which included an NK-1 inhibitor. 7<br />
An atypical antipsychotic, olanzapine blocks several neurotransmitters<br />
including serotonin, dopamine, acetylcholine,<br />
and histamine. The proposed antiemetic mechanism is<br />
thought to be D2 and 5HT3 inhibition, used in an <strong>of</strong>f-label<br />
indication for nausea. 7,8 Two recent large randomized studies<br />
demonstrate efficacy for patients receiving highly emetogenic<br />
chemotherapy. The largest effect <strong>of</strong> olanzapine<br />
appears to be on delayed nausea with complete control as<br />
late as day 5 or 7. 7,8 Additional improvements in global<br />
functioning, insomnia, appetite, and fatigue have been<br />
found to be significant over placebo (p � 0.01). 7,8 Weight<br />
KEY POINTS<br />
● Celiac plexus neurolysis alleviates pain in the majority<br />
<strong>of</strong> patients with pancreatic cancer and should be<br />
performed early—even at time <strong>of</strong> cancer diagnosis.<br />
● With highly emetogenic chemotherapy indicated for<br />
advanced pancreatic cancer, nausea and vomiting<br />
should be managed rigorously per evidence in the<br />
available literature.<br />
● Malignant bowel obstruction can be palliated quickly<br />
with optimal medical management: gastric decompression,<br />
octreotide, parenteral opioids, and standing<br />
antiemetics.<br />
● In patients with pancreatic cancer, exocrine insufficiency<br />
contributes to weight loss and failure to thrive,<br />
but is reversible with pancreatic enzyme replacement<br />
therapy.<br />
● Depression occurs in the majority <strong>of</strong> patients with<br />
advanced pancreatic cancer and should be treated<br />
aggressively and holistically for improved quality <strong>of</strong><br />
life.<br />
244<br />
gain and hearty appetite are described side effects <strong>of</strong> olanzapine.<br />
Olanzapine appears to be well-tolerated in patients undergoing<br />
chemotherapy. 7,8 In 123 patients, no grade 3 or 4<br />
toxicity was found by Navari and colleagues, though patients<br />
noted fatigue, drowsiness, and dry mouth most <strong>of</strong>ten. 7<br />
Prescribers should familiarize themselves with potential<br />
toxicities. Prolonged administration, for months to years,<br />
increases the risk <strong>of</strong> diabetes and pancreatitis in the psychiatric<br />
population, though this is less concerning with the<br />
prognosis <strong>of</strong> APC. Finally, olanzapine carries a black box<br />
warning for increased mortality in elderly patients with<br />
dementia-related psychosis.<br />
Malignant Gastroparesis<br />
Gastroparesis associated with pancreatic cancer is a welldescribed<br />
phenomenon. Up to one-half <strong>of</strong> patients with APC<br />
experience slowed gastric emptying with no anatomic obstruction.<br />
9 Cancer in the pancreas alters normal gut motility<br />
by direct infiltration <strong>of</strong> autonomic nerve fibers and by<br />
altering neurohormonal pathways within the bowel. 9 Prior<br />
abdominal surgery, radiation, or comorbid diabetes also<br />
contribute to gastroparesis. 9<br />
A careful medication review and esophagogastroduodenoscopy<br />
may rule out reversible causes <strong>of</strong> gastroparesis. The<br />
gold standard for diagnosis is functional imaging, but APC<br />
patients with refractory nausea, vomiting, bloating, and<br />
early satiety should be started on prokinetics empirically—<br />
either metoclopramide 5–10 mg four times daily or erythromycin.<br />
9 Small high-calorie, low-fiber meals with ample<br />
liquids are best. Patients feel better sitting upright postprandially,<br />
followed by ambulation. Palliation <strong>of</strong> severe<br />
cases can be achieved with decompressive gastrostomy or<br />
Roux-en-Y, although overall prognosis and the presence <strong>of</strong><br />
ascites should be considered carefully before interventions. 9<br />
Gastric Outlet Obstruction<br />
LAUREN A. WIEBE<br />
Approximately 15% <strong>of</strong> patients with APC develop gastric<br />
outlet obstruction (GOO) during the disease trajectory and<br />
present with refractory vomiting. 10 <strong>Clinical</strong> suspicion can be<br />
confirmed with a plain radiograph alone, although patients<br />
<strong>of</strong>ten undergo ultrasound, CT scan, or endoscopy to confirm<br />
the diagnosis. Historically, an open, palliative gastrojejunostomy<br />
(GJ) was the definitive treatment for malignant GOO,<br />
but surgery can result in substantial morbidity and even<br />
mortality. 10,11 More recently, self-expanding metallic stents<br />
(SEMS) have become a safe and easier alternative with<br />
minimal morbidity and mortality.<br />
In a systematic review <strong>of</strong> the literature, endoscopic stent<br />
placement was feasible in 96% <strong>of</strong> patients with malignancyrelated<br />
GOO, only limited by inability to transverse the<br />
obstruction or stent deployment failure. After 1 week, 72% <strong>of</strong><br />
patients who underwent SEMS were eating s<strong>of</strong>t or regular<br />
foods compared with 14% <strong>of</strong> patients post GJ. Mean hospital<br />
stay is shorter with stent placement as compared with<br />
surgery (7 vs. 13 days). Repeat obstructive symptoms occurred<br />
in 18% <strong>of</strong> patients with a stent compared with 1% <strong>of</strong><br />
surgical cases, but overall, few complications are seen in<br />
patients with SEMS. 10 GJ, particularly laparoscopic, remains<br />
a viable palliative option, but only for a select group <strong>of</strong><br />
healthy patients with a longer prognosis. 10,11
OPTIMIZING PALLIATIVE CARE FOR APC<br />
Fig. 1. Algorithm for assessing and managing a patient with malignant bowel obstruction (MBO). Reprinted from Ripamonti CI, Easson AM,<br />
Gerdes H. Management <strong>of</strong> malignant bowel obstruction. European Journal <strong>of</strong> Cancer. 2008;44(8):1105-1115, with permission from Elsevier.<br />
Malignant Bowel Obstruction<br />
Many patients with APC develop peritoneal implants and<br />
carcinomatosis. Unfortunately, this can result in malignant<br />
bowel obstruction (MBO) with abdominal distension, pr<strong>of</strong>ound<br />
cramping, and refractory vomiting causing the inability<br />
to eat and <strong>of</strong>ten death. Palliative surgical or endoscopic<br />
interventions can be considered but may be limited by<br />
location <strong>of</strong> disease, multifocal obstruction, presence <strong>of</strong> ascites,<br />
and nutritional status <strong>of</strong> the patient. 11 Detailed in<br />
Figure 1, complex decision making in an urgent setting is<br />
difficult, and MBO can quickly change the trajectory <strong>of</strong> a<br />
patient’s life.<br />
A combined approach <strong>of</strong> gastric decompression with optimal<br />
pharmacologic therapy provides relief <strong>of</strong> suffering as<br />
well as reversal <strong>of</strong> MBO in the majority <strong>of</strong> cases. Adequate<br />
pain control with strong opioids is critical by either intravenous<br />
or subcutaneous route; transdermal medications<br />
should be reserved for later when a patient has stable opioid<br />
requirements. Antisecretory agents are critical to reduce<br />
splanchnic blood flow, bowel secretions, fluid loss, and<br />
cramping pain. Considered the standard <strong>of</strong> care in MBO,<br />
octreotide has been shown to be effective for palliation in<br />
prospective trials. An acceptable total dose <strong>of</strong> octreotide<br />
ranges from 300–1,200 mcg per day given via intravenous or<br />
245
subcutaneous bolus or infusion. Haloperidol is the gold<br />
standard to relieve nausea and vomiting associated with<br />
MBO, although a combination <strong>of</strong> antiemetics may be required.<br />
Corticosteroids may help with nausea and bowel<br />
edema but are less well studied in this setting. 11<br />
The goal for managing MBO is complete relief <strong>of</strong> vomiting,<br />
nausea, and pain. Gastric decompression via nasogastric<br />
tube provides immediate symptom relief, but should be<br />
temporarily used, on the order <strong>of</strong> days. With aggressive<br />
around-the-clock pharmacologic therapy detailed above, the<br />
goal is to eliminate symptoms within 24 hours. Without<br />
clear signs <strong>of</strong> returning bowel function in 2 or 3 days, a<br />
venting gastrostomy should be placed early for optimal<br />
palliation and return <strong>of</strong> patient mobility. 11 Hospice and<br />
homecare agencies are familiar with home drainage protocols,<br />
which the patient and/or family can learn.<br />
Malignant Ascites<br />
Malignant ascites causes discomfort, early satiety, bowel<br />
stasis, orthopnea, and diminished mobility. In APC, peritoneal<br />
fluid accumulation occurs by several mechanisms, including<br />
portal hypertension from intrahepatic metastases,<br />
portal vein compression and/or thrombus, peritoneal tumor<br />
implants with increased capillary permeability, or direct<br />
disruption <strong>of</strong> lymphatic drainage from the peritoneum. 12<br />
Treatment goals are to palliate symptoms and preserve<br />
normal function as long as possible.<br />
For exudative ascites, dietary sodium restriction and<br />
246<br />
diuretics may provide relief. These patients have a serumascites<br />
albumin gradient (SAAG) greater than 1.1 and<br />
elevated portal pressures from venous occlusion or diffuse<br />
hepatic metastases. However, patients with carcinomatosis<br />
resulting in transudative ascites and a low SAAG (� 1.1) are<br />
<strong>of</strong>ten primarily refractory to diuretic combinations and<br />
should undergo early paracentesis for palliation. 12 For all<br />
patients with malignant ascites, therapeutic paracentesis<br />
relieves symptoms 90% <strong>of</strong> the time with minimal risk <strong>of</strong><br />
adverse events. 12<br />
The majority <strong>of</strong> patients (88% to 100%) with anticipated<br />
survival greater than 2 to 3 months will benefit from<br />
placement <strong>of</strong> an indwelling pigtail or tunneled catheter to<br />
control ascites with easy drainage at home or clinic. In a<br />
meta-analysis, tunneled catheters have a low risk <strong>of</strong> infectious<br />
complications (2.5%), but no randomized study has<br />
been performed. Nontunneled devices have a reported complication<br />
rate as high as 30% in some series. 12 Although an<br />
<strong>of</strong>f-label use in the United States, PleurX catheters have<br />
been studied in this setting and are commonly used with<br />
success for palliation <strong>of</strong> refractory malignant ascites. 12<br />
Pancreatic Exocrine Insufficiency<br />
LAUREN A. WIEBE<br />
Fig. 2. Common symptoms and complications<br />
<strong>of</strong> advanced pancreatic cancer with<br />
corresponding recommendations.<br />
Abbreviations: WHO, World Health Organization;<br />
SEMS, self-expanding metal<br />
stent; NG, nasogastric; SAAG, serumascites<br />
albumin gradient; IU, international<br />
units; SSRI, selective serotonin reuptake<br />
inhibitor.<br />
Patients with pancreatic cancer are at high risk for<br />
pancreatic exocrine insufficiency (PEI). When rigorously<br />
assessed 1 year after partial pancreatectomy, 55% <strong>of</strong> patients<br />
have PEI by diagnostic criteria. 13 Radiation and<br />
surgery may cause ductal fibrosis resulting in poor delivery
OPTIMIZING PALLIATIVE CARE FOR APC<br />
<strong>of</strong> lipase and trypsin into the gut lumen. In situ pancreatic<br />
tumors block secretion <strong>of</strong> enzymes, leading to failure absorbing<br />
fat-soluble vitamins such as A, D, E, and K. Although<br />
patients may not manifest clinical steatorrhea until the<br />
lipase concentration falls below 10% <strong>of</strong> normal, symptoms <strong>of</strong><br />
maldigestion can be debilitating with abdominal cramping,<br />
bloating, and urgent stools. 14 PEI also contributes greatly<br />
to the weight loss and malnutrition seen in patients with<br />
APC. 14<br />
Diagnostic testing can be pursued but is expensive, difficult<br />
to perform, and usually unnecessary in this population.<br />
Patients with pancreatic cancer and weight loss should be<br />
given empiric supplementation with pancreatic enzymes.<br />
When taken with each meal or shortly afterward, 40,000–<br />
50,000 IU <strong>of</strong> lipase results in substantial weight gain over<br />
placebo in several studies. 14 If not clinically effective after<br />
several weeks, the dose should be increased as tolerated.<br />
Because a low gastric pH irreversibly inhibits lipase, enzymes<br />
should never be taken on an empty stomach. H2blockade<br />
and proton-pump inhibition both augment the<br />
effect <strong>of</strong> enzyme supplementation; either acid-suppressant<br />
can be added for augmented absorption <strong>of</strong> fat. Low-fat<br />
diets have not been shown to improve symptoms, but<br />
moderate-fat diets are acceptable paired with enzyme supplements.<br />
14 Medium-chain fatty acids are absorbed without<br />
lipase, so referral to a specialized dietician is helpful for<br />
guidance.<br />
Anorexia-Cachexia Syndrome<br />
One <strong>of</strong> the most striking symptoms observed in patients<br />
with pancreatic cancer is the anorexia-cachexia syndrome<br />
(ACS). With circulating proinflammatory cytokines, patients<br />
demonstrate negative protein-energy balance along with<br />
minimal desire to eat, resulting in pr<strong>of</strong>ound muscle wasting<br />
and functional decline. The hallmark <strong>of</strong> cachexia is that the<br />
syndrome is incompletely reversed by appetite stimulants<br />
and nutritional intervention. 15 ACS can be seen in earlystage<br />
pancreatic cancer, but occurs in more than one-half <strong>of</strong><br />
patients with advanced disease leading to poor quality <strong>of</strong> life<br />
and survival. 15<br />
The ability to stabilize the weight loss can improve quality<br />
<strong>of</strong> life, but interventions remain limited with no FDAapproved<br />
therapy for ACS. 15 Currently, options include<br />
targeted dietary counseling and appetite stimulants, typically<br />
with oral progestin. 16 Berenstein and Ortiz performed<br />
a rigorous review <strong>of</strong> the literature evaluating megestrol<br />
acetate compared with placebo in patients with advanced<br />
cancer. In total, megestrol was found to have a relative risk<br />
<strong>of</strong> 3.03 (95% CI 1.83 to 5.01) for improved appetite compared<br />
with placebo. Most trials utilized a dose between 400 and<br />
800 mg daily. Both physicians and patients should be aware<br />
<strong>of</strong> the prothrombotic tendency <strong>of</strong> progestins. 17 No other<br />
orexigenics have a documented benefit in ACS, though<br />
individual patient response is variable to drugs such as<br />
dronabinol. 16<br />
Recent research on ACS is promising and may be clinically<br />
translatable in the near future. Recent phase II interventions<br />
targeted components <strong>of</strong> the inflammatory pathway<br />
such as IL-6, COX-2, TNF-alpha, NF-kappaB, and growth<br />
hormone. 15 A phase II dose-escalation study <strong>of</strong> thalidomide<br />
in 35 patients with cancer-related ACS resulted in significantly<br />
improved appetite in 64% <strong>of</strong> patients after 2 weeks<br />
(p � 0.001). 18 Finally, dietary omega-3 fatty acids suppress<br />
inflammatory and angiogenic pathways in pancreatic cancer<br />
models, suggesting a potential future role for patients with<br />
APC and ACS. 19<br />
Depression<br />
Patients with pancreatic cancer have the highest incidence<br />
<strong>of</strong> depression seen in any cancer population. 20 Studies<br />
report rates <strong>of</strong> depression in APC ranging from 33% to 76%<br />
with resultant poor quality <strong>of</strong> life and difficulty achieving<br />
pain control. 21 Elevated levels <strong>of</strong> circulating cytokines such<br />
as IL-6 and TNF-alpha are thought to alter neurohormonal<br />
pathways in the brain causing depressive symptoms even<br />
before the diagnosis <strong>of</strong> cancer is suspected. 21 In particular,<br />
men with pancreatic cancer have a death rate from suicide<br />
11 times that <strong>of</strong> the general population, reflecting the need<br />
for an improved holistic approach to care <strong>of</strong> these patients.<br />
2,20,22<br />
In patients with advanced cancer, it may be difficult to tell<br />
clinical depression from normal sadness experienced with a<br />
daunting prognosis. Common somatic symptoms <strong>of</strong> APC<br />
such as fatigue, anorexia, and weight loss can overlap with<br />
the signs and symptoms <strong>of</strong> depression. 21 Simply asking a<br />
patient whether he or she has “felt depressed most <strong>of</strong> the<br />
time” is a validated tool with good sensitivity and specificity<br />
for depression, even in patients with terminal illness. 2<br />
Timely treatment <strong>of</strong> depression is critical in APC, as patients<br />
with depression are more likely to endorse a desire for<br />
hastened death. 2<br />
Antidepressant medications—mainly selective serotonin<br />
reuptake inhibitors (SSRIs)—have been shown to be effective<br />
in patients with advanced cancer. Selection <strong>of</strong> SSRI<br />
should be dictated by the side effect pr<strong>of</strong>ile. 21 In APC<br />
patients suffering marked anorexia, mirtazapine may be an<br />
advantageous choice, although not studied for this indication.<br />
21 Anxiolytics are helpful in patients with underlying<br />
anxiety disorders. In clinical practice, lorazepam is used<br />
frequently without clear evidence to guide ongoing administration.<br />
Patients requiring more than twice daily dosing <strong>of</strong><br />
a short-acting benzodiazepine could be switched to clonazepam<br />
with a longer half-life to avoid rebound anxiety,<br />
usually starting at 0.5 mg orally twice daily.<br />
In addition to pharmacologic interventions, patients<br />
with APC benefit greatly from supportive counseling to<br />
strengthen innate coping strategies and help with anticipatory<br />
grief as an integral part <strong>of</strong> their cancer care. 2,21 Additional<br />
support will be provided by early referral to a<br />
palliative care provider. Depression and anxiety <strong>of</strong>ten occur<br />
because <strong>of</strong> distress from unaddressed fears <strong>of</strong> death or the<br />
symptoms that may arise in the process <strong>of</strong> dying. 2 Early and<br />
frequent discussion <strong>of</strong> symptom concerns and quality <strong>of</strong> life<br />
preserves hope for patients with APC. Active interventions<br />
such as Dignity Therapy have shown promise in relief <strong>of</strong><br />
suffering in patients with limited prognosis. 2 ASCO recently<br />
released a statement <strong>of</strong> provisional clinical opinion that<br />
palliative care leads to better patient and caregiver outcomes.<br />
23 In the end, patients want to be individuals acknowledged<br />
with compassion and healing at this time in<br />
their lives.<br />
Conclusion<br />
Patients with advanced pancreatic cancer suffer numerous<br />
symptoms throughout the illness. It is critical that these<br />
247
patients are cared for completely with aggressive palliation<br />
<strong>of</strong> symptoms to maximize their remaining time. Celiac<br />
plexus neurolysis improves pain in the majority <strong>of</strong> APC<br />
patients and should be moved earlier in the analgesic<br />
paradigm. Antiemetics continue to evolve for maximal prevention<br />
<strong>of</strong> CINV, including olanzapine in recent studies.<br />
Nutritional status remains important to patients and families;<br />
weight loss related to undiagnosed pancreas exocrine<br />
Author’s Disclosures <strong>of</strong> Potential Conflicts <strong>of</strong> Interest<br />
Author<br />
Lauren A. Wiebe*<br />
*No relevant relationships to disclose.<br />
Employment or<br />
Leadership<br />
Positions<br />
Consultant or<br />
Advisory Role<br />
1. Siegel R, Naishadham D, Jemal A. Cancer Statistics, <strong>2012</strong>. CA Cancer<br />
J Clin. <strong>2012</strong>;62:10-29.<br />
2. Chochinov HM. Dying, dignity and new horizons in palliative end-<strong>of</strong>-life<br />
care. CA Cancer J Clin. 2006;56:84-103.<br />
3. Yan BM, Myers RP. Neurolytic celiac plexus block for pain control in<br />
unresectable pancreatic cancer. Am J Gastroenterol. 2007;102:430-438.<br />
4. Wyse JM, Carone M, Paquin SC, et al. Randomized, double-blind,<br />
controlled trial <strong>of</strong> early endoscopic ultrasound-guided celiac plexus neurolysis<br />
to prevent pain progression in patients with newly diagnosed, painful,<br />
inoperable pancreatic cancer. J Clin Oncol. 29:3541-3546.<br />
5. Basch E, Prestrud AA, Hesketh PJ, et al. Antiemetics: <strong>American</strong> <strong>Society</strong><br />
<strong>of</strong> <strong>Clinical</strong> <strong>Oncology</strong> Practice Guideline Update. J Clin Oncol. 2011;29:4189-<br />
4198.<br />
6. National Comprehensive Cancer Network. <strong>Clinical</strong> Practice Guidelines<br />
in <strong>Oncology</strong>. http://www.nccn.org/pr<strong>of</strong>essionals/physician_gls/f_guidelines.<br />
asp#supportive. Accessed February 29, <strong>2012</strong>.<br />
7. Navari RM, Gray SE, Kerr AC. Olanzapine versus aprepitant for the<br />
prevention <strong>of</strong> chemotherapy-induced nausea and vomiting: A randomized<br />
phase III trial. J Support Oncol. 2011;9:188-195.<br />
8. Tan L, Liu J, Liu X, et al. <strong>Clinical</strong> research <strong>of</strong> olanzapine for prevention<br />
<strong>of</strong> chemotherapy-induced nausea and vomiting. J Exp Clin Cancer Res.<br />
2009;29:131.<br />
9. Donthireddy KR, Ailawadhi S, Nasser E, et al. Malignant gastroparesis:<br />
Pathogenesis and management <strong>of</strong> an underrecognized disorder. J Support<br />
Oncol. 2007;5:355-363.<br />
10. Juernink SM, van Eijck CHJ, Steyerberg EW, et al. Stent versus<br />
gastrojejunostomy for the palliation <strong>of</strong> gastric outlet obstruction: A systematic<br />
review. BMC Gastroenterol. 2007;7:18.<br />
11. Ripamonti CT, Easson AM, Gerdes H. Management <strong>of</strong> malignant bowel<br />
obstruction. Eur J Cancer. 2008;44:1105-1115.<br />
12. Chung M and Kozuch P. Treatment <strong>of</strong> malignant ascites. Curr Treat<br />
Options Oncol. 2008;9:215-233.<br />
248<br />
insufficiency improves with aggressive enzyme supplementation.<br />
Bowel dysfunction from gastroparesis or malignant<br />
obstruction is treatable with palliative interventions. Finally,<br />
the existential and psychosocial concerns <strong>of</strong> patients<br />
facing death from pancreatic cancer should be addressed in<br />
a holistic manner. Early integration <strong>of</strong> palliative care will<br />
help patients achieve the best quality <strong>of</strong> life in the face <strong>of</strong><br />
this unfortunate diagnosis.<br />
Stock<br />
Ownership Honoraria<br />
REFERENCES<br />
Research<br />
Funding<br />
Expert<br />
Testimony<br />
LAUREN A. WIEBE<br />
Other<br />
Remuneration<br />
13. Halloran CM, Cox TF, Chauhan S, et al. Partial pancreatic resection for<br />
pancreatic malignancy is associated with sustained pancreatic exocrine<br />
failure and reduced quality <strong>of</strong> life: A prospective study. Pancreatology.<br />
2011;11:535-545.<br />
14. Dominguez-Munoz JE. Pancreatic exocrine insufficiency: Diagnosis<br />
and treatment. J Gastroenterol Hepatol. 2011;26 Suppl. 2:12-16.<br />
15. Dodson S, Baracos VE, Jatoi A, et al. Muscle wasting in cancer<br />
cachexia: <strong>Clinical</strong> implications, diagnosis, and emerging treatment strategies.<br />
Annu Rev Med. 2011;62:265-279.<br />
16. Dy SM, Lorenz KA, Naeim A, et al. Evidence-based recommendations<br />
for cancer fatigue, anorexia, depression and dyspnea. J Clin Oncol. 2008;26:<br />
3886-3895.<br />
17. Berenstein G, Ortiz Z. Megestrol acetate for treatment <strong>of</strong> anorexiacachexia<br />
syndrome. Cochrane Database <strong>of</strong> Systematic Reviews 2005, Issue 2.<br />
Art. No.: CD004310.<br />
18. Davis M, Lasheen W, Walsh D, et al. A phase II dose titration study <strong>of</strong><br />
thalidomide for cancer-associated anorexia. J Pain Symptom Manage. <strong>2012</strong>;<br />
43:78-86.<br />
19. Boutros C, Somasundar P, Razzak A, et al. Omega-3 fatty acids:<br />
Investigations from cytokine regulation to pancreatic cancer gene suppression.<br />
Arch Surg. 2010;145:515-520.<br />
20. Clark KL, Loscalzo M, Trask PC, et al. Pyschological distress in<br />
patients with pancreatic cancer—an understudied group. Pychooncology.<br />
2010;19:1313-1320.<br />
21. Mayr M, Schmid RM. Pancreatic cancer and depression: Myth and<br />
truth. BMC Cancer. 2010;10:569.<br />
22. Turaga KT, Malafa MP, Jacobsen PB, et al. Suicide in patients with<br />
pancreatic cancer. Cancer. 2011;117:642-647.<br />
23. Smith TJ, Temin S, Alesi ER, et al. <strong>American</strong> <strong>Society</strong> <strong>of</strong> <strong>Clinical</strong><br />
<strong>Oncology</strong> Provisional <strong>Clinical</strong> Opinion: The Integration <strong>of</strong> Palliative Care into<br />
Standard <strong>Oncology</strong> Care. http://jco.ascopubs.org/cgi/doi/10.1200/JCO.2011.<br />
38.5161. Accessed February 29, <strong>2012</strong>.
GLOBAL PERSPECTIVE OF LOCALLY ADVANCED<br />
GASTRIC CANCER: DIFFERENT TREATMENT<br />
PARADIGMS AND THEIR RATIONALE<br />
CHAIR<br />
Manish A. Shah, MD<br />
Weill Cornell Medical College<br />
New York, NY<br />
SPEAKERS<br />
Sam S. Yoon, MD<br />
Massachusetts General Hospital<br />
Boston, MA<br />
Yoon-Koo Kang, MD, PhD<br />
Asan Medical Center<br />
Seoul, South Korea
Varying Lymphadenectomies for Gastric<br />
Adenocarcinoma in the East Compared<br />
with the West: Effect on Outcomes<br />
By Benjamin Schmidt, MD, and Sam S. Yoon, MD<br />
Overview: There are notable differences in surgical approaches<br />
to gastric adenocarcinoma throughout the world,<br />
particularly in terms <strong>of</strong> the extent <strong>of</strong> lymphadenectomy (LAD).<br />
In high-incidence countries such as Japan and South Korea,<br />
more extensive (e.g., D2) lymphadenectomies are standard,<br />
and these surgeries are generally done by experienced surgeons<br />
with low morbidity and mortality. In countries such as<br />
the United States, where the incidence <strong>of</strong> gastric adenocarcinoma<br />
is 10-fold lower, the majority <strong>of</strong> patients are treated at<br />
nonreferral centers with less extensive (e.g., D1 or D0) lymphadenectomy.<br />
There is little disagreement among gastric cancer<br />
(GC) experts that the minimum lymphadenectomy that<br />
should be performed for gastric adenocarcinoma should be at<br />
least a D1 lymphadenectomy, and many <strong>of</strong> these experts<br />
recommend a D2 lymphadenectomy. More extensive lymphadenectomies<br />
provide better staging <strong>of</strong> patient disease and<br />
IT IS estimated that there are more than one million cases<br />
<strong>of</strong> GC worldwide per year, making it the fourth most<br />
common cancer. 1 Nearly three-quarters <strong>of</strong> cases occur in<br />
developing countries, and nearly half <strong>of</strong> cases occur in<br />
eastern Asia (mainly in China). GC is the second leading<br />
worldwide cause <strong>of</strong> cancer death for both men and women,<br />
with a total <strong>of</strong> more than 700,000 deaths each year. The<br />
incidence <strong>of</strong> gastric adenocarcinoma varies tremendously<br />
throughout the world and country by country, with the<br />
highest incidence occurring in South Korea at 66.5 to 72.5<br />
per 100,000 males and 19.5 to 30.4 per 100,000 females. 2<br />
The incidence <strong>of</strong> GC in the United States is only one-tenth<br />
that <strong>of</strong> South Korea. The estimated number <strong>of</strong> new GC cases<br />
in the United States in <strong>2012</strong> was 21,320, and the estimated<br />
number <strong>of</strong> deaths was 10,540. 3<br />
In addition to the global differences in GC epidemiology,<br />
there are also appreciable differences in the surgical treatment<br />
<strong>of</strong> GC, particularly in the extent <strong>of</strong> LAD. This article<br />
will examine the effect <strong>of</strong> varying LADs in Eastern and<br />
Western countries on patient outcomes. Institutional studies<br />
from two countries, Japan and South Korea, will be used<br />
to represent two high-incidence Eastern countries, and institutional<br />
and national database studies from the United<br />
States will be used to represent low-volume Western countries.<br />
Definitions<br />
Before discussion <strong>of</strong> differences in LAD for gastric adenocarcinoma,<br />
one should define the terms to be used. The node<br />
stations surrounding the stomach were precisely defined by<br />
the Japanese Gastric Cancer Association (JGCA), formerly<br />
known as the Japanese Research <strong>Society</strong> for Gastric Cancer,<br />
in 1973 4 (Fig. 1 and Table 1). In its most recent GC<br />
treatment guidelines, the JGCA again changed the definitions<br />
for D levels <strong>of</strong> LAD such that they are now defined<br />
according to the type <strong>of</strong> gastrectomy performed rather than<br />
the location <strong>of</strong> the tumor (Table 2). 5 To broadly summarize,<br />
a D1 LAD removes the first tier <strong>of</strong> perigastric nodes and the<br />
left gastric artery nodes whereas a D2 LAD removes the<br />
250<br />
likely reduce locoregional recurrence rates. Two large, prospective<br />
randomized trials performed in the United Kingdom<br />
and the Netherlands in the 1990s failed to demonstrate a<br />
survival benefit <strong>of</strong> D2 over D1 lymphadenectomy, but these<br />
trials have been criticized for inadequate surgical training and<br />
high surgical morbidity and mortality rates (10% to 13%) in the<br />
D2 group. More recent studies have demonstrated that Western<br />
surgeons can be trained to perform D2 lymphadenectomies<br />
on Western patients with low morbidity and mortality.<br />
The 15-year follow-up <strong>of</strong> the Netherlands trial now demonstrates<br />
an improved disease-specific survival and locoregional<br />
recurrence in the D2 group. Retrospective analyses and<br />
one prospective, randomized trial suggest that there may be a<br />
survival benefit to more extensive lymphadenectomies when<br />
performed safely, but this assertion requires further validation.<br />
second tier <strong>of</strong> nodes that generally fall along primary and<br />
secondary branches <strong>of</strong> the celiac axis (i.e., splenic artery,<br />
common hepatic artery, proper hepatic artery). The JGCA<br />
guidelines recommend a D2 LAD for all gastric carcinomas<br />
beyond a clinical T1 tumor (e.g., tumor invades lamina<br />
propria, muscularis mucosa, or submucosa).<br />
Differences in Surgical Volume and Extent <strong>of</strong> LAD<br />
Japan and South Korea have two <strong>of</strong> the highest incidences<br />
<strong>of</strong> gastric adenocarcinoma in the world, but despite the high<br />
incidence <strong>of</strong> gastric adenocarcinoma in these countries,<br />
patients are <strong>of</strong>ten referred to tertiary centers for treatment.<br />
Two-thirds <strong>of</strong> all GC surgeries in South Korea are performed<br />
at 16 high-volume institutions, which perform at least 200<br />
GC surgeries per year. Thus GC surgeons at high-volume<br />
institutions in South Korea gain tremendous experience in<br />
the surgical management <strong>of</strong> GC. As noted earlier, the<br />
minimum LAD performed by Japanese and Korean surgeons<br />
for gastric adenocarcinoma (except for T1 tumors) is a D2<br />
LAD. 5 Despite performing extensive LADs, the morbidity<br />
and mortality rates are quite low. For example, Seoul<br />
National University Hospital (SNUH), which performs almost<br />
1,000 GC operations per year, recently reported a<br />
morbidity rate <strong>of</strong> 18% and a mortality rate <strong>of</strong> 0.5%. 6<br />
In contrast, the majority <strong>of</strong> GC surgeries in the United<br />
States are performed at nonreferral centers. A “high volume”<br />
institution in the United States has been defined in<br />
some studies as centers with as low as more than 15 to 20<br />
surgeries per year. 7,8 Birkmeyer and colleagues reviewed a<br />
database <strong>of</strong> Medicare patients and found that hospitals with<br />
more than 20 gastrectomies per year had one-third less risk<br />
From the Department <strong>of</strong> Surgery, Massachusetts General Hospital, Boston, MA; Harvard<br />
Medical School, Boston, MA.<br />
Authors’ disclosures <strong>of</strong> potential conflicts <strong>of</strong> interest are found at the end <strong>of</strong> this article.<br />
Address reprint requests to Sam S. Yoon, MD, Division <strong>of</strong> Surgical <strong>Oncology</strong>, Department<br />
<strong>of</strong> Surgery, Massachusetts General Hospital, Yawkey 7B, 55 Fruit St., Boston, MA 02114;<br />
email: syoon@partners.org.<br />
© <strong>2012</strong> by <strong>American</strong> <strong>Society</strong> <strong>of</strong> <strong>Clinical</strong> <strong>Oncology</strong>.<br />
1092-9118/10/1-10
GASTRIC CANCER LYMPHADENECTOMY AND OUTCOMES<br />
<strong>of</strong> perioperative death (odds ratio 0.55–0.74), yet more than<br />
80% <strong>of</strong> patients were operated on at centers that performed<br />
20 or fewer gastrectomies per year. 7 As most U.S. surgeons<br />
see only a few GC patients a year, they likely err on the side<br />
<strong>of</strong> more limited LADs in order to avoid excess morbidity and<br />
mortality. In the Intergroup 0116 trial published in 2001,<br />
patients were randomly selected after GC surgery to undergo<br />
no further therapy or chemoradiation, and more than<br />
50% <strong>of</strong> operations were less aggressive than a D1 LAD (aka<br />
D0 LAD). 9 Despite the performance <strong>of</strong> less extensive LADs<br />
in the United States, surgical morbidity and mortality rates<br />
for gastric adenocarcinoma are generally much higher in the<br />
United States than in South Korea and Japan. An analysis<br />
<strong>of</strong> the Nationwide Inpatient Sample from 1998 to 2003 <strong>of</strong><br />
more than 50,000 patents with GC found the overall mortality<br />
rate following gastric surgery was 6%. 10 Singleinstitution<br />
series have reported morbidity rates following<br />
gastrectomy <strong>of</strong> up to 40%. 11<br />
Differences in Survival<br />
Gastric adenocarcinoma frequently metastasizes to regional<br />
nodes. For T1 lesions invading the submucosa, node<br />
involvement is found in approximately 20% <strong>of</strong> patients. 12<br />
For T2 lesions (invading muscularis propria), the node<br />
metastasis rate increases to more than 50%. There is some<br />
evidence that at least some patients with node metastases<br />
beyond the immediate perigastric (D1) nodes and into D2<br />
nodes can be cured with surgical resection alone. 13 A sizable<br />
minority <strong>of</strong> GC patients with positive D2 nodes survive for<br />
more than 5 years following D2 lymphadenectomy at the<br />
Japanese National Cancer Center in Tokyo.<br />
Numerous studies have demonstrated decreased overall<br />
KEY POINTS<br />
● More extensive D2 lymphadenectomies are standard<br />
in high-incidence Eastern countries such as Japan<br />
and South Korea, leading to better staging <strong>of</strong> disease<br />
and likely lower rates <strong>of</strong> locoregional recurrence.<br />
● In the United States (a low-incidence Western country),<br />
the vast majority <strong>of</strong> gastric resections are performed<br />
at low-volume (less than 20 cases per year)<br />
centers with generally less extensive lymphadenectomies<br />
and higher morbidity and mortality.<br />
● Stage for stage, overall survival is worse in the<br />
United States than in Japan and South Korea, but<br />
much <strong>of</strong> this difference could be explained by stage<br />
migration and clinicopathologic differences between<br />
gastric cancers in Eastern versus Western countries.<br />
● The Dutch and U.K. D1 versus D2 lymphadenectomy<br />
randomized trials were flawed, and further prospective<br />
randomized trials <strong>of</strong> lymphadenectomies performed by<br />
well-trained Western surgeons on Western patients are<br />
needed to determine if there is an overall survival<br />
benefit to more extensive lymphadenectomies.<br />
● Strategies to improve the surgical outcomes <strong>of</strong> patients<br />
with gastric cancer in low-incidence Western<br />
countries include referral to tertiary centers and<br />
improved training <strong>of</strong> surgeons.<br />
Fig 1. Location <strong>of</strong> node stations according to the Japanese Gastric<br />
Cancer Association. 38 (A) Perigastric nodes stations 1 to 7. (B) Second<br />
tier node stations 8 to 12 and 14.<br />
survival (OS) after potentially curative gastrectomy for<br />
gastric adenocarcinoma in the West compared with the East.<br />
Table 3 demonstrates 5-year OS results stage-for-stage from<br />
four large databases based on the sixth <strong>American</strong> Joint<br />
Committee on Cancer (AJCC) staging system. D2 LADs are<br />
generally performed at the National Cancer Center (NCC) in<br />
Tokyo, Japan, and at SNUH is South Korea. The median<br />
number <strong>of</strong> examined nodes at both these institutions is<br />
greater than 30, and there is a remarkable similarity in the<br />
5-year survival figures from these two institutions. In the<br />
U.S. Surveillance Epidemiology and End Results (SEER)<br />
database, most patients had either a D0 or D1 LAD, and the<br />
median number <strong>of</strong> examined nodes is 10 to 11. 14 For stages<br />
I to III, the 5-year survival rate is 14% to 30% lower for<br />
SEER database patients. At Memorial Sloan-Kettering Cancer<br />
Center, where approximately 80% <strong>of</strong> patients receive a<br />
D2 LAD, 15 stage-for-stage OS is intermediate between<br />
SEER database patients and NCC/SNUH patients. 16 Some<br />
<strong>of</strong> these differences in 5-year survival can clearly be attributed<br />
to stage migration. 17 As more nodes are harvested,<br />
more malignant nodes can be found, leading to a higher<br />
staging <strong>of</strong> patients.<br />
There are also clearly some differences in the clinical and<br />
pathologic presentation and adjuvant treatment <strong>of</strong> GC in the<br />
West compared with the East that make comparison <strong>of</strong><br />
outcomes difficult. In terms <strong>of</strong> patient demographics, West-<br />
251
Table 1. Regional Lymph Nodes <strong>of</strong> the Stomach a<br />
Number Description<br />
1 Right paracardial<br />
2 Left paracardial<br />
3 Lesser curvature<br />
a Along branches <strong>of</strong> left gastric artery<br />
b Along second branch and distal part <strong>of</strong> right gastric artery<br />
4 Greater curvature<br />
sa Along short gastric vessels<br />
sb Along left gastroepiploic vessels<br />
d Along second branch and distal part <strong>of</strong> right gastroepiploic artery<br />
5 Suprapyloric along first branch and proximal part <strong>of</strong> right gastric artery<br />
6 Infrapyloric along first branch and proximal part <strong>of</strong> right gastroepiploic<br />
artery<br />
7 Left gastric artery<br />
8 Common hepatic artery<br />
a Anterosuperior group<br />
p Posterior group<br />
9 Celiac artery<br />
10 Splenic hilum<br />
11 Along splenic artery<br />
p Along proximal splenic artery<br />
d Along distal splenic artery<br />
12 Hepatoduodenal ligament<br />
a Along proper hepatic artery<br />
b Along bile duct<br />
p Along portal vein<br />
14 Along superior mesenteric vessels<br />
v Along superior mesenteric vein<br />
a Along superior mesenteric artery<br />
a<br />
Adapted from the Japanese Gastric Cancer Association’s 2011 classifications.<br />
38<br />
ern patients compared to Eastern patients are generally: (1)<br />
older, (2) have a higher body mass index, (3) have a lower<br />
incidence <strong>of</strong> H pylori infection, (4) have more proximal<br />
tumors, (5) present with later stage disease, and (6) receive<br />
different adjuvant therapies. Many <strong>of</strong> the factors more<br />
common in Western patients are negative prognostic factors<br />
for gastric adenocarcinoma. Verdecchia and colleagues compared<br />
GC incidence and survival in four regions (Campina,<br />
Brazil; Iowa, United States; Varese province, Italy; and<br />
Osaka Japan). 18 U.S. patients were older, had more proximal<br />
cancers, more commonly presented with metastatic<br />
disease, and had worse survival rates than Japanese patients.<br />
One <strong>of</strong> the best attempts to determine if Eastern patients<br />
generally have a better OS than Western patients following<br />
more extensive LADs was performed by Strong and colleagues.<br />
They analyzed 711 U.S. patients treated at Memorial<br />
Sloan-Kettering Cancer Center and 1,646 Korean<br />
patients treated at Seoul St. Mary’s Hospital. 19 In this<br />
study, the median age <strong>of</strong> U.S. patients was 10 years older<br />
than that <strong>of</strong> Korean patients (69 vs. 59 years old). Thirtynine<br />
percent <strong>of</strong> U.S. patients had upper third or gastroesophageal<br />
junction tumors compared to only 9.4% <strong>of</strong><br />
Korean patients, and 59% <strong>of</strong> U.S. patients had intestinal<br />
type tumors compared to 49% <strong>of</strong> Korean patients. D2 LAD<br />
was performed in 84% and 89% <strong>of</strong> U.S. and Korean patients,<br />
respectively, but there were more nodes examined<br />
in Korean patients than U.S. patients (97% <strong>of</strong> Korean<br />
patients with � 15 nodes examined compared to 78% <strong>of</strong> U.S.<br />
patients). The T stage, N stage, and overall stage <strong>of</strong> U.S.<br />
patients were significantly more advanced than those <strong>of</strong><br />
252<br />
Table 2. Extent <strong>of</strong> Lymphadenectomy a<br />
Extent <strong>of</strong> Gastrectomy D1 Dissection D1� Dissection D2 Dissection<br />
Total gastrectomy 1–7 D1 � 8a, 9p,<br />
11p<br />
Distal/subtotal<br />
gastrectomy<br />
Proximal<br />
gastrectomy<br />
1, 3, 4sb, 4d, 5,<br />
6, 7<br />
1,2,3a, 4sa, 4sb,<br />
7<br />
Korean patients (p � 0.0001, p � 0.008, p � 0.0001,<br />
respectively). Survival was worse, stage-for-stage, in U.S.<br />
patients compared to Korean patients for AJCC stages I to<br />
III. Interestingly, the survival <strong>of</strong> U.S. patients with middle<br />
or upper tumors was worse than that <strong>of</strong> Korean patients, but<br />
U.S. and Korean patients had similar OS for distal tumors.<br />
After adjusting for clinically important prognostic factors,<br />
Korean patients still had a 30% better disease-specific survival<br />
rate than U.S. patients. There are some potential<br />
confounding variables in this study, including the fact that<br />
316 U.S. patients were excluded because they received<br />
neoadjuvant treatment. However, this study does suggest<br />
that after controlling for prognostic factors and following<br />
relatively uniform D2 LAD, Eastern patients still have a<br />
better survival rate compared with Western patients.<br />
Potential Benefits <strong>of</strong> More Extensive LADs<br />
D1 � 8a, 9p, 11p,<br />
11d, 12a<br />
D1 � 8a, 9 D1 � 8a, 9, 11p,<br />
12a<br />
D1 � 8a, 9, N/A<br />
11p<br />
a Adapted from the Japanese Gastric Cancer Association’s 2010 guidelines. 5<br />
LAD for cancer can serve three primary purposes: staging<br />
<strong>of</strong> disease, prevention <strong>of</strong> locoregional recurrence, and improvement<br />
in OS. There is little doubt that more extensive<br />
LADs for gastric adenocarcinoma can lead to better staging<br />
<strong>of</strong> disease. The 2010 (seventh edition) AJCC Cancer Staging<br />
Manual for gastric adenocarcinoma recommends that at<br />
least 16 nodes be examined for correct assessment <strong>of</strong> the N<br />
category. 20 Despite this, our analysis <strong>of</strong> the SEER database<br />
found that only one-third <strong>of</strong> 18,043 resected GC patients had<br />
16 or more nodes examined. 14 It is difficult to be confident<br />
that a GC is truly node-negative when fewer than 10 nodes<br />
are examined, 21,22 and N1 tumors can be upstaged to N2 or<br />
even N3 tumors as more nodes are harvested. 22,23 Furthermore,<br />
it is impossible to be categorized as N3b if less than 16<br />
nodes are harvested. Thus many patients are understaged<br />
following surgical resection <strong>of</strong> their GCs because <strong>of</strong> inadequate<br />
node sampling.<br />
In the Unites States, the pathologist is usually the one<br />
who finds and examines the dissected nodes. Thus a coordi-<br />
Table 3. Five-Year Overall Survival Rates <strong>of</strong> Patients with GC<br />
by AJCC 6th Edition Stage<br />
SEER, 14<br />
n � 32,531<br />
MSKCC, 16<br />
n � 1,038<br />
SCHMIDT AND YOON<br />
NCC,<br />
Japan, a<br />
n � 6,730<br />
SNUH,<br />
South Korea, 39<br />
n � 12,026<br />
Lymphadenectomy D0, D1 D2 � D1 � D2 � D2<br />
Median nodes examined 10–11 22 � 30 � 30<br />
IA 78% 92% 91.5% 94.4%<br />
IB 58% 85% 84.6% 84.2%<br />
II 34% 49% 69.3% 68.5%<br />
IIIA 20% 32% 50.4% 47.3%<br />
IIIB 8% 11% 30.6% 31.6%<br />
IV 7% 11% 5.4% 17.2%<br />
Abbreviations: AJCC, <strong>American</strong> Joint Committee on Cancer; MSKCC, Memorial<br />
Sloan Kettering Cancer Center; NCC, National Cancer Center; SEER, Surveillance<br />
Epidemiology and End Results; SNUH, Seoul National University Hospital.<br />
a From a written communication with H.K. Yang, SNUH (May, 2011).
GASTRIC CANCER LYMPHADENECTOMY AND OUTCOMES<br />
nated effort is required between the surgeon and pathologist<br />
if more extensive LADs are to result in improved staging <strong>of</strong><br />
patients. In Japan and South Korea, following the en bloc<br />
dissection <strong>of</strong> the stomach and nodes, surgeons generally<br />
dissect out the individual nodal stations from the surgical<br />
specimen, allowing the pathologist to examine and report<br />
the number <strong>of</strong> positive and negative nodes for each nodal<br />
station. Thus more extensive LAD must be combined with<br />
more rigorous pathologic analysis to optimize the node<br />
staging <strong>of</strong> GC.<br />
There is some evidence that more extensive LADs result<br />
in lower rates <strong>of</strong> locoregional recurrence. Locoregional recurrence<br />
after potentially curative surgery for gastric adenocarcinoma<br />
can be quite high. In a 1982 series from the<br />
University <strong>of</strong> Minnesota, 107 patients with gastric adenocarcinoma<br />
underwent second-look laparotomy, and 80% had<br />
recurrence. 24 Of these recurrences, 88% were locoregional,<br />
54% were peritoneal, and 29% were distant. More recently in<br />
the U.S. Intergroup 0116 trial, 177 <strong>of</strong> 275 patients (64%) in<br />
the surgery-only group developed recurrent disease. 9 In<br />
terms <strong>of</strong> the site <strong>of</strong> first relapse, 29% had local recurrence,<br />
72% had regional recurrence, and only 18% had distant<br />
recurrence. Rates <strong>of</strong> locoregional recurrence are generally<br />
lower in reports from both Western and Eastern institutions<br />
that perform more extensive LADs. In a series <strong>of</strong> 367<br />
patients with recurrent gastric adenocarcinoma from Memorial<br />
Sloan-Kettering Cancer Center over 15 years, 81% <strong>of</strong><br />
patients had a D2 or greater LAD, and the median number<br />
<strong>of</strong> nodes removed was 22. 15 Of patients in whom disease<br />
recurred, locoregional recurrence was the initial and only<br />
site <strong>of</strong> recurrence in 26% <strong>of</strong> patients and was a component <strong>of</strong><br />
initial recurrence in 54% <strong>of</strong> patients. Yoo and colleagues<br />
examined 508 patients in whom recurrent disease developed<br />
after curative gastrectomy at Yonsei University in South<br />
Korea. Nineteen percent <strong>of</strong> patients had locoregional recurrence<br />
only as the first site <strong>of</strong> recurrence, and 32.5% <strong>of</strong><br />
patients had locoregional recurrence combined with peritoneal<br />
or distant recurrence as the initial site <strong>of</strong> recurrent<br />
disease. In the Japanese prospective randomized trial <strong>of</strong><br />
adjuvant S-1 chemotherapy, 188 (35.5%) <strong>of</strong> 530 patients<br />
treated with surgery suffered a recurrence. 25 The site <strong>of</strong> first<br />
recurrence in these 188 patients was local in 7.9% and in<br />
nodes in 24.5%.<br />
The effect <strong>of</strong> more extensive LADs on OS for GC is still<br />
controversial. Two large, prospective randomized trials in<br />
Western countries have failed to identify a survival advantage<br />
for D2 over D1 LAD. 26,27 However, these two trials had<br />
fairly high morbidity (43% to 46%) and mortality rates (10%<br />
to 13%) for D2 LAD. In these trials, the distal pancreas and<br />
spleen were <strong>of</strong>ten resected during dissection <strong>of</strong> station 10<br />
and 11 nodes, which likely increased morbidity. Of note in<br />
the Dutch trial, if patients with in-hospital mortality are<br />
excluded, patients with N2 disease had a survival advantage<br />
when treated with a D2 LAD. 28 Several studies have now<br />
demonstrated that D2 LADs can be performed without the<br />
need for distal pancreatectomy 29 or splenectomy. 30,31 Furthermore,<br />
a recent randomized trial in Taiwan demonstrated<br />
an OS advantage <strong>of</strong> more extensive LAD over D1<br />
LAD, with the overall 5-year survival rate being 59.5%<br />
compared with 53.6%, respectively (p � 0.041). 32 However,<br />
the applicability <strong>of</strong> this trial to Western patients has been<br />
called into question. 33 The long-term follow-up <strong>of</strong> the Dutch<br />
trial was recently reported. 34 After a median follow-up <strong>of</strong><br />
15 years, D2 LAD was associated with lower locoregional<br />
recurrence and gastric cancer–related death rates (37% vs.<br />
48%) than D1 LAD. Degiuli and colleagues in Italy have<br />
demonstrated that Western surgeons, following extensive<br />
training, can perform D2 LADs on Western patients with<br />
low morbidity and almost no mortality, 35,36 and survival<br />
results from a prospective randomized trial <strong>of</strong> D1 compared<br />
with D2 LAD from this group are pending. 37<br />
Can Western Surgeons Perform More Extensive<br />
LADs Safely?<br />
Italian GC Study Group approached the issue <strong>of</strong> Western<br />
surgeons performing more extensive LADs in Western patients<br />
in a series <strong>of</strong> two prospective clinical trials. 35,36<br />
Following extensive training <strong>of</strong> 16 surgeons in D2 LAD, a<br />
phase II trial <strong>of</strong> D2 LAD was instituted in which all<br />
surgeries were performed by the two attending surgeons. Of<br />
the 191 patients enrolled in the study, 106 patients (55%)<br />
were ultimately found to be ineligible, usually as a result <strong>of</strong><br />
more extensive disease. The mean number <strong>of</strong> nodes removed<br />
was 39 (range: 22 to 93). Overall postoperative morbidity<br />
and mortality were impressively low at 20.9% and 3.1%,<br />
respectively. Subsequent to this study, the surgeons from<br />
the five highest-volume centers performed a randomized<br />
trial <strong>of</strong> D1 compared with D2 LAD. 37 Of 267 randomly<br />
selected patients, total morbidity and mortality was 12.0%<br />
and 3.0%, respectively, in the D1 group and 17.9% and 2.2%,<br />
respectively, in the D2 group. Survival results are pending.<br />
The experience <strong>of</strong> the Italian GC Study Group clearly<br />
demonstrates that following a period <strong>of</strong> fairly rigorous training,<br />
Western surgeons can perform D2 LADs on Western<br />
patients with morbidity and mortality results similar to that<br />
<strong>of</strong> high-volume centers in Korea and Japan.<br />
Several tertiary referral centers in Western countries<br />
routinely perform D2 LADs for GC, 15 but as noted earlier,<br />
LADs for GC in Western countries are limited and <strong>of</strong>ten do<br />
not even reach the D1 LAD threshold. There are several<br />
reasons why more extensive LADs are not more commonly<br />
performed. First and foremost is the lack <strong>of</strong> a proven benefit<br />
in OS <strong>of</strong> D2 over D1 LAD based on the Dutch and U.K. trials.<br />
Unfortunately, many Western surgeons have interpreted<br />
the results <strong>of</strong> these trials to mean any LAD does not improve<br />
OS. Certainly some patients with node-positive disease are<br />
cured by surgical resection alone, and these patients would<br />
undoubtedly not have been cured if diseased nodes were left<br />
undissected without additional therapy. Another important<br />
obstacle to the performance <strong>of</strong> more extensive LADs is the<br />
relative paucity <strong>of</strong> gastric adenocarcinomas seen at any<br />
given institution. In order for more extensive LADs to<br />
benefit GC patients, they must be performed without excess<br />
morbidity and mortality, and this can only be achieved with<br />
adequate surgical training and adequate case volume. Contributing<br />
to the lack <strong>of</strong> high-volume centers for GC surgery<br />
is a potential reluctance <strong>of</strong> general surgeons to refer GC<br />
patients to tertiary referral centers given that gastric surgery<br />
has been historically the realm <strong>of</strong> the general surgeon.<br />
38 Finally, there are geographical and language<br />
barriers between different countries that make dissemination<br />
<strong>of</strong> information and techniques on the surgical treatment<br />
<strong>of</strong> GC difficult.<br />
253
Conclusion<br />
There are clear differences in the extent <strong>of</strong> LAD performed<br />
between Eastern and Western countries. D2 LAD is the<br />
standard LAD performed in Japan and South Korea for all<br />
resectable tumors except for T1 tumors. Significantly less<br />
extensive LADs are generally performed in the United<br />
States. How does this difference in extent <strong>of</strong> LAD affect GC<br />
patient outcomes? There is no doubt that less extensive<br />
LADs result in understaging <strong>of</strong> patients. 17 Many U.S. medical<br />
oncologists and radiation oncologists already factor this<br />
understaging into their medical decision making. For any<br />
patient in the SEER database who is stage IB or greater, the<br />
5-year OS rate is 58% or less and adjuvant therapy is likely<br />
warranted. However, a patient treated at MSKCC who is<br />
stage IB has an 85% 5-year OS and may not warrant<br />
adjuvant therapy. Our analysis <strong>of</strong> the SEER database found<br />
wide variations in survival <strong>of</strong> each AJCC stage based on<br />
subgroup analysis, with more than half <strong>of</strong> patients being<br />
Authors’ Disclosures <strong>of</strong> Potential Conflicts <strong>of</strong> Interest<br />
Author<br />
Benjamin Schmidt*<br />
Sam S. Yoon*<br />
*No relevant relationships to disclose.<br />
Employment or<br />
Leadership<br />
Positions<br />
Consultant or<br />
Advisory Role<br />
1. Jemal A, Bray F, Center MM, et al. Global cancer statistics. CA Cancer<br />
J Clin. 2011;61:69-90.<br />
2. Lee J, Demissie K, Lu SE, et al. Cancer incidence among Korean-<br />
<strong>American</strong> immigrants in the United States and native Koreans in South<br />
Korea. Cancer Control. 2007;14:78-85.<br />
3. Siegel R, Naishadham D, Jemal A. Cancer statistics, <strong>2012</strong>. CA Cancer<br />
J Clin. <strong>2012</strong>;62:10-29.<br />
4. Japanese Research <strong>Society</strong> for Gastric Cancer. The general rules for the<br />
gastric cancer study in surgery. Jpn J Surg. 1973;3:61.<br />
5. Japanese Gastric Cancer Association. Japanese gastric cancer treatment<br />
guidelines 2010 (ver. 3). Gastric Cancer. 2011;14:113-123.<br />
6. Park DJ, Lee HJ, Kim HH, et al. Predictors <strong>of</strong> operative morbidity and<br />
mortality in gastric cancer surgery. Br J Surg. 2005;92:1099-1102.<br />
7. Birkmeyer JD, Siewers AE, Finlayson EV, et al. Hospital volume and<br />
surgical mortality in the United States. N Engl J Med. 2002;346:1128-1137.<br />
8. Smith DL, Elting LS, Learn PA, et al. Factors influencing the volumeoutcome<br />
relationship in gastrectomies: a population-based study. Ann Surg<br />
Oncol. 2007;14:1846-1852.<br />
9. Macdonald JS, Smalley SR, Benedetti J, et al. Chemoradiotherapy after<br />
surgery compared with surgery alone for adenocarcinoma <strong>of</strong> the stomach or<br />
gastroesophageal junction. N Engl J Med. 2001;345:725-730.<br />
10. Smith JK, McPhee JT, Hill JS, et al. National outcomes after gastric<br />
resection for neoplasm. Arch Surg. 2007;142:387-393.<br />
11. Marcus SG, Cohen D, Lin K, et al. Complications <strong>of</strong> gastrectomy<br />
following CPT-11-based neoadjuvant chemotherapy for gastric cancer. J<br />
Gastrointest Surg. 2003;7:1015-1022.<br />
12. Gotoda T, Yanagisawa A, Sasako M, et al. Incidence <strong>of</strong> lymph node<br />
metastasis from early gastric cancer: estimation with a large number <strong>of</strong> cases<br />
at two large centers. Gastric Cancer. 2000;3:219-225.<br />
13. Yoon SS, Yang HK. Lymphadenectomy for gastric adenocarcinoma:<br />
should west meet east? Oncologist. 2009;14:871-882.<br />
14. Wang J, Dang P, Raut CP, et al. Comparison <strong>of</strong> a lymph node<br />
ratio-based staging system with the 7th AJCC System for Gastric Cancer:<br />
analysis <strong>of</strong> 18,043 patients from the SEER database. Ann Surg. In press.<br />
15. D’Angelica M, Gonen M, Brennan MF, et al. Patterns <strong>of</strong> initial recurrence<br />
in completely resected gastric adenocarcinoma. Ann Surg. 2004;240:<br />
808-816.<br />
16. Karpeh MS, Leon L, Klimstra D, et al. Lymph node staging in gastric<br />
cancer: is location more important than number? An analysis <strong>of</strong> 1,038<br />
patients. Ann Surg. 2000;232:362-371.<br />
17. Bunt AM, Hermans J, Smit VT, et al. Surgical/pathologic-stage migra-<br />
254<br />
misclassified. 14 Thus there are inherent problems with inaccurately<br />
staged patients.<br />
Less extensive LADs also likely result in increased locoregional<br />
recurrence, making the decisions between adjuvant<br />
chemotherapy versus chemoradiation more difficult. GC<br />
patients are <strong>of</strong>ten limited in the extent <strong>of</strong> neoadjvuant and<br />
adjuvant treatment that can be delivered, and one must<br />
<strong>of</strong>ten choose between multiagent chemotherapy regimens<br />
versus 5-fluorouracil or capecitabine-based chemoradiation.<br />
In terms <strong>of</strong> OS, the effects <strong>of</strong> more extensive LAD are<br />
difficult to discern. The Dutch and U.K. trials <strong>of</strong> D1 compared<br />
with D2 LAD demonstrated that when D2 LAD is<br />
performed with excess morbidity and mortality, there is no<br />
survival benefit compared to D1 LAD. If D2 LAD is performed<br />
with low morbidity and mortality, there also may be<br />
a benefit in OS, at least in Chinese patients, 32 but this<br />
potential benefit needs to be demonstrated by future prospective,<br />
randomized trials <strong>of</strong> Western patients.<br />
Stock<br />
Ownership Honoraria<br />
REFERENCES<br />
Research<br />
Funding<br />
Expert<br />
Testimony<br />
SCHMIDT AND YOON<br />
Other<br />
Remuneration<br />
tion confounds comparisons <strong>of</strong> gastric cancer survival rates between Japan<br />
and Western countries. J Clin Oncol. 1995;13:19-25.<br />
18. Verdecchia A, Mariotto A, Gatta G, et al. Comparison <strong>of</strong> stomach cancer<br />
incidence and survival in four continents. Eur J Cancer. 2003;39:1603-1609.<br />
19. Strong VE, Song KY, Park CH, et al. Comparison <strong>of</strong> gastric cancer<br />
survival following R0 resection in the United States and Korea using an<br />
internationally validated nomogram. Ann Surg. 2010;251:640-646.<br />
20. <strong>American</strong> Joint Committee on Cancer. AJCC Cancer Staging Manual.<br />
7th edition. New York: Springer; 2010.<br />
21. Bouvier AM, Haas O, Piard F, et al. How many nodes must be examined<br />
to accurately stage gastric carcinomas? Results from a population based<br />
study. Cancer. 2002;94:2862-2866.<br />
22. Smith DD, Schwarz RR, Schwarz RE. Impact <strong>of</strong> total lymph node count<br />
on staging and survival after gastrectomy for gastric cancer: data from a large<br />
US-population database. J Clin Oncol. 2005;23:7114-7124.<br />
23. Estes NC, Macdonald JS, Touijer K, et al. Inadequate documentation<br />
and resection for gastric cancer in the United States: a preliminary report. Am<br />
Surg. 1998;64:680-685.<br />
24. Gunderson LL, Sosin H. Adenocarcinoma <strong>of</strong> the stomach: areas <strong>of</strong><br />
failure in a re-operation series (second or symptomatic look) clinicopathologic<br />
correlation and implications for adjuvant therapy. Int J Radiat Oncol Biol<br />
Phys. 1982;8:1-11.<br />
25. Sakuramoto S, Sasako M, Yamaguchi T, et al. Adjuvant chemotherapy<br />
for gastric cancer with S-1, an oral fluoropyrimidine. N Engl J Med. 2007;<br />
357:1810-1820.<br />
26. Cuschieri A, Fayers P, Fielding J, et al. Postoperative morbidity and<br />
mortality after D1 and D2 resections for gastric cancer: preliminary results <strong>of</strong><br />
the MRC randomised controlled surgical trial. The Surgical Cooperative<br />
Group. Lancet. 1996;347:995-999.<br />
27. Bonenkamp JJ, Songun I, Hermans J, et al. Randomised comparison <strong>of</strong><br />
morbidity after D1 and D2 dissection for gastric cancer in 996 Dutch patients.<br />
Lancet. 1995;345:745-748, 1995.<br />
28. Jansen EP, Boot H, Verheij M, et al. Optimal locoregional treatment in<br />
gastric cancer. J Clin Oncol. 2005;23:4509-4517.<br />
29. Maruyama K, Sasako M, Kinoshita T, et al. Pancreas-preserving total<br />
gastrectomy for proximal gastric cancer. World J Surg. 1995;19:532-536.<br />
30. Uyama I, Ogiwara H, Takahara T, et al. Spleen- and pancreaspreserving<br />
total gastrectomy with superextended lymphadenectomy including<br />
dissection <strong>of</strong> the para-aortic lymph nodes for gastric cancer. J Surg Oncol.<br />
1996;63:268-270.
GASTRIC CANCER LYMPHADENECTOMY AND OUTCOMES<br />
31. Biffi R, Chiappa A, Luca F, et al. Extended lymph node dissection<br />
without routine spleno-pancreatectomy for treatment <strong>of</strong> gastric cancer: low<br />
morbidity and mortality rates in a single center series <strong>of</strong> 250 patients. J Surg<br />
Oncol. 2006;93:394-400.<br />
32. Wu CW, Hsiung CA, Lo SS, et al. Nodal dissection for patients with<br />
gastric cancer: a randomised controlled trial. Lancet Oncol. 2006;7:309-315.<br />
33. Roggin KK, Posner MC. D3 or not D3. That is not the question. Lancet<br />
Oncol. 2006;7:279-280.<br />
34. Songun I, Putter H, Kranenbarg EM, et al. Surgical treatment <strong>of</strong><br />
gastric cancer: 15-year follow-up results <strong>of</strong> the randomised nationwide Dutch<br />
D1D2 trial. Lancet Oncol. 2010;11:439-449.<br />
35. Degiuli M, Sasako M, Ponti A, et al. Morbidity and mortality after D2<br />
gastrectomy for gastric cancer: results <strong>of</strong> the Italian Gastric Cancer Study Group<br />
prospective multicenter surgical study. J Clin Oncol. 1998;16:1490-1493.<br />
36. Degiuli M, Sasako M, Calgaro M, et al. Morbidity and mortality after<br />
D1 and D2 gastrectomy for cancer: interim analysis <strong>of</strong> the Italian Gastric<br />
Cancer Study Group (IGCSG) randomised surgical trial. Eur J Surg Oncol.<br />
2004;30:303-308.<br />
37. Degiuli M, Sasako M, Ponti A. Morbidity and mortality in the Italian<br />
Gastric Cancer Study Group randomized clinical trial <strong>of</strong> D1 versus D2<br />
resection for gastric cancer. Br J Surg. 2010;97:643-649.<br />
38. Karpeh MS Jr. Should gastric cancer surgery be performed in community<br />
hospitals? Semin Oncol. 2005;32:S94-S96.<br />
39. Japanese Gastric Cancer Association. Japanese classification <strong>of</strong> gastric<br />
carcinoma: 3rd English edition. Gastric Cancer. 2011;14:101-112.<br />
40. Ahn HS, Lee HJ, Yoo MW, et al. Changes in clinicopathological features<br />
and survival after gastrectomy for gastric cancer over a 20-year period. Br J<br />
Surg. 2011;98:255-260.<br />
255
Will Disease Heterogeneity Help Define<br />
Treatment Paradigms for Gastroesophageal<br />
Adenocarcinoma? A Global Perspective<br />
Overview: Cancers <strong>of</strong> the upper gastrointestinal (GI) tract<br />
form a heterogeneous group <strong>of</strong> diseases for which treatment<br />
paradigms for localized disease continue to emerge. Recently,<br />
several phase III studies in esophagus and gastric cancer that<br />
have attempted to define new standards <strong>of</strong> care have been<br />
reported. However, controversy still persists and treatment<br />
algorithms <strong>of</strong>ten depend on individual preference, patient<br />
DIFFERENT TREATMENT paradigms characterize upper<br />
GI malignancies across the globe. Among the most<br />
varied is the approach to patients with locally advanced, but<br />
resectable, esophagogastric carcinoma. National Comprehensive<br />
Cancer Network (NCCN) guidelines for esophageal<br />
carcinoma in patients with localized disease who are medically<br />
fit for resection allow for several standard treatment<br />
options, including preoperative chemoradiotherapy (CRT),<br />
definitive CRT, preoperative chemotherapy (for adenocarcinoma),<br />
and resection followed by postoperative CRT. 1 For<br />
localized, resectable gastric cancer, acceptable treatment<br />
options are nearly equally varied, with the notable exception<br />
<strong>of</strong> definitive CRT. 2 Several important phase III studies were<br />
recently reported at the last two annual ASCO meetings. We<br />
will attempt to place these studies into context <strong>of</strong> the current<br />
accepted treatment paradigms on the basis <strong>of</strong> current best<br />
evidence.<br />
Results <strong>of</strong> Recent Phase III Studies: Gastric Cancer<br />
The management <strong>of</strong> localized gastric cancer has become<br />
increasingly complicated with emerging data <strong>of</strong> the survival<br />
advantage <strong>of</strong> systemic chemotherapy alone, as well as the<br />
potential benefit <strong>of</strong> CRTin specific settings. Recently, investigators<br />
presented the results <strong>of</strong> the CLASSIC study, a<br />
randomized phase III evaluation <strong>of</strong> post-operative chemotherapy<br />
with capecitabine and oxaliplatin compared with<br />
observation in resected gastric cancer. 3 In this study, 1,035<br />
patients with stage II to III gastric cancer who had a D2 (i.e.,<br />
extended lymphadenectomy) surgical dissection were randomly<br />
assigned to receive either observation alone (n � 515)<br />
or eight cycles <strong>of</strong> chemotherapy (n � 520) consisting <strong>of</strong><br />
oxaliplatin 130 mg/m 2 plus capecitabine 1,000 mg/m 2 twice<br />
daily for 14 days repeated every 3 weeks. The investigators<br />
met their prespecified primary end point <strong>of</strong> improving 3-year<br />
disease-free survival (DFS), demonstrating 74% 3-year DFS<br />
with adjuvant chemotherapy versus 60% with observation<br />
alone (hazard ratio [HR] � 0.56; Table 1). These data are<br />
consistent with the previous adjuvant S1 study reported by<br />
Sakuramoto and colleagues 4 and, bolstered by the pooled<br />
individual patient-data metaanalysis, 5 further support the<br />
use <strong>of</strong> postoperative chemotherapy alone. The applicability<br />
<strong>of</strong> the recent large phase III studies from Korea and Japan<br />
to Western patients remains a question, especially with<br />
markedly different epidemiology <strong>of</strong> increased proximal and<br />
GEJ tumors in the West compared with Asia. Even when<br />
controlling for the extent <strong>of</strong> surgical resection, apparent<br />
differences in outcomes persist. 6 Furthermore, a recent<br />
256<br />
By Manish A. Shah, MD<br />
referral patterns, and treatment biases. In the current era <strong>of</strong><br />
improving quality control and standardization <strong>of</strong> care, such<br />
variations in practice present a substantial challenge for both<br />
patients and physicians. In this article, I will highlight differences<br />
in disease biology for upper GI diseases, and in particular,<br />
gastric cancer.<br />
well-performed, although underpowered, study examined<br />
postoperative chemotherapy with cisplatin, epirubicin, fluorouracil<br />
(FU), and leucovorin (LV; PELF) in a randomized<br />
study and did not show a survival advantage in a European<br />
patient population <strong>of</strong> resected advanced gastric cancer. 7 In<br />
this study, 258 patients were randomly assigned to receive<br />
four cycles <strong>of</strong> PELF chemotherapy or observation alone, and<br />
with a median follow-up <strong>of</strong> more than 72 months, no differences<br />
in patient outcomes were observed. Specifically, at the<br />
end <strong>of</strong> the study period, 47% <strong>of</strong> the patients who received<br />
chemotherapy were still alive compared with 45.3% <strong>of</strong> the<br />
surgery-alone arm. 7 Thus, despite the recent compelling<br />
phase III data, treatment paradigms in the West for locally<br />
advanced gastric/gastroesophageal junction (GEJ) adenocarcinoma<br />
currently involve either preoperative or perioperative<br />
chemotherapy, 8,9 or postoperative CRT. 10,11 The role <strong>of</strong><br />
postoperative chemotherapy in the West remains controversial.<br />
Regarding CRT after curative-intent resection in gastric<br />
cancer, the addition <strong>of</strong> epirubicin and cisplatin importantly<br />
did not demonstrate an improvement over standard adjuvant<br />
FU/radiotherapy (RT) in the CALGB national phase III<br />
study. 11 This was a 546-patient study in which patients<br />
were randomly assigned to receive standard postoperative<br />
CRT (n � 280) consisting <strong>of</strong> bolus FU/LV (cycle 1, 3 and 4),<br />
along with infusional FU (200 mg/m 2 /day as an intravenous<br />
continuous infusion � 5 weeks, cycle 2) or experimental<br />
postoperative CRT (n � 266) consisting <strong>of</strong> epirubicin, cisplatin,<br />
and FU (ECF) chemotherapy (cycle 1, 3 and 4), and the<br />
same CRT (infusional FU � RT), to demonstrate an improvement<br />
in overall survival <strong>of</strong> 30% with the addition <strong>of</strong><br />
ECF. There was no difference in overall survival (HR �<br />
1.03), suggesting that epirubicin and cisplatin do not add<br />
significantly to FU adjuvant therapy. Investigators also<br />
noted remarkably little improvement in RT planning over<br />
the last two large United States–based phase III studies<br />
(15% major deviations in the RT plan). 11 In another study<br />
(the ARTIST Trial), investigators from Korea investigated<br />
From the Weill Cornell Medical College; New York-Presbyterian Hospital, New York, NY.<br />
Author’s disclosures <strong>of</strong> potential conflicts <strong>of</strong> interest are found at the end <strong>of</strong> this article.<br />
Address reprint requests to: Manish A. Shah, MD, Weill Cornell Medical College/New<br />
York-Presbyterian Hospital, 1305 York Avenue, 12 th Floor, New York, NY 10065; email:<br />
mas9313@med.cornell.edu.<br />
© <strong>2012</strong> by <strong>American</strong> <strong>Society</strong> <strong>of</strong> <strong>Clinical</strong> <strong>Oncology</strong>.<br />
1092-9118/10/1-10
TREATMENT PARADIGMS FOR GASTROESOPHAGEAL ADENOCARCINOMA<br />
the role <strong>of</strong> postoperative RT in a patient population that<br />
underwent a standard D2 gastric dissection. 12 In this study,<br />
458 patients were randomly assigned to receive postoperative<br />
chemotherapy alone (capecitabine/cisplatin [XP]) or<br />
CRT (XP 3 capecitabine/RT 3 XP), with the aim to identify<br />
a 45% improvement in 3-year DFS. These investigators did<br />
observe a modest improvement in the addition <strong>of</strong> RT, particularly<br />
in the large subset <strong>of</strong> patients who were node<br />
positive (3-year DFS, 72% vs. 77%; p � 0.0365; Table 1).<br />
These data support the role <strong>of</strong> adjuvant RT in this disease,<br />
as suggested initially by INT-0116, and are currently being<br />
prospectively validated in a node-positive resected cohort.<br />
Results <strong>of</strong> Recent Phase III Studies:<br />
Esophageal Cancer<br />
In esophageal carcinoma, the CROSS study was presented<br />
at the 2011 ASCO Annual Meeting, and examined the<br />
effects <strong>of</strong> preoperative CRTin advanced esophageal cancer.<br />
This study found that a combination regimen <strong>of</strong> CRT before<br />
resection is superior to surgery alone. 13 In this multicenter<br />
phase III randomized study, 364 patients in the Netherlands<br />
with resectable esophageal adenocarcinoma or squamous<br />
cell carcinoma (SCC) were randomly assigned to receive<br />
combined-modality therapy <strong>of</strong> CRTfollowed by surgery or<br />
KEY POINTS<br />
Table 1. Recent Phase III Studies in Patients with Localized Esophagogastric Carcinoma<br />
Study Year Disease Type Subgroup Comparison<br />
Esophagus cancer<br />
Van Der Gaast<br />
et al 13<br />
2010 Esophageal<br />
adenocarcinoma (74%)<br />
and SCC (23%), stage<br />
II/III<br />
Mariette et al 18 2010 Esophageal<br />
adenocarcinoma (30%)<br />
and SCC (70%), stage<br />
I/II<br />
CRT (41.4 Gy) �<br />
surgery surgery<br />
Gastric cancer<br />
Bang et al 3 2011 Gastric adenocarcinoma Post-op<br />
capecitabine/oxaliplatin<br />
vs. surgery alone<br />
Fuchs et al 11 2011 Gastric and GEJ<br />
adenocarcinoma<br />
● Significant heterogeneity in treatment paradigms for<br />
upper GI malignancies.<br />
● Several recent studies have attempted to redefine the<br />
standard <strong>of</strong> care.<br />
● Global disease heterogeneity make broad applicability<br />
somewhat questionable.<br />
● Our challenge is to recognize differences in disease<br />
biology to optimize treatment paradigms.<br />
No. <strong>of</strong><br />
Patients Survival<br />
Hazard<br />
Ratio 95% CI p value<br />
364 49 26 mo 0.67 0.49 to 0.91 0.008<br />
Adeno CA 0.82 0.58 to 1.16<br />
SCC 0.34 0.17 to 0.68<br />
CRT (45 Gy) � surgery vs.<br />
surgery alone<br />
195 31.8 vs.44.5 months* 0.92 0.63 to 1.35 0.68<br />
Post-op FU/LV � FU/RT vs.<br />
post-op ECF � FU/RT<br />
1035 74% vs.60%† 0.56 0.44 to 0.72 � 0.0001<br />
540 36.6 vs.37.8 months* 1.03 0.80 to 1.34 0.80<br />
Lee et al 12 <strong>2012</strong> Gastric adenocarcinoma Post-op XP vs. XP–RT–XP 458 74% vs.78.2%† 0.6865 0.47 to 0.995 0.047<br />
Node-positive patients 396 72% vs. 77.5%† 0.0365<br />
Abbreviations: ECF, epirubicin, cisplatin, and fluorouracil; FU, fluourouracil; GEJ, gastroesophageal junction; LV, leucovorin; post-op, postoperative; RT,<br />
radiotherapy; SCC, squamous cell carcinoma; XP, capecitabine/cisplatin.<br />
* Overall survival.<br />
† 3-year disease-free survival.<br />
surgery alone. Preoperative CRT consisted <strong>of</strong> weekly paclitaxel<br />
50 mg/m 2 and carboplatin dosed at area under the<br />
curve (AUC) 2 for 5 weeks with concurrent 41.4 Gy RT<br />
administered in 23 fractions. After CRT, patients underwent<br />
resection within 6 weeks <strong>of</strong> completion <strong>of</strong> preoperative<br />
therapy. This study suggests that most patients with T1N1<br />
or T2–3Nx esophageal carcinoma should consider preoperative<br />
CRT as a standard care option. The median survival <strong>of</strong><br />
patients who received CRT and surgery was 49 months,<br />
compared with 26 months for those who received surgery<br />
alone (HR � 0.67; p � 0.011; Table 1). With a median<br />
follow-up <strong>of</strong> 32 months, 70 patients had died in the CRT<br />
group compared with 97 in the surgery-alone group, and<br />
3-year overall survival was also superior in the CRT arm.<br />
However, although the majority <strong>of</strong> patients (74% in both<br />
arms) had adenocarcinoma, it appears that the benefit <strong>of</strong><br />
CRT was primarily derived in patients with esophageal<br />
SCC. Patients with esophageal SCC observed an HR <strong>of</strong> 0.34,<br />
representing a dramatic 66% reduction in risk <strong>of</strong> death with<br />
preoperative CRT, whereas in the subset <strong>of</strong> patients with<br />
esophageal adenocarcinoma, the HR for survival in patients<br />
receiving CRT was 0.82 (Table 1).<br />
To place these data into context with other randomized<br />
studies that predominantly included distal esophageal and<br />
GEJ carcinoma as well as a recent updated metaanalysis,<br />
9,14-17 if surgery is identified as part <strong>of</strong> the treatment<br />
plan for a patient with localized disease, applying<br />
preoperative therapy does confer a survival advantage. In<br />
addition, esophageal SCC seems to be more sensitive to<br />
CRT. There are data that support either combined preoperative<br />
CRT or preoperative chemotherapy alone for<br />
esophageal adenocarcinoma, and the data supporting the<br />
superiority <strong>of</strong> trimodal therapy (CRT � surgery) over bimodal<br />
therapy (chemotherapy � surgery) remains debatable.<br />
A recent study by Stahl and colleagues suggested an<br />
improved survival with trimodality therapy in adenocarcinoma<br />
<strong>of</strong> the esophagus/GEJ, although the study was closed<br />
257
prematurely due to poor accrual. 18 In addition, Mariette and<br />
colleagues demonstrated that for early-stage esophageal<br />
adenocarcinoma, there appears to be no added benefit <strong>of</strong><br />
CRT to surgical resection. 19<br />
Influence <strong>of</strong> Disease Heterogeneity on Treatment<br />
Gastric cancer is a heterogeneous disease and subtypes<br />
<strong>of</strong> gastric cancer exist. 20,21 More than 95% <strong>of</strong> all cancers <strong>of</strong><br />
the stomach are adenocarcinomas. A common distinguishing<br />
feature is histopathology, such as the Lauren’s classification,<br />
which distinguishes intestinal and diffuse gastric cancer<br />
subtypes. The well-differentiated intestinal type tends<br />
to expand through the stomach wall, whereas the Lauren’s<br />
diffuse-type is more commonly poorly differentiated and<br />
spreads as individual discohesive cells in an infiltrative<br />
pattern. Diffuse gastric cancer is associated with loss <strong>of</strong> the<br />
cell-surface protein E-cadherin and germ-line mutations in<br />
CDH1 are associated with the familial form <strong>of</strong> diffuse gastric<br />
cancer, hereditary diffuse gastric cancer. 22 Intestinal gastric<br />
cancers predominate in high-incidence areas (e.g., China),<br />
and this histology is responsible for much <strong>of</strong> the ethnic<br />
variation across the globe.<br />
Gastric cancers may have different outcomes depending<br />
on disease subtype. More proximal GEJ and cardia tumors<br />
tend to have a worse prognosis compared with distal pyloric,<br />
antral, and curvature cancers. 23 Data are also just now<br />
emerging on the potential influence <strong>of</strong> disease subtype on<br />
treatment outcome. For example, HER2 amplification and<br />
Author’s Disclosure <strong>of</strong> Potential Conflicts <strong>of</strong> Interest<br />
Author<br />
Manish A. Shah*<br />
*No relevant relationships to disclose.<br />
Employment or<br />
Leadership<br />
Positions<br />
Consultant or<br />
Advisory Role<br />
1. Ajani JA, Barthel JS, Bentrem D, et al. Esophageal and esophagogastric<br />
junction cancers. National Comp Canc Netw. 2011;2.2011:1-97.<br />
2. Ajani JA, Barthel JS, Bentrem D, et al. Gastric cancer. National Comp<br />
Canc Netw. 2011;2.2011:1-84.<br />
3. Bang YJ, Kim YW, Yang HK, et al. Adjuvant capecitabine and oxaliplatin<br />
for gastric cancer after D2 gastrectomy (CLASSIC): a phase 3 open-label,<br />
randomised controlled trial. Lancet. <strong>2012</strong>;379:315-321.<br />
4. Sakuramoto S, Sasako M, Yamaguchi T, et al. Adjuvant chemotherapy<br />
for gastric cancer with S-1, an oral fluoropyrimidine. N Engl J Med. Nov 1<br />
2007;357:1810-1820.<br />
5. GASTRIC (Global Advanced/Adjuvant Stomach Tumor Research International<br />
Collaboration) Group, Paoletti X, Oba K, et al. Benefit <strong>of</strong> adjuvant<br />
chemotherapy for resectable gastric cancer: a meta-analysis. JAMA. 2010;<br />
303:1729-1737.<br />
6. Strong VE, Song KY, Park CH, et al. Comparison <strong>of</strong> gastric cancer<br />
survival following R0 resection in the United States and Korea using an<br />
internationally validated nomogram. Ann Surg. 2010;251:640-646.<br />
7. Di Costanzo F, Gasperoni S, Manzione L, et al. Adjuvant chemotherapy<br />
in completely resected gastric cancer: a randomized phase III trial conducted<br />
by GOIRC. J Natl Cancer Inst. 2008;100:388-398.<br />
8. Cunningham D, Allum WH, Stenning SP, et al. Perioperative chemotherapy<br />
versus surgery alone for resectable gastroesophageal cancer. N Engl<br />
J Med. 2006;355:11-20.<br />
9. Ychou M, Boige V, Pignon JP, et al. Perioperative chemotherapy compared<br />
with surgery alone for resectable gastroesophageal adenocarcinoma: an<br />
FNCLCC and FFCD Multicenter phase III trial. J Clin Oncol. 2011;29:1715-<br />
1721.<br />
10. MacDonald JS, Smalley S, Benedetti J, et al. Chemoradiotherapy after<br />
surgery compared with surgery alone for adenocarcinoma <strong>of</strong> the stomach or<br />
gastroesophageal junction. N Engl J Med. 2001;345:725-730.<br />
258<br />
overexpression is far more prevalent in proximal/GEJ adenocarcinoma<br />
than in diffuse gastric cancer. 24,25 In an exploratory<br />
analysis, these proximal/GEJ tumors appeared to be<br />
less sensitive to bevacizumab therapy than are diffuse and<br />
distal nondiffuse gastric cancers. 26 Thus, disease biology<br />
may indeed influence patient outcomes with specific treatments.<br />
Implications for the Future<br />
Where do we go from here? Because treatment paradigms<br />
are defined on a global basis, it will be important to understand<br />
the global heterogeneity <strong>of</strong> these diseases. The epidemiology,<br />
risk factors, and patterns <strong>of</strong> care for cancers <strong>of</strong> the<br />
upper GI tract are substantially different across the globe.<br />
Disease biology in the various regions around the world may<br />
also be different, although this needs to be more thoroughly<br />
investigated. For example, proximal/GEJ adenocarcinomas<br />
seem less frequent in the far East compared with Europe<br />
and the Americas. However, it is not clear whether this<br />
difference solely contributes to differences in patient outcome<br />
between these two regions. Could it be practice patterns<br />
(screening and early disease identification), use <strong>of</strong><br />
second-line therapy, 27 or disease biology 6 ? Our challenge is<br />
to recognize the influence <strong>of</strong> disease biology and heterogeneity<br />
on treatment paradigms in this disease. If we are able to<br />
do so, we will be able to provide better treatment guidelines<br />
for specific disease subtypes, and improve patient outcomes<br />
in a more directed approach—a laudable goal, indeed!<br />
Stock<br />
Ownership Honoraria<br />
REFERENCES<br />
Research<br />
Funding<br />
Expert<br />
Testimony<br />
MANISH A. SHAH<br />
Other<br />
Remuneration<br />
11. Fuchs CS, Tepper JE, Niedzwiecki D, et al. Postoperative adjuvant<br />
chemoradiation for gastric or gastroesophageal junction (GEJ) adenocarcinoma<br />
using epirubicin, cisplatin, and infusional (CI) 5-FU (ECF) before and<br />
after CI 5-FU and radiotherapy (CRT) compared with bolus 5-FV/LV before<br />
and after CRT: Intergroup trial CALGB 80101. J Clin Oncol. 2011;29(suppl;<br />
abstr 4003).<br />
12. Lee J, Lim do H, Kim S, et al. Phase III trial comparing capecitabine<br />
plus cisplatin versus capecitabine plus cisplatin with concurrent capecitabine<br />
radiotherapy in completely resected gastric cancer with D2 lymph node<br />
dissection: the ARTIST Trial. J Clin Oncol. <strong>2012</strong>;30:268-273.<br />
13. van der Gaast A, van Hagen P, Hulsh<strong>of</strong> M, et al. Effect <strong>of</strong> preoperative<br />
concurrent chemoradiotherapy on survival <strong>of</strong> patients with resectable esophageal<br />
or esophagogastric junctino cancer: results from a multi-center randomized<br />
phase III study. J Clin Oncol. 2010;28(suppl; abstr 4004).<br />
14. Bedenne L, Michel P, Bouche O, et al. Chemoradiation followed by<br />
surgery compared with chemoradiation alone in squamous cancer <strong>of</strong> the<br />
esophagus: FFCD 9102. J Clin Oncol. 2007;25:1160-1168.<br />
15. Stahl M, Stuschke M, Lehmann N, et al. Chemoradiation with and<br />
without surgery in patients with locally advanced squamous cell carcinoma <strong>of</strong><br />
the esophagus. J Clin Oncol. 2005;23:2310-2317.<br />
16. Tepper J, Krasna MJ, Niedzwiecki D, et al. Phase III trial <strong>of</strong> trimodality<br />
therapy with cisplatin, fluorouracil, radiotherapy, and surgery compared<br />
with surgery alone for esophageal cancer: CALGB 9781. J Clin Oncol.<br />
2008;26:1086-1092.<br />
17. Sjoquist KM, Burmeister BH, Smithers BM, et al. Survival after<br />
neoadjuvant chemotherapy or chemoradiotherapy for resectable oesophageal<br />
carcinoma: an updated meta-analysis. Lancet Oncol. 2011;12:681-692.<br />
18. Stahl M, Walz MK, Stuschke M, et al. Phase III comparison <strong>of</strong><br />
preoperative chemotherapy compared with chemoradiotherapy in patients
TREATMENT PARADIGMS FOR GASTROESOPHAGEAL ADENOCARCINOMA<br />
with locally advanced adenocarcinoma <strong>of</strong> the esophagogastric junction. J Clin<br />
Oncol. 2009;27:851-856.<br />
19. Mariette C, Seitz JF, Mailliard E, et al. Surgery alone versus chemoradiotherapy<br />
followed by surgery for localized esophageal cancer: analysis <strong>of</strong> a<br />
randomized controlled phase III trial FFCD 9901. J Clin Oncol. 2010;<br />
28(suppl; abstr 4005).<br />
20. Shah MA, Kelsen DP. Gastric cancer: a primer on the epidemiology<br />
and biology <strong>of</strong> the disease and an overview <strong>of</strong> the medical management <strong>of</strong><br />
advanced disease. J Natl Compr Canc Netw. 2010;8:437-447.<br />
21. Shah MA, Khanin R, Tang LH, et al. Molecular classsification <strong>of</strong> gastric<br />
cancer: a new paradigm. Clin Cancer Res. 2011;17:2693-2701.<br />
22. Guilford P, Hopkins J, Harraway J, et al. E-cadherin germline mutations<br />
in familial gastric cancer. Nature. 1998;392:402-405.<br />
23. Catalano V, Labianca R, Beretta GD, et al. Gastric cancer. Crit Rev<br />
Oncol Hematol. 2009;19:127-164.<br />
24. Bang YJ, Van Cutsem E, Feyereislova A, et al. Trastuzumab in<br />
combination with chemotherapy versus chemotherapy alone for treatment<br />
<strong>of</strong> Her2-positive advanced gastric or gastro-oesophageal junction cancer<br />
(ToGA): a phase 3, open-label, randomised controlled trial. Lancet. 2010<br />
28;376:687-97.<br />
25. Tafe LJ, Janjigian YY, Zaidinski M, et al. Human epidermal growth<br />
factor receptor 2 testing in gastroesophageal cancer: correlation between<br />
immunohistochemistry and fluorescence in situ hybridization. Arch Pathol<br />
Lab Med. 2011;135:1460-1465.<br />
26. Shah MA, Van Cutsem E, Kang YK, et al. Survival analysis according<br />
to disease subtype in AVAGAST: first-line capecitabine and cisplatin plus<br />
bevacizumab (bev) or placebo in patients with advanced gastric cancer. J Clin<br />
Oncol. <strong>2012</strong>;30:(suppl; abstr 5).<br />
27. Park SH, Lim DH, Park S, et al. A multicenter, randomized phase III<br />
trial comparing second-line chemotherapy (SLC) plus best supportive care<br />
(BSC) with BSC alone for pretreated advanced gastric cancer (AGC). J Clin<br />
Oncol. 2011;29:(suppl; abstr 4004).<br />
259
Adjuvant Treatments for Localized Advanced<br />
Gastric Cancer: Differences among<br />
Geographic Regions<br />
By Yoon-Koo Kang, MD, PhD, and Changhoon Yoo, MD<br />
Overview: After much debate, adjuvant therapy has become<br />
the standard <strong>of</strong> care worldwide for resected localized gastric<br />
cancer. However, geographic differences exist in standard<br />
adjuvant treatments: postoperative chemoradiation in North<br />
America, perioperative chemotherapy in the United Kingdom,<br />
and postoperative chemotherapy in East Asia. Now that D2<br />
gastrectomy has been recognized as the optimal surgery for<br />
localized gastric cancer in the West as well as in Asia, the<br />
standard adjuvant treatments used in the West may need to be<br />
reconsidered. One <strong>of</strong> the most important issues in adjuvant<br />
therapy for localized gastric cancer is how to improve the<br />
clinical outcomes <strong>of</strong> current standard treatments. Recent<br />
ALTHOUGH SURGERY is the only curative treatment<br />
option for patients with localized advanced gastric<br />
cancer, many patients experience recurrence even after<br />
complete resection, leading to poor survival. 1 To improve<br />
survival outcomes, many clinical trials have evaluated adjuvant<br />
treatments over the decades; however, these studies<br />
have produced conflicting results, mainly because <strong>of</strong> modest<br />
sample size and problematic study design. Meta-analyses<br />
have consistently described a small but significant survival<br />
benefit associated with adjuvant chemotherapy when<br />
compared to surgery alone. 2,3 This finding was recently<br />
confirmed by a large patient-level meta-analysis <strong>of</strong> 17 randomized<br />
controlled trials conducted by the Global Advanced/<br />
Adjuvant Stomach Tumor Research International Collaboration<br />
(GASTRIC) group. 4 Since multiple phase III studies<br />
with large numbers <strong>of</strong> patients have demonstrated the<br />
survival benefits <strong>of</strong> adjuvant treatments in localized gastric<br />
cancer compared with surgery alone, there is now global<br />
agreement that adjuvant therapy improves outcomes in<br />
patients with stage II–IV (M0) gastric cancer undergoing<br />
curative surgical resection.<br />
Current Standard Adjuvant Treatments<br />
Even though adjuvant treatment in localized advanced<br />
gastric cancer has become the standard <strong>of</strong> care worldwide,<br />
no single regimen has been accepted as the global standard.<br />
In addition, geographical differences still exist in the treatment<br />
<strong>of</strong> resectable gastric cancer (see Table 1). The<br />
Intergroup-0116 study revealed that postoperative chemoradiation<br />
consisting <strong>of</strong> bolus 5-fluorouracil/leucovorin<br />
(5-FU/LV) and concurrent radiotherapy significantly prolonged<br />
survival compared with surgery alone. Based on the<br />
results <strong>of</strong> this trial, adjuvant chemoradiotherapy has been<br />
adopted as the standard adjuvant treatment for curatively<br />
resected gastric cancer in North America. 5 Perioperative<br />
chemotherapy is currently the standard practice across<br />
Europe for patients with resectable gastric cancer. This<br />
treatment is based on the results <strong>of</strong> the Medical Research<br />
Council Adjuvant Gastric Infusional Chemotherapy (MAGIC)<br />
trial 6 in the United Kingdom. In this study, the perioperative<br />
chemotherapy arm consisting <strong>of</strong> three preoperative<br />
and three postoperative cycles <strong>of</strong> epirubicin, cisplatin, and<br />
5-fluorouracil (ECF) demonstrated longer survival than the<br />
Cancer and Leukemia Group B (CALGB) and AMC studies<br />
suggest that simply intensifying chemotherapy by adding<br />
more agents or prolonging treatment duration is insufficient.<br />
However, new strategies like early initiation <strong>of</strong> chemotherapy<br />
and/or intraperitoneal chemotherapy may further improve the<br />
current standard adjuvant therapy. In the era <strong>of</strong> targeted<br />
therapy, the role <strong>of</strong> biologic agents for gastric cancer should<br />
also be explored in the adjuvant setting. With a deeper<br />
understanding <strong>of</strong> the molecular biology <strong>of</strong> gastric cancer,<br />
adjuvant therapy for patients with localized gastric cancer can<br />
be optimized and individualized.<br />
surgery alone arm. In Asia, postoperative chemotherapy<br />
with fluoropyrimidine-based regimens has been adopted as<br />
standard adjuvant therapy, based on the results <strong>of</strong> the<br />
Adjuvant Chemotherapy Trial <strong>of</strong> TS-1 for Gastric Cancer<br />
(ACTS-GC) 7 and the recent CLASSIC trials 8 conducted in<br />
Japan and Korea, respectively. These trials demonstrated<br />
improved survival in patients with resected gastric cancer<br />
given S-1 for 1 year or the combination <strong>of</strong> capecitabine and<br />
oxaliplatin for 6 months, compared with surgery alone.<br />
Why Do Standard Treatments Differ among<br />
Geographic Regions?<br />
Geographic differences in the standard adjuvant therapy<br />
for resectable gastric cancer can be explained primarily by<br />
differences in the standard surgery. Extended (D2) lymph<br />
node dissection is well established as the standard <strong>of</strong> care<br />
in East Asia, whereas Western surgeons have been skeptical<br />
<strong>of</strong> the benefit <strong>of</strong> D2 resection over D1 surgery because <strong>of</strong><br />
conflicting results reported by previous randomized trials.<br />
However, the 15-year follow-up results <strong>of</strong> a Dutch D1D2<br />
trial show that D2 surgery is associated with lower rate <strong>of</strong><br />
disease-related death than D1 surgery in Western patients. 9<br />
This finding suggests that gastric cancer has been surgically<br />
undertreated in the West, which may explain why radiation<br />
(in North America) and intensive perioperative chemotherapy<br />
(in Europe) have improved outcomes in Western<br />
countries.<br />
Another potential reason for regional differences is the<br />
heterogeneity <strong>of</strong> study populations in previous clinical trials<br />
for gastric cancer. The ACTS-GC and CLASSIC trials, which<br />
were conducted in East Asia, included only patients with<br />
gastric cancer. However, Western trials, especially in the<br />
United Kingdom, developed therapeutic strategies for localized<br />
gastric cancer by including considerable numbers <strong>of</strong><br />
From the Department <strong>of</strong> <strong>Oncology</strong>, Asan Medical Center, University <strong>of</strong> Ulsan College <strong>of</strong><br />
Medicine, Seoul, Korea.<br />
Authors’ disclosures <strong>of</strong> potential conflicts <strong>of</strong> interest are found at the end <strong>of</strong> this article.<br />
Address reprint requests to Yoon-Koo Kang, MD, PhD, Department <strong>of</strong> <strong>Oncology</strong>, Asan<br />
Medical Center, University <strong>of</strong> Ulsan College <strong>of</strong> Medicine, 88, Olympic-ro 43-gil, Songpa-gu,<br />
Seoul, South Korea 138-736;email: ykkang@amc.seoul.kr.<br />
© <strong>2012</strong> by <strong>American</strong> <strong>Society</strong> <strong>of</strong> <strong>Clinical</strong> <strong>Oncology</strong>.<br />
1092-9118/10/1-10<br />
e31
patients with adenocarcinoma <strong>of</strong> the esophagogastric junction<br />
or lower esophagus, primarily because <strong>of</strong> the increasing<br />
incidence <strong>of</strong> these cancers and decreasing incidence <strong>of</strong> gastric<br />
cancer. However, unlike gastric cancer, esophageal cancer<br />
tends to easily invade surrounding tissue and regional<br />
lymph nodes because <strong>of</strong> the lack <strong>of</strong> serosa and abundance<br />
<strong>of</strong> lymphatics in the esophagus. Thus, long-term survival<br />
rarely exceeds 20% even after successful resection in advanced<br />
disease. For this reason, multimodality therapy<br />
including chemotherapy, radiotherapy, and surgery has<br />
been widely investigated for treatment <strong>of</strong> esophageal cancer,<br />
and Western trials also used this strategy for treatment <strong>of</strong><br />
gastric cancer. Given the significant differences between<br />
esophageal cancer and gastric cancer in terms <strong>of</strong> etiology,<br />
biology, and clinical characteristics, including patients with<br />
adenocarcinoma <strong>of</strong> the esophagogastric junction or lower<br />
esophagus in clinical trials for gastric cancer does not seem<br />
appropriate.<br />
The standard adjuvant therapies currently used in the<br />
KEY POINTS<br />
Table 1. Major Pivotal Phase III Trials for Adjuvant Treatments <strong>of</strong> Gastric Cancer<br />
Study Name Treatment Arms<br />
● After much debate over the past decades, adjuvant<br />
therapy has become the standard <strong>of</strong> care worldwide<br />
for resected localized gastric cancer.<br />
● However, standard adjuvant treatments vary among<br />
geographic regions: postoperative chemoradiation in<br />
North America, perioperative chemotherapy in the<br />
United Kingdom, and postoperative chemotherapy in<br />
East Asia.<br />
● Standard adjuvant treatments in the West may need<br />
to be reconsidered as D2 gastrectomy is now the<br />
standard surgical treatment for localized gastric cancer<br />
in the West, as well as in the East.<br />
● Recent Cancer and Leukemia Group B and AMC<br />
studies suggest that simply intensifying chemotherapy<br />
by adding more agents or prolonging treatment<br />
duration is insufficient. However, new strategies like<br />
early initiation <strong>of</strong> chemotherapy and/or intraperitoneal<br />
chemotherapy may further improve the current<br />
standard adjuvant therapy.<br />
● The role <strong>of</strong> targeted agents proven effective in metastatic<br />
or recurrent disease should be explored in the<br />
adjuvant setting.<br />
Total<br />
Patients<br />
Patients with<br />
EGJ or Lower<br />
Esophageal Cancer<br />
Patients Who<br />
Underwent<br />
D2 Surgery<br />
Hazard Ratio for<br />
OS (95% CI), p<br />
Intergroup-0116 (United States) Surgery alone versus postoperative chemoradiation 556 20% 10% 1.35 (1.09–1.66)*<br />
p � .005<br />
MAGIC (United Kingdom) Surgery alone versus perioperative chemotherapy 503 26% 38% 0.75 (0.60–0.93)<br />
p � .009<br />
ACTS-GC (Japan) Surgery alone versus postoperative S-1 1059 0% 100% 0.68 (0.52–0.87)<br />
p � .003<br />
CLASSIC (East Asia) Surgery alone versus postoperative capecitabine and oxaliplatin 1035 0% 100% 0.72 (0.52–1.00)<br />
p � .0493<br />
Abbreviations: EGJ, esophagogastric junction; OS, overall survival; CI, confidence interval.<br />
* Hazard ratio <strong>of</strong> surgery only group.<br />
Overall survival data are not yet mature; a primary endpoint <strong>of</strong> this study was disease-free survival.<br />
e32<br />
KANG AND YOO<br />
West were established before D2 gastrectomy became the<br />
standard surgery. Although the superiority <strong>of</strong> D2 surgery<br />
over D0/1 surgery has been consistently demonstrated, it<br />
will take considerable time and effort before D2 surgery is<br />
widely performed in the West, because <strong>of</strong> the lack <strong>of</strong> expertise<br />
in this procedure and insufficient number <strong>of</strong> patients to<br />
implement new surgical technique. Nevertheless, we argue<br />
that the standard adjuvant treatments used in the West<br />
need to be reconsidered, because the D2 gastrectomy has<br />
finally become the standard surgery for localized gastric<br />
cancer in the West as well as in the East. Adjuvant therapy<br />
should be determined according to whether the patient<br />
undergoes optimal surgery (D2) or suboptimal surgery (D0/<br />
1), not according to where the patient is treated. Despite the<br />
success <strong>of</strong> the Intergroup-0116 trial, the role <strong>of</strong> radiotherapy<br />
was seldom investigated in countries where D2 gastrectomy<br />
is the standard <strong>of</strong> surgery, because many investigators are<br />
reluctant to add another local therapy to an “optimal”<br />
surgery (extended lymph node dissection). Based on the<br />
positive results <strong>of</strong> a retrospective study, the ARTIST trial, 10<br />
which evaluated adjuvant chemotherapy with or without<br />
radiation, was conducted in Korea. In contrast to the<br />
Intergroup-0116 study, the control arm in the ARTIST trial<br />
underwent chemotherapy rather than observation, and D2<br />
surgery was mandatory. However, this study failed to show<br />
that adding radiation to adjuvant chemotherapy improved<br />
outcomes for patients who underwent D2 gastrectomy. Although<br />
the inclusion <strong>of</strong> too many patients with early-stage<br />
cancer (approximately 60% in stage IB or II) may have<br />
limited the power <strong>of</strong> the study, the negative results in the<br />
ARTIST trial suggest that radiation does not improve the<br />
efficacy <strong>of</strong> adjuvant chemotherapy following optimal surgery.<br />
The results <strong>of</strong> the latest trials strongly suggest that<br />
all patients with localized gastric cancer should undergo<br />
D2 surgery, if technically feasible, and subsequently undergo<br />
adjuvant chemotherapy, regardless <strong>of</strong> ethnicity or<br />
geographic location.<br />
Future Perspective: How to Improve the Current<br />
Standard Adjuvant Treatment?<br />
The most important issue in adjuvant therapy for localized<br />
gastric cancer is how to improve the clinical outcomes <strong>of</strong><br />
current standard treatments. The results <strong>of</strong> the following<br />
studies <strong>of</strong>fer timely suggestions. Cancer and Leukemia<br />
Group B (CALGB) 80101 study in North America and the<br />
AMC 0201 study in Korea failed to demonstrate that intensification<br />
<strong>of</strong> adjuvant chemotherapy improves outcomes. The<br />
CALGB 80101 trial 11 intensified the regimen used in the
ADJUVANT TREATMENT FOR GASTRIC CANCER<br />
Intergroup-0116 trial by adding two more drugs (epirubicin<br />
and cisplatin) to the bolus 5-FU/LV before and after 5-FU/<br />
radiotherapy for resected gastric or esophagogastric junction<br />
adenocarcinoma. The AMC 0201 study 12 increased the duration<br />
<strong>of</strong> oral fluoropyrimidine treatment (12 months) and<br />
added cisplatin to the combination <strong>of</strong> mitomycin-C and 3<br />
months <strong>of</strong> oral fluoropyrimidine; the control regimen was<br />
based on prolonged survival <strong>of</strong> patients with resected stage<br />
III gastric cancer compared with surgery alone in a Spanish<br />
phase III study. 13 However, both phase III trials failed to<br />
show increased survival, suggesting that simply intensifying<br />
the adjuvant chemotherapy (with or without radiation) by<br />
adding an agent or prolonging treatment duration does not<br />
always enhance its efficacy in patients with localized gastric<br />
cancer. However, the AMC 0101 study, 14 a companion trial<br />
<strong>of</strong> AMC 0201, evaluated the efficacy <strong>of</strong> two more strategies<br />
in patients with D2-resected macroscopically serosa-positive<br />
gastric cancer than the AMC 0201 with same control arm:<br />
intraoperative intraperitoneal chemotherapy using cisplatin<br />
and early initiation (day after surgery) <strong>of</strong> systemic chemotherapy.<br />
In this phase III study, which included 521 patients,<br />
these strategies significantly improved recurrencefree<br />
survival and overall survival compared with the control<br />
regimen (mitomycin-C and 3 months <strong>of</strong> oral fluoropyrimidine);<br />
this finding was verified by follow-up data (median<br />
6.6 years). 15 In light <strong>of</strong> the negative results <strong>of</strong> AMC 0201,<br />
improved survival in AMC 0101 is attributable to early<br />
initiation <strong>of</strong> systemic chemotherapy and/or intraperitoneal<br />
cisplatin. These two strategies may enhance the efficacy <strong>of</strong><br />
current standard regimens for gastric cancer such as postoperative<br />
chemoradiation and perioperative chemotherapy<br />
as well as postoperative chemotherapy. Regarding the potential<br />
benefits <strong>of</strong> early initiation <strong>of</strong> systemic chemotherapy<br />
in localized gastric cancer, neoadjuvant chemotherapy is, in<br />
a sense, the earliest possible adjuvant chemotherapy. Despite<br />
poor compliance in the postoperative phase <strong>of</strong> the<br />
MAGIC trial, the successful outcome associated with perioperative<br />
chemotherapy also suggests the potential benefits<br />
<strong>of</strong> neoadjuvant chemotherapy. However, in the MAGIC trial,<br />
perioperative chemotherapy was compared with surgery<br />
alone 6 ; therefore, it is not clear whether improved survival<br />
was because <strong>of</strong> preoperative chemotherapy, postoperative<br />
chemotherapy, or both. Furthermore, only 40% <strong>of</strong> the patients<br />
in the MAGIC trial underwent D2 surgery. Therefore,<br />
the efficacy <strong>of</strong> neoadjuvant chemotherapy in countries where<br />
D2 surgery and postoperative chemotherapy is the standard<br />
<strong>of</strong> care remains to be determined. The PRODIGY trial<br />
(preoperative docetaxel, oxaliplatin, and S-1 followed by<br />
postoperative S-1 versus postoperative S-1 for patients with<br />
D2 resection; NCT01515748) aims to answer this question.<br />
In the era <strong>of</strong> targeted therapy in oncology, biologic agents<br />
have also been investigated for adjuvant treatment <strong>of</strong> gas-<br />
Authors’ Disclosures <strong>of</strong> Potential Conflicts <strong>of</strong> Interest<br />
Author<br />
Employment or<br />
Leadership<br />
Positions<br />
Consultant or<br />
Advisory Role<br />
tric cancer. Based on the success <strong>of</strong> the MAGIC trial in the<br />
United Kingdom, MAGIC-B trial (MRC-ST03) is ongoing to<br />
determine the efficacy <strong>of</strong> adding bevacizumab to adjuvant<br />
therapy (perioperative epirubicin, capecitabine, and cisplatin<br />
with or without bevacizumab), and recently reported the<br />
feasibility <strong>of</strong> this regimen. The recent AVAGAST trial 16<br />
(conventional chemotherapy with or without bevacizumab)<br />
failed to demonstrate the benefit <strong>of</strong> bevacizumab in a global<br />
patient population with locally advanced and metastatic<br />
gastric cancer. However, it will be interesting to see whether<br />
adding bevacizumab benefits Western patients with locally<br />
advanced esophagogastric cancer. With the success <strong>of</strong> the<br />
ToGA trial, 17 the door to targeted therapy for gastric cancer<br />
has finally opened. In locally advanced or metastatic gastric<br />
cancer with overexpressed human epidermal growth factor<br />
receptor (HER)-2, trastuzumab in combination with chemotherapy<br />
significantly prolonged survival outcomes and is<br />
therefore considered the new standard <strong>of</strong> care. The efficacy<br />
<strong>of</strong> adjuvant trastuzumab in HER2-positive breast cancer led<br />
us to expect benefit from adjuvant trastuzumab in HER2positive<br />
gastric cancer, although prognostic impact <strong>of</strong> HER2<br />
expression in resectable gastric cancer was shown to be<br />
limited. 18 Multiple biologic agents are currently under investigation<br />
for use in gastric cancer, especially in metastatic<br />
or recurrent disease. If the results are promising, these<br />
agents should also be explored in the adjuvant setting. In<br />
the future, we should work to improve our understanding <strong>of</strong><br />
the molecular biology <strong>of</strong> gastric cancer in order to provide<br />
optimized and individualized therapy for patients with localized<br />
gastric cancer.<br />
Conclusion<br />
After much debate over the past decades, adjuvant therapy<br />
has become the standard <strong>of</strong> care worldwide for resected<br />
localized gastric cancer. However, geographic differences<br />
exist in standard adjuvant treatments: postoperative chemoradiation<br />
in North America, perioperative chemotherapy in<br />
the United Kingdom, and postoperative chemotherapy in<br />
East Asia. Standard adjuvant treatments in the West may<br />
need to be reconsidered as D2 gastrectomy has finally<br />
become the standard surgery for localized gastric cancer in<br />
the West, as well as in the East. Results <strong>of</strong> the recent<br />
CALGB and AMC studies suggest that simply intensifying<br />
chemotherapy by adding more cytotoxic agents or prolonging<br />
duration <strong>of</strong> treatment is insufficient. Instead, new strategies<br />
such as early initiation <strong>of</strong> chemotherapy and/or<br />
intraperitoneal chemotherapy may further improve the current<br />
standard adjuvant therapy. The role <strong>of</strong> targeted agents<br />
in adjuvant treatment for localized gastric cancer should be<br />
investigated in future based on the experiences in recurrent<br />
or metastatic disease.<br />
Stock<br />
Ownership Honoraria<br />
Yoon-Koo Kang<br />
Changhoon Yoo*<br />
Roche Taiho<br />
Pharmaceuticals;<br />
Roche<br />
*No relevant relationships to disclose.<br />
Research<br />
Funding<br />
Jeil<br />
Pharmaceutical;<br />
Roche<br />
Expert<br />
Testimony<br />
Other<br />
Remuneration<br />
e33
1. D’Angelica M, Gonen M, Brennan MF, et al. Patterns <strong>of</strong> initial recurrence<br />
in completely resected gastric adenocarcinoma. Ann Surg. 2004;240:<br />
808-816.<br />
2. Earle CC, Maroun JA. Adjuvant chemotherapy after curative resection<br />
for gastric cancer in non-Asian patients: Revisiting a meta-analysis <strong>of</strong><br />
randomised trials. Eur J Cancer. 1999;35:1059-1064.<br />
3. Nakajima T, Ota K, Ishihara S, et al. [Meta-analysis <strong>of</strong> 10 postoperative<br />
adjuvant chemotherapies for gastric cancer in CIH]. Gan To Kagaku Ryoho.<br />
1994;21:1800-1805.<br />
4. Paoletti X, Oba K, Burzykowski T, et al. Benefit <strong>of</strong> adjuvant chemotherapy<br />
for resectable gastric cancer: A meta-analysis. JAMA. 2010;303:1729-<br />
1737.<br />
5. Macdonald JS, Smalley SR, Benedetti J, et al. Chemoradiotherapy after<br />
surgery compared with surgery alone for adenocarcinoma <strong>of</strong> the stomach or<br />
gastroesophageal junction. N Engl J Med. 2001;345:725-730.<br />
6. Cunningham D, Allum WH, Stenning SP, et al. Perioperative chemotherapy<br />
versus surgery alone for resectable gastroesophageal cancer. N Engl<br />
J Med. 2006;355:11-20.<br />
7. Sakuramoto S, Sasako M, Yamaguchi T, et al. Adjuvant chemotherapy<br />
for gastric cancer with S-1, an oral fluoropyrimidine. N Engl J Med. 2007;<br />
357:1810-1820.<br />
8. Bang YJ, Kim YW, Yang HK, et al. Adjuvant capecitabine and oxaliplatin<br />
for gastric cancer after D2 gastrectomy (CLASSIC): a phase 3 open-label,<br />
randomised controlled trial. Lancet. <strong>2012</strong>;379:315-321.<br />
9. Songun I, Putter H, Kranenbarg EM, et al. Surgical treatment <strong>of</strong> gastric<br />
cancer: 15-year follow-up results <strong>of</strong> the randomised nationwide Dutch D1D2<br />
trial. Lancet Oncol. 2010;11:439-449.<br />
10. Lee J, Lim DH, Kim S, et al. Phase III Trial Comparing Capecitabine<br />
Plus Cisplatin Versus Capecitabine Plus Cisplatin With Concurrent Capecitabine<br />
Radiotherapy in Completely Resected Gastric Cancer With D2 Lymph<br />
Node Dissection: The ARTIST Trial. J Clin Oncol. <strong>2012</strong>;30:268-273.<br />
11. Fuchs CS, Tepper JE, Niedzwiecki D, et al. Postoperative adjuvant<br />
chemoradiation for gastric or gastroesophageal junction (GEJ) adenocarcinoma<br />
using epirubicin, cisplatin, and infusional (CI) 5-FU (ECF) before and<br />
after CI 5-FU and radiotherapy (CRT) compared with bolus 5-FU/LV before<br />
e34<br />
REFERENCES<br />
KANG AND YOO<br />
and after CRT: Intergroup trial CALGB 80101. J Clin Oncol. 2011;29(suppl;<br />
abstr 4003).<br />
12. Chang H-M, Kang Y-K, Min YJ, et al. A randomized phase III trial<br />
comparing mitomycin-C plus short-term doxifluridine (Mf) versus<br />
mitomycin-C plus long-term doxifluridine plus cisplatin (MFP) after curative<br />
resection <strong>of</strong> advanced gastric cancer (AMC 0201) (NCT00296335). J Clin<br />
Oncol. 2008;26(suppl; abstr 4531).<br />
13. Cirera L, Balil A, Batiste-Alentorn E, et al. Randomized clinical trial <strong>of</strong><br />
adjuvant mitomycin plus tegafur in patients with resected stage III gastric<br />
cancer. J Clin Oncol. 1999;17:3810-3815.<br />
14. Kang Y-K, Change H-M, Zang DY, et al. Postoperative adjuvant<br />
chemotherapy for grossly serosa-positive advanced gastric cancer: A randomized<br />
phase III trial <strong>of</strong> intraperitoneal cisplatin and early mitomycin-C plus<br />
long-term doxifluridine puls cisplatin (iceMFP) versus mitomycin-C plus<br />
short-term doxifluridine (Mf) (AMC0101) (NCT00296322). J Clin Oncol.<br />
2008;28(suppl; abstr LBA4511).<br />
15. Kang Y-K, Ryoo B-Y, Chang H-M, et al. Update <strong>of</strong> AMC 0101 study: A<br />
phase III trial <strong>of</strong> intraperitoneal cisplatin and early mitomycin-C plus<br />
long-term doxifluridine plus cisplatin (iceMFP) versus mitomycin-C plus<br />
short-term doxifluridine (Mf) as adjuvant chemotherapy for grossly serosapositive<br />
advanced gastric cancer (NCT00296322). J Clin Oncol. <strong>2012</strong>;30(suppl<br />
4; abstr 4).<br />
16. Ohtsu A, Shah MA, Van Cutsem E, et al. Bevacizumab in combination<br />
with chemotherapy as first-line therapy in advanced gastric cancer: A<br />
randomized, double-blind, placebo-controlled phase III study. J Clin Oncol.<br />
2011;29:3968-3976.<br />
17. Bang YJ, Van Cutsem E, Feyereislova A, et al. Trastuzumab in<br />
combination with chemotherapy versus chemotherapy alone for treatment <strong>of</strong><br />
HER2-positive advanced gastric or gastro-oesophageal junction cancer<br />
(ToGA): a phase 3, open-label, randomised controlled trial. Lancet. 2010;376:<br />
687-697.<br />
18. Kataoka Y, Okabe H, Yoshizawa A, et al. HER2 expression and its<br />
clinicopathological features in resectable gastric cancer. Gastric Cancer. Epub<br />
<strong>2012</strong> Mar 14.
LIVER-DIRECTED THERAPEUTIC OPTIONS FOR<br />
HEPATOCELLULAR CARCINOMA: PATIENT<br />
SELECTION AND CLINICAL OUTCOMES<br />
CHAIR<br />
Laura A. Dawson, MD<br />
Princess Margaret Hospital<br />
Toronto, ON, Canada<br />
SPEAKERS<br />
Kenneth K. Tanabe, MD<br />
Massachusetts General Hospital Cancer Center<br />
Boston, MA<br />
Fred T. Lee, MD<br />
University <strong>of</strong> Wisconsin<br />
Madison, WI
Stereotactic Body Radiation Therapy for<br />
Hepatocellular Carcinoma<br />
By Laura A. Dawson, MD, Sameh Hashem, MD, and Alexis Bujold, MD<br />
Overview: Stereotactic body radiotherapy (SBRT), in which<br />
highly conformal potent radiation doses are delivered in fewer<br />
fractions than traditional radiation therapy (RT), is an increasingly<br />
popular treatment for hepatocellular carcinoma (HCC).<br />
The great majority <strong>of</strong> HCCs smaller than 6 cm and with<br />
Child-Pugh A liver function are controlled with SBRT with<br />
limited toxicity. Long-term local control is reduced in larger<br />
GLOBALLY, HCC IS THE sixth most common cancer<br />
and the fourth most common cause <strong>of</strong> cancer-related<br />
death. The overall 5-year survival is poor (approximately<br />
5%), and its incidence is increasing. 1 While resection and<br />
transplant can cure HCC, only a minority <strong>of</strong> patients are<br />
suitable for surgery because <strong>of</strong> multifocal or extrahepatic<br />
cancer, inadequate liver function, and/or involvement <strong>of</strong><br />
large vessels. Radi<strong>of</strong>requency ablation and percutaneous<br />
ethanol injection are associated with excellent local control<br />
in small HCCs, but outcomes are reduced in HCCs larger<br />
than 4 cm or adjacent to large vessels.<br />
RT is an effective local therapy that has the potential to<br />
benefit patients unsuitable for and/or at high risk <strong>of</strong> complications<br />
following standard local-regional therapies. All <strong>of</strong><br />
the following have facilitated the safe delivery <strong>of</strong> tumorcidal<br />
doses to focal HCCs using conformal RT: advances in imaging,<br />
RT planning techniques (to produce three-dimensional<br />
conformal RT plans minimizing dose to surrounding tissues),<br />
image-guided radiotherapy (IGRT; to localize the<br />
tumor at time <strong>of</strong> treatment), tumor immobilization (to account<br />
for breathing-related organ motion), and improved<br />
knowledge <strong>of</strong> what volume <strong>of</strong> liver is required to be spared<br />
from radiation to preserve function. Protons and carbon<br />
ions—available at specialized centers—have the ability to<br />
spare more liver parenchyma than photons.<br />
SBRT—which is shorter radiotherapy schedules (hyp<strong>of</strong>ractionation),<br />
with higher very conformal doses delivered at<br />
each radiation fraction—has more recently been used to<br />
treat focal HCC. This has been used when the majority <strong>of</strong> the<br />
liver can be spared from irradiation. Initially stereotactic<br />
immobilization body frames were used to aid in patient<br />
SBRT positioning, but with recent advancements in IGRT,<br />
the requirement for such frames has disappeared, as the<br />
liver can be directly visualized before or during RT deliverly.<br />
SBRT is widely available (unlike protons or carbon ions) and<br />
more convenient for patients than conventionally fractionated<br />
RT, as it is delivered in far fewer fractions (typically 10<br />
or less) than standard fractionated RT. This article will<br />
focus primarily on SBRT for HCC.<br />
Liver Tolerance <strong>of</strong> RT<br />
Just as a portion <strong>of</strong> the liver may be resected with surgery<br />
or ablated with radi<strong>of</strong>requency ablation, portions <strong>of</strong> the liver<br />
can tolerate high doses <strong>of</strong> radiation, without liver toxicity.<br />
With long-term follow-up, atrophy <strong>of</strong> the irradiated portion<br />
<strong>of</strong> the liver and hypertrophy <strong>of</strong> the spared portions <strong>of</strong> the<br />
liver are commonly seen. Objective response assessment is<br />
challenging within 4 months following RT because <strong>of</strong><br />
changes in the irradiated liver volume during that time.<br />
tumors, and toxicity is increased in patients with Child-Pugh B<br />
or C liver function. SBRT is an effective treatment for tumor<br />
vascular thrombi and can lead to sustained vascular recanalization.<br />
The first site <strong>of</strong> recurrence following SBRT is most<br />
<strong>of</strong>ten within the liver, away from the high dose volume,<br />
providing rationale for combining SBRT with regional or systemic<br />
therapies. Randomized trials <strong>of</strong> SBRT are warranted.<br />
The classical described toxicity following RT for liver<br />
cancer is a clinical syndrome <strong>of</strong> anicteric hepatomegaly,<br />
ascites, and elevated liver enzymes (particularly serum<br />
alkaline phosphatase) occurring within three months after<br />
RT, referred to as classic radiation-induced liver disease<br />
(RILD). 2 For the most part, this toxicity is preventable, as<br />
long as a sufficient fraction <strong>of</strong> liver can be spared from RT.<br />
For example, the risk <strong>of</strong> classic RILD is less than 5% when<br />
the mean liver doses are kept less than 14 Gy (in 6 fractions)<br />
and 28 Gy (in 1.5 Gy fractions).<br />
Other hepatic toxicities, referred to as “nonclassic RILD”<br />
are more challenging to prevent. Such toxicities include<br />
reactivation <strong>of</strong> viral hepatitis, elevation <strong>of</strong> liver enzymes,<br />
and a general decline in liver function. The partial volume<br />
tolerance <strong>of</strong> the liver following RT is less clearly defined for<br />
nonclassic RILD, with different dose-volume tolerances and<br />
risk factors observed for different types <strong>of</strong> liver toxicity. In<br />
Taiwan, 17 <strong>of</strong> 89 patients with HCC treated with up to 66 Gy<br />
(in 1.8–3 Gy per fraction) developed liver toxicity. The risk<br />
was increased in hepatitis B carriers and in Child-Pugh B<br />
liver function. 3 Treatment <strong>of</strong> viral hepatitis B is important<br />
before initiating RT. Other studies have shown that the risk<br />
<strong>of</strong> toxicity is increased if the spared volume <strong>of</strong> liver is too<br />
small. For example, in a study <strong>of</strong> 48 patients with HCC<br />
treated with three-fraction SBRT (30–39 Gy), 11% <strong>of</strong> patients<br />
had a decline in Child-Pugh class, and this was more<br />
likely if less than 800 mL <strong>of</strong> liver could be spared from 18 Gy<br />
or more. 4<br />
Outcomes: Non-SBRT RT<br />
The first experience in RT for HCC was with hyper- or<br />
conventional fractionation, using conformal RT. The median<br />
survival <strong>of</strong> patients with locally advanced HCC treated with<br />
conventional fractionation ranges from 6 to 19 months. 5-9 A<br />
majority <strong>of</strong> series demonstrated that patients treated with<br />
higher RT doses had better local control and survival than<br />
those treated with lower doses.<br />
Protons or carbon ions produced from highly specialized<br />
treatment units not widely available have also been used to<br />
treat HCC. An advantage <strong>of</strong> protons is that they are associ-<br />
From the Department <strong>of</strong> Radiation <strong>Oncology</strong>, Princess Margaret Hospital, University <strong>of</strong><br />
Toronto, Toronto, Ontario, Canada, and the Département de Radio-oncologie Clinique-<br />
Enseignement-Recherche, Hôpital Maisonneuve-Rosemont, Montreal, Quebec, Canada.<br />
Authors’ disclosures <strong>of</strong> potential conflicts <strong>of</strong> interest are found at the end <strong>of</strong> this article.<br />
Address reprint requests to Laura Dawson, MD, 610 University Ave., Toronto, ON, M5G<br />
2M9, Canada; email: laura.dawson@rmp.uhn.on.ca.<br />
© <strong>2012</strong> by <strong>American</strong> <strong>Society</strong> <strong>of</strong> <strong>Clinical</strong> <strong>Oncology</strong>.<br />
1092-9118/10/1-10<br />
261
ated with unique dose distributions that allow rapid fall<strong>of</strong>f<br />
in dose and substantial sparing <strong>of</strong> dose to tissues adjacent to<br />
tumors. Proton or carbon RT has the best reported outcomes<br />
postradiation in patients with HCC. 10,11 In one prospective<br />
study, patients with Child-Pugh A liver disease and potentially<br />
resectable single HCCs, had a 5-year survival <strong>of</strong> 56%<br />
following proton therapy. 10 In another series, in six patients<br />
with HCC who went on to have liver transplantation 6 to<br />
18 months after proton therapy (63 Gy equivalent in 15<br />
fractions), two complete pathologic responses were observed,<br />
demonstrating pro<strong>of</strong> that high-dose RT may ablate HCC. 12<br />
Proton and photon therapy have been used to successfully<br />
treat HCC with portal vein or inferior vena cava<br />
thrombosis. 11<br />
Outcomes: SBRT<br />
More recently, SBRT has been used to treat patients with<br />
HCC, with a summary <strong>of</strong> outcomes shown in Table 1.<br />
Blomgren et al first reported on the use <strong>of</strong> SBRT on extracranial<br />
sites in 1995 and 1998. 13,14 In this series that included<br />
patients with HCC, 15 to 45 Gy was delivered over one to<br />
five fractions. The majority <strong>of</strong> patients had objective responses.<br />
Subsequently, Herfarth et al treated 60 liver tumors<br />
(which included 3 primary HCCs) in 37 patients with<br />
KEY POINTS<br />
● Stereotactic body radiation therapy (SBRT)—highly<br />
conformal, potent-dose radiation therapy delivered in<br />
fewer fractions than usual (usually � 10)—is being<br />
used more commonly in hepatocellular carcinoma<br />
(HCC).<br />
● SBRT is most effective in HCC smaller than 6 cm,<br />
with local control rates from 70% to 95% at 2 years.<br />
● SBRT can also lead to sustained control <strong>of</strong> HCCs<br />
larger than 6 cm and the recanalization <strong>of</strong> vascular<br />
tumor thrombi from HCC.<br />
● Toxicity risks are increased in patients with Child-<br />
Pugh B and C baseline liver function.<br />
● Randomized trials <strong>of</strong> SBRT for HCC are required.<br />
Table 1. Selected Outcomes from HCC SBRT Series<br />
Number <strong>of</strong> Patients Dose/Fraction<br />
Median Follow-up<br />
(Months) Tumor Size<br />
Overall<br />
Response Rate Survival<br />
Blomgren, 1998 14 9 pts HCC, 1 pt IHC 5–15 Gy/1–3 # NR NR 70% NR<br />
Choi, 2006 23 20 pts HCC 50 Gy/5–10 # 23 3.8 cm (2–6.5 cm) 80% overall OS 1 yr: 70%<br />
20% CR/60% PR OS 2 yr: 43.1%<br />
Mendez Romero, 2006 17 11 pts HCC 25 Gy/5 # NR � 7 cm 1-yr LC: 82% OS 1 yr: 75%<br />
CP A and B 30 �37.5 Gy/3 #<br />
Tse, 2008 18 31 pts HCC 36 Gy (median)/6 # 18 Median: 173 cm 3 1-yr infield LC: 65% OS 1 yr: 48%<br />
All CP A Range 24–54 Gy<br />
Louis, 2010 26 25 pts HCC 45 Gy/3 # 13 Median: 150 cm 3 86% overall OS 1 yr: 79%<br />
CP A and B OS 2 yr: 52%<br />
Kwon, 2010 24 42 pts HCC 30–39 Gy/3# 29 15.4 cm 3 (3–82 cm 3 ) Overall: 86% OS 3 yr: 59%<br />
CP A 90% 3-yr LC: 68%<br />
Facciuto, 2011 22 27 pts HCC 24–36/2–4 # 22 2.0 cm � 0.8 cm Overall: 37% OS 2 yr: 82%<br />
All CP A<br />
Andolino, 2011 20 60 pts HCC 44 Gy (median)/3 # CP A 27 Median: 3 cm Overall: 90% OS 2 yr: 67%<br />
CP A/B: 36/24 40 Gy (median)/5 # CP B<br />
Abbreviations: HCC, hepatocellular carcinoma; SBRT, stereotactic body radiotherapy; Pts, patients; IHC, immunohistochemical; #, fractions; NR, not reported; OS,<br />
overall survival; CR, complete response; PR, partial response; LC, local control; CP, Child-Pugh.<br />
262<br />
DAWSON, HASHEM, AND BUJOLD<br />
14 to 26 Gy SBRT in one fraction. The treatment was well<br />
tolerated. 15 Wulf et al then reported on 39 patients treated<br />
with three-fraction SBRT to the liver, five <strong>of</strong> whom had<br />
HCC. No more than 50% <strong>of</strong> the total functional liver tissue<br />
could receive more than 5 Gy, and no more than 30% could<br />
receive more than 7 Gy. The doses delivered ranged from 30<br />
Gy to 37.5 Gy in three fractions. No local failures were seen<br />
in two HCC patients after 15 and 48 months. Three <strong>of</strong> the<br />
HCC patients died <strong>of</strong> disease progression in the liver outside<br />
the RT field at 2, 7, and 17 months. No acute or late serious<br />
toxicity was reported in any patient. 16<br />
Mendez Romero et al treated eight patients with 11 HCCs<br />
<strong>of</strong> Child-Pugh A or B (in addition to 17 patients with liver<br />
metastases) with 25 Gy in five fractions, 30 Gy in three<br />
fractions, or 37.5 Gy in three fractions over 5 to 10 days. The<br />
maximum tumor size was 7 cm. Two <strong>of</strong> the patients with<br />
HCC who received 25 Gy in five fractions failed locally at 4<br />
and 7 months and were retreated with 24 Gy in three<br />
fractions. The local control rate <strong>of</strong> the patients with HCC<br />
was reported as 82%, and the 1- and 2-year actuarial<br />
survival rates were 75% and 40%, respectively. One patient<br />
in the HCC group with Child-Pugh B liver function developed<br />
grade 5 toxicity because <strong>of</strong> liver failure and infection. 17<br />
In Toronto, a prospective dose-escalation study <strong>of</strong> sixfraction<br />
SBRT was conducted in 31 patients with locally<br />
advanced HCC unsuitable for standard therapies. A typical<br />
plan is shown in Fig. 1. The dose per fraction was determined<br />
based on the effective volume <strong>of</strong> normal liver irradiated<br />
(Veff). When the effective liver volume irradiated was<br />
low (Veff � 25%), doses <strong>of</strong> 54 Gy (9 Gy � 6) were delivered<br />
safely to HCCs, with excellent local control. For patients<br />
requiring moderate volume liver irradiation (Veff 25% to<br />
80%), doses from 24 to 54 Gy (4 to 9 Gy � 6) were delivered<br />
safely. All patients had Child-Pugh liver function A, and the<br />
majority <strong>of</strong> patients had portal vein thrombosis. No classic<br />
RILD was seen. Five patients had a decline in their Child-<br />
Pugh class at 3 months. These patients had extensive<br />
tumors, and three had tumor progression that likely contributed<br />
to the decline in liver function. The median survival<br />
was 11.7 months (95% CI, 9.2–21.6 months). 18 One hundred<br />
and two eligible patients were included in an update <strong>of</strong> the<br />
Toronto phase I study and a subsequent phase II study,<br />
using the same dose-allocation approach. Underlying liver
SBRT FOR HEPATOCELLULAR CARCINOMA<br />
Fig. 1. Axial slice <strong>of</strong> HCC SBRT plan showing a<br />
conformal isodose distribution. The red and blue<br />
solid lines represent the gross target volume (GTV)<br />
and planning target volume (PTV) to be irradiated,<br />
to account for uncertainties, respectively. The HCC<br />
was treated with 42 Gy in six fractions (pink line),<br />
which tightly surrounds the PTV, with a rapid<br />
fall<strong>of</strong>f in dose.<br />
Abbreviations: HCC, hepatocellular carcinoma;<br />
SBRT, stereotactic body radiotherapy.<br />
disease included hepatitis B, 39%; hepatitis C, 40%; and<br />
alcohol, 25%. Prior therapies were delivered in 50% <strong>of</strong><br />
patients. Baseline Barcelona Clinic Liver Cancer stage was<br />
class C in 66%. Multiple lesions were present in 59% <strong>of</strong><br />
patients, and the median sum <strong>of</strong> liver lesion diameters was<br />
10 cm (1.8–43 cm). Tumor vascular thrombosis was present<br />
in 55% and extrahepatic disease in 14%. At one year, the<br />
local control <strong>of</strong> irradiated HCC was 79% (CI 95%, 66–87%).<br />
Median overall survival was 17.0 months (CI 95%, 10.6–21.8<br />
months). 19<br />
The Indiana University School <strong>of</strong> Medicine has also conducted<br />
prospective studies <strong>of</strong> HCC SBRT. 20,21 In a phase I<br />
study by Cardenes et al, in 17 patients with HCC with 25<br />
lesions (median diameter 3.2 cm), dose was escalated in<br />
patients with Child-Pugh A from 36 Gy in three fractions to<br />
48 Gy in three fractions, without dose-limiting toxicity. In<br />
patients with Child-Pugh B, two developed grade 3 hepatic<br />
toxicity at 42 Gy in three fractions. Subsequently, five<br />
patients with Child-Pugh B were treated with 40 Gy in five<br />
fractions, and one patient developed liver failure. 21 Overall,<br />
20% <strong>of</strong> patients experienced an increase in Child-Pugh class:<br />
seven <strong>of</strong> 36 patients with Child-Pugh A experienced progression<br />
to Child-Pugh B, and five <strong>of</strong> 24 patients with Child-<br />
Pugh B progressed to Child-Pugh C, demonstrating an<br />
increased risk <strong>of</strong> any toxicity in patients with worse Child-<br />
Pugh class at baseline. After a median follow-up <strong>of</strong> 27<br />
months, 2-year local control was 90%, progression-free survival<br />
was 48%, and overall survival was 67%.<br />
Retrospective series <strong>of</strong> SBRT have also been published.<br />
22,23 Kwon et al treated 42 patients with HCC with 30<br />
to 39 Gy in three fractions. With a median follow-up <strong>of</strong> 29<br />
months, the response rate was 86% (60% complete response<br />
and 26% partial response). 24 Seo et al treated 38 patients<br />
with HCC with 33 to 57 Gy in three to four fractions, with a<br />
61% 2-year survival and 79% local control rate. 25 Doses <strong>of</strong><br />
more than 42 Gy in three fractions were associated with<br />
improved local control. Twenty-five European patients with<br />
HCC have also been treated with SBRT (45 Gy in 3 fractions),<br />
with a 1-year local control rate <strong>of</strong> 95%. 26<br />
SBRT has been used to effectively treat HCC tumor<br />
thrombi and has been used as a bridge to liver transplant.<br />
21,27,28 In most series, following SBRT, the first site <strong>of</strong><br />
recurrence is in the liver outside the irradiated volume,<br />
providing rationale for studies combining regional or systemic<br />
therapies with SBRT.<br />
Toxicity: SBRT<br />
In addition to the potential for liver toxicity (including<br />
classic and nonclassic RILD), as previously described, SBRT<br />
is associated with the possibility for other late toxicity.<br />
Gastric, duodenum, and small and large bowel late toxicity<br />
(e.g., ulcer, fistula, bleed) are more likely to occur following<br />
SBRT potent doses than conventional radiation fractionations.<br />
Therefore, HCC tumors best suited for SBRT are<br />
located at least 1 cm away from luminal gastrointestinal<br />
structures. Proton pump inhibitors may reduce the risk <strong>of</strong><br />
luminal gastrointestinal toxicity, and strategies to move<br />
gastrointestinal structures away from the tumor itself may<br />
be beneficial to patients.<br />
A potential toxicity that is unique to SBRT is chest pain<br />
and rib fracture. This has not commonly been reported but is<br />
a possible late sequelae from therapy for peripherally located<br />
HCCs. The risk <strong>of</strong> toxicities can be reduced by ensuring<br />
that radiation “hot spots” are within the target volume<br />
and not near adjacent critical normal tissues.<br />
Another potential toxicity is biliary. For caudate lesions<br />
and HCC invading the biliary track, edema may occur<br />
following SBRT, so care is required to stent the patient<br />
before therapy and/or to premedicate patients with steroids<br />
to reduce this risk. In the long term, there is a risk <strong>of</strong> late<br />
biliary stenosis, which has been rarely reported following<br />
proton therapy and not yet reported following SBRT. Nonetheless,<br />
being cautious to reduce the dose per fraction and/or<br />
overall maximal dose is reasonable for caudate lesions to<br />
reduce the risk <strong>of</strong> this potential late toxicity.<br />
Conclusion<br />
Technical advances in radiation oncology have made it<br />
possible for SBRT to be used safely for the treatment <strong>of</strong><br />
263
HCC, with encouraging outcomes in early and locally<br />
advanced HCC. The risk <strong>of</strong> liver toxicity is increased<br />
in patients with Child-Pugh B or C cirrhosis, hepatitis B<br />
carrier status, and large tumor size relative to the<br />
liver. Improved tumor control and survival are seen in<br />
patients who can be treated with higher doses. Randomized<br />
Authors’ Disclosures <strong>of</strong> Potential Conflicts <strong>of</strong> Interest<br />
Employment or<br />
Leadership Consultant or Stock<br />
Research<br />
Author<br />
Positions Advisory Role Ownership Honoraria Funding<br />
Laura A. Dawson<br />
Sameh Hashem*<br />
Alexis Bujold*<br />
*No relevant relationships to disclose.<br />
Bayer<br />
1. Jemal A, Siegel R, Xu J, et al. Cancer statistics, 2010. CA Cancer J Clin.<br />
2010;60:277-300.<br />
2. Pan CC, Kavanagh BD, Dawson LA, et al. Radiation-associated liver<br />
injury. Int J Radiat Oncol Biol Phys. 2010;76:S94-S100.<br />
3. Cheng JC, Liu HS, Wu JK, et al. Inclusion <strong>of</strong> biological factors in<br />
parallel-architecture normal-tissue complication probability model for<br />
radiation-induced liver disease. Int J Radiat Oncol Biol Phys. 2005;62:1150-<br />
1156.<br />
4. Son SH, Choi BO, Ryu MR, et al. Stereotactic body radiotherapy for<br />
patients with unresectable primary hepatocellular carcinoma: Dosevolumetric<br />
parameters predicting the hepatic complication. Int J Radiat<br />
Oncol Biol Phys. 2010;78:1073-1080.<br />
5. Ben-Josef E, Normolle D, Ensminger WD, et al. Phase II trial <strong>of</strong><br />
high-dose conformal radiation therapy with concurrent hepatic artery floxuridine<br />
for unresectable intrahepatic malignancies. J Clin Oncol. 2005;23:<br />
8739-8747.<br />
6. Mornex F, Girard N, Beziat C, et al. Feasibility and efficacy <strong>of</strong> high-dose<br />
three-dimensional radiotherapy in cirrhotic patients with small-size hepatocellular<br />
carcinoma non-eligible for curative therapies—mature results <strong>of</strong> the<br />
French phase II RTF-1 trial. Int J Radiat Oncol Biol Phys 2006;66:1152-1158.<br />
7. Seong J, Shim SJ, Lee IJ, et al. Evaluation <strong>of</strong> the prognostic value <strong>of</strong><br />
Okuda, Cancer <strong>of</strong> the Liver Italian Program, and Japan Integrated Staging<br />
systems for hepatocellular carcinoma patients undergoing radiotherapy. Int J<br />
Radiat Oncol Biol Phys. 2007;67:1037-1042.<br />
8. Seong J, Park HC, Han KH, et al. <strong>Clinical</strong> results <strong>of</strong> 3-dimensional<br />
conformal radiotherapy combined with transarterial chemoembolization for<br />
hepatocellular carcinoma in the cirrhotic patients. Hepatol Res. 2003;27:30-<br />
35.<br />
9. McIntosh A, Hagspiel KD, Al-Osaimi AM, et al. Accelerated treatment<br />
using intensity-modulated radiation therapy plus concurrent capecitabine for<br />
unresectable hepatocellular carcinoma. Cancer. 2009;115:5117-5125.<br />
10. Fukumitsu N, Sugahara S, Nakayama H, et al. A prospective study <strong>of</strong><br />
hyp<strong>of</strong>ractionated proton beam therapy for patients with hepatocellular carcinoma.<br />
Int J Radiat Oncol Biol Phys. 2009;74:831-836.<br />
11. Mizumoto M, Tokuuye K, Sugahara S, et al. Proton beam therapy for<br />
hepatocellular carcinoma with inferior vena cava tumor thrombus: Report <strong>of</strong><br />
three cases. Jpn J Clin Oncol. 2007;37:459-462.<br />
12. Bush DA, Hillebrand DJ, Slater JM, et al. High-dose proton beam<br />
radiotherapy <strong>of</strong> hepatocellular carcinoma: Preliminary results <strong>of</strong> a phase II<br />
trial. Gastroenterology. 2004;127:S189-S193.<br />
13. Blomgren H, Lax I, Naslund I, et al. Stereotactic high dose fraction<br />
radiation therapy <strong>of</strong> extracranial tumors using an accelerator. <strong>Clinical</strong><br />
experience <strong>of</strong> the first thirty-one patients. Acta Oncol. 1995;34:861-870.<br />
14. Blomgren H, Lax I, Goranson II, et al. Radiosurgery for tumors in the<br />
body: <strong>Clinical</strong> experience using a new method. J Radiosurg. 1998;1:63-74.<br />
264<br />
REFERENCES<br />
DAWSON, HASHEM, AND BUJOLD<br />
trials are required to better understand the benefits<br />
and toxicities <strong>of</strong> SBRT. An international phase III study<br />
(RTOG 1112, in development) <strong>of</strong> sorafenib versus SBRT<br />
followed by sorafenib in locally advanced HCC should<br />
provide some insight into the benefits <strong>of</strong> SBRT in this<br />
setting.<br />
Expert<br />
Testimony<br />
Other<br />
Remuneration<br />
15. Herfarth KK, Debus J, Lohr F, et al. Stereotactic single-dose radiation<br />
therapy <strong>of</strong> liver tumors: Results <strong>of</strong> a phase I/II trial. J Clin Oncol. 2001;19:<br />
164-170.<br />
16. Wulf J, Guckenberger M, Haedinger U, et al. Stereotactic radiotherapy<br />
<strong>of</strong> primary liver cancer and hepatic metastases. Acta Oncologica. 2006;45:<br />
838-847.<br />
17. Mendez Romero A, Wunderink W, Hussain SM, et al. Stereotactic body<br />
radiation therapy for primary and metastatic liver tumors: A single institution<br />
phase I-II study. Acta Oncologica. 2006;45:831-837.<br />
18. Tse R, Hawkins M, Lockwood G, et al. Phase I study <strong>of</strong> individualized<br />
stereotactic body radiotherapy for hepatocellular carcinoma and intrahepatic<br />
cholangiocarcinoma. J Clin Oncol. 2008;26:657-664.<br />
19. Bujold A, Massey C, Kim JJ, et al. Outcomes <strong>of</strong> stereotactic body<br />
radiotherapy (SBRT) for hepatocellular carcinoma (HCC). Int J Radiat Oncol<br />
Biol Phys. 2011;81:S70-S71.<br />
20. Andolino DL, Johnson CS, Maluccio M, et al. Stereotactic body radiotherapy<br />
for primary hepatocellular carcinoma. Int J Radiat Oncol Biol Phys.<br />
2011;81:447-453.<br />
21. Cardenes HR, Price TR, Perkins SM, et al. Phase I feasibility trial <strong>of</strong><br />
stereotactic body radiation therapy for primary hepatocellular carcinoma.<br />
Clin Transl Oncol. 2010;12:218-225.<br />
22. Facciuto ME, Singh MK, Rochon C, et al. Stereotactic body radiation<br />
therapy in hepatocellular carcinoma and cirrhosis: evaluation <strong>of</strong> radiological<br />
and pathological response. J Surg Oncol. Epub 2011 Sep 29.<br />
23. Choi BO, Jang HS, Kang KM, et al. Fractionated stereotactic radiotherapy<br />
in patients with primary hepatocellular carcinoma. Jpn J Clin Oncol.<br />
2006;36:154-158.<br />
24. Kwon JH, Bae SH, Kim JY, et al. Long-term effect <strong>of</strong> stereotactic body<br />
radiation therapy for primary hepatocellular carcinoma ineligible for local<br />
ablation therapy or surgical resection. Stereotactic radiotherapy for liver<br />
cancer. BMC Cancer. 2010;10:475.<br />
25. Seo YS, Kim MS, Yoo SY, et al. Preliminary result <strong>of</strong> stereotactic body<br />
radiotherapy as a local salvage treatment for inoperable hepatocellular<br />
carcinoma. J Surg Oncol. 2010;102:209-214.<br />
26. Louis C, Dewas S, Mirabel X, et al. Stereotactic radiotherapy <strong>of</strong><br />
hepatocellular carcinoma: Preliminary results. Technol Cancer Res Treat.<br />
2010;9:479-487.<br />
27. Zeng ZC, Fan J, Tang ZY, et al. A comparison <strong>of</strong> treatment combinations<br />
with and without radiotherapy for hepatocellular carcinoma with portal<br />
vein and/or inferior vena cava tumor thrombus. Int J Radiat Oncol Biol Phys.<br />
2005;61:432-443.<br />
28. Sandroussi C, Dawson LA, Lee M, et al. Radiotherapy as a bridge to<br />
liver transplantation for hepatocellular carcinoma. Transpl Int. 2010;23:299-<br />
306.
Patient Selection, Resection, and Outcomes<br />
for Hepatocellular Carcinoma<br />
By Claudius Conrad, MD, PhD, and Kenneth K. Tanabe, MD<br />
Overview: Hepatocellular carcinoma (HCC) is an aggressive<br />
malignancy <strong>of</strong> the liver that most <strong>of</strong>ten arises in patients with<br />
cirrhosis and other chronic liver diseases. Worldwide, it is the<br />
sixth most common cancer and the third most common cause<br />
<strong>of</strong> cancer-related death. Median survival is poor, ranging from<br />
6 to 20 months. Definitive treatment options for HCC are<br />
surgical resection, ablation, or transplantation. The selection<br />
<strong>of</strong> patients for surgical resection is based on clinical findings,<br />
laboratory data, and imaging. Although a number <strong>of</strong> staging<br />
systems exist, all have their limitations. A multidisciplinary<br />
approach to patient selection for surgery that includes the<br />
input <strong>of</strong> an experienced liver surgeon assures optimal outcomes.<br />
Sound understanding <strong>of</strong> liver segmentation, modern<br />
surgical techniques, and the use <strong>of</strong> intraoperative ultrasound<br />
have led to a reported perioperative mortality rate below 3%,<br />
THE INCIDENCE <strong>of</strong> HCC is rising. HCC represents the<br />
fastest growing cause <strong>of</strong> cancer-related death in men in<br />
the United States, with reported overall survival rates <strong>of</strong><br />
20% to 60% for resected patients. 1 In contrast to other<br />
malignancies, the resectability <strong>of</strong> this tumor is not only<br />
affected by its anatomic location, the extent <strong>of</strong> disease, and<br />
the overall medical condition <strong>of</strong> the patient but also by the<br />
degree <strong>of</strong> underlying chronic cirrhosis <strong>of</strong> the liver that is<br />
present in more than 80% <strong>of</strong> patients. The etiology <strong>of</strong> the<br />
liver cirrhosis is chronic hepatitis B and C in the majority <strong>of</strong><br />
cases, although alcoholic liver disease; cryptogenic cirrhosis;<br />
and, increasingly, nonalcoholic steatohepatitis secondary to<br />
the increasing incidence <strong>of</strong> obesity are clinically relevant as<br />
well. 2 The increased incidence <strong>of</strong> HCC in the United States<br />
has been primarily attributed to the concomitant increase in<br />
hepatitis C infection. 3 However, one must keep in mind that<br />
15% to 20% <strong>of</strong> patients with HCC in the United States<br />
develop HCC without any known risk factors. 4 HCC is now<br />
identified earlier with screening <strong>of</strong> high-risk patients. 2 For<br />
patients with cirrhosis, portal hypertension, and tumors<br />
confined to the liver, orthotopic liver transplantation has<br />
been considered the most effective treatment. However,<br />
there are no prospective randomized controlled trials that<br />
directly compare liver transplantation with liver resection<br />
for HCC. Most studies compare either transplantation or<br />
resection with historical controls and do not analyze according<br />
to “intent to treat,” and a significant selection bias <strong>of</strong><br />
these studies cannot be excluded. Transplantation for HCC<br />
is limited by the donor organ shortage, which results in<br />
disease progression and death among patients with HCC<br />
awaiting a donor liver. 5 In addition to the limitations <strong>of</strong> liver<br />
transplantation, surgical resection for HCC has its challenges<br />
as well. Only 20% to 30% <strong>of</strong> all patients with HCC in<br />
the United States will be candidates for either resection or<br />
even locoregional therapies, including radi<strong>of</strong>requency and<br />
cryoablation or transarterial chemoembolization. This article<br />
will focus on patient selection for surgical resection,<br />
which is mainly determined by extent <strong>of</strong> disease and liver<br />
function, advances in operative technique, and prognostic<br />
factors that determine outcome.<br />
blood transfusion requirements <strong>of</strong> less than 10%, and 5-year<br />
survival rates <strong>of</strong> at least 50%. Advances in laparoscopic<br />
technique and technology have expanded the indications for a<br />
safe and oncologically appropriate minimally invasive resection.<br />
Deciding which treatment option to employ depends on<br />
tumor resectability and the degree <strong>of</strong> underlying liver disease,<br />
which is present in 80% to 85% <strong>of</strong> patients with HCC; however,<br />
despite these surgical advances, a high recurrence rate <strong>of</strong><br />
70% in patients with cirrhosis and a survival rate <strong>of</strong> 65% to<br />
80% in well-selected transplant patients are expected. This<br />
article will focus on the evaluation and selection <strong>of</strong> patients<br />
for surgical intervention, considerations in selecting the appropriate<br />
type <strong>of</strong> resection, and expected outcomes following<br />
liver resection.<br />
Patient Selection and Assessment <strong>of</strong> Hepatic Reserve<br />
The existence <strong>of</strong> a multitude <strong>of</strong> scoring and staging systems<br />
(e.g., Okuda, Cancer <strong>of</strong> the Liver Italian Program, and<br />
<strong>American</strong> Joint Committee on Cancer) suggests that the<br />
ideal one has yet to be identified. Thorough clinical, laboratory,<br />
and imaging assessment and careful preoperative<br />
patient selection by experienced liver surgeons are necessary<br />
for optimal surgical outcomes. Assessment <strong>of</strong> HCC<br />
resectability and extent <strong>of</strong> resection requires evaluation <strong>of</strong><br />
patient functional status and comorbidities, tumor location,<br />
and underlying liver function. Routine grayscale ultrasound<br />
has value for screening patients with cirrhosis but rarely<br />
has value in surgical planning. Multidetector computed<br />
tomography (CT) with three-dimensional reconstruction and<br />
magnetic resonance imaging (MRI) are superior for surgical<br />
planning, and they provide information on the morphologic<br />
characteristics <strong>of</strong> the tumor, presence <strong>of</strong> intrahepatic metastasis<br />
and secondary lesions, extent <strong>of</strong> chronic liver disease,<br />
and vascular involvement. Special focus should be given to<br />
the number and location <strong>of</strong> suspicious lesions and any<br />
suspicious regional nodes, as well as any signs <strong>of</strong> advanced<br />
liver disease (ascites, nodular hepatic contour, enlarged<br />
caudate lobe, splenomegaly, gastrosplenic and umbilical<br />
venous collateral vessels, and fatty hepatic infiltration). 1 In<br />
addition, the anatomic relationship <strong>of</strong> the tumor to important<br />
vascular structures as well as the presence <strong>of</strong> a portal or<br />
hepatic vein thrombus are important findings on CT and<br />
MRI. A vascular thrombus that is bulging in appearance and<br />
enhances with contrast is suspicious for tumor thrombus,<br />
whereas nonenhancing thrombus may represent venous clot<br />
rather than tumor. Macrovascular involvement portends a<br />
poor prognosis, unlikely to be improved with surgical resection,<br />
and represents a contraindication to liver transplanta-<br />
From the Massachusetts General Hospital, Department <strong>of</strong> Surgery, Division <strong>of</strong> Surgical<br />
<strong>Oncology</strong>, Harvard Medical School, Boston, MA.<br />
Authors’ disclosures <strong>of</strong> potential conflicts <strong>of</strong> interest are found at the end <strong>of</strong> this article.<br />
Address reprint requests to Kenneth K. Tanabe, MD, Massachusetts General Hospital, 55<br />
Fruit St., Yawkey 7.924, Boston, MA 02114; email: ktanabe@partners.org.<br />
© <strong>2012</strong> by <strong>American</strong> <strong>Society</strong> <strong>of</strong> <strong>Clinical</strong> <strong>Oncology</strong>.<br />
1092-9118/10/1-10<br />
265
tion. Advanced computer-based image reconstruction<br />
provides a virtual three-dimensional model for accurate<br />
quantitative assessment <strong>of</strong> areas at risk for devascularization<br />
or venous congestion, thus influencing the extent <strong>of</strong><br />
resection or need for vascular reconstruction. Liver volumetry<br />
using axial images from two-dimensional CT scans is<br />
used to build a virtual model <strong>of</strong> the liver to measure the<br />
future liver remnant (FLR) as a predictor <strong>of</strong> postoperative<br />
hepatic dysfunction. 6 The following formula for the estimation<br />
<strong>of</strong> the total liver volume (TLV) in adults was found to be<br />
the most precise 7 :<br />
TLV (cm 3 ) ��794.41 � 1267.28 � body-surface area (m 2 )<br />
Preoperative portal vein embolization <strong>of</strong> the branches<br />
supplying the portion <strong>of</strong> the liver to be resected may be used<br />
to induce hypertrophy to increase the FLR and reduce the<br />
risk <strong>of</strong> morbidity and mortality. This is particularly helpful<br />
if the future remnant volume is below 40% in a noncirrhotic<br />
liver or below 60% in a cirrhotic liver. 8 This technique allows<br />
liver regeneration (and therefore an increase in FLR) to take<br />
place before definitive major liver resection. Failure <strong>of</strong> the<br />
liver to undergo hypertrophy and regenerative hyperplasia<br />
in response to portal vein embolization should be a warning<br />
sign <strong>of</strong> a decreased ability <strong>of</strong> the liver to regenerate and an<br />
increased risk <strong>of</strong> liver failure after liver resection. Portal<br />
vein ligation (rather than percutaneous embolization) can<br />
KEY POINTS<br />
● A multidisciplinary approach to patient selection for<br />
surgery that includes the input <strong>of</strong> an experienced<br />
liver surgeon is necessary for optimal surgical outcomes,<br />
which are a perioperative mortality rate below<br />
3%, blood transfusion requirements in less than 10%<br />
<strong>of</strong> cases, and 5-year survival rates <strong>of</strong> 50%.<br />
● Underlying liver disease is present in more than 80%<br />
<strong>of</strong> patients with hepatocellular carcinoma. Thorough<br />
preoperative clinical, laboratory, and imaging assessment<br />
is necessary to optimize patient selection and<br />
avoid small-for-size future liver remnant leading to<br />
liver failure.<br />
● Intraoperative ultrasound with full liver mobilization<br />
is an essential component <strong>of</strong> every liver cancer operation.<br />
Anterior approach, hanging maneuver, and<br />
diverse parenchymal transaction devices have improved<br />
surgical outcome.<br />
● Laparoscopic resection is a viable alternative to open<br />
resection with an improved perioperative period and<br />
similar oncologic outcomes. A laparoscopic approach<br />
may decrease morbidity <strong>of</strong> salvage liver transplantation.<br />
● Risk factors associated with early recurrence are<br />
tumor size, microvascular invasion, satellite nodules,<br />
alpha-fetoprotein levels, and nonanatomical resection.<br />
Risk factors associated with late recurrence<br />
include presence <strong>of</strong> cirrhosis, active hepatitis, vascular<br />
invasion, moderate or poor differentiation, and<br />
multinodularity.<br />
266<br />
Table 1. Overall and Disease-Free Survival Rates after<br />
Resection According to Prognostic Factors a<br />
Prognosticator Factor 5-yr OS (%) 5-yr DFS (%)<br />
Cirrhosis HCC and cirrhosis 23–48 22–36<br />
HCC, no cirrhosis 44–58 24–45<br />
Tumor Size HCC � 3 cm 55–78 30–51<br />
HCC � 5 cm 41–67 21–44<br />
HCC � 5 cm 29–56 22–23<br />
Number <strong>of</strong> lesions Single 35–68 19–46<br />
Multiple 21–58 6–25<br />
Abbreviations: DFS, disease-free survival; HCC, hepatocellular carcinoma;<br />
OS,overall survival.<br />
a Adapted from Rahbari and colleagues. 30 The three most important prognosticators<br />
for long-term survival after liver resection for HCC are the absence or<br />
degree <strong>of</strong> liver cirrhosis; size <strong>of</strong> the lesion below 3 cm, below 5 cm, or above 5<br />
cm; and the presence <strong>of</strong> single or multiple lesions.<br />
also be performed laparoscopically in a safe manner and<br />
may be particularly useful in patients undergoing staging<br />
laparoscopy to rule out disseminated disease. 9 Before any<br />
regional therapy including resection or transplantation,<br />
chest imaging should be performed to rule out distant<br />
metastases.<br />
In addition to imaging and clinical evaluation, laboratory<br />
studies (total bilirubin, albumin, International normalized<br />
ratio) provide information required to determine Child-Pugh<br />
classification. Patients with class A cirrhosis may be good<br />
surgical candidates, whereas almost no patients with class B<br />
cirrhosis are appropriate for resection. Thrombocytopenia,<br />
especially when combined with splenomegaly, is generally a<br />
sign <strong>of</strong> portal hypertension and excludes a patient from<br />
consideration <strong>of</strong> resection (though liver transplantation remains<br />
a potential option). Preoperative percutaneous biopsy<br />
to confirm a diagnosis <strong>of</strong> HCC is not typically required for<br />
lesions that meet radiographic criteria for HCC in the<br />
setting <strong>of</strong> underlying liver disease.<br />
Advances in Operative Techniques<br />
CONRAD AND TANABE<br />
It is well-accepted that HCC resection should be performed<br />
with at least 1-cm margins and sound oncologic<br />
principles (avoidance <strong>of</strong> tumor spillage). Controversy exists<br />
as to whether there is a survival benefit <strong>of</strong> anatomic resections<br />
according to the Couinaud classification <strong>of</strong> liver segmentation<br />
over nonanatomic resection. A series from an<br />
Asian center demonstrated significantly improved survival<br />
<strong>of</strong> the group that underwent resection according to<br />
Couinaud classification over nonanatomic resection: 66%<br />
compared with 35%, p � 0.01 for 5-year survival, and 34%<br />
compared with 16%, p � 0.006 for disease-free survival. 10 A<br />
comparison <strong>of</strong> the outcome with anatomic and nonanatomic<br />
resection for HCC, using a nationwide Japanese database <strong>of</strong><br />
72,744 patients, demonstrated an improved disease-free<br />
survival rate with anatomic resection but no difference in<br />
the overall survival rate. When survival was stratified by<br />
tumor size, the disease-free survival rate was significantly<br />
improved with an anatomic resection for HCC with a diameter<br />
<strong>of</strong> 2 to 5 cm (p � 0.0005). 11 This finding was not<br />
confirmed by Western studies, where only tumor size and<br />
presence <strong>of</strong> vascular invasion affected survival. 12<br />
Full liver mobilization and ultrasound examination allow<br />
full ultrasonic inspection <strong>of</strong> the liver and delineation <strong>of</strong> the<br />
intrahepatic anatomy and may spare the patient from undergoing<br />
a potentially noncurative operation or one in which<br />
a tumor is unknowingly left behind. Depending on the
SURGICAL MANAGEMENT OF HCC<br />
quality <strong>of</strong> preoperative imaging, new lesions are detected in<br />
up to 30% <strong>of</strong> cases by intraoperative ultrasound, <strong>of</strong> which<br />
one-third will turn out to be malignant, underscoring the<br />
importance <strong>of</strong> this diagnostic tool. 13 Historically, resection <strong>of</strong><br />
the right liver has involved mobilization <strong>of</strong> the right liver<br />
with extreme left displacement to expose the retrohepatic<br />
inferior vena cava followed by extrahepatic control <strong>of</strong> the<br />
right hepatic vein. Potential deleterious consequences <strong>of</strong> this<br />
approach are iatrogenic rupture <strong>of</strong> the tumor and hemodynamic<br />
compromise <strong>of</strong> the left lobe. The so-called anterior<br />
approach entails initial vascular inflow control and parenchymal<br />
transection before mobilization <strong>of</strong> the right liver.<br />
This technique minimizes manipulation <strong>of</strong> the tumorbearing<br />
liver and spillage <strong>of</strong> tumor cells. Results from a<br />
prospective randomized controlled study <strong>of</strong> anterior compared<br />
with a conventional approach for patients with HCC<br />
measuring 5 cm or more demonstrated significantly less<br />
major operative blood loss <strong>of</strong>2Lormore (28.3% vs. 8.3%, p �<br />
0.005) and superior overall survival (median 68.1 months vs.<br />
22.6 months, p � 0.006) with the anterior approach. 14<br />
The liver-hanging maneuver originally described by Belighiti<br />
and colleagues 15 consists <strong>of</strong> passing a tape in the<br />
avascular retrohepatic, precaval space to suspend the liver<br />
during parenchymal transaction. Elevation <strong>of</strong> the tape compresses<br />
the liver and reduces bleeding <strong>of</strong> the deeper parenchymal<br />
plane while simultaneously guiding the plane <strong>of</strong> the<br />
parenchymal transection. The hanging maneuver approach,<br />
which greatly facilitates the anterior approach described<br />
above, has been modified based on the three Glisson’s<br />
pedicles and hepatic veins facilitating right and left anatomic<br />
liver resections. 16<br />
The best technique <strong>of</strong> parenchymal transection remains a<br />
matter <strong>of</strong> debate. A Cochrane Collaborative meta-analysis <strong>of</strong><br />
seven trials evaluated 556 randomly selected patients who<br />
had undergone liver resection using the most common liver<br />
parenchymal transection devices available today. 17 The<br />
comparisons include cavitron ultrasound surgical aspirator<br />
(CUSA) compared with the clamp-crush technique (two<br />
trials); radi<strong>of</strong>requency dissecting sealer (RFDS) compared<br />
with the clamp-crush technique (two trials); sharp dissection<br />
compared with the clamp-crush technique (one trial); and<br />
hydrojet compared with CUSA (one trial). 15 The clampcrush<br />
technique is a liver parenchymal transsection technique<br />
that involves crushing the liver parenchyma with a<br />
clamp, which leaves blood vessels and bile ducts behind.<br />
Those structures can subsequently be ligated or clipped. One<br />
trial compared CUSA, RFDS, hydrojet, and the clamp-crush<br />
technique. The report found that the clamp-crush technique<br />
was 2 to 6 times less expensive than the other methods<br />
depending on the number <strong>of</strong> surgeries performed each<br />
year and therefore was favored by the authors. An additional<br />
technique that is now commonly used for parenchymal<br />
liver transection is stapling with a vascular stapler.<br />
Currently, there is an open prospective trial comparing<br />
the clamp-crush technique to vascular stapler hepatectomy<br />
for parenchymal transection in elective hepatic resection<br />
(CRUNSH, NCT01049607). Weber and colleagues reported<br />
on liver tumor resections using radi<strong>of</strong>requency energy in 15<br />
patients between January 2000 and June 2001. 18 The device<br />
creates zones <strong>of</strong> necrosis that are subsequently transected<br />
with a scalpel, thereby rendering this approach suitable for<br />
laparoscopic liver resections. 19<br />
A growing body <strong>of</strong> literature has confirmed the safety and<br />
good long-term outcome <strong>of</strong> laparoscopic liver resection. The<br />
largest series <strong>of</strong> laparoscopic liver resection for HCC was a<br />
recent multicenter European trial in which 163 laparoscopic<br />
liver resections were performed in a population <strong>of</strong> 74%<br />
cirrhotics. 20 Inclusion criteria were predefined and included<br />
patients with compensated cirrhosis, esophageal varices <strong>of</strong><br />
grade 1 or less, a platelet count <strong>of</strong> at least 80,000/mm 3 ,<br />
tumors less than 10 cm in size, an absence <strong>of</strong> major vascular<br />
invasion, and an <strong>American</strong> <strong>Society</strong> <strong>of</strong> Anesthesiologists<br />
(ASA) score <strong>of</strong> 3 or less, as well as those without evidence <strong>of</strong><br />
cirrhosis. Median follow-up was 30.4 months after resection<br />
with 1- and 3-year overall survival rates <strong>of</strong> 93% and 69%,<br />
respectively. Since this series was reported, operative time,<br />
blood loss, number <strong>of</strong> transfused packed red blood cells, and<br />
open conversion rates have continued to decline, suggesting<br />
a learning curve <strong>of</strong> this relatively novel technique. 21 A small<br />
study has suggested patients who had previously undergone<br />
laparoscopic resection compared with those who had undergone<br />
open surgery for HCC had decreased morbidity following<br />
salvage liver transplantation. 22 In this study, 24 total<br />
patients underwent salvage liver transplant after either<br />
prior laparoscopic (12 patients) or open resection (12 patients).<br />
The laparoscopy group demonstrated shorter resection<br />
and total operative time, less blood loss, and reduced<br />
need for blood transfusions. However, the most important<br />
question in comparing open resection with laparoscopic<br />
resection is whether oncologic outcomes are the same. A<br />
recent meta-analysis evaluated 10 nonrandomized controlled<br />
studies with 494 subjects, <strong>of</strong> whom 213 underwent<br />
laparoscopic and 281 underwent open resection for HCC. In<br />
addition to the improved morbidity among patients undergoing<br />
laparoscopic compared with open resection, there was<br />
no difference between the groups with respect to surgical<br />
margins, overall survival rates, and disease-free survival<br />
rates. 23 These findings were similar to those observed in a<br />
recent meta-analysis <strong>of</strong> 10 studies looking at 627 patients<br />
from China. 24 Despite the absence <strong>of</strong> higher-level evidence,<br />
laparoscopic liver resection for HCC is rapidly becoming<br />
preferable to open resection in well-selected patients.<br />
Outcomes and Prognostic Factors<br />
Hepatic resection for HCC has become a safer operation,<br />
with a reported in-hospital mortality rate <strong>of</strong> around 2% and<br />
a 90-day mortality rate <strong>of</strong> 5%, largely due to advances in<br />
surgical technique and improved patient selection, as demonstrated<br />
in a recent series <strong>of</strong> 129 patients with HCC from<br />
Toronto. 25 These marked improvements in outcomes can, in<br />
part, be attributed to increased utilization <strong>of</strong> segmental and<br />
parenchymal-sparing resections and decreased intraoperative<br />
blood loss. Nevertheless, morbidity for liver resections<br />
rates remain high at 20% to 50% and include complications<br />
such as pleural effusion (9%), perihepatic abscess (6%), ileus<br />
(6%), sterile perihepatic fluid collections (5%), wound infection<br />
(5%), urinary tract infection (4%), bile leak/biloma (3%),<br />
pneumonia (3%), and deep venous thrombosis (2%), as<br />
reported in a series <strong>of</strong> 1,803 patients consisting <strong>of</strong> 21%<br />
patients with HCC. 26 Expected 1-, 3-, 5-, and 10-year survival<br />
rates following HCC resection are 85%, 64%, 45%, and<br />
21%, respectively, as reported by the Liver Cancer Study<br />
Group in Japan. These data are derived from the largest<br />
report to date, which includes 6,785 patients with cirrhosis<br />
operated on between 1988 and 1999. 27 Patients without<br />
significant portal hypertension and normal bilirubin achieve<br />
267
a 5-year survival rate <strong>of</strong> 70% following liver resection for<br />
HCC. In those with portal hypertension, however, the rate <strong>of</strong><br />
5-year survival is 50% and even lower in the setting <strong>of</strong> an<br />
abnormal bilirubin. Similar results have been reported from<br />
Western and other Asian centers. Survival rates as high as<br />
60% at 5 years may be achieved in Child class A patients<br />
with well-encapsulated tumors <strong>of</strong> 2 cm in diameter, which<br />
are equivalent to results following liver transplantation.<br />
Unfortunately, few patients (� 10%) meet these selection<br />
criteria. 28 Furthermore, recurrence after resection is common,<br />
occurring in up to 80% <strong>of</strong> the patients at 5 years.<br />
Approximately two-thirds <strong>of</strong> recurrences occur within 2<br />
years after treatment; these are considered early recurrence<br />
by convention. 29<br />
The factors associated with early recurrence are vascular<br />
invasion, tumor size (only 25% <strong>of</strong> patients with HCC <strong>of</strong> less<br />
than 2 cm have vascular invasion), satellite nodules, alphafetoprotein<br />
levels, nonanatomic resection, and extent <strong>of</strong><br />
underlying disease. There has been some debate as to<br />
whether tumor recurrence 2 or more years following resection<br />
represents a true recurrence <strong>of</strong> the initial lesion or a de<br />
novo HCC in the oncogenic cirrhotic liver. Risk factors<br />
associated with late recurrence include presence <strong>of</strong> cirrhosis,<br />
active hepatitis, vascular invasion, moderate or poorly differentiated<br />
HCC, and multinodularity. 29 Genomic analyses<br />
to better define recurrence risk based on primary tumor<br />
characteristics and molecular evaluation <strong>of</strong> the cirrhotic<br />
liver may be valuable for prognostication in the future, 30<br />
and future research will show whether neoadjuvant therapy<br />
with radiation, retinoids, chemoembolization, interferon, or<br />
sorafenib (an inhibitor <strong>of</strong> several tyrosine protein kinases<br />
[VEGFR, PDGFR, Raf]) leads to delayed recurrence. Despite<br />
the fact that recurrence following resection <strong>of</strong> HCC is a poor<br />
prognosticator, there is evidence that some patients will<br />
benefit from aggressive surgical approaches if recurrence<br />
Authors’ Disclosures <strong>of</strong> Potential Conflicts <strong>of</strong> Interest<br />
Author<br />
Employment or<br />
Leadership<br />
Positions<br />
Consultant or<br />
Advisory Role<br />
Claudius Conrad*<br />
Kenneth K. Tanabe Best Doctors;<br />
Covidien; Health<br />
Advances; LEK<br />
Consulting;<br />
UpToDate<br />
*No relevant relationships to disclose.<br />
1. Duffy JP, Hiatt JR, Busuttil RW. Surgical resection <strong>of</strong> hepatocellular<br />
carcinoma. Cancer J. 2008;14:100-110.<br />
2. Bruix J, Hessheimer AJ, Forner A, et al. New aspects <strong>of</strong> diagnosis and<br />
therapy <strong>of</strong> hepatocellular carcinoma. Oncogene. 2006;25:3848-3856.<br />
3. Davila JA, Morgan RO, Shaib Y, et al. Hepatitis C infection and the<br />
increasing incidence <strong>of</strong> hepatocellular carcinoma: a population-based study.<br />
Gastroenterology. 2004;127:1372-1380.<br />
4. El-Serag HB, Mason AC. Risk factors for the rising rates <strong>of</strong> primary liver<br />
cancer in the United States. Arch Intern Med. 2000;160:3227-3230.<br />
5. Duffy JP, Vardanian A, Benjamin E, et al. Liver transplantation criteria<br />
for hepatocellular carcinoma should be expanded: a 22-year experience with<br />
467 patients at UCLA. Ann Surg. 2007;246:502-509; discussion 509-511.<br />
6. Schindl MJ, Redhead DN, Fearon KC, et al. The value <strong>of</strong> residual liver<br />
volume as a predictor <strong>of</strong> hepatic dysfunction and infection after major liver<br />
resection. Gut. 2005;54:289-296.<br />
7. Johnson TN, Tucker GT, Tanner MS, et al. Changes in liver volume from<br />
birth to adulthood: a meta-analysis. Liver Transpl. 2005;11:1481-1493.<br />
268<br />
occurs. It is our practice to <strong>of</strong>fer patients regional therapy—<br />
occasionally including resection—for recurrent HCC when<br />
the recurrence is limited to the liver. If recurrence occurs, it<br />
has been reported that multimodality therapy including<br />
transarterial chemoembolization, percutaneous ablations,<br />
and surgery results in an overall 5-year survival rate <strong>of</strong> 20%.<br />
Conclusion<br />
Surgical resection, ablation, and liver transplantation<br />
<strong>of</strong>fer the highest likelihood <strong>of</strong> long-term survival or cure in<br />
carefully selected patients. A multidisciplinary approach by<br />
a team consisting <strong>of</strong> surgical oncologists, transplant surgeons,<br />
medical oncologists, interventional radiologists, radiation<br />
oncologists, social workers, and other ancillary staff<br />
provides best outcomes. Advances in surgical techniques,<br />
technology, and understanding <strong>of</strong> liver anatomy allow experienced<br />
liver surgeons to perform resections that are safer<br />
than before, and resections are available to more patients<br />
than before. Whereas neoadjuvant therapy has not been<br />
shown to convert unresectable tumors to resectable ones,<br />
preoperative portal vein embolization can increase the FLR<br />
and convert patients from unresectable to resectable. Ablative<br />
techniques can be valuable adjuncts in the surgical<br />
therapy <strong>of</strong> patients with HCC. And in some instances<br />
ablation can be the sole local treatment modality during an<br />
open or laparoscopic operation, recognizing that ablation is<br />
similarly effective as resection for early HCC. Improvements<br />
in surgical techniques have contributed to decreased perioperative<br />
morbidity and mortality. Important prognostic factors<br />
associated with survival after resection include vascular<br />
invasion, satellite nodules, and extent <strong>of</strong> underlying liver<br />
disease. Further improvements in survival after hepatectomy<br />
for HCC will depend on strategies for early detection as<br />
well as effective neoadjuvant or adjuvant therapies.<br />
Stock<br />
Ownership Honoraria<br />
REFERENCES<br />
Springer<br />
Science and<br />
Business Media<br />
Research<br />
Funding<br />
AstraZeneca<br />
Expert<br />
Testimony<br />
CONRAD AND TANABE<br />
Other<br />
Remuneration<br />
8. Makuuchi M, Sano K. The surgical approach to HCC: our progress and<br />
results in Japan. Liver Transpl. 2004;10:S46-S52.<br />
9. Are C, Iacovitti S, Prete F, et al. Feasibility <strong>of</strong> laparoscopic portal vein<br />
ligation prior to major hepatectomy. HPB (Oxford). 2008;10:229-233.<br />
10. Hasegawa K, Kokudo N, Imamura H, et al. Prognostic impact <strong>of</strong> anatomic<br />
resection for hepatocellular carcinoma. Ann Surg. 2005;242:252-259.<br />
11. Eguchi S, Kanematsu T, Arii S, et al. Comparison <strong>of</strong> the outcomes<br />
between an anatomical subsegmentectomy and a non-anatomical minor<br />
hepatectomy for single hepatocellular carcinomas based on a Japanese<br />
nationwide survey. Surgery. 2008;143:469-475.<br />
12. Eltawil KM, Kidd M, Giovinazzo F, et al. Differentiating the impact <strong>of</strong><br />
anatomic and non-anatomic liver resection on early recurrence in patients<br />
with Hepatocellular Carcinoma. World J Surg Oncol. 2010;8:43.<br />
13. Takigawa Y, Sugawara Y, Yamamoto J, et al. New lesions detected by<br />
intraoperative ultrasound during liver resection for hepatocellular carcinoma.<br />
Ultrasound Med Biol. 2001;27:151-156.<br />
14. Liu CL, Fan ST, Cheung ST, et al. Anterior approach versus conven-
SURGICAL MANAGEMENT OF HCC<br />
tional approach right hepatic resection for large hepatocellular carcinoma: a<br />
prospective randomized controlled study. Ann Surg. 2006;244:194-203.<br />
15. Belghiti J, Guevara OA, Noun R, et al. Liver hanging maneuver: a safe<br />
approach to right hepatectomy without liver mobilization. J Am Coll Surg.<br />
2001;193:109-111.<br />
16. Kim SH, Park SJ, Lee SA, et al. Various liver resections using hanging<br />
maneuver by three Glisson’s pedicles and three hepatic veins. Ann Surg.<br />
2007;245:201-205.<br />
17. Gurusamy KS, Pamecha V, Sharma D, et al. Techniques for liver<br />
parenchymal transection in liver resection. Cochrane Database Syst Rev.<br />
2009;(1):CD006880.<br />
18. Weber JC, Navarra G, Jiao LR, et al. New technique for liver resection<br />
using heat coagulative necrosis. Ann Surg. 2002;236:560-563.<br />
19. Jiao LR, Ayav A, Navarra G, et al. Laparoscopic liver resection assisted<br />
by the laparoscopic Habib Sealer. Surgery. 2008;144:770-774.<br />
20. Dagher I, Belli G, Fantini C, et al. Laparoscopic hepatectomy for hepatocellular<br />
carcinoma: a European experience. J Am Coll Surg. 2010;211:16-23.<br />
21. Reddy SK, Tsung A, Geller DA. Laparoscopic liver resection. World<br />
J Surg. 35:1478-1486.<br />
22. Gigot JF, Glineur D, Santiago Azagra J, et al. Laparoscopic liver<br />
resection for malignant liver tumors: preliminary results <strong>of</strong> a multicenter<br />
European study. Ann Surg. 2002;236:90-97.<br />
23. Zhou YM, Shao WY, Zhao YF, et al. Meta-analysis <strong>of</strong> laparoscopic versus<br />
open resection for hepatocellular carcinoma. Dig Dis Sci. 2011;56:1937-1943.<br />
24. Li N, Wu YR, Wu B, et al. Surgical and oncologic outcomes following<br />
laparoscopic versus open liver resection for hepatocellular carcinoma: a<br />
meta-analysis. Hepatol Res. <strong>2012</strong>;42:51-59.<br />
25. Greco E, Nanji S, Bromberg IL, et al. Predictors <strong>of</strong> peri-opertative<br />
morbidity and liver dysfunction after hepatic resection in patients with<br />
chronic liver disease. HPB (Oxford). 2011;13:559-565.<br />
26. Jarnagin WR, Gonen M, Fong Y, et al. Improvement in perioperative<br />
outcome after hepatic resection: analysis <strong>of</strong> 1,803 consecutive cases over the<br />
past decade. Ann Surg. 2002;236:397-406; discussion 406-407.<br />
27. Ikai I, Itai Y, Okita K, et al. Report <strong>of</strong> the 15th follow-up survey <strong>of</strong><br />
primary liver cancer. Hepatol Res. 2004;28:21-29.<br />
28. Cha CH, Ruo L, Fong Y, et al. Resection <strong>of</strong> hepatocellular carcinoma in<br />
patients otherwise eligible for transplantation. Ann Surg. 2003;238:315-321;<br />
discussion 321-323.<br />
29. Imamura H, Matsuyama Y, Tanaka E, et al. Risk factors contributing<br />
to early and late phase intrahepatic recurrence <strong>of</strong> hepatocellular carcinoma<br />
after hepatectomy. J Hepatol. 2003;38:200-207.<br />
30. Hoshida Y, Villanueva A, Kobayashi M, et al. Gene expression in fixed<br />
tissues and outcome in hepatocellular carcinoma. N Engl J Med. 2008;359:<br />
1995-2004.<br />
31. Rahbari NN, Mehrabi A, Mollberg NM, et al. Hepatocellular carcinoma:<br />
current management and perspectives for the future. Ann Surg. 2011;253:<br />
453-469.<br />
269
THE MANAGEMENT OF LESS COMMON BUT<br />
COMPLEX UPPER GASTROINTESTINAL<br />
MALIGNANCIES: HEPATOCELLULAR CARCINOMA,<br />
PANCREATIC NEUROENDOCRINE TUMORS, AND<br />
BILIARY TRACT TUMORS<br />
CHAIR<br />
Pamela L. Kunz, MD<br />
Stanford University Medical Center<br />
Stanford, CA<br />
SPEAKERS<br />
Andrew X. Zhu, MD, PhD<br />
Massachusetts General Hospital<br />
Boston, MA<br />
Jordan D. Berlin, MD<br />
Vanderbilt-Ingram Cancer Center<br />
Nashville, TN
A Renaissance in Therapeutic Options for<br />
Pancreatic Neuroendocrine Tumors<br />
Overview: The field <strong>of</strong> pancreatic neuroendocrine tumors<br />
(NETs) has seen a remarkable renaissance in recent years with<br />
exponential increases in published research, clinical trials,<br />
and U.S. Food and Drug Administration (FDA)-approved treatments.<br />
Surgical resection remains the foundation for management<br />
<strong>of</strong> locoregional disease. However, for patients with<br />
advanced disease, novel therapeutic options have emerged.<br />
NETS ARISE from neuroendocrine cells throughout<br />
the body, most commonly in the lungs, gastrointestinal<br />
tract, and pancreas. The field <strong>of</strong> pancreatic NETs has seen<br />
a remarkable renaissance in recent years, with exponential<br />
increases in published research, clinical trials, and FDAapproved<br />
treatments (Fig. 1). Pancreatic NETs have an<br />
estimated incidence <strong>of</strong> 0.32 cases/100,000 people 1 and account<br />
for 1% <strong>of</strong> all pancreatic cancers by incidence and 10%<br />
<strong>of</strong> pancreatic cancers by prevalence. 2 The median age at<br />
diagnosis is age 60, with a white male predominance. The<br />
majority <strong>of</strong> NETs are functionally inactive but some produce<br />
hormones that lead to the clinical syndromes associated<br />
with hormone excess. Though most pancreatic NETs are<br />
indolent, the spectrum <strong>of</strong> clinical behavior is quite variable<br />
and sometimes difficult to predict. The nomenclature, classification,<br />
and grading systems for NETs have historically<br />
been inconsistent. Through critical evaluations <strong>of</strong> these<br />
systems, common principles have emerged. 3,4 NETs are<br />
generally divided into two main categories: welldifferentiated<br />
(low and intermediate grade) and poorly differentiated<br />
(high grade). The proliferative rate (mitotic<br />
index and Ki67) is considered a critical component <strong>of</strong> pathologic<br />
evaluation. Well-differentiated tumors usually have<br />
less than 20 mitoses/high powered field (hpf) and a Ki67<br />
index <strong>of</strong> less than 20%, and poorly differentiated tumors<br />
more than 20 mitoses/hpf and a Ki67 more than 20%.<br />
This discussion will focus on the management <strong>of</strong> welldifferentiated<br />
pancreatic NETs.<br />
Approach to Treatment<br />
Locoregional Disease<br />
At the time <strong>of</strong> diagnosis, approximately 40% <strong>of</strong> patients<br />
with pancreatic NETs have locoregional disease. 2 Surgical<br />
resection remains the mainstay <strong>of</strong> treatment for these patients.<br />
There is currently no role for adjuvant treatment<br />
following resection <strong>of</strong> a primary pancreatic NET, as little is<br />
known about patient and tumor characteristics that predict<br />
for recurrence.<br />
Unresectable and Metastatic Disease<br />
Nonsurgical therapies play a role in the setting <strong>of</strong> unresectable<br />
and metastatic disease, which account for approximately<br />
50% <strong>of</strong> newly diagnosed patients. 2 Patient selection<br />
is a critical first step in the treatment algorithm. For<br />
patients with functional tumors, somatostatin analogs<br />
should be considered to alleviate symptoms <strong>of</strong> peptide release.<br />
For patients with nonfunctional tumors, systemic<br />
treatments are generally indicated for patients with pro-<br />
By Pamela L. Kunz, MD<br />
Two separate randomized placebo-controlled studies have<br />
shown prolonged progression-free survival (PFS) with everolimus<br />
or sunitinib. Future studies are designed to answer<br />
questions about the role <strong>of</strong> somatostatin analogs as antiproliferative<br />
agents, combinations <strong>of</strong> biologic therapies, and new<br />
cytotoxic chemotherapy backbones.<br />
gressive, bulky, or symptomatic disease. Otherwise, patients<br />
with low-volume, stable, and asymptomatic disease may be<br />
monitored closely without treatment.<br />
Mediating Symptoms <strong>of</strong> Hormone Excess<br />
An estimated 40% to 53% <strong>of</strong> pancreatic NETs are functional.<br />
5,6 Functional tumors are defined as having inappropriate<br />
elevation <strong>of</strong> serum hormone markers combined with<br />
clinical evidence <strong>of</strong> hormone oversecretion, including insulinomas,<br />
gastrinomas, VIPomas, glucagonomas, and somatostatinomas.<br />
Somatostatin analogs are the foundation <strong>of</strong><br />
symptom management for patients with functional pancreatic<br />
NETs and can both decrease the secretion <strong>of</strong> such<br />
hormones and inhibit their end-organ effects. Somatostatin<br />
is a naturally occurring polypeptide produced by paracrine<br />
cells that are scattered throughout the gastrointestinal tract<br />
and inhibits gastrointestinal endocrine and exocrine function.<br />
Its effects are mediated through G-coupled protein<br />
somatostatin receptors (SSTR 1–5). Short- and long-acting<br />
octreotide (with high affinity for SSTR 2) is available in the<br />
United States. Lanreotide, available in Europe, is a longacting<br />
analog with similar binding affinity to octreotide.<br />
Pasireotide, a novel somatostatin analog with a different<br />
binding affinity pr<strong>of</strong>ile compared to octreotide or lanreotide,<br />
is currently in development. Note that somatostatin analogues<br />
should be used with caution for patients with insulinomas,<br />
as it may precipitate or worsen hypoglycemia.<br />
Oncologic Control<br />
Biologic Agents<br />
Recent studies with inhibitors <strong>of</strong> signaling pathways<br />
that target vascular endothelial growth factor (VEGF) and<br />
the mammalian target <strong>of</strong> rapamycin (mTOR) have demonstrated<br />
considerable activity in pancreatic NETs.<br />
RADIANT-3 was a randomized phase III study that evaluated<br />
the efficacy <strong>of</strong> everolimus, an mTOR inhibitor, in<br />
advanced pancreatic NETs. 7 In this international, multisite<br />
study, 410 patients with low- or intermediate-grade progressive<br />
advanced pancreatic NETs were randomly selected to<br />
From the Stanford University School <strong>of</strong> Medicine, Stanford Cancer Institute, Stanford,<br />
CA.<br />
Author’s disclosure <strong>of</strong> potential conflicts <strong>of</strong> interest are found at the end <strong>of</strong> this article.<br />
Address reprint requests to Pamela L. Kunz, MD, Stanford University School <strong>of</strong> Medicine,<br />
Stanford Cancer Institute, 875 Blake Wilbur Drive, Stanford, CA 94305; email:<br />
pkunz@stanford.edu.<br />
© <strong>2012</strong> by <strong>American</strong> <strong>Society</strong> <strong>of</strong> <strong>Clinical</strong> <strong>Oncology</strong>.<br />
1092-9118/10/1-10<br />
271
Fig 1. PubMed publications and clinical trial postings by year for<br />
pancreatic NETs.<br />
receive everolimus 10 mg orally daily or placebo. The primary<br />
endpoint was PFS; secondary endpoints were overall<br />
survival (OS), response rate (RR), and safety. The median<br />
PFS was 11.0 months with everolimus as compared to 4.6<br />
months with placebo (hazard ratio [HR] 0.35; 95% CI 0.27 to<br />
0.45; p � 0.001). The RR was 5% in the everolimus arm<br />
compared to 2% in the placebo arm. Updated OS showed no<br />
difference between arms; however, patients in the placebo<br />
arm crossed over to treatment at progression, likely reducing<br />
the chance <strong>of</strong> observing a survival difference (Table 1).<br />
Another randomized study evaluated the efficacy <strong>of</strong><br />
sunitinib, an inhibitor <strong>of</strong> the VEGF pathway, in advanced<br />
pancreatic NETs. One hundred seventy one patients with<br />
advanced, well-differentiated, progressive pancreatic NETs<br />
were randomly selected to receive sunitinib 37.5 mg orally<br />
daily or placebo. 8 Primary endpoints were PFS; secondary<br />
endpoints were RR, OS, and safety. The study was discontinued<br />
before the preplanned interim efficacy analysis, after<br />
an independent data and safety monitoring committee observed<br />
more serious adverse events and deaths in the<br />
placebo arm and a difference in PFS that favored the<br />
sunitinib arm. At the time <strong>of</strong> study closure, median PFS was<br />
11.4 months in the sunitinib arm compared with 5.5 months<br />
in the placebo arm (HR 0.42; 95% CI 0.26 to 0.66; p � 0.001).<br />
KEY POINTS<br />
● Surgical resection continues to be the mainstay <strong>of</strong><br />
treatment for patients with locoregional disease.<br />
● Somatostatin analogs are indicated for patients experiencing<br />
symptoms <strong>of</strong> hormone excess; there are<br />
currently no prospective data to support the use <strong>of</strong><br />
somatostatin analogs as antitumor agents for pancreatic<br />
neuroendocrine tumors.<br />
● For patients with advanced disease, single-agent<br />
everolimus and sunitinib have recently been shown to<br />
improve progression-free survival.<br />
● Temozolomide-based chemotherapy regimens are<br />
emerging as an acceptable alternative to streptozocin;<br />
prospective randomized studies are in<br />
development.<br />
272<br />
RRs in the sunitinib and placebo arms were 9.3% and 0%,<br />
respectively. A high proportion <strong>of</strong> patients in the placebo<br />
arm received sunitinib at progression through a continuation<br />
study, and although an early survival analysis appeared<br />
to favor the sunitinib arm, this difference did not<br />
remain significant with additional follow-up (Table 1).<br />
Cytotoxic Chemotherapy<br />
Patients with advanced pancreatic NETs currently have<br />
few treatment options that yield objective radiographic<br />
tumor regression. Recent studies evaluating everolimus 7<br />
and sunitinib 8 in this patient population have demonstrated<br />
prolongation <strong>of</strong> PFS compared to placebo, but overall tumor<br />
RRs (by RECIST) with these agents are less than 10%.<br />
Historical studies reporting the highest RRs include regimens<br />
with cytotoxic chemotherapy. In an initial randomized<br />
study <strong>of</strong> 106 patients, Moertel and colleagues reported<br />
activity associated with the combination <strong>of</strong> streptozocin and<br />
doxorubicin in patients with advanced islet-cell tumors<br />
(Table 1). 9 The RR associated with this regimen was reported<br />
to be 69%; however, because <strong>of</strong> the use <strong>of</strong> nonstandard<br />
response criteria, the true objective radiologic RR was<br />
likely much lower. Retrospective series have reported overall<br />
RRs <strong>of</strong> 6% to 39% associated with streptozocin-based<br />
regimens in pancreatic NETs. 10,11 Although clearly associated<br />
with activity, the combination <strong>of</strong> streptozocin and<br />
doxorubicin is also associated with considerable toxicity,<br />
including myelosuppression, asthenia, and renal insufficiency,<br />
precluding its routine use in this disease. Additionally,<br />
recent issues with drug availability have made routine<br />
use difficult.<br />
Recent retrospective series and small, prospective phase<br />
II studies suggest that temozolomide is similarly active but<br />
less toxic than streptozocin-based therapy when treating<br />
patients with pancreatic NETs. In a retrospective series <strong>of</strong><br />
36 patients treated with temozolomide monotherapy, tumor<br />
regression was observed in 31% <strong>of</strong> bronchial carcinoid tumors<br />
and 8% <strong>of</strong> pancreatic NETs. 12 In a series <strong>of</strong> 97 patients,<br />
18 <strong>of</strong> 53 patients with pancreatic NETs (34%) achieved a<br />
partial or complete response to temozolomide-based therapies.<br />
13 A third retrospective series <strong>of</strong> 21 patients treated<br />
with temozolomide monotherapy demonstrated responses in<br />
25% <strong>of</strong> pancreatic NETs. 14<br />
Temozolomide has also been investigated prospectively in<br />
phase II studies <strong>of</strong> patients with NETs, though always in<br />
combination with other agents. In an initial study, 29<br />
patients with metastatic NETs were treated with a combination<br />
<strong>of</strong> temozolomide, administered at a dose <strong>of</strong> 150 mg/m 2<br />
Table 1. Studies Leading to FDA Approval <strong>of</strong> Agents in<br />
Advanced Pancreatic NETs<br />
Regimen<br />
Number <strong>of</strong><br />
Patients<br />
RR<br />
(%)<br />
TTP or PFS<br />
(mo)<br />
PAMELA L. KUNZ<br />
OS<br />
(mo) Reference<br />
Strep/dox vs. 36 69 20 26.4 Moertel et al. 9<br />
Strep/5FU vs. 33 45 6.9 16.8<br />
Chlorotozocin 33 30 6.9 16.8<br />
Everolimus vs. 207 5 11.0 NR Yao et al. 7<br />
Placebo 203 2 4.6 36.6<br />
Sunitinib vs. 86 9 11.4 30.5 Raymond et al. 8<br />
Placebo 85 0 5.5 24.4<br />
Abbreviations: mo, months; NET, pancreatic neuroendocrine tumor; NR, not<br />
reached; OS, overall survival; PFS, progression-free survival; RR, response rate;<br />
TTP, time to progression.
THERAPEUTIC OPTIONS FOR PANCREATIC NETS<br />
for 7 days, every other week, and thalidomide at doses <strong>of</strong><br />
50 to 400 mg daily; this combination was associated with<br />
objective tumor responses in 5 <strong>of</strong> 11 patients (45%). 15 A<br />
subsequent phase II study evaluated the combination <strong>of</strong><br />
temozolomide and bevacizumab in the treatment <strong>of</strong> 34<br />
patients with NETs (15 pancreatic, 19 carcinoid). 16 Patients<br />
received 150 mg/m 2 /day <strong>of</strong> temozolomide orally for 7 days,<br />
every other week, and 5 mg/kg <strong>of</strong> bevacizumab intravaneously<br />
every other week. Because <strong>of</strong> anticipated lymphopenia,<br />
patients also received prophylaxis with trimethoprim/<br />
sulfamethoxazole (1 double strength tablet every Monday,<br />
Wednesday, Friday). Objective tumor responses were observed<br />
in 33% <strong>of</strong> patients with pancreatic NETs, but not in<br />
patients with carcinoid tumors. In a third, prospective<br />
study, a regimen <strong>of</strong> everolimus and temozolomide was associated<br />
with an overall tumor RR <strong>of</strong> 35% for patients with<br />
advanced pancreatic NETs. 17 Taken together, these prospective<br />
and retrospective studies suggest that temozolomidebased<br />
therapy is comparable to streptozocin-based regimens<br />
and might reasonably be associated with an overall tumor<br />
RR <strong>of</strong> 30% to 40% for patients with advanced pancreatic<br />
NETs.<br />
Interestingly, preclinical and early clinical evidence suggests<br />
that capecitabine may be synergistic with temozolomide.<br />
18,19 In a series <strong>of</strong> 17 patients with pancreatic NETs,<br />
combination therapy with temozolomide and capecitabine<br />
was associated with a tumor RR <strong>of</strong> 59%. 20 A recent singleinstitution<br />
retrospective study by Strosberg and colleagues<br />
reported RRs <strong>of</strong> 70% and median PFS <strong>of</strong> 18 months for 30<br />
patients with advanced pancreatic NETs. 21 In this study,<br />
temozolomide was administered at a dose <strong>of</strong> 200 mg/m 2<br />
on days 10 to 14, and capecitabine was administered at a<br />
dose <strong>of</strong> 750 mg/m 2 twice daily on days 1 to 14. The combination<br />
and specific dosing schedules were well-tolerated,<br />
with only four patients experiencing grade 3 or 4 adverse<br />
events (anemia, thrombocytopenia, elevated aspartate aminotransferase,<br />
and elevated alanine aminotranferease). The<br />
most common grade 1 and 2 adverse events were fatigue,<br />
nausea, myelosuppression, and hand-foot syndrome. Based<br />
on this study, the combination <strong>of</strong> temozolomide and capecitabine<br />
has become popular and commonly used in patients<br />
with advanced pancreatic NETs.<br />
Other Therapies for Advanced Disease<br />
Hepatic metastases commonly occur in patients with<br />
pancreatic NETs and adversely affect overall prognosis and<br />
quality <strong>of</strong> life. Therapies directed at locoregional control <strong>of</strong><br />
hepatic disease may be necessary to decrease symptoms<br />
associated with hormone excess. Surgery for hepatic metastases<br />
should be considered whenever the metastases are<br />
considered resectable and when there is no evidence <strong>of</strong><br />
extrahepatic disease. Thermal ablation or cryoablation may<br />
be considered as an adjunct to surgery or in settings where<br />
extrahepatic disease or comorbidities might favor a less<br />
aggressive intervention. Selective catheterization <strong>of</strong> the<br />
hepatic artery and embolization <strong>of</strong> vessels perfusing the<br />
tumors can result in clinically significant responses. Embolization<br />
options include bland embolization, chemoembolization,<br />
and radioembolization. Retrospective studies report<br />
benefit, though no prospective studies have been conducted.<br />
22<br />
Radiolabeled somatostatin analogs with therapeutic doses<br />
<strong>of</strong> the radioactive isotope can also provide disease control in<br />
patients with advanced disease and are used routinely in<br />
Europe. 23 The most commonly used radionuclides are indium<br />
( 111 I), yttrium ( 90 Y), and lutetium ( 177 Lu) and are only<br />
available in Europe. The largest retrospective study <strong>of</strong><br />
patients with gastroenteropancreatic NETs demonstrated<br />
clinical responses in 46% <strong>of</strong> patients at 3 months (complete<br />
2%, partial 28%, and minor 16%) and stable disease in 36%;<br />
the minority had progressive disease (20%). Median time to<br />
progression (TTP) was 40 months; median OS was 128<br />
months. The bone marrow and kidneys are the most important<br />
dose-limiting organs in peptide receptor radionuclide<br />
therapy. 24 Prospective randomized studies are needed to<br />
confirm these observed benefits.<br />
Ongoing and Future Directions<br />
The renaissance <strong>of</strong> clinical research in the field <strong>of</strong> NETs<br />
has only just begun. Ongoing and proposed studies are<br />
poised to answer important questions.<br />
First, do somatostatin analogs have antitumor activity in<br />
pancreatic NETs? Though many physicians extrapolate the<br />
findings from the PROMID study, 25 there is currently no<br />
prospective data to support the use <strong>of</strong> somatostatin analogs<br />
as antitumor agents for pancreatic NETs. An ongoing international<br />
study was designed to answer this question by<br />
randomly selecting patients with advanced nonfunctioning<br />
pancreatic NETs to receive lanreotide autogel versus placebo<br />
(CLARINET, NCT00353496). Primary endpoints are<br />
TTP and death. The study opened in June 2006, and accrual<br />
is ongoing.<br />
Second, does the combination <strong>of</strong> two biologic agents improve<br />
efficacy outcomes in pancreatic NETs? A CALGB<br />
study (80701, NCT01229943) addresses this question and<br />
randomly selects patients with advanced pancreatic NETs to<br />
receive everolimus and octreotide long-acting release (LAR)<br />
versus everolimus, bevacizumab, and octreotide LAR. The<br />
primary endpoint is PFS. The study opened in October 2010,<br />
and accrual is ongoing.<br />
The development <strong>of</strong> thoughtful future studies is critical<br />
in moving the field forward. In fact, a National Cancer<br />
Institute Neuroendocrine Tumor <strong>Clinical</strong> Trials Planning<br />
Meeting 26 held in 2009 identified key unmet needs, recommended<br />
appropriate study endpoints and inclusion criteria,<br />
and formulated priorities for future NET studies. The prospective<br />
evaluation <strong>of</strong> single-agent temozolomide and a<br />
comparision to temozolomide-based combinations for pancreatic<br />
NETs was a key recommendation <strong>of</strong> this meeting.<br />
ECOG 2211 is a proposed study in which patients with<br />
advanced pancreatic NETs will be randomly selected to<br />
receive temozolomide versus temozolomide and capecitabine.<br />
The principal objective <strong>of</strong> the study will be to assess<br />
whether the addition <strong>of</strong> capecitabine to temozolomide improves<br />
RRs when compared to temozolomide alone. It is<br />
anticipated that the superior arm will serve as a building<br />
block in future studies.<br />
Conclusion<br />
The recent advances in the field <strong>of</strong> pancreatic NETs<br />
are truly exciting. We now have additional agents in our<br />
arsenal <strong>of</strong> therapeutic options for patients with this disease,<br />
including everolimus and sunitinib, both <strong>of</strong> which have<br />
been shown to prolong PFS. Given the increasing number<br />
273
<strong>of</strong> therapeutic options, patient selection and treatment<br />
sequence can be complicated. A number <strong>of</strong> tools are available<br />
for the treating physician, including the National<br />
Comprehensive Cancer Network Neuroendocrine Tumor<br />
Guidelines 27 and guidelines from the North <strong>American</strong><br />
Neuroendocrine Tumor <strong>Society</strong>. 28 Additionally, a multidis-<br />
Author’s Disclosure <strong>of</strong> Potential Conflicts <strong>of</strong> Interest<br />
ciplinary approach can be helpful as modalities from medical<br />
oncology, interventional radiology, and surgery need to be<br />
weighed. Future studies are aimed at answering questions<br />
about the utility <strong>of</strong> somatostatin analogs for tumor control,<br />
use <strong>of</strong> combined biologics, and temozolomide-based chemotherapy<br />
regimens.<br />
Employment or<br />
Leadership Consultant or Stock<br />
Research<br />
Author<br />
Positions Advisory Role Ownership Honoraria Funding<br />
Pamela L. Kunz Roche/Genentech<br />
1. Yao JC, Hassan M, Phan A, et al. One hundred years after “carcinoid”:<br />
epidemiology <strong>of</strong> and prognostic factors for neuroendocrine tumors in 35,825<br />
cases in the United States. J Clin Oncol. 2008;26:3063-3072.<br />
2. Yao JC, Eisner MP, Leary C, et al. Population-based study <strong>of</strong> islet cell<br />
carcinoma. Ann Surg Oncol. 2007;14:3492-3500.<br />
3. Klimstra DS, Modlin IR, Adsay NV, et al. Pathology reporting <strong>of</strong><br />
neuroendocrine tumors: application <strong>of</strong> the Delphic consensus process to the<br />
development <strong>of</strong> a minimum pathology data set. Am J Surg Pathol. 2010;34:<br />
300-313.<br />
4. Klimstra DS, Modlin IR, Coppola D, et al. The pathologic classification<br />
<strong>of</strong> neuroendocrine tumors: a review <strong>of</strong> nomenclature, grading, and staging<br />
systems. Pancreas. 2010;39:707-712.<br />
5. Ito T, Tanaka M, Sasano H, et al. Preliminary results <strong>of</strong> a Japanese<br />
nationwide survey <strong>of</strong> neuroendocrine gastrointestinal tumors. J Gastroenterol.<br />
2007;42:497-500.<br />
6. Panzuto F, Nasoni S, Falconi M, et al. Prognostic factors and survival in<br />
endocrine tumor patients: comparison between gastrointestinal and pancreatic<br />
localization. Endocr Relat Cancer. 2005;12:1083-1092.<br />
7. Yao JC, Shah MH, Ito T, et al. Everolimus for advanced pancreatic<br />
neuroendocrine tumors. N Engl J Med. 2011;364:514-523.<br />
8. Raymond E, Dahan L, Raoul JL, et al. Sunitinib malate for the<br />
treatment <strong>of</strong> pancreatic neuroendocrine tumors. N Engl J Med. 2011;364:501-<br />
513.<br />
9. Moertel CG, Lefkopoulo M, Lipsitz S, et al. Streptozocin-doxorubicin,<br />
streptozocin-fluorouracil or chlorozotocin in the treatment <strong>of</strong> advanced isletcell<br />
carcinoma. N Engl J Med. 1992;326:519-523.<br />
10. Kouvaraki M, Ajani J, H<strong>of</strong>f P, et al. Fluorouracil, doxorubicin, and<br />
streptozocin in the treatment <strong>of</strong> patients with locally advanced and metastatic<br />
pancreatic endocrine carcinomas. J Clin Oncol. 2004;22:4762-4771.<br />
11. Cheng P, Saltz L. Failure to confirm major objective antitumor activity<br />
<strong>of</strong> streptozocin and doxorubicin in the treatment <strong>of</strong> patients with advanced<br />
islet cell carcinoma. Cancer. 1999;86:944-948.<br />
12. Ekeblad S, Sundin A, Janson ET, et al. Temozolomide as monotherapy<br />
is effective in treatment <strong>of</strong> advanced malignant neuroendocrine tumors. Clin<br />
Cancer Res. 2007;13:2986-2991.<br />
13. Kulke MH, Hornick JL, Frauenh<strong>of</strong>fer C, et al. O6-methylguanine DNA<br />
methyltransferase deficiency and response to temozolomide-based therapy in<br />
patients with neuroendocrine tumors. Clin Cancer Res. 2009;15:338-345.<br />
14. Maire F, Hammel P, Faivre S, et al. Temozolomide: a safe and effective<br />
treatment for malignant digestive endocrine tumors. Neuroendocrinology.<br />
2009;90:67-72.<br />
15. Kulke M, Stuart K, Enzinger P, et al. Phase II study <strong>of</strong> temozolomide<br />
and thalidomide in patients with metastatic neuroendocrine tumors. J Clin<br />
Oncol. 2006;24:401-406.<br />
274<br />
REFERENCES<br />
Expert<br />
Testimony<br />
PAMELA L. KUNZ<br />
Other<br />
Remuneration<br />
16. Kulke M, Stuart K, Earle C, et al. A phase II study <strong>of</strong> temozolomide and<br />
bevacizumab in patients with advanced neuroendocrine tumors. J Clin Oncol.<br />
2006;24:189s (suppl; abstr 4044).<br />
17. Kulke M, Blaszkowsky L, Zhu A, et al. Phase I/II study <strong>of</strong> everolimus<br />
(RAD001) in combination with temozolomide (TMZ) in patients (pts) with<br />
advanced pancreatic neuroendocrine tumors (NET). Gastrointestinal Cancers<br />
Symposium Proceedings. 2010 (abstr 223).<br />
18. Fine R, Fogelman D, Schreibman S. Effective treatment <strong>of</strong> neuroendocrine<br />
tumors with temozolomide and capecitabine. J Clin Oncol. 2005;23:361s<br />
(suppl; abstr 4216).<br />
19. Murakami J, Lee YJ, Kokeguchi S, et al. Depletion <strong>of</strong> O6methylguanine-DNA<br />
methyltransferase by O6-benzylguanine enhances 5-FU<br />
cytotoxicity in colon and oral cancer cell lines. Oncol Rep. 2007;17:1461-1467.<br />
20. Isac<strong>of</strong>f W, Moss R, Pecora A, et al. Temozolomide/capecitabine therapy<br />
for metastatic neuroendocrine tumors <strong>of</strong> the pancreas: a retrospective review.<br />
J Clin Oncol. 2006;24:625s (suppl; abstr 14023).<br />
21. Strosberg JR, Fine RL, Choi J, et al. First-line chemotherapy with<br />
capecitabine and temozolomide in patients with metastatic pancreatic endocrine<br />
carcinomas. Cancer. 2011;117:268-275. Epub 2010 Sep 7.<br />
22. Kvols LK, Turaga KK, Strosberg J, et al. Role <strong>of</strong> interventional<br />
radiology in the treatment <strong>of</strong> patients with neuroendocrine metastases in the<br />
liver. J Natl Compr Canc Netw. 2009;7:765-772.<br />
23. van Vliet EI, Teunissen JJ, Kam BL, et al. Treatment <strong>of</strong> gastroenteropancreatic<br />
neuroendocrine tumors with peptide receptor radionuclide therapy.<br />
Neuroendocrinology. Epub <strong>2012</strong> Jan 10.<br />
24. Kwekkeboom DJ, de Herder WW, Kam BL, et al. Treatment with the<br />
radiolabeled somatostatin analog [ 177 Lu-DOTA 0 ,Tyr 3 ] octreotate: toxicity,<br />
efficacy, and survival. J Clin Oncol. 2008;26:2124-2130.<br />
25. Rinke A, Muller HH, Schade-Brittinger C, et al. Placebo-controlled,<br />
double-blind, prospective, randomized study on the effect <strong>of</strong> octreotide LAR<br />
in the control <strong>of</strong> tumor growth in patients with metastatic neuroendocrine<br />
midgut tumors: a report from the PROMID Study Group. J Clin Oncol.<br />
2009;27:4656-4663.<br />
26. Kulke MH, Siu LL, Tepper JE, et al. Future directions in the treatment<br />
<strong>of</strong> neuroendocrine tumors: consensus report <strong>of</strong> the National Cancer Institute<br />
Neuroendocrine Tumor clinical trials planning meeting. J Clin Oncol. 2011;<br />
29:934-943.<br />
27. Kulke M. NCCN <strong>Clinical</strong> Practice Guidelines in <strong>Oncology</strong> (NCCN<br />
Guidelines): Neuroendocrine Tumors. Version 1. National Comprehensive<br />
Cancer Network; 2011.<br />
28. Kulke MH, Anthony LB, Bushnell DL, et al. NANETS treatment<br />
guidelines: well-differentiated neuroendocrine tumors <strong>of</strong> the stomach and<br />
pancreas. Pancreas. 2010;39:735-752.
A Review <strong>of</strong> the Management <strong>of</strong><br />
Hepatocellular Carcinoma: Standard Therapy<br />
and a Look to New Targets<br />
Overview: Management <strong>of</strong> hepatocellular carcinoma (HCC)<br />
continues to be challenging, but new treatment options are<br />
evolving. A multidisciplinary evaluation will help make the best<br />
treatment decisions for each patient. Although we continue to<br />
improve the outcomes <strong>of</strong> curative treatment with resection,<br />
liver transplant, and radi<strong>of</strong>requency ablation (RFA), many new<br />
liver-directed regional therapies including drug-eluting beads,<br />
HCC IS a malignancy <strong>of</strong> global importance: it is the sixth<br />
most common cancer and the third most common<br />
cause <strong>of</strong> cancer-related mortality worldwide. 1 In the United<br />
States, the incidence <strong>of</strong> HCC has tripled, and the 5-year<br />
survival rate has remained below 12% during the past two<br />
decades. 2 HCC is the most common primary liver cancer,<br />
accounting for 90% <strong>of</strong> all liver malignancies and has become<br />
the fastest-rising cause <strong>of</strong> cancer-related death in the United<br />
States. Management <strong>of</strong> HCC continues to present with many<br />
challenges. First, despite the well-recognized risk factors,<br />
including hepatitis B infection (HBV), hepatitis C infection<br />
(HCV), alcoholic liver disease, and nonalcoholic fatty liver<br />
disease, as well as the increased use <strong>of</strong> surveillance programs,<br />
most patients still present with unresectable or<br />
advanced-stage disease. Unfortunately, these patients do<br />
not benefit from curative treatment options including surgical<br />
resection or liver transplantation. Second, despite the<br />
availability <strong>of</strong> many local-regional therapies, including radi<strong>of</strong>requency<br />
ablation (RFA) and transarterial chemoembolization<br />
(TACE), there is a paucity <strong>of</strong> data from definitive<br />
phase III studies. It is <strong>of</strong>ten confusing to the community<br />
oncologist how to select the most appropriate treatment for<br />
HCC patients, underscoring the importance <strong>of</strong> multidisciplinary<br />
evaluation. Third, most patients with HCC have<br />
underlying cirrhosis, which adversely affects the overall<br />
survival (OS) <strong>of</strong> these patients and greatly limits the treatment<br />
options. Fourth, despite the approval <strong>of</strong> sorafenib<br />
based on phase III trials demonstrating OS benefits 3,4 and<br />
its extensive application in clinical practice during the past<br />
few years, it is increasingly clear that the benefits <strong>of</strong><br />
sorafenib remain modest. More importantly, the mechanisms<br />
<strong>of</strong> sorafenib’s efficacy, toxicity, and resistance remain<br />
elusive. On the other hand, we have witnessed an unparalleled<br />
time period <strong>of</strong> active drug development in HCC. Many<br />
molecularly targeted agents that inhibit different pathways<br />
<strong>of</strong> hepatocarcinogenesis are under various phases <strong>of</strong> clinical<br />
development. This review will attempt to summarize briefly<br />
the current status <strong>of</strong> management <strong>of</strong> HCC with a focus on<br />
the advanced stage <strong>of</strong> disease and potential new targeted<br />
agents under investigation in HCC.<br />
Staging and General Principles for the Management<br />
<strong>of</strong> HCC<br />
Once HCC is diagnosed, clinicians would aim to assess the<br />
extent <strong>of</strong> disease, the presence and severity <strong>of</strong> underlying<br />
cirrhosis, the performance status, and the access to institutional<br />
or local-regional expertise in formulating a treatment<br />
By Andrew X. Zhu, MD, PhD<br />
radioembolization, and radiation are emerging. Sorafenib remains<br />
the only approved agent for advanced HCC, and its<br />
role in the adjuvant setting following resection or RFA, with<br />
transarterial chemoembolization, or in combination with other<br />
targeted agents or chemotherapy in the advanced stage is<br />
under investigation. Many molecularly targeted agents with<br />
novel mechanisms <strong>of</strong> action are under active development.<br />
decision. Every effort should be made to ensure a careful<br />
evaluation <strong>of</strong> whether the patient is a candidate for surgical<br />
resection or liver transplant. A referral to a tertiary medical<br />
center is <strong>of</strong>ten helpful to render a definitive decision. A<br />
multidisciplinary team evaluation is critical to assess the<br />
pros and cons <strong>of</strong> each treatment option for individual patients,<br />
particularly in the setting <strong>of</strong> assessing local-regional<br />
therapy. The clinicians should be familiar with the indication<br />
<strong>of</strong> sorafenib and how to manage the sorafenib-related<br />
side effects. It is important to be aware <strong>of</strong> the ongoing<br />
clinical trials available for HCC and to support the enrollment<br />
<strong>of</strong> HCC-specific clinical trials. When the patients have<br />
advanced HCC with poor performance status or worsening<br />
cirrhosis, symptomatic management and preserving the<br />
quality <strong>of</strong> life <strong>of</strong> these patients should be the primary<br />
considerations.<br />
Numerous staging systems for HCC have been developed.<br />
These systems are useful for stratification <strong>of</strong> patients<br />
based on their prognosis, allocating specific treatment<br />
based on the stage, and allowing comparison <strong>of</strong> clinical<br />
outcomes from different clinical studies. Many different<br />
staging systems have been developed–Barcelona Clinic<br />
Liver Cancer (BCLC), 5 Cancer <strong>of</strong> the Liver Italian Program<br />
(CLIP), 6 tumor–node–metastasis (tumor, node, metastasis<br />
staging system), 7 Groupe d’Etude et de Traitement du<br />
Carcinome Hépatocellulaire (GRETCH), 8 Chinese University<br />
Prognostic Index (CUPI), 9 and Japanese Integrated<br />
Staging (JIS) 10 –and there is currently no universally accepted<br />
staging system. The BCLC staging classification<br />
is increasingly used in western regions, and it tries to<br />
capture the tumor features, severity <strong>of</strong> cirrhosis, performance<br />
status, and a recommended treatment algorithm for<br />
each stage. However, because <strong>of</strong> the geographic variation <strong>of</strong><br />
different risk factors, one staging system may perform<br />
better than others in certain regions. In addition, depending<br />
on the stage <strong>of</strong> the disease, certain staging system may<br />
be more prognostic, as suggested by one report comparing<br />
the various staging systems for patients with advanced<br />
disease. 11 In this study, the BCLC system was found to be<br />
From the Massachusetts General Hospital Cancer Center, Harvard Medical School,<br />
Boston, MA.<br />
Author’s disclosures <strong>of</strong> potential conflicts <strong>of</strong> interest are found at the end <strong>of</strong> this article.<br />
Address reprint requests to Dr. Andrew X. Zhu, Massachusetts General Hospital Cancer<br />
Center, 55 Fruit Street, LH/POB 232, Boston, MA 02114; email: azhu@partners.org.<br />
© <strong>2012</strong> by <strong>American</strong> <strong>Society</strong> <strong>of</strong> <strong>Clinical</strong> <strong>Oncology</strong>.<br />
1092-9118/10/1-10<br />
275
less informative than the GRETCH and CLIP classifications.<br />
11 Clinicians should become familiar with some <strong>of</strong><br />
these staging systems, their limitations, and controversies<br />
in the assessment <strong>of</strong> HCC. Although genomic analysis and<br />
molecular markers have been used to identify possible<br />
sub-classification <strong>of</strong> HCC, these findings would require additional<br />
validation and have not been incorporated in clinical<br />
staging.<br />
Management <strong>of</strong> Early-Stage HCC<br />
Despite some variation <strong>of</strong> practice patterns worldwide, it<br />
is clear that early-stage HCC can be cured by several<br />
treatment options including surgical resection, liver transplantation,<br />
and ablative therapies (Table 1).<br />
The aim <strong>of</strong> surgical resection is to remove the entire portal<br />
territory <strong>of</strong> the neoplastic segment(s) with a clear margin,<br />
while preserving maximum liver parenchyma to avoid hepatic<br />
failure. Because <strong>of</strong> the presence <strong>of</strong> extrahepatic disease,<br />
severe underlying cirrhosis, anatomic location <strong>of</strong><br />
tumor, and vascular invasion, less than 20% <strong>of</strong> patients with<br />
HCC are suitable for surgical resection. Resection and<br />
transplantation achieve the best outcomes in well-selected<br />
candidates (5-year survival <strong>of</strong> 60% to 80%) and compete as<br />
the first option in patients with early tumors on an<br />
intention-to-treat perspective. Hepatic resection is the treatment<br />
<strong>of</strong> choice for HCC in noncirrhotic patients, 12 where<br />
major resections can be performed with low rates <strong>of</strong> lifethreatening<br />
complications and acceptable outcome (5-year<br />
survival: 30% to 50%).<br />
Orthotopic liver transplantation is considered to be the<br />
first-line treatment option for patients with single tumors<br />
less than 5 cm or 3 nodules or less with each measuring 3 cm<br />
or less (Milan criteria) and advanced liver dysfunction not<br />
suitable for resection. In a landmark study by Mazzaferro<br />
KEY POINTS<br />
● Management <strong>of</strong> hepatocellular carcinoma (HCC) requires<br />
a multidisciplinary approach with careful<br />
evaluation <strong>of</strong> the extent <strong>of</strong> the tumor, the presence<br />
and severity <strong>of</strong> underlying cirrhosis, and performance<br />
status before a treatment decision can be made.<br />
● Surgical resection, liver transplantation, and radi<strong>of</strong>requency<br />
represent curative treatment options for<br />
HCC. There is no established role <strong>of</strong> adjuvant therapy.<br />
● Transarterial chemoembolization and other localregional<br />
therapies including drug-eluting beads, radioembolizaton,<br />
and radiation are evolving treatment<br />
options for unresectable HCC. The role <strong>of</strong> sorafenib,<br />
when given in combination with these treatments, is<br />
under investigation.<br />
● Sorafenib remains the only agent approved for advanced<br />
HCC. Careful selection <strong>of</strong> patients and timely<br />
management <strong>of</strong> side effects are important to optimize<br />
the efficacy <strong>of</strong> sorafenib.<br />
● Molecularly targeted agents are under active development<br />
and hold promise to improve the outcomes in<br />
patients with HCC.<br />
276<br />
Table 1. Management <strong>of</strong> Hepatocellular Carcinoma<br />
Early stage:<br />
● Surgical resection<br />
● Transplantation<br />
● RFA, PEI<br />
Intermediate stage:<br />
● TACE<br />
● drug-eluting beads<br />
● radioembolization<br />
● radiation<br />
Advanced stage:<br />
● Sorafenib<br />
● <strong>Clinical</strong> trials<br />
● Best supportive care<br />
Abbreviations: RFA, radi<strong>of</strong>requency ablation; PEI, percutaneous ethanol injection;<br />
TACE, transarterial chemoembolization.<br />
and colleges, patients with HCC who meet the Milan criteria<br />
have an expected 4-year overall survival rate <strong>of</strong> 85% and a<br />
recurrence-free survival rate <strong>of</strong> 92%. 13 Therefore, Milan<br />
criteria have been adopted by the United Network for Organ<br />
Sharing (UNOS) and widely around the world. Efforts have<br />
been made to expand the transplant criteria, for example,<br />
the University <strong>of</strong> San Francisco criteria (solitary HCC measuring<br />
up to 6.5 cm in diameter or up to three lesions, each<br />
measuring no more than 4.5 cm in diameter, with a total<br />
combined measurement <strong>of</strong> less than 8 cm) has been proposed<br />
and used in select centers. 14<br />
For patients with small HCCs who are poor surgical<br />
candidates because <strong>of</strong> impaired liver function or serious<br />
comorbid medical conditions, local ablative therapy represents<br />
another attractive treatment option. RFA has become<br />
the most commonly used local ablation therapy, as recent<br />
randomized trials have shown RFA to be more effective than<br />
percutaneous ethanol injection (PEI) in treating patients<br />
with small HCC (2–3 cm in diameter), with lower rates <strong>of</strong><br />
local recurrence and higher rates <strong>of</strong> overall and disease-free<br />
survival. 15 In addition, a randomized controlled trial has<br />
compared RFA with surgical resection and shown no significant<br />
differences in overall or recurrence-free survival, with<br />
lower rates <strong>of</strong> complications and hospitalization associated<br />
with RFA. 16<br />
An ongoing study is assessing the benefits <strong>of</strong> sorafenib<br />
in the adjuvant setting following surgical resection and<br />
RFA. Various bridging therapies including RFA and TACE<br />
are continuing to be used while patients are waiting for<br />
transplant, although the definitive benefits remain to be<br />
defined (p � 0.05).<br />
Management <strong>of</strong> Intermediate-Stage HCC<br />
ANDREW X. ZHU<br />
For patients with intermediate-stage disease with multifocal<br />
lesions and without vascular invasion, TACE has<br />
been the default treatment option (Table 1). TACE takes<br />
advantage <strong>of</strong> the fact that HCCs derive their blood supply<br />
almost entirely from the hepatic artery. The experience<br />
with TACE has been mixed, leaving many unanswered<br />
questions and controversies. Although several studies<br />
have shown negative survival benefits, two studies have<br />
demonstrated improved OS compared with best supportive<br />
care (BSC) alone in highly selective patient populations.<br />
In a randomized controlled trial, Llovet and colleagues<br />
demonstrated that patients (more than 80% with underlying<br />
HCV-related cirrhosis) who received doxorubicin-based
EVOLVING TREATMENT OF HCC<br />
Endpoint<br />
Table 2. Phase III Studies <strong>of</strong> Sorafenib in Hepatocellular<br />
Carcinoma<br />
SHARP Sorafenib<br />
versus Placebo 3<br />
Hazard Ratio<br />
(95% CI) P value<br />
TACE had improved OS compared with those who received<br />
BSC (p � 0.009). 17 Survival probabilities at 1 year and 2<br />
years were 82% and 63% for chemoembolization and 63%<br />
and 27% for control. In another single-center study conducted<br />
in Hong Kong where the majority <strong>of</strong> patients had<br />
underlying HBV infection, Lo and colleagues showed that<br />
patients with unresectable HCC who received cisplatinbased<br />
TACE had improved survival (1 year, 57%; 2 years,<br />
31%; 3 years, 26%) compared with those who received only<br />
symptomatic control (1 year, 32%; 2 years, 11%; 3 years,<br />
3%; p � 0.002). 18 A meta-analysis <strong>of</strong> randomized, controlled<br />
trials assessing the use <strong>of</strong> arterial embolization, chemoembolization,<br />
or both as primary palliative treatment for<br />
HCC showed that these procedures were associated with an<br />
improved 2-year survival rate as compared with conservative<br />
treatment. 19<br />
In addition to conventional TACE, several novel regional<br />
therapies, including drug-eluting beads TACE, radioembolization,<br />
and external beam radiation, are under active investigation<br />
in HCC. Whether certain techniques will perform<br />
better than others and how to incorporate regional therapy<br />
for patients with portal vein invasion requires additional<br />
studies. Current clinical studies are also assessing the role<br />
<strong>of</strong> sorafenib given concurrently or following TACE or other<br />
regional treatment modalities. Despite the early studies<br />
demonstrating the tolerability <strong>of</strong> combining sorafenib with<br />
TACE, 20,21 this approach has not definitively shown clinical<br />
benefits <strong>of</strong> sorafenib either following TACE or concurrently<br />
with TACE. 22,23<br />
Management <strong>of</strong> Advanced-Stage HCC<br />
Asia-Pacific Sorafenib<br />
versus Placebo 4<br />
Hazard Ratio<br />
(95% CI) P value<br />
OS 10.7 vs. 7.9 mo �0.001 6.5 vs. 4.2 mo 0.014<br />
0.69 (0.55–0.87) 0.68 (0.50–0.93)<br />
TTSP 1.08 (0.88–1.31) 0.768 0.90 (0.67–1.22) 0.50<br />
TTP 5.5 vs. 2.8 mo �0.001 2.8 vs. 1.4 mo �0.001<br />
0.58 (0.45–0.74) 0.57 (0.42–0.79)<br />
RR 2% vs. 1% 3.3% vs. 1.3%<br />
Abbreviations: OS, overall survival; TTSP, time to symptomatic progression;<br />
TTP, time to tumor progression; RR, response rate.<br />
Following the initial experience in a phase II study, 24 two<br />
randomized phase III trials definitively demonstrated the<br />
improved OS benefit <strong>of</strong> sorafenib in advanced HCC (Table 2).<br />
The approval and wide application <strong>of</strong> sorafenib has changed<br />
the treatment paradigm for HCC. However, as sorafenib is<br />
gaining more clinical experience, several important findings<br />
and related questions have emerged. First, the clinical<br />
benefits are modest and only seen in certain patients. This<br />
highlights the importance <strong>of</strong> understanding the mechanism<br />
<strong>of</strong> action <strong>of</strong> sorafenib and identifying potential predictive<br />
markers. Second, as sorafenib-related toxicities, including<br />
hand and foot skin reaction, diarrhea, and fatigue can be<br />
challenging to manage and affect the quality <strong>of</strong> life <strong>of</strong><br />
patients, many clinicians and investigators have asked the<br />
relevant question: what is the optimal dose <strong>of</strong> sorafenib?<br />
Although the targeted dose <strong>of</strong> sorafenib should be 400 mg<br />
twice daily based on phase III data, many clinicians have<br />
adopted a dose titration step-up approach, starting at 400<br />
mg daily and increase the dosage by 200 mg every 1–2 weeks<br />
to the targeted 800 mg daily as tolerated. Experience form<br />
the observational GIDEON study showed that 34% <strong>of</strong> patients<br />
in the United States started sorafenib at 400 mg daily.<br />
Third, because the agent was tested only in patients with<br />
underlying Child A cirrhosis in the phase III trials, the<br />
benefits <strong>of</strong> sorafenib in patients with worsening hepatic<br />
dysfunction remains uncertain. Several institutional-based<br />
retrospective studies have examined the use <strong>of</strong> sorafenib<br />
in patients with Child B cirrhosis. Although these studies<br />
have their inherent limitation, these findings suggest that<br />
sorafenib can be safely given to most patients with underlying<br />
Child B cirrhosis. However, increased hyperbilirubinemia<br />
and other side effects can be encountered at higher<br />
frequency. The pharmacokinetic (PK) parameters are similar<br />
or modestly different in patients with Child B in comparison<br />
with those with Child A. The treatment duration<br />
and OS are generally shorter in Child B than those with<br />
underlying Child A cirrhosis. Based on the available data,<br />
the starting dose for patients with good performance status<br />
and compensated hepatic function should be 800 mg daily.<br />
However, in patients with borderline performance status<br />
and compromised hepatic function, a reduced dose <strong>of</strong><br />
sorafenib at 200–400 mg can be started with the goal <strong>of</strong><br />
escalating to full dosage if tolerated. The PK study <strong>of</strong><br />
sorafenib in patients with hepatic and renal dysfunction also<br />
provides some guideline for dosing for these patients.<br />
New Targets and Regimens under Development for<br />
Advanced HCC<br />
The approval <strong>of</strong> sorafenib has greatly stimulated research<br />
in drug development in HCC. Many molecularly targeted<br />
agents that inhibit different pathways <strong>of</strong> hepatocarcinogenesis<br />
are under various phases <strong>of</strong> clinical development, and<br />
novel targets are being assessed in HCC. There are three<br />
main strategies in this area: (1) identifying and testing<br />
targeted agents with novel mechanisms <strong>of</strong> action; (2) combining<br />
various targeted agents that blocks specific targets in<br />
different or same pathways; (3) combining targeted agents<br />
with chemotherapy. Despite the excitement for these extensive<br />
efforts in the past few years, we have observed a few<br />
worrisome trends. First, all ongoing phase III studies with<br />
targeted agents are conducted in unselected populations.<br />
Second, most agents that are in phase III testing are not<br />
based on robust data from randomized phase II studies.<br />
Third, there is a high rate <strong>of</strong> failure <strong>of</strong> phase III trials in<br />
HCC. Table 3 lists some <strong>of</strong> the key studies in various phases<br />
<strong>of</strong> clinical development.<br />
Antiangiogenic Agents<br />
HCCs are vascular tumors, and increased levels <strong>of</strong> vascular<br />
endothelial growth factor (VEGF) and microvessel density<br />
(MVD) have been observed. 25 VEGF is one <strong>of</strong> the main<br />
inducers <strong>of</strong> liver tumor angiogenesis. High VEGF expression<br />
has been associated with lower survival. Inhibition <strong>of</strong> angiogenesis<br />
represents a potential therapeutic strategy and has<br />
been extensively tested in HCC.<br />
Several VEGF-R inhibitors have moved to phase III development<br />
based on the initial phase II data. 25 Sunitinib is<br />
an oral multikinase inhibitor that targets receptor tyrosine<br />
277
Table 3. New Agents/Regimens under Development in Advanced<br />
Hepatocellular Carcinoma 25<br />
Agents/Regimen<br />
Phase III, first-line Phase III, second-line<br />
● Sorafenib/erlotinib ● Brivanib (failed)<br />
● Sorafenib/doxorubicin ● Everolimus<br />
● Sunitinib (failed) ● Ramucirumab<br />
● Brivanib ● ADI-PEG 20<br />
● Linifanib<br />
Phase I-II<br />
Single-agent study:<br />
Antiangiogenic agents<br />
● Sunitinib, brivanib, bevacizumab, ramucirumab, TSU-68, linifanib,<br />
cediranib, pazopanib, lenvatinib, lenalidomide, and axitinib<br />
Epidermal growth factor receptor inhibitors<br />
● Erlotinib, gefitinib, lapatinib, cetuximab<br />
mTOR inhibitors<br />
● Everolimus, temsirolimus, and sirolimus<br />
c-Met inhibitors<br />
● Tivantinib (ARQ197), and cabozantinib (XL184)<br />
MEK inhibitors<br />
● Selumetinib (AZD6244)<br />
Histone deacetylase inhibitor<br />
● Belinostat<br />
HSP-90 inhibitor<br />
● STA-9090<br />
Combined targeted agents<br />
● Bevacizumab�erlotinib, sorafenib�everolimas, sorafenib�AZD6244,<br />
sorafenib�bevacizumab, sorafenib� temsirolimus, sorafenib and vorinostat,<br />
sorafenib �GC33, sorafenib� OSI-906<br />
kinases (RTKs) including VEGFR-1, VEGFR-2, PDGFRalpha/beta,<br />
c-KIT, FLT3, and RET kinases. Following the<br />
initial experience <strong>of</strong> sunitinib from several single-arm phase<br />
II studies that used three different dose schedules, a randomized<br />
phase III study comparing sunitinib at 37.5 mg<br />
continuous daily dosing with sorafenib at 400 mg twice daily<br />
in advanced HCC was conducted. Unfortunately, in this<br />
large study <strong>of</strong> 1,073 patients, sunitinib failed to demonstrate<br />
either superiority or noninferiority in OS when compared<br />
with sorafenib. 26 Toxicities including thrombocytopenia and<br />
neutropenia were seen with sunitinib leading to discontinuation.<br />
Brivanib alaninate is a dual inhibitor <strong>of</strong> VEGFR and<br />
fibroblast growth factor receptor (FGFR) signaling pathways<br />
that can induce tumor growth inhibition in mouse HCC<br />
xenograft models. A phase II study was conducted to assess<br />
the efficacy and safety <strong>of</strong> brivanib in patients with unresectable,<br />
locally advanced, or metastatic HCC who had received<br />
either no prior systemic therapy (Cohort A, 55 patients) or<br />
one prior regimen <strong>of</strong> angiogenesis inhibitor (Cohort B, 46<br />
patients). Treatment schedule consisted <strong>of</strong> continuous daily<br />
dosing <strong>of</strong> brivanib at 800 mg. Both schedules reported<br />
preliminary evidence <strong>of</strong> antitumor activity. Median<br />
progression-free survival (PFS) and OS were 2.7 (95% CI,<br />
1.4–3.0) and 10 months (95% CI, 6.8–15.2) in the first-line<br />
study. 27 Median OS and time to progression (TTP) as assessed<br />
by study investigators following second-line treatment<br />
with brivanib were 9.79 and 2.7 months,<br />
respectively. 28 Despite the ambitious phase III development<br />
program with brivanib in HCC, a recent press release<br />
reported that brivanib failed to demonstrate improved OS<br />
when compared with placebo in patients with advanced<br />
HCC who failed sorafenib. Linifanib (ABT-869) is a potent<br />
and selective inhibitor <strong>of</strong> VEGFR and PDGFR. Preliminary<br />
278<br />
results from an open label, multicenter phase II study <strong>of</strong><br />
linifanib in advanced HCC were reported. Linifanib was<br />
given at 0.25 mg/kg daily in Child-Pugh A. For all 34<br />
patients, median TTP was 112 days and median OS was<br />
295 days (95% CI, 182–333). A phase III study comparing<br />
linifanib with sorafenib should complete in the near future.<br />
Ramucirumab (IMC-1121B), a recombinant human monoclonal<br />
antibody against VEGFR-2, has been examined in a<br />
phase II study. This demonstrated a response rate (RR) <strong>of</strong><br />
10%, PFS <strong>of</strong> 4.0 months, and OS <strong>of</strong> 12.0 months in patients<br />
who have not received prior systemic therapy. Remucirumab<br />
is undergoing phase III development in the secondline<br />
setting against placebo in patients where sorafenib<br />
failed.<br />
Several other antiagiogenic agents are at early stages <strong>of</strong><br />
clinical development in HCC. These include bevacizumab, 29<br />
cediranib, pazopanib, lenvatinib, and axitinib. The abundance<br />
<strong>of</strong> VEGF inhibitors that entered HCC clinical trials<br />
and the failure <strong>of</strong> sunitinib and brivanib in phase III trials<br />
have prompted us to reconsider the relevance <strong>of</strong> targeting<br />
VEGF receptors in HCC. The challenge <strong>of</strong> developing additional<br />
antiangiogenic agents in HCC is to understand the<br />
mechanisms <strong>of</strong> action and develop potential surrogate and<br />
predictive markers to identify patients likely to benefit from<br />
treatment. Although we aim to select the most potent<br />
inhibitors, the safety pr<strong>of</strong>iles <strong>of</strong> these agents including<br />
bleeding, thromboembolic events, skin rashes, and hypertension<br />
should be carefully evaluated.<br />
Epidermal Growth Factor Receptor (EGFR) Inhibitors<br />
Increasing evidence has highlighted the importance <strong>of</strong><br />
EGFR/HER1 and its ligands EGF and transforming growth<br />
factor-alpha (TGF-alpha) in hepatocarcinogenesis. The expression<br />
<strong>of</strong> several EGF family members, specifically EGF,<br />
TGF-alpha, and heparin-binding–epidermal growth factor,<br />
as well as EGFR, has been described in several HCC cell<br />
lines and tissues. Multiple strategies to target EGFR signaling<br />
pathways have been developed, and two classes <strong>of</strong><br />
anti-EGFR agents are tested in HCC: monoclonal antibodies<br />
that competitively inhibit extracellular endogenous ligand<br />
binding and small molecules that inhibit the intracellular<br />
tyrosine kinase (TK) domain. With the exception <strong>of</strong> erlotinib<br />
showing modest activity in single-arm studies, the other<br />
EGFR inhibitors (gefinitib, lapatinib, and cetuximab) have<br />
not demonstrated convincing antitumor activity as single<br />
agents. Only erlotinib is being examined in an ongoing<br />
phase III study in combination with sorafenib.<br />
mTOR Inhibitors<br />
ANDREW X. ZHU<br />
Mammalian target <strong>of</strong> rapamycin (mTOR) functions to<br />
regulate protein translation, angiogenesis, and cell cycle<br />
progression in many cancers including HCC. Preclinical<br />
data have demonstrated that mTOR inhibitors were effective<br />
in inhibiting cell growth and tumor vascularity in HCC<br />
cell lines and HCC tumor models. Aberrant mTOR signaling<br />
was present in half <strong>of</strong> the HCC cases and correlated with<br />
HCC recurrence following resection.<br />
A number <strong>of</strong> mTOR inhibitors (sirolimus, temsirolimus,<br />
and everolimus) are under clinical development in HCC.<br />
Early evidence <strong>of</strong> tolerability and efficacy has emerged from<br />
phase I and II studies with everolimus, which has led to the<br />
ongoing randomized phase III study comparing everolimus
EVOLVING TREATMENT OF HCC<br />
against placebo for patients who failed or could not tolerate<br />
sorafenib.<br />
Hepatocyte Growth Factor (HGF)/c-Met Inhibitors<br />
Dysregulation <strong>of</strong> c-Met is seen in HCC, and silencing the<br />
expression <strong>of</strong> c-Met inhibits HCC growth in HCC cell lines<br />
and tumor models. Tivantinib (ARQ197), a selective, non-<br />
ATP-competitive inhibitor targeting MET tyrosine kinase, is<br />
under early clinical evaluation. In a press release, tivantinib<br />
reportedly met the primary endpoint <strong>of</strong> improving TTP in a<br />
randomized phase II study comparing tivantinib with placebo<br />
in previously treated patients. Cabozantinib (XL184), a<br />
dual c-Met/VEGFR-2 inhibitor, also demonstrated early evidence<br />
<strong>of</strong> antitumor activity in a randomized discontinuation<br />
phase II study with a median PFS <strong>of</strong> 4.2 months.<br />
MEK Inhibitors<br />
HCC is characterized by frequent MEK/ERK activation in<br />
the absence <strong>of</strong> RAS or RAF mutation. A multicenter, singlearm<br />
study with a two-stage design was conducted using<br />
selumetinib (AZD6244), a specific inhibitor <strong>of</strong> MEK, in<br />
advanced HCC. 30 The primary endpoint was RR. No radiographic<br />
responses were seen and TTP was only 8 weeks<br />
suggesting minimal single-agent activity despite evidence <strong>of</strong><br />
inhibition <strong>of</strong> ERK phosphorylation.<br />
Author’s Disclosures <strong>of</strong> Potential Conflicts <strong>of</strong> Interest<br />
Author<br />
Employment or<br />
Leadership<br />
Positions<br />
Consultant or<br />
Advisory Role<br />
Andrew X. Zhu Bristol-Myers<br />
Squibb; Pfizer;<br />
San<strong>of</strong>i<br />
1. Parkin DM, Bray F, Ferlay J, et al. Global cancer statistics, 2002. CA<br />
Cancer J Clin. 2005;55:74-108.<br />
2. El-Serag HB. Hepatocellular carcinoma. N Engl J Med. 365:1118-1127.<br />
3. Llovet JM, Ricci S, Mazzaferro V, et al. Sorafenib in advanced hepatocellular<br />
carcinoma. N Engl J Med. 2008;359:378-390.<br />
4. Cheng AL, Kang YK, Chen Z, et al. Efficacy and safety <strong>of</strong> sorafenib in<br />
patients in the Asia-Pacific region with advanced hepatocellular carcinoma: A<br />
phase III randomised, double-blind, placebo-controlled trial. Lancet Oncol.<br />
2009;10:25-34.<br />
5. Llovet JM, Bru C, Bruix J. Prognosis <strong>of</strong> hepatocellular carcinoma: The<br />
BCLC staging classification. Semin Liver Dis. 1999;19:329-338.<br />
6. A new prognostic system for hepatocellular carcinoma: A retrospective<br />
study <strong>of</strong> 435 patients: The Cancer <strong>of</strong> the Liver Italian Program (CLIP)<br />
investigators. Hepatology. 1998;28:751-755.<br />
7. Sobin LH. TNM, sixth edition: New developments in general concepts<br />
and rules. Semin Surg Oncol. 2003;21:19-22.<br />
8. Chevret S, Trinchet JC, Mathieu D, et al. A new prognostic classification<br />
for predicting survival in patients with hepatocellular carcinoma. Groupe<br />
d’Etude et de Traitement du Carcinome Hepatocellulaire. J Hepatol. 1999;<br />
31:133-141.<br />
9. Leung TW, Tang AM, Zee B, et al. Construction <strong>of</strong> the Chinese University<br />
Prognostic Index for hepatocellular carcinoma and comparison with the<br />
TNM staging system, the Okuda staging system, and the Cancer <strong>of</strong> the Liver<br />
Italian Program staging system: A study based on 926 patients. Cancer.<br />
2002;94:1760-1769.<br />
10. Kudo M, Chung H, Osaki Y. Prognostic staging system for hepatocellular<br />
carcinoma (CLIP score): its value and limitations, and a proposal for<br />
a new staging system, the Japan Integrated Staging Score (JIS score).<br />
J Gastroenterol. 2003;38:207-215.<br />
11. Huitzil-Melendez FD, Capanu M, O’Reilly EM, et al. Advanced hepa-<br />
Other Molecularly Targeted Agents under<br />
Development in HCC<br />
Multiple genetic and epigenetic changes occur during<br />
hepatocarcinogenesis. These pathways include the PI3K/Akt<br />
pathway, insulin growth factor (IGF) and its receptor<br />
(IGFR), as well as the Wnt/beta-catenin pathway. Multiple<br />
agents targeting these key pathways are under early-stage<br />
evaluation in HCC. The major challenge for future development<br />
<strong>of</strong> these targeted agents is the identification <strong>of</strong> predictive<br />
markers for specific drugs so that the targeted<br />
population can be enriched.<br />
Conclusion<br />
Management <strong>of</strong> HCC continues to be challenging, but<br />
potential treatment options are evolving. A multidisciplinary<br />
evaluation will allow physicians to make the best<br />
treatment decisions for patients. As we continue to improve<br />
the outcomes <strong>of</strong> curative treatment with resection, liver<br />
transplant, and RFA, many new liver-directed regional<br />
therapies are emerging. Sorafenib remains the only approved<br />
agent for advanced HCC, and its role in the adjuvant<br />
setting following resection or RFA, with TACE, or in combination<br />
with other targeted agents or chemotherapy in the<br />
advanced stage is under investigation. Many targeted<br />
agents with novel mechanisms <strong>of</strong> action are under active<br />
development.<br />
Stock<br />
Ownership Honoraria<br />
REFERENCES<br />
Research<br />
Funding<br />
Bayer/Onyx;<br />
ImClone Systems<br />
Expert<br />
Testimony<br />
Other<br />
Remuneration<br />
tocellular carcinoma: Which staging systems best predict prognosis? J Clin<br />
Oncol. 2010;28:2889-2895.<br />
12. Lang H, Sotiropoulos GC, Domland M, et al. Liver resection for<br />
hepatocellular carcinoma in non-cirrhotic liver without underlying viral<br />
hepatitis. Br J Surg. 2005;92:198-202.<br />
13. Mazzaferro V, Regalia E, Doci R, et al. Liver transplantation for the<br />
treatment <strong>of</strong> small hepatocellular carcinomas in patients with cirrhosis.<br />
N Engl J Med. 1996;334:693-699.<br />
14. Yao FY, Ferrell L, Bass NM, et al. Liver transplantation for hepatocellular<br />
carcinoma: Expansion <strong>of</strong> the tumor size limits does not adversely impact<br />
survival. Hepatology. 2001;33:1394-1403.<br />
15. Cho YK, Kim JK, Kim MY, et al. Systematic review <strong>of</strong> randomized<br />
trials for hepatocellular carcinoma treated with percutaneous ablation therapies.<br />
Hepatology. 2009;49:453-459.<br />
16. Chen MS, Li JQ, Zheng Y, et al. A prospective randomized trial<br />
comparing percutaneous local ablative therapy and partial hepatectomy for<br />
small hepatocellular carcinoma. Ann Surg. 2006;243:321-328.<br />
17. Llovet JM, Real MI, Montana X, et al. Arterial embolisation or<br />
chemoembolisation versus symptomatic treatment in patients with unresectable<br />
hepatocellular carcinoma: A randomised controlled trial. Lancet. 2002;<br />
359:1734-1739.<br />
18. Lo CM, Ngan H, Tso WK, et al. Randomized controlled trial <strong>of</strong><br />
transarterial lipiodol chemoembolization for unresectable hepatocellular carcinoma.<br />
Hepatology. 2002;35:1164-1171.<br />
19. Bruix J, Sala M, Llovet JM. Chemoembolization for hepatocellular<br />
carcinoma. Gastroenterology. 2004;127:S179-188.<br />
20. Dufour JF, Hoppe H, Heim MH, et al. Continuous administration <strong>of</strong><br />
sorafenib in combination with transarterial chemoembolization in patients<br />
with hepatocellular carcinoma: Results <strong>of</strong> a phase I study. Oncologist.<br />
2010;15:1198-1204.<br />
21. Pawlik TM, Reyes DK, Cosgrove D, et al. Phase II trial <strong>of</strong> sorafenib<br />
279
combined with concurrent transarterial chemoembolization with drug-eluting<br />
beads for hepatocellular carcinoma. J Clin Oncol. 2011;29:3960-3967.<br />
22. Kudo M, Imanaka K, Chida N, et al. Phase III study <strong>of</strong> sorafenib after<br />
transarterial chemoembolisation in Japanese and Korean patients with<br />
unresectable hepatocellular carcinoma. Eur J Cancer. 2011;47:2117-2127.<br />
23. Lencioni R LJ, Han G, Tak WY, et al. Sorafenib or placebo in<br />
combination with transarterial chemoembolization (TACE) with doxorubicineluting<br />
beads (DEBDOX) for intermediate-stage hepatocellular carcinoma<br />
(HCC): Phase II, randomized, double-blind SPACE trial.. J Clin Oncol.<br />
<strong>2012</strong>;30(suppl 4; abstr LBA154).<br />
24. Abou-Alfa GK, Schwartz L, Ricci S, et al. Phase II Study <strong>of</strong> Sorafenib in<br />
Patients With Advanced Hepatocellular Carcinoma. J Clin Oncol. 2006;10:<br />
4293-4300.<br />
25. Zhu AX, Duda DG, Sahani DV, et al. HCC and angiogenesis: Possible<br />
targets and future directions. Nat Rev Clin Oncol. 2011;8:292-301.<br />
280<br />
ANDREW X. ZHU<br />
26. Cheng A, Kang YK, Lin D, et al. Phase III trial <strong>of</strong> sunitinib (Su) versus<br />
sorafenib (So) in advanced hepatocellular carcinoma (HCC). J Clin Oncol.<br />
2011;29(suppl; abstr 4000).<br />
27. Park JW, Finn RS, Kim JS, et al. Phase II, open-label study <strong>of</strong> brivanib<br />
as first-line therapy in patients with advanced hepatocellular carcinoma. Clin<br />
Cancer Res. 2011;17:1973-1983.<br />
28. Finn RS, Kang YK, Mulcahy M, et al. Phase II, Open-label Study <strong>of</strong><br />
Brivanib as Second-line Therapy in Patients with Advanced Hepatocellular<br />
Carcinoma. Clin Cancer Res. <strong>2012</strong>; Epub ahead <strong>of</strong> print.<br />
29. Siegel AB, Cohen EI, Ocean A, et al. Phase II trial evaluating the<br />
clinical and biologic effects <strong>of</strong> bevacizumab in unresectable hepatocellular<br />
carcinoma. J Clin Oncol. 2008;26:2992-2998.<br />
30. O’Neil BH, G<strong>of</strong>f LW, Kauh JS, et al. Phase II study <strong>of</strong> the mitogenactivated<br />
protein kinase 1/2 inhibitor selumetinib in patients with advanced<br />
hepatocellular carcinoma. J Clin Oncol. 2011;29:2350-2356.
Advanced Biliary Tract Cancers<br />
By Laura Williams G<strong>of</strong>f, MD, and Jordan D. Berlin, MD<br />
Overview: Single-agent management <strong>of</strong> metastatic biliary<br />
tract cancers with 5-fluorouracil (5-FU) or gemcitabine has<br />
shown limited efficacy, although 5-FU has been shown to be<br />
more effective than best supportive care alone. An analysis<br />
<strong>of</strong> phase II trials has suggested that platinums enhanced<br />
the efficacy <strong>of</strong> single-agent fluoropyrimidines. In a phase III<br />
randomized trial comparing single-agent gemcitabine with<br />
gemcitabine plus cisplatin, the gemcitabine/cisplatin combination<br />
significantly improved median overall survival (OS)<br />
BILIARY TRACT cancers consist <strong>of</strong> a somewhat heterogeneous<br />
group <strong>of</strong> tumors that include gallbladder cancer,<br />
extrahepatic biliary tract cancer, and intrahepatic<br />
cholangiocarcinoma. The exact incidence <strong>of</strong> each <strong>of</strong> these<br />
cancers is difficult to discern from annual cancer statistics<br />
because, for example, intrahepatic cholangiocarcinoma is<br />
still combined with hepatocellular carcinoma despite the<br />
fact that these are biologically distinct entities. In 2011,<br />
9,250 new cases and 3,300 deaths from biliary tract cancers<br />
and gallbladder cancer (excluding intrahepatic cholangiocrcinoma)<br />
were anticipated. 1 Cancers <strong>of</strong> the biliary tract are<br />
<strong>of</strong>ten found at late stages when resection is not feasible and<br />
treatment options are limited. Overall, 5-year survival rates<br />
are estimated to be approximately 15%.<br />
Cytotoxic Chemotherapy<br />
The systemic treatment <strong>of</strong> biliary tract cancers has largely<br />
been based on cytotoxic chemotherapy. Data comparing<br />
5-FU–based chemotherapy with best supportive care demonstrated<br />
that median survival times are significantly prolonged<br />
(6.0 months vs. 2.5 months) with treatment. 2 In<br />
addition, the time before declines in quality <strong>of</strong> life was<br />
prolonged with 5-FU–based chemotherapy. Rates <strong>of</strong> response<br />
to either single-agent 5-FU or 5-FU with leucovorin<br />
range up to approximately 20% with survival times up to<br />
8 months. Similarly, single-agent gemcitabine has been<br />
explored as an alternative treatment option for biliary tract<br />
cancers. Response rates for single-agent gemcitabine have<br />
ranged from 16% to 36% and survival times from 6 months<br />
to 16 months. 3 Phase II study <strong>of</strong> combinations <strong>of</strong> gemcitabine<br />
with either 5-FU or its oral prodrug, capecitabine, has<br />
not clearly improved on the results <strong>of</strong> single-agent phase II<br />
data. 4 The most promising results came from the combination<br />
<strong>of</strong> gemcitabine and capecitabine with a response rate<br />
<strong>of</strong> 31% in 45 patients and an OS <strong>of</strong> 14 months. 5<br />
However, studies combining 5-FU or gemcitabine with<br />
platinums have yielded very promising response rates. 4<br />
Therefore, a study was undertaken to evaluate the effects <strong>of</strong><br />
gemcitabine plus cisplatin compared to gemcitabine alone. 6<br />
This study had an initial randomized phase II portion<br />
(ABC-01) and, when its endpoints were met, the trial proceeded<br />
to phase III (ABC-02). The ABC-02 trial demonstrated<br />
a significant benefit in both response rate and PFS<br />
favoring the gemcitabine/cisplatin arm. Most importantly,<br />
OS increased from 8.1 months for those treated with singleagent<br />
gemcitabine to 11.7 months for those treated with<br />
gemctiabine/cisplatin. These results have established the<br />
standard <strong>of</strong> care for biliary tract cancers at the current time.<br />
and progression-free survival (PFS), which established a new<br />
option for standard <strong>of</strong> care. However, the future <strong>of</strong> cancer<br />
medicine lies in newer, targeted agents. In the management <strong>of</strong><br />
biliary tract cancers, preliminary evidence with epidermal<br />
growth factor receptor inhibitors has already demonstrated<br />
activity. This article reviews systemic therapies for metastatic<br />
biliary tract cancers as they relate to current and emerging<br />
standards <strong>of</strong> care.<br />
KEY POINTS<br />
● Single-agent management <strong>of</strong> advanced biliary cancers<br />
with 5-fluorouracil or gemcitabine yields short<br />
survival times.<br />
● Adding cisplatin to gemcitabine significantly improves<br />
progression-free survival and overall survival<br />
for patients with biliary tract cancers.<br />
● Targeted therapies are being evaluated in patients<br />
with biliary tract cancers, showing some preliminary<br />
evidence <strong>of</strong> activity for epidermal growth factor receptor<br />
inhibition.<br />
Targeted Agents<br />
Because <strong>of</strong> the limited likelihood that cytotoxic therapy<br />
will substantially change the natural history <strong>of</strong> biliary tract<br />
cancers, investigation is now focusing on the use <strong>of</strong> targeted<br />
agents. One <strong>of</strong> the first agents studied in biliary tract<br />
cancers was erlotinib, an oral tyrosine kinase inhibitor <strong>of</strong><br />
the epidermal growth factor receptor (EGFR). EGFR is<br />
overexpressed in most biliary tract cancers. 7 In a singleagent,<br />
multi-institutional study, erlotinib produced rare<br />
responses (8%) in 42 patients. 8 Although the median time to<br />
tumor progression was only 2.6 months, 17% <strong>of</strong> patients<br />
were progression-free at 6 months and OS was 7.5 months.<br />
Because adding EGFR inhibition to chemotherapy has<br />
worked in other diseases such as colorectal cancer, investigators<br />
have added cetuximab to gemcitabine and oxaliplatin<br />
in a phase II study. A small randomized trial<br />
comparing the combination <strong>of</strong> gemcitabine plus oxaliplatin<br />
with gemcitabine/oxaliplatin plus erlotinib has been reported.<br />
9 In the 133 randomly assigned patients, there was a<br />
trend toward a PFS benefit from adding erlotinib, but OS<br />
was identical for both arms at 9.5 months. However, the<br />
EGFR pathway is actually a complex network and has<br />
interactions with several other pathways. Others have tried<br />
From the Vanderbilt Ingram Cancer Center, Nashville, TN.<br />
Authors’ disclosures <strong>of</strong> potential conflicts <strong>of</strong> interest are found at the end <strong>of</strong> this article.<br />
Address reprint requests to Jordan D. Berlin, MD, Vanderbilt Ingram Cancer Center, 777<br />
Preston Research Building, 2220 Pierce Ave., Nashville, TN 37232-6307, email: jordan.<br />
berlin@vanderbilt.edu.<br />
© <strong>2012</strong> by <strong>American</strong> <strong>Society</strong> <strong>of</strong> <strong>Clinical</strong> <strong>Oncology</strong>.<br />
1092-9118/10/1-10<br />
281
to optimize the use <strong>of</strong> erlotinib by adding other targeted<br />
agents. One phase II study <strong>of</strong> combined EGFR and vascular<br />
endothelial growth factor inhibition with erlotinib and bevacizumab<br />
showed promising results. 10 In 53 patients, six<br />
(12%) patients had confirmed partial responses. Median<br />
time to progression was 4.4 months and OS was 9.9 months.<br />
Other investigators have evaluated targeting different proteins<br />
downstream <strong>of</strong> EGFR for better results. Some evidence<br />
<strong>of</strong> activity has been seen with MEK inhibition and biliary<br />
tract cancers. 11 However, better understanding <strong>of</strong> these<br />
pathways in biliary tract cancer will help us to better use the<br />
new, targeted therapies for patients with biliary tract cancers.<br />
Authors’ Disclosures <strong>of</strong> Potential Conflicts <strong>of</strong> Interest<br />
Author<br />
Employment or<br />
Leadership<br />
Positions<br />
Consultant or<br />
Advisory Role<br />
Jordan D. Berlin NCI (L) Amgen; Arno<br />
Therapeutics;<br />
AstraZeneca;<br />
Celgene; Clovis<br />
<strong>Oncology</strong>;<br />
Genentech;<br />
Novartis; Otsuka;<br />
Roche; San<strong>of</strong>i<br />
Laura Williams G<strong>of</strong>f Abbott<br />
Laboratories<br />
1. Siegel R, Ward E, Brawley O, et al. Cancer statistics, 2011: the impact <strong>of</strong><br />
eliminating socioeconomic and racial disparities on premature cancer deaths.<br />
CA: Cancer J Clin. 2011;61:212-236.<br />
2. Glimelius B, H<strong>of</strong>fman K, Sjoden PO, et al. Chemotherapy improves<br />
survival and quality <strong>of</strong> life in advanced pancreatic and biliary cancer. Ann<br />
Oncol. 1996;7:593-600.<br />
3. Anderson CD, Pinson CW, Berlin J, et al. Diagnosis and treatment <strong>of</strong><br />
cholangiocarcinoma. Oncologist. 2004;9:43-57.<br />
4. Eckel F, Schmid RM. Chemotherapy in advanced biliary tract carcinoma:<br />
a pooled analysis <strong>of</strong> clinical trials. Br J Cancer. 2007;96:896-902.<br />
5. Knox JJ, Hedley D, Oza A, et al. Combining gemcitabine and capecitabine<br />
in patients with advanced biliary cancer: a phase II trial. J Clin Oncol.<br />
2005;23:2332-2338.<br />
6. Valle JW, Wasan HS, Palmer DD, et al. Gemcitabine with or without<br />
cisplatin in patients (pts) with advanced or metastatic biliary tract cancer<br />
(ABC): Results <strong>of</strong> a multicenter, randomized phase III trial (the UK ABC-02<br />
trial). J Clin Oncol. 2009;27:15s (suppl; abstr 4503).<br />
282<br />
Conclusion<br />
Management <strong>of</strong> biliary tract cancers has previously been<br />
limited to fluoropyrimidines; however, over the last several<br />
years, new two-drug combinations have been developed with<br />
evidence <strong>of</strong> better activity than single-agent regimens. Although<br />
gemcitabine/cisplatin have been established as a<br />
standard option for these cancers, other two-drug regimens<br />
appear to have similar results. Now that the era <strong>of</strong> targeted<br />
agents for biliary tract cancers is starting, it is important<br />
that we improve our understanding <strong>of</strong> the biology <strong>of</strong> the<br />
disease and how best to use these new classes <strong>of</strong> agents to<br />
improve the outcomes for patients with these deadly diseases.<br />
Stock<br />
Ownership Honoraria<br />
REFERENCES<br />
Research<br />
Funding<br />
Bristol Myers<br />
Squibb; OSI<br />
Pharmaceuticals;<br />
San<strong>of</strong>i<br />
Expert<br />
Testimony<br />
BERLIN AND GOFF<br />
Other<br />
Remuneration<br />
7. Lee CS, Pirdas A. Epidermal growth factor receptor immunoreactivitiy<br />
in gallbladder and extrahepatic biliary tract tumours. Pathol Res Pract.<br />
1995;191:1087-1091.<br />
8. Philip PA, Mahoney MR, Allmer C, et al. Phase II study <strong>of</strong> erlotinib<br />
in patients with advanced biliary cancer. J Clin Oncol. 2006;24:3069-<br />
3074.<br />
9. Lee J, Park SH, Chang HM, et al. Gemcitabine and oxaliplatin<br />
with or without eroltinib in advanced biliary-tract cancer: A multicentre,<br />
open-label, randomised, phase 3 study. Lancet Oncol. <strong>2012</strong>;13:181-<br />
188<br />
10. Lubner SJ, Mahoney MR, Kolesar JL et al. Report <strong>of</strong> a multicenter<br />
phase II trial testing a combination <strong>of</strong> biweekly bevacizumab and daily<br />
erlotinib in pateints with unresectable biliary cancer: A phase II consortium<br />
study. J Clin Oncol. 2010;28:3491-3497.<br />
11. Bekaii-Saab T, Phelps M, Li X, et al. A multi-institutional study <strong>of</strong><br />
AZD6244 in patients with advanced biliary cancers. Cancer Res. 2009; (suppl;<br />
abstr LB 129)
BEST USE OF IMAGING TECHNIQUES, OLD AND<br />
NEW, FOR GENITOURINARY CANCERS<br />
IN CLINICAL PRACTICE<br />
CHAIR<br />
Walter M. Stadler, MD<br />
The University <strong>of</strong> Chicago<br />
Chicago, IL<br />
SPEAKERS<br />
Charles J. Ryan, MD<br />
University <strong>of</strong> California, San Francisco<br />
San Francisco, CA<br />
Josef J. Fox, MD<br />
Memorial Sloan-Kettering Cancer Center<br />
New York, NY
Optimal Use <strong>of</strong> Imaging to Guide Treatment<br />
Decisions for Kidney Cancer<br />
Overview: Treatment monitoring for solid tumors in general<br />
and for metastatic renal cancer in particular has been dominated<br />
by assessment <strong>of</strong> tumor burden via cross-sectional<br />
imaging. This poses a special problem for the mammalian<br />
target <strong>of</strong> rapamycin and vascular endothelial growth factor<br />
pathway-directed agents used in this disease. The standard<br />
RECIST metrics used to categorize “response” and “progression”<br />
are arbitrary and do not adequately capture the effect <strong>of</strong><br />
these agents. Other approaches, including use <strong>of</strong> relative<br />
RECIST measures as a continuous variable, volumetric measurements,<br />
and functional assessments, such as dynamic<br />
THE USE <strong>of</strong> imaging for therapeutic decision making in<br />
oncology has traditionally been limited to assessment <strong>of</strong><br />
changes in tumor burden with treatment. The underlying<br />
presumption is that cancer will continue to grow in the<br />
absence <strong>of</strong> treatment and any tumor shrinkage is thus a<br />
bona fide treatment effect. The original challenges were thus<br />
considered to simply be a determination <strong>of</strong> the degree <strong>of</strong><br />
tumor shrinkage that can be reliably detected. These tumor<br />
burden changes were then codified and categorized into<br />
complete response, partial response, and progressive disease,<br />
with the remaining category labeled “stable disease.”<br />
Over the past decade, however, this rather simplistic view<br />
has been challenged by changes in both available therapies<br />
and available imaging technologies. Renal cancer is a microcosm<br />
<strong>of</strong> these challenges, and the issues are not only important<br />
for management <strong>of</strong> patients with this disease but are<br />
also illustrative <strong>of</strong> the challenges in oncology therapeutic<br />
monitoring in general.<br />
We discuss tumor burden monitoring in the context <strong>of</strong><br />
available renal cancer therapies, novel approaches to tumor<br />
burden assessment, the additional data available from contrast<br />
enhanced imaging, and radioisotope imaging with an<br />
emphasis on technetium bone scans and fluorodeoxyglucosepositron<br />
emission tomography(FDG-PET). In the context <strong>of</strong><br />
these discussions, we also briefly mention data on use <strong>of</strong><br />
baseline imaging data to select among various therapeutic<br />
options.<br />
Two-Dimensional Tumor Burden Assessment<br />
In the early days <strong>of</strong> medical oncology and oncologic therapeutics,<br />
reliable detection <strong>of</strong> tumor shrinkage in the context<br />
<strong>of</strong> limited technology was a major challenge. In a now<br />
classic study by Mortel and colleagues, 1 the investigators<br />
asked 16 experienced oncologists at an early ASCO meetings<br />
to palpate 12 solid spheres <strong>of</strong> various sizes, two <strong>of</strong> which<br />
were the same, that were beneath a thin mattress. The<br />
principal objective was to determine the size difference that<br />
could be reliably detected. The authors demonstrated that<br />
the product <strong>of</strong> the orthogonal bi-dimensional diameter <strong>of</strong> a<br />
sphere had to be at least 50% smaller in order for the error<br />
in detecting a difference to be �10%. They thus proposed<br />
that this 50% shrinkage would be necessary before a “therapeutic<br />
response” could be declared. As cross-sectional imaging<br />
was developed, this metric was arbitrarily translated<br />
to these new modalities. The metric was further simplified<br />
284<br />
By Walter M. Stadler, MD<br />
contrast-enhanced magnetic resonance imaging-based quantitative<br />
variables and fluorodeoxyglucose-positron emission<br />
tomography, have been proposed as alternatives, but the data<br />
do not support their routine clinical use. Even fewer data are<br />
available on the use <strong>of</strong> baseline imaging characteristics to<br />
choose a specific therapy. Therefore, until further research on<br />
imaging predictive and intermediate biomarkers matures, a<br />
combination <strong>of</strong> standard cross-sectional imaging and clinical<br />
judgment is the most pragmatic option for treatment decision<br />
making for patients with metastatic renal cancer.<br />
by the RECIST guidelines to translate the original required<br />
bi-dimensional measurements to uni-dimensional measurements.<br />
2<br />
Interestingly, these standards were promulgated without<br />
any formal assessment as to the degree <strong>of</strong> tumor shrinkage<br />
that could be reliably detected by these newer methods.<br />
Even more importantly, there were essentially no formal,<br />
statistically rigorous studies at the time that were designed<br />
to determine whether any specified degree <strong>of</strong> tumor shrinkage<br />
could be used as an intermediate endpoint in trials.<br />
More recently, studies have demonstrated the variability <strong>of</strong><br />
assessing uni-dimensional changes in the size <strong>of</strong> nodules in<br />
well-aerated lung. In one such study, approximately 84% <strong>of</strong><br />
the measures were within a range <strong>of</strong> �10% to �10%, with<br />
much greater relative variability among smaller tumors. 3<br />
Few if any studies have been performed with tumors in<br />
other organs, and no studies have been done, to my knowledge,<br />
to specifically evaluate metastatic renal cancer. Perhaps<br />
more concerning is the fact that the clinical trial metric<br />
<strong>of</strong> a “partial response,” which is based on a reliability<br />
measurement <strong>of</strong> simulated palpation, has become a metric<br />
for clinical decision making. Thus, a patient with 31% tumor<br />
shrinkage is deemed to have a “response” and continues<br />
therapy, whereas one with 29% shrinkage is deemed as<br />
having “stable disease” and as such, discontinuation <strong>of</strong><br />
therapy might be considered.<br />
The development <strong>of</strong> agents directed at the vascular endothelial<br />
growth factor (VEGF) pathway and the mammalian<br />
target <strong>of</strong> rapamycin (mTOR) in renal cancer furthermore<br />
made it clear that inhibition <strong>of</strong> tumor growth is not only a<br />
therapeutic outcome but is also an outcome with likely<br />
clinical relevance. To this end, it is notable that the standard<br />
tumor burden criteria for progressive disease (20% increase<br />
in uni-dimensional measurements) are completely arbitrary<br />
and are not based on any experimental derivation. It is<br />
perhaps thus not surprising that progression-free survival,<br />
despite being the endpoint for the majority <strong>of</strong> phase III renal<br />
cancer trials, and the basis <strong>of</strong> U.S. Food and Drug Admin-<br />
From the University <strong>of</strong> Chicago, Chicago, IL.<br />
Authors’ disclosures <strong>of</strong> potential conflicts <strong>of</strong> interest are found at the end <strong>of</strong> this article.<br />
Address reprint requests to Walter M. Stadler, MD, University <strong>of</strong> Chicago, 5814 S.<br />
Maryland MC2115, Chicago, IL 60637; email: wstadler@medicine.bsd.uchicago.edu.<br />
© <strong>2012</strong> by <strong>American</strong> <strong>Society</strong> <strong>of</strong> <strong>Clinical</strong> <strong>Oncology</strong>.<br />
1092-9118/10/1-10
IMAGING BIOMARKERS IN RENAL CANCER<br />
istration (FDA) approval for several drugs, has never been<br />
demonstrated to be a surrogate endpoint for survival in this<br />
disease. One <strong>of</strong> the few studies in which this progression<br />
metric was formally evaluated (which has been reported in<br />
abstract form only) demonstrated that there is a correlation<br />
between progression-free survival and overall survival in<br />
renal cancer but that the Kendall’s tau statistic was 0.50,<br />
which is considered a modest correlation and not sufficient<br />
for general use as a surrogate endpoint. 4<br />
A number <strong>of</strong> new immune therapy regimens being developed<br />
might have activity in renal cancer as well. An important<br />
issue to consider with regard to therapeutic monitoring<br />
<strong>of</strong> response to these drugs is that tumor shrinkage and<br />
inhibition <strong>of</strong> tumor growth might be delayed, presumably<br />
while the immune function is activated. As such, one needs<br />
to allow for a period <strong>of</strong> growth before declaring such an agent<br />
ineffective. Whether the criteria suggested for melanoma in<br />
the context <strong>of</strong> ipilimumab therapy will apply in renal cancer<br />
as well remains to be determined. 5<br />
Use <strong>of</strong> Tumor Size Measurements as a<br />
Continuous Variable<br />
Traditional tumor burden measurements using either<br />
bi-dimensional or uni-dimensional criteria, and which were<br />
derived from reliability measurements, categorized changes<br />
in tumor size into variables labeled “complete response,”<br />
“partial response,” “stable disease,” and “progressive disease.”<br />
The questionable clinical relevance <strong>of</strong> these categories<br />
has already been noted. Just as importantly, categorization<br />
<strong>of</strong> an otherwise continuous variable leads to loss <strong>of</strong> statistical<br />
information. Using actual RECIST-based measurements<br />
rather than their aforementioned categorization, one can<br />
compare two groups (e.g., a treated and an untreated group)<br />
with use <strong>of</strong> standard statistical approaches for group comparisons.<br />
Because RECIST measurements are not normally<br />
distributed, an initial transformation <strong>of</strong> the data (typically<br />
log-transformed) and formal rules for how to manage missing<br />
measurements and difficult-to-measure lesions need to<br />
be constructed. We undertook this exercise, and our simulations<br />
suggest that differences between an experimental<br />
and a control group in terms <strong>of</strong> changes in tumor size can be<br />
KEY POINTS<br />
● RECIST-based metrics <strong>of</strong> response, stable disease,<br />
and progression are arbitrary.<br />
● In the context <strong>of</strong> the vascular endothelial growth<br />
factor pathway and mammalian target <strong>of</strong> rapamycindirected<br />
therapy in renal cancer, the correlation<br />
between RECIST-based disease progression and survival<br />
is modest.<br />
● Volumetric tumor measurements, dynamic contrastenhanced<br />
imaging quantitative parameters, and<br />
functional imaging with radionuclides have all been<br />
proposed as novel predictive and intermediate biomarkers<br />
for renal cancer therapy.<br />
● Standard cross-sectional imaging and clinical judgment<br />
remain the best pragmatic approach to therapeutic<br />
monitoring in renal cancer.<br />
reliably detected with a sample <strong>of</strong> approximately 150 patients.<br />
6<br />
Although a detectable difference between an experimental<br />
and a control group using tumor size measurements as a<br />
continuous variable could be reliably ascribed to the intervention<br />
with use <strong>of</strong> standard statistical tests, such a difference<br />
clearly does not demonstrate that the difference has<br />
clinical relevance. Thus, a continuous tumor size metric<br />
could theoretically be used as a phase II clinical trial<br />
endpoint but not a phase III trial endpoint without additional<br />
data elucidating the relationship between changes<br />
in tumor size and clinically relevant endpoints. To this end,<br />
we have conducted additional simulations on data from<br />
phase III renal cancer trials <strong>of</strong> sorafenib and pazopanib and<br />
suggest that such an approach, if used in a phase II trial,<br />
would be highly predictive for the phase III results. 7 Obviously,<br />
further work assessing the relationship between tumor<br />
burden and clinical outcome is necessary.<br />
Novel Approaches to Tumor Burden Assessments<br />
Challenges with analysis and interpretation <strong>of</strong> unidimensional<br />
and bi-dimensional measurements <strong>of</strong> tumor<br />
size have led some authors to contend that volumetric<br />
measurements would be more predictive <strong>of</strong> clinical outcome.<br />
This belief is based on the observation that many tumors,<br />
including metastatic renal cancer, do not always grow or<br />
shrink in a spherical fashion. This observation raises the<br />
hypothesis that the poor correlation between measurements<br />
<strong>of</strong> tumor burden and clinically relevant outcomes such as<br />
survival are due to the fact that the uni-dimensional and<br />
bi-dimensional measurements do not accurately reflect tumor<br />
burden. Just as important, the manual extraction <strong>of</strong><br />
tumor burden data is cumbersome and, thus, the RECIST<br />
guidelines exclude many lesions. More recently, a number <strong>of</strong><br />
algorithms have been developed to more accurately determine<br />
tumor volume from standard cross-sectional images. 8<br />
It is now becoming technically possible to address this<br />
underlying hypothesis, and several studies are attempting<br />
to do so.<br />
Contrast-Enhanced Imaging<br />
With the advent <strong>of</strong> therapies directed at the VEGF pathway<br />
in renal cancer, it became quickly apparent that<br />
contrast-enhanced imaging provides additional information<br />
beyond tumor size alone. A number <strong>of</strong> investigators realized<br />
that these targeted agents lead to a hypoperfused appearance<br />
<strong>of</strong> lesions, consistent with the mechanism <strong>of</strong> action <strong>of</strong><br />
the agents. It has thus been proposed that more formal<br />
measurements <strong>of</strong> this hypoperfusion could not only be a<br />
measure <strong>of</strong> drug effect (a pharmacodynamic endpoint) but<br />
also an indication <strong>of</strong> clinical benefit (an intermediate or<br />
surrogate endpoint). A number <strong>of</strong> criteria for assessing these<br />
changes has been used, <strong>of</strong> which the most common are the<br />
“Choi Criteria,” which classify a “response” as tumor shrinkage<br />
<strong>of</strong> at least 10% or a decrease in tumor density <strong>of</strong> at least<br />
15%. 9 Some studies have suggested a correlation <strong>of</strong> “Choi<br />
Response” to VEGF pathway inhibitors with clinical outcome<br />
in patients with renal cancer. 10 A major challenge with<br />
these criteria, however, is that computerized tomography<br />
(CT) scans are not always well standardized with regard<br />
to the timing <strong>of</strong> image acquisition in relation to contrast<br />
administration or in the contrast administration rate, which<br />
are both critical variables in determining the degree <strong>of</strong><br />
285
enhancement seen. Furthermore, the criteria are categoric<br />
in nature, raising the same issues as discussed earlier for<br />
categorization <strong>of</strong> tumor size metrics.<br />
Investigators have thus attempted to conduct a more<br />
formal analysis <strong>of</strong> uptake, distribution, and washout <strong>of</strong><br />
contrast agent in lesions using pharmacokinetic approaches.<br />
Although this can be done with both contrast-enhanced CT<br />
and magnetic resonance imaging (MRI), the requirement for<br />
multiple images and associated radiation exposure, as well<br />
as the limited resolution traditionally associated with CT<br />
scans, has led to MRI being used in most <strong>of</strong> these studies. A<br />
number <strong>of</strong> quantitative variables describing uptake, distribution,<br />
and washout <strong>of</strong> contrast agent can be derived, with<br />
K trans , a complex variable dependent on both tumor blood<br />
flow and perfusion, being the most commonly used variable.<br />
Studies in patients with renal cancer have demonstrated<br />
that VEGF pathway inhibitors decrease K trans in a reliably<br />
detectable manner, but that there is little to no correlation<br />
between this decrease and clinical outcome. 11,12 Some have<br />
suggested that baseline K trans might be predictive <strong>of</strong> outcome<br />
in the context <strong>of</strong> these therapies, but that hypothesis<br />
remains to be proven. 11,12<br />
MRI-based measurements also have a number <strong>of</strong> additional<br />
limitations, including imaging time required to obtain<br />
the parameters, difficulty in determining these quantitative<br />
parameters in more than one or a few lesions, heterogeneity<br />
in the quantitative parameter even within one<br />
tumor, (which is technically challenging to measure), and<br />
the nonlinear relationship between contrast enhancement<br />
and contrast concentration, which is required for assessing<br />
K trans and other parameters.<br />
Alternative approaches using blood flow determined by<br />
nuclear medicine techniques and molecular probes are being<br />
developed, but the technology is currently immature, and<br />
correlations with clinical outcome remains incompletely<br />
determined.<br />
Technetium Bone Scans<br />
Bone scans have been used extensively in oncology imaging<br />
to determine the presence <strong>of</strong> metastases in bone. Given<br />
the limitations <strong>of</strong> standard CT-based cross-sectional imaging<br />
in x-ray dense bone and the ability to image the entire<br />
skeleton with a bone scan, a bone scan has been a valuable<br />
modality for clinical decision making. Nevertheless, it needs<br />
to be remembered that technetium is taken up at sites <strong>of</strong><br />
bone remodeling and thus the bone scan detects bone remodeling<br />
not the tumor per se. Other pathologic conditions,<br />
including inflammation, Paget’s disease, and fractures, can<br />
all lead to increased bone remodeling and positive bone<br />
scans. Furthermore, in renal cancer, sensitivity <strong>of</strong> bone<br />
scintigraphy has been reported to be 62%. 13 Lastly, dramatic<br />
antitumor effects might actually lead to an increase in<br />
bone modeling and a worse-appearing bone scan. This has<br />
been most prominently demonstrated in prostate cancer<br />
(see article in the ASCO <strong>2012</strong> Educational Book by Biting<br />
and Armstrong). Given these limitations, bone scans are a<br />
poor modality for monitoring response to therapy in renal<br />
cancer.<br />
FDG-PET<br />
The Warburg effect is one <strong>of</strong> the most ubiquitous and<br />
earliest recognized features <strong>of</strong> malignancy. In this phenomenon,<br />
malignant cells preferentially utilize an aerobic glyco-<br />
286<br />
lysis pathway, which in turn leads to dramatic upregulation<br />
<strong>of</strong> glucose transport. Fluorodeoxyglucose is a nonmetabolizable<br />
radioactive glucose analog that is taken up in cells by<br />
these same glucose transporters. It is therefore a useful<br />
imaging diagnostic marker for many cancers. Interestingly,<br />
renal cancers, and especially clear cell renal cancers, do not<br />
uniformly have elevated glucose uptake; in fact the sensitivity<br />
<strong>of</strong> FDG-PET for small pulmonary nodules that were<br />
subsequently histologically demonstrated to be metastatic<br />
renal cancer is 64%. 14<br />
The mTOR inhibitors are known to lead to hyperglycemia<br />
in part because <strong>of</strong> a more general inhibition <strong>of</strong> glucose<br />
uptake, and glucose uptake has been proposed to be a<br />
pharmacodynamic biomarker for mTOR therapy. Perhaps<br />
more interesting and related to the variable glucose uptake<br />
in renal cancer specifically, it has been proposed that renal<br />
cancers that have elevated glucose uptake would be preferentially<br />
sensitive to mTOR inhibition. 15 We tested this<br />
concept in a small single-arm trial and demonstrated that<br />
treatment with everolimus reliably leads to a decrease in<br />
FDG uptake, and is thus a pharmacodynamic biomarker.<br />
However, baseline FDG measurements did not correlate<br />
with patient outcome, and the correlation between the<br />
degree <strong>of</strong> decrease in FDG uptake and patient outcome was<br />
quite modest as well. 16<br />
Summary and <strong>Clinical</strong> Implications<br />
WALTER M. STADLER<br />
Many <strong>of</strong> the discussions here are more directly applicable<br />
to the use <strong>of</strong> imaging and imaging biomarkers in clinical<br />
trial monitoring. Although their findings may not be directly<br />
applicable to the clinical care scenario, the studies conducted<br />
to date provide some guidance. Most importantly,<br />
these studies have strongly suggested that more sophisticated<br />
imaging methods such as FDG-PET or quantitative<br />
parameters derived from measurements <strong>of</strong> contrast uptake<br />
and washout are unlikely to be useful for monitoring response<br />
to treatment directed at the VEGF pathway or<br />
mTOR or in selecting specific therapies. In addition, bone<br />
scans, although useful from a diagnostic perspective, have<br />
rather limited utility in therapeutic monitoring as well.<br />
Thus, for clinical monitoring <strong>of</strong> response to therapies directed<br />
at VEGF and mTOR in renal cancer, we are left<br />
essentially with standard cross-sectional imaging.<br />
Because agents directed at the VEGF pathway and mTOR<br />
lead to tumor shrinkage that does not always meet standard<br />
RECIST criteria, or lead only to tumor growth inhibition,<br />
“lack <strong>of</strong> growth” is typically used as a clinical indicator <strong>of</strong><br />
benefit. The obvious clinical challenge is that renal cancer<br />
can be indolent, and it is <strong>of</strong>ten impossible to determine<br />
whether any observed lack <strong>of</strong> growth is because <strong>of</strong> or despite<br />
<strong>of</strong> the administered agent. Conversely, growth exceeding<br />
the standard 20% above nadir used for determining progressive<br />
disease in clinical trials might still reflect a certain<br />
degree <strong>of</strong> growth inhibition by the agent. <strong>Clinical</strong> management<br />
therefore requires an understanding <strong>of</strong> the patient’s<br />
history, clinical disease symptoms, and clinical toxicities,<br />
and a thorough review <strong>of</strong> the images themselves. Careful<br />
clinical judgment is required to provide the patient with the<br />
maximum treatment duration <strong>of</strong> an agent from which he or<br />
she is benefiting, but no longer. Such judgment may become<br />
even more challenging if immune-stimulating agents, from<br />
which delayed responses may occur, become clinically available.
IMAGING BIOMARKERS IN RENAL CANCER<br />
Authors’ Disclosures <strong>of</strong> Potential Conflicts <strong>of</strong> Interest<br />
Author<br />
Employment or<br />
Leadership<br />
Positions<br />
Consultant or<br />
Advisory Role<br />
Walter M. Stadler UpToDate AVEO; Caremark;<br />
Genentech;<br />
Millennium;<br />
Novartis; Pfizer<br />
Stock<br />
Ownership Honoraria<br />
Abbott<br />
Laboratories (I)<br />
1. Moertel CG, Hanley JA. The effect <strong>of</strong> measuring error on the results <strong>of</strong><br />
therapeutic trials in advanced cancer. Cancer. 1976;38:388-394.<br />
2. Therasse P, Arbuck SG, Eisenhauer EA, et al. New guidelines to<br />
evaluate the response to treatment in solid tumors. European Organization<br />
for Research and Treatment <strong>of</strong> Cancer, National Cancer Institute <strong>of</strong> the<br />
United States, National Cancer Institute <strong>of</strong> Canada. J Natl Cancer Inst.<br />
2000;92:205-216.<br />
3. Oxnard GR, Zhao B, Sima CS, et al. Variability <strong>of</strong> lung tumor measurements<br />
on repeat computed tomography scans taken within 15 minutes. J Clin<br />
Oncol. 2011;29:3114-3119.<br />
4. Halabi S, Rini BI, Stadler WM, et al: Use <strong>of</strong> progression-free survival<br />
(PFS) to predict overall survival (OS) in patients with metastatic renal cell<br />
carcinoma. J Clin Oncol. 2010;28;15s (suppl; abstract 4525).<br />
5. Wolchok JD, Hoos A, O’Day S, et al. Guidelines for the evaluation <strong>of</strong><br />
immune therapy activity in solid tumors: immune-related response criteria.<br />
Clin Cancer Res. 2009;15:7412-7420.<br />
6. Karrison TG, Maitland ML, Stadler WM, et al. Design <strong>of</strong> phase II cancer<br />
trials using a continuous endpoint <strong>of</strong> change in tumor size: application to a<br />
study <strong>of</strong> sorafenib and erlotinib in non small-cell lung cancer. J Natl Cancer<br />
Inst. 2007;99:1455-1461.<br />
7. Sharma MR, Karrison TG, Jin Y, et al. Resampling phase III data to<br />
assess phase II trial designs and endpoints. Clin Cancer Res. Epub <strong>2012</strong> Jan<br />
27.<br />
8. Zhao B, Oxnard GR, Moskowitz CS, et al. A pilot study <strong>of</strong> volume<br />
measurement as a method <strong>of</strong> tumor response evaluation to aid biomarker<br />
development. Clin Cancer Res. 2010;16:4647-4653.<br />
9. Choi H, Charnsangavej C, Faria SC, et al. Correlation <strong>of</strong> computed<br />
REFERENCES<br />
Bayer/Onyx;<br />
<strong>Clinical</strong> Care<br />
Options; CME<br />
Innovations;<br />
Imedex; Pfizer;<br />
Research-to-<br />
Practice<br />
Research<br />
Funding<br />
Active Biotech;<br />
Bayer;<br />
Boehringer<br />
Ingelheim;<br />
Bristol-Myers<br />
Squibb;<br />
Genentech;<br />
GlaxoSmithKline;<br />
ImClone<br />
Systems; Infinity;<br />
Lilly; Medarex;<br />
Medivation;<br />
Millennium;<br />
Novartis; Pfizer;<br />
Solvay; Takeda<br />
Expert<br />
Testimony<br />
Other<br />
Remuneration<br />
tomography and positron emission tomography in patients with metastatic<br />
gastrointestinal stromal tumor treated at a single institution with imatinib<br />
mesylate: proposal <strong>of</strong> new computed tomography response criteria. J Clin<br />
Oncol. 2007;25:1753-1759.<br />
10. Krajewski KM, Guo M, Van den Abbeele AD, et al. Comparison <strong>of</strong> four<br />
early posttherapy imaging changes (EPTIC; RECIST 1.0, tumor shrinkage,<br />
computed tomography tumor density, Choi criteria) in assessing outcome to<br />
vascular endothelial growth factor-targeted therapy in patients with advanced<br />
renal cell carcinoma. Eur Urol. 2011;59:856-862.<br />
11. Flaherty KT, Rosen MA, Heitjan DF, et al. Pilot study <strong>of</strong> DCE-MRI to<br />
predict progression-free survival with sorafenib therapy in renal cell carcinoma.<br />
Cancer Biol Ther. 2008;7:496-501.<br />
12. Hahn OM, Yang C, Medved M, et al. Dynamic contrast-enhanced<br />
magnetic resonance imaging pharmacodynamic biomarker study <strong>of</strong> sorafenib<br />
in metastatic renal carcinoma. J Clin Oncol. 2008;26:4572-4578.<br />
13. Sohaib SA, Cook G, Allen SD, et al. Comparison <strong>of</strong> whole-body MRI and<br />
bone scintigraphy in the detection <strong>of</strong> bone metastases in renal cancer. Br J<br />
Radiol. 2009;82:632-639.<br />
14. Majhail NS, Urbain JL, Albani JM, et al. F-18 fluorodeoxyglucose<br />
positron emission tomography in the evaluation <strong>of</strong> distant metastases from<br />
renal cell carcinoma. J Clin Oncol. 2003;21:3995-4000.<br />
15. Thomas GV, Tran C, Mellingh<strong>of</strong>f IK, et al. Hypoxia-inducible factor<br />
determines sensitivity to inhibitors <strong>of</strong> mTOR in kidney cancer. Nat Med.<br />
2006;12:122-127.<br />
16. Chen JL, Appelbaum DE, Kocherginsky M, et al. FDG-PET as a<br />
predictive marker for therapy with everolimus in metastatic renal cell cancer.<br />
J Clin Oncol. 2011;29 (suppl; abstr e1504).<br />
287
CASTRATION-RESISTANT PROSTATE CANCER: NEW<br />
INSIGHTS INTO BIOLOGY AND TREATMENT (eQ&A)<br />
CHAIR<br />
Oliver Sartor, MD<br />
Tulane Cancer Center<br />
New Orleans, LA<br />
SPEAKERS<br />
Robert Evan Reiter, MD<br />
University <strong>of</strong> California, Los Angeles<br />
Los Angeles, CA<br />
Andrew J. Armstrong, MD, MSc<br />
Duke Cancer Institute<br />
Durham, NC
State-<strong>of</strong>-the-Art Management for the Patient<br />
with Castration-Resistant Prostate Cancer<br />
in <strong>2012</strong><br />
Overview: Much progress has been made in metastatic<br />
castration-resistant prostate cancer (CRPC), and multiple new<br />
U.S. Food and Drug Administration (FDA)-approved survivalprolonging<br />
drugs are now available. In 2004, docetaxel/prednisone<br />
was the first therapy shown to prolong survival. In 2010<br />
and 2011, sipuleucel-T, cabazitaxel/prednisone, and abiraterone/prednisone<br />
were FDA approved. Two new agents,<br />
radium-223 and MDV-3100, have recently reported large phase<br />
III trials prolonging overall survival and will be submitted for<br />
regulatory approval in <strong>2012</strong>. One can now begin to ask, is there<br />
an optimal sequence for therapies in metastatic CRPC? Despite<br />
the recent progress, there is much we do not know and<br />
virtually no information on this important question. We know<br />
that abiraterone/prednisone and cabazitaxel/prednisone are<br />
appropriate choices for a patient after receiving docetaxel,<br />
but we do not know what, if anything, represents the optimal<br />
sequence for abiraterone and cabazitaxel. In fact we do not<br />
understand how one therapy may affect the response to a<br />
THE LAST several years have seen extraordinary progress<br />
in the management <strong>of</strong> patients with CRPC, with<br />
multiple FDA approvals for agents that extend patients’<br />
overall survival. In addition there are several new anticipated<br />
FDA approvals that might occur in <strong>2012</strong>. From a<br />
perspective <strong>of</strong> understanding the current state <strong>of</strong> affairs, I<br />
would first like to review the history <strong>of</strong> FDA approvals for<br />
patients with metastatic CRPC. As seen in Fig. 1, all the<br />
FDA drug approvals that occurred before 2004 involved<br />
endpoints other than overall survival. In 2004, the combination<br />
<strong>of</strong> docetaxel/prednisone was approved on the basis <strong>of</strong><br />
a prolongation in survival (as compared with mitoxantrone/<br />
prednisone), using data from the pivotal TAX327 study. 1 A<br />
separate study with docetaxel in 2004 (SWOG 9916), using<br />
docetaxel/estramustine, also demonstrated superior survival<br />
in comparison to mitoxantrone/prednisone. 2 Because <strong>of</strong><br />
the importance <strong>of</strong> these findings, from both a pragmatic and<br />
regulatory perspective, the post-2004 world <strong>of</strong> metastatic<br />
CRPC began to be divided into patients who had or had not<br />
received prior docetaxel.<br />
For the next 6 years, no substantial progress was made in<br />
prolonging survival in metastatic CRPC. Since 2010, rather<br />
remarkably, there have been four new FDA approvals and<br />
three <strong>of</strong> these have involved agents that prolong survival.<br />
Unlike a number <strong>of</strong> other diseases, the underlying mechanisms<br />
for these agents are distinct with a novel immunotherapy<br />
termed sipuleucel-T, a novel taxane called<br />
cabazitaxel, and novel hormone therapy in the form <strong>of</strong><br />
abiraterone prolonging survival in various pivotal phase III<br />
randomized trials. 3-5 Two <strong>of</strong> these agents were FDA approved<br />
in the postdocetaxel setting (cabazitaxel/prednisone<br />
and abiraterone/prednisone). No comparisons between any<br />
<strong>of</strong> these agents have been performed and no direct comparison<br />
between docetaxel/prednisone plus an additional agent<br />
has ever proven superior to docetaxel/prednisone alone.<br />
A number <strong>of</strong> trials have tried to improve on docetaxel/<br />
prednisone but none have succeeded. Combinations <strong>of</strong> do-<br />
By Oliver Sartor, MD<br />
subsequent therapy. We are also aware that the pre- and<br />
postdocetaxel spaces represent regulatory rather than biologic<br />
divisions. In addition, despite the proven role <strong>of</strong> docetaxel/prednisone,<br />
many patients with CRPC are not<br />
considered to be suitable for chemotherapy, and worldwide<br />
many never receive any form <strong>of</strong> chemotherapy. What is the<br />
optimal management for these patients? Taken together it is<br />
reasonable to assess patient preferences, prior therapies and<br />
response/tolerance to prior therapies, burden <strong>of</strong> disease,<br />
comorbidities, current symptoms, drug toxicities, out-<strong>of</strong>pocket<br />
costs, etc., in clinical decision making. Given the many<br />
factors we do not know, it is hard to be dogmatic in approaching<br />
the therapeutic options for the patient with CRPC. We will<br />
likely soon move beyond the current sequencing paradigm and<br />
begin to assess new combinations in a systematic and rational<br />
fashion. Perhaps one day, in the not too distant future, we will<br />
develop molecular “stratification systems” to better guide<br />
therapeutic choices in CRPC.<br />
cetaxel and DN-101 (calcitriol), GVAX, bevacizumab,<br />
atrasentan, zibotentan, and lenalidomide all have failed in<br />
phase III trials against docetaxel/prednisone. Dasatinib,<br />
aflibercept, and custirsen are in current combination trials<br />
with docetaxel/prednisone, and in each case the control arm<br />
is docetaxel/prednisone. These agents target the src kinase<br />
(dasatinib), vascular endothelial growth factor (aflibercept),<br />
and clusterin (custirsen), and each <strong>of</strong> these targets are<br />
supported by preclinical data in various systems. Whether<br />
or not these combinations will be an improvement on the<br />
current standard <strong>of</strong> care is unknown at the time this was<br />
written, but some results may be available during the <strong>2012</strong><br />
ASCO Annual Meeting. Both the dasatinib and aflibercept<br />
trials are fully accrued and awaiting maturation <strong>of</strong> survival<br />
data at this time. The trial with custirsen is still accruing as<br />
<strong>of</strong> spring <strong>2012</strong>.<br />
One trial is going head to head with docetaxel/prednisone.<br />
That trial will compare cabazatizel/prednisone (two doses <strong>of</strong><br />
cabazitaxel are being tested, 20 or 25 mg/m 2 ) compared with<br />
docetaxel/prednisone. The primary endpoint is overall survival.<br />
This trial (FIRSTANA) is currently accruing patients.<br />
In 2011 and early <strong>2012</strong>, two new agents have resulted in<br />
prolonged survival in large phase III trials. These agents<br />
include the novel targeted alpha-particle emitter radium-<br />
223 and a novel antiandrogen MDV-3100. 6,7 Taken together<br />
there are now six agents that have prolonged survival in<br />
phase III trials conducted in metastatic CRPC (Fig. 2).<br />
As noted above, the disease state for these therapies has<br />
varied. For instance, sipuleucel-T is indicated for use in the<br />
asymptomatic or minimally symptomatic metastatic CRPC<br />
From the Tulane Cancer Center, Tulane University School <strong>of</strong> Medicine, New Orleans, LA.<br />
Author’s disclosures <strong>of</strong> potential conflicts <strong>of</strong> interest are found at the end <strong>of</strong> this article.<br />
Address reprint requests to Oliver Sartor, MD, Tulane School <strong>of</strong> Medicine, 1430 Tulane<br />
Ave., Box SL-42, New Orleans, LA 70112; email: osartor@tulane.edu.<br />
© <strong>2012</strong> by <strong>American</strong> <strong>Society</strong> <strong>of</strong> <strong>Clinical</strong> <strong>Oncology</strong>.<br />
1092-9118/10/1-10<br />
289
Fig. 1. FDA regulatory approvals in metastatic CRPC in the United<br />
States.<br />
state. The eligibility requirements for the phase III trials<br />
with MDV3100, cabazitaxel, and abiraterone all required<br />
docetaxel pretreatment. The radium-223 study was unique,<br />
though the majority <strong>of</strong> patients had received prior docetaxel,<br />
the trial was also open to patients with bone-metastatic<br />
CRPC, patients who were not deemed ineligible for docetaxel,<br />
and for those patients who refused docetaxel. 6 This<br />
is the first trial to evaluate this group <strong>of</strong> patients. Because<br />
not all patients with metastatic CRPC go on to receive<br />
docetaxel, this is an important group <strong>of</strong> patients to include in<br />
prospective studies. 8<br />
There are several questions about the management <strong>of</strong> the<br />
patients with metastatic CRPC. The first is a very practical<br />
one: is there an optimal sequence for the FDA-approved<br />
therapies? Is there a proper way to pick the right therapy for<br />
the right patient at the right time? Is there any way to risk<br />
stratify patients to maximize their opportunity for response?<br />
The short answer to these reasonable questions is that we<br />
are poorly informed on what might constitute an optimal<br />
sequencing strategy for using new agents.<br />
KEY POINTS<br />
● Patients with metastatic castration-resistant prostate<br />
cancer (CRPC) have more options than ever<br />
before, including new agents that have been shown to<br />
prolong survival.<br />
● New Food and Drug Administration–approved agents<br />
for CRPC that prolong survival include sipuleucel-T,<br />
cabazitaxel, and abiraterone, and more are on the<br />
way.<br />
● Optimal choices and sequences are much discussed in<br />
CRPC but comparative data are absent; thus, dogmatic<br />
views <strong>of</strong> sequences are inappropriate.<br />
● Patient preferences, current symptoms, responses<br />
and tolerance <strong>of</strong> prior therapies, pace <strong>of</strong> the disease,<br />
burden <strong>of</strong> disease, drug toxicities, performance status,<br />
comorbidities, out-<strong>of</strong>-pocket costs, compliance,<br />
and various clinical trial options are some <strong>of</strong> the<br />
choices that guide practical considerations in treatment<br />
management.<br />
● Combination therapies are beginning to be explored<br />
in a systematic fashion.<br />
290<br />
OLIVER SARTOR<br />
Fig. 2. Phase III trials in metastatic CRPC with survival advantages.<br />
One can perhaps imagine that sipuleucel-T could be administered<br />
to a patient with small-volume asymptomatic<br />
disease, followed by docetaxel at progression, followed by<br />
cabazitaxel and/or abiraterone at progression. This would be<br />
a logical series. Despite the excellent logic <strong>of</strong> such an<br />
approach, there is very limited clinical data on patients that<br />
might be treated in this manner. We have reports <strong>of</strong> docetaxel<br />
following sipuleucel-T, and certainly that appears to<br />
be a safe combination, but we have no real data on abiraterone<br />
response rates postcabazitaxel or cabazitaxel response rates<br />
postabiraterone. Though MDV-3100 has yet to be approved by<br />
regulatory authorities, it is not too soon to ask what will be the<br />
activity <strong>of</strong> this agent postabiraterone (and vice versa). Will<br />
resistance to one <strong>of</strong> the newer hormonal therapies result in<br />
resistance to the other? We do not know.<br />
At the same time that therapeutic decisions are made, it is<br />
an important reminder that a bone-targeted therapy, such<br />
as zoledronic acid or denosumab, might be integrated into<br />
the treatment regimen for appropriate patients. External<br />
beam radiation could be used to palliate focal pain on an “as<br />
needed” basis. Growth factors and various other supportive<br />
care options are also considerations in selected settings.<br />
Taken together, there is little that we are certain <strong>of</strong> when<br />
it comes to optimal drug sequencing in this disease state.<br />
Further, the best choice <strong>of</strong> therapies—when choices clearly<br />
exist (such as the current postdocetaxel setting)—is not at<br />
all clear and one is simply left to conjecture. It is reasonable<br />
to assess patient preferences, prior therapies and response<br />
to prior therapies, performance status, pace <strong>of</strong> progression,<br />
burden <strong>of</strong> disease, comorbidities, tolerance/intolerance <strong>of</strong><br />
prior therapies, drug toxicities, neuroendocrine status, and<br />
current symptoms in clinical decision making. It is also<br />
critical to understand the availability <strong>of</strong> clinical trials,<br />
patient compliance, travel distances, and out-<strong>of</strong>-pocket<br />
costs. Taken together, at this time, no simple algorithm can<br />
suffice when it comes to making clinical decisions, and any<br />
dogmatic approach to this problem cannot be justified.<br />
We all realize that none <strong>of</strong> the therapies for CRPC to date<br />
are curative, and in my practice, I try to ensure that patients<br />
have the opportunity to be treated with as many <strong>of</strong> the active<br />
therapies as possible during the course <strong>of</strong> their disease. Our<br />
ability to predict who will and who will not respond to<br />
various therapies is poor, and many <strong>of</strong> my grateful patients<br />
have benefited from a treatment plan that was not judged a<br />
priori to have a high rate <strong>of</strong> success. Perhaps one day our<br />
molecular markers will guide us to the right choice <strong>of</strong>
THERAPEUTIC OPTIONS FOR PROSTATE CANCER<br />
therapy at each juncture, but today we are <strong>of</strong>ten humbled by<br />
how poorly we predict patient outcomes.<br />
What about combination therapy? Is the metastatic CRPC<br />
state to be restricted to a sequencing therapeutic paradigm?<br />
Though combination therapy has a certain attraction in<br />
CRPC, other than the agents designed to inhibit skeletalrelated<br />
events (zoledronic acid and denosumab), there are<br />
minimal data on combination therapies using the survivalprolonging<br />
agents. Until new data exist, the current sequencing<br />
paradigm may be optimal for patients in nonclinical trial<br />
settings.<br />
It is clear from phase I/II studies that patients with<br />
nondocetaxel-pretreated metastatic CRPC are quite<br />
responsive to the newer hormonal agents, such as abiraterone<br />
and MDV-3100. 9,10 This emphasizes that our division<br />
<strong>of</strong> metastatic CRPC into pre- and postdocetaxel spaces<br />
Author’s Disclosures <strong>of</strong> Potential Conflicts <strong>of</strong> Interest<br />
Author<br />
Employment or<br />
Leadership<br />
Positions<br />
Consultant or<br />
Advisory Role<br />
Oliver Sartor Algeta; Amgen;<br />
Bayer; Bellicum;<br />
Bristol-Myers<br />
Squibb; Celgene;<br />
Dendreon;<br />
Exelixis;<br />
GlaxoSmithKline;<br />
Johnson &<br />
Johnson;<br />
Medivation;<br />
Oncogenex;<br />
Pfizer; San<strong>of</strong>i;<br />
Takeda<br />
Authors’ Disclosures <strong>of</strong> Potential Conflicts <strong>of</strong> Interest<br />
1. Tannock IF, de Wit R, Berry WR, et al. Docetaxel plus prednisone or<br />
mitoxantrone plus prednisone for advanced prostate cancer. N Engl J Med.<br />
2004;351:1502-1512.<br />
2. Petrylak DP, Tangen CM, Hussain MH, et al. Docetaxel and estramustine<br />
compared with mitoxantrone and prednisone for advanced refractory<br />
prostate cancer. N Engl J Med. 2004;351:1513-1520.<br />
3. Kant<strong>of</strong>f PW, Higano CS, Shore ND, et al. Sipuleucel-T immunotherapy<br />
for castration-resistant prostate cancer. N Engl J Med. 2010;363:411-422.<br />
4. de Bono JS, Oudard S, Ozguroglu M, et al. Prednisone plus cabazitaxel<br />
or mitoxantrone for metastatic castration-resistant prostate cancer progressing<br />
after docetaxel treatment: A randomised open-label trial. Lancet. 2010;<br />
376:1147-1154.<br />
5. de Bono JS, Logothetis CJ, Molina A, et al. Abiraterone and increased<br />
survival in metastatic prostate cancer. N Engl J Med. 2011;364:1995-2005.<br />
6. Parker C, Heinrich D, O’Sullivan JM, et al. Overall survival benefit <strong>of</strong><br />
radium-223 chloride (Alpharadin) in the treatment <strong>of</strong> patients with symptom-<br />
is one based on regulatory and not biologic concerns. Trials<br />
are ongoing to further assess both abiraterone and MDV-<br />
3100 in the predoctaxel space and perhaps there will be<br />
results from one <strong>of</strong> these trials by the <strong>2012</strong> ASCO Annual<br />
Meeting.<br />
All <strong>of</strong> the FDA-approved agents to date have been approved<br />
in a metastatic CRPC setting, yet many patients<br />
present to a physician with a rising prostate-specific antigen,<br />
a castrate level <strong>of</strong> testosterone, and no radiographic<br />
evidence <strong>of</strong> metastatic disease. What do you do in this case?<br />
The answer is not yet clear and there are no FDA-approved<br />
treatments. In my practice, a variety <strong>of</strong> secondary hormonal<br />
agents, such as nilutamide, bicalutamide, low-dose diethylstilbestrol,<br />
and ketoconazole are used despite no phase III<br />
evidence to support their use. Good clinical trials are always<br />
an important consideration as well.<br />
Stock<br />
Ownership Honoraria<br />
REFERENCES<br />
Research<br />
Funding<br />
San<strong>of</strong>i Algeta;<br />
AstraZeneca;<br />
Cougar<br />
Biotechnology;<br />
Exelixis;<br />
GlaxoSmithKline;<br />
Johnson &<br />
Johnson; San<strong>of</strong>i<br />
Expert<br />
Testimony<br />
Other<br />
Remuneration<br />
atic bone metastases in castration-resistant prostate cancer (CRPC): a phase<br />
III randomized trial (ALSYMPCA). Eur J Cancer. 2011;47 (suppl 2).<br />
7. Scher HI, Fizazi K, Saad F, et al. Effect <strong>of</strong> MDV3100, an androgen<br />
receptor signaling inhibitor (ARSI), on overall survival in patients with<br />
prostate cancer postdocetaxel: Results from the phase III AFFIRM study.<br />
J Clin Oncol. <strong>2012</strong>;30 (suppl 5; abstr LBA1).<br />
8. Perlroth DJ, Thompson SF, Luna Y, et al. Time to ADT and chemotherapy<br />
initiation for treatment <strong>of</strong> metastatic prostate cancer (mPC). J Clin<br />
Oncol. <strong>2012</strong>;30 (suppl 5; abstr 41).<br />
9. Attard G, Reid AH, Yap TA, et al. Phase I clinical trial <strong>of</strong> a selective<br />
inhibitor <strong>of</strong> CYP17, abiraterone acetate, confirms that castration-resistant<br />
prostate cancer commonly remains hormone driven. J Clin Oncol. 2008;26:<br />
4563-4571.<br />
10. Scher HI, Beer TM, Higano CS, et al. Antitumour activity <strong>of</strong> MDV3100<br />
in castration-resistant prostate cancer: A phase 1-2 study. Lancet. 2010;375:<br />
1437-1446.<br />
291
Prognostic, Predictive, and Surrogate Factors<br />
for Individualizing Treatment for Men with<br />
Castration-Resistant Prostate Cancer<br />
By Rhonda L. Bitting, MD, and Andrew J. Armstrong, MD, MSc<br />
Overview: With thesurge in therapeutic options for castrationresistant<br />
prostate cancer (CRPC) comes increasingly complicated<br />
treatment decision making, highlighting the need for<br />
biomarkers that can identify appropriate patients for specific<br />
WITH THE increasing number <strong>of</strong> U.S. Food and Drug<br />
Administration (FDA) approvals <strong>of</strong> new systemic<br />
agents for men with metastatic CRPC over the past 2 years,<br />
the optimal treatment strategy and sequence <strong>of</strong> therapies<br />
are evolving quickly. With novel immunotherapies, hormonal<br />
therapies, bone-targeted radioisotopes, bone microenvironment–targeted<br />
agents, and chemotherapies emerging<br />
in both the predocetaxel and postdocetaxel space, it becomes<br />
imperative to develop rational combinations and sequences<br />
<strong>of</strong> these agents, as well as to identify groups <strong>of</strong> men that are<br />
most likely to benefit from a particular treatment. Thus,<br />
there is an increasing need for biomarkers to help guide<br />
clinical decision making, especially considering the number<br />
<strong>of</strong> emerging agents currently in development and the high<br />
cost <strong>of</strong> these therapies. Here we discuss the role, context <strong>of</strong><br />
use, and the limitations <strong>of</strong> existing biomarkers in CRPC. We<br />
also examine approaches for the evaluation <strong>of</strong> novel biomarkers<br />
in clinical trials depending on the clinical or research<br />
context <strong>of</strong> use.<br />
A biomarker is defined as a “characteristic that is objectively<br />
measured and evaluated as an indicator <strong>of</strong> normal<br />
biologic processes, pathogenic processes, or pharmacologic<br />
responses to a therapeutic intervention.” 1 Biomarkers are<br />
intended to guide the treatment <strong>of</strong> patients, and they can be<br />
prognostic, predictive, pharmacodynamic, correlative, and/or<br />
surrogate in nature. A prognostic biomarker reflects a patient’s<br />
disease outcome independent <strong>of</strong> therapy (natural<br />
history), whereas a predictive biomarker identifies the likelihood<br />
<strong>of</strong> benefit from a specific therapy. 2 A number <strong>of</strong><br />
prognostic biomarkers have been reported in CRPC (Table<br />
1); however, there are no validated predictive biomarkers in<br />
this disease. Although predictive biomarkers exist in oncology<br />
(e.g., HER2 overexpression in breast cancer predicts for<br />
benefit with anti-HER2 agents, BRAF mutations predict for<br />
improved outcomes with BRAF inhibitors in melanoma,<br />
anaplastic lymphoma kinase [ALK] fusions predict for benefit<br />
to ALK tyrosine kinase inhibitors in lung adenocarcinoma),<br />
prostate cancer has lagged behind in the clinical<br />
application <strong>of</strong> predictive biomarkers in drug development.<br />
These predictive biomarkers <strong>of</strong>fer the hope <strong>of</strong> individualized<br />
approaches to therapy, breaking down a generally heterogeneous<br />
disease into more homogeneous molecularly defined<br />
subsets <strong>of</strong> patients in order to maximize benefit and minimize<br />
harm.<br />
In addition, surrogate biomarkers <strong>of</strong> overall survival (OS)<br />
remain uncertain in CRPC, in which measures such as<br />
progression-free survival (PFS), radiographic or pain responses,<br />
and prostate-specific antigen (PSA) declines have<br />
generally fallen short <strong>of</strong> a generalizable surrogate. A surrogate<br />
biomarker is a biomarker that can substitute for a<br />
292<br />
treatments and accurately assess disease response. Here we<br />
discuss existing and potential prognostic, predictive, and<br />
surrogate biomarkers in CRPC.<br />
clinically important endpoint such as OS but must meet<br />
several statistical criteria (such as Prentice’s criteria) and be<br />
validated across multiple trials examining agents with a<br />
range <strong>of</strong> mechanisms and clinical contexts. 3 Ongoing work<br />
in CRPC to identify optimal surrogate biomarkers, particularly<br />
in the fields <strong>of</strong> circulating tumor cell (CTC) characterization<br />
and optimal classifications <strong>of</strong> radiographic<br />
progression in CRPC, holds some promise in providing early<br />
evidence <strong>of</strong> activity for novel therapies and thus accelerated<br />
drug approvals and delivery <strong>of</strong> active agents into the clinic.<br />
However, much work is to be done on this front.<br />
Certain biomarkers are entirely used for determining<br />
drug mechanism or metabolism (pharmacogenomic or pharmacogenetic)<br />
and will not be further discussed in this<br />
review. These biomarkers are essential, however, in identifying<br />
pathophysiology and drug mechanism and in suggesting<br />
predictive and surrogate measures <strong>of</strong> clinical efficacy,<br />
and they are typically explored in the initial characterization,<br />
validation, and qualification <strong>of</strong> a biomarker. As biomarkers<br />
emerge in clinical settings, these predictive or<br />
surrogate factors can potentially guide treatment decision<br />
making (e.g., when to start a therapy, who to <strong>of</strong>fer a therapy<br />
to, when to stop a treatment and declare failure or response),<br />
underscoring the need for a rigorous approach to the development<br />
<strong>of</strong> predictive and surrogate biomarkers in parallel<br />
with drug development in CRPC.<br />
Prognostic Biomarkers in CRPC<br />
The sections below provide advantages, limitations, and<br />
evidence for use <strong>of</strong> common serologic and urinary biomarkers<br />
in CRPC. Imaging response, pain, and quality-<strong>of</strong>-life<br />
changes are not addressed here, though Table 1 provides a<br />
comprehensive list <strong>of</strong> prognostic biomarkers in CRPC.<br />
PSA<br />
Serum PSA <strong>of</strong>ten reflects the disease burden in men with<br />
CRPC and has been included in prognostic models as an<br />
independent risk factor for disease-related mortality. 4<br />
Changes in PSA over time during treatment may be informative<br />
regarding a patient’s response to therapy; however,<br />
this response is variable depending on the mechanism <strong>of</strong><br />
action <strong>of</strong> the therapy. For example, treatment with the<br />
immunotherapy sipuleucel-T improves survival without<br />
From the Duke Cancer Institute and the Duke Prostate Center, Durham, NC.<br />
Authors’ disclosures <strong>of</strong> potential conflicts <strong>of</strong> interest are found at the end <strong>of</strong> this article.<br />
Address reprint requests to Andrew Armstrong, MD, MSc, Duke University Medical<br />
Center 102002, Durham, NC 27710; email: andrew.armstrong@duke.edu..<br />
© <strong>2012</strong> by <strong>American</strong> <strong>Society</strong> <strong>of</strong> <strong>Clinical</strong> <strong>Oncology</strong>.<br />
1092-9118/10/1-10
INDIVIDUALIZING TREATMENT FOR MEN WITH CRPC<br />
changing PSA levels, whereas a 30% or greater PSA decline<br />
within 3 months <strong>of</strong> treatment initiation with the cytotoxic<br />
agent docetaxel correlates with a survival benefit in retrospective<br />
analyses. 5-7 However, there is no percentage <strong>of</strong> PSA<br />
decline that consistently correlates with survival across<br />
trials <strong>of</strong> a range <strong>of</strong> drug mechanisms, and therefore PSA<br />
decline is not a surrogate for survival in itself and should not<br />
be a primary endpoint for systemic therapy trials in CRPC<br />
or for regulatory approval. Prospective evaluation is needed<br />
for PSA declines in certain contexts where the correlation<br />
appears strongest, particularly for hormonal therapies. 8<br />
Likewise, PSA progression during therapy conveys a poor<br />
prognosis but is not a surrogate for survival. 9,10 Early but<br />
transient rises in PSA occur in up to 20% <strong>of</strong> men during<br />
chemotherapy but do not affect survival. 11 PSA declines can<br />
<strong>of</strong>ten lag or be slow to occur. Therefore, the Prostate Cancer<br />
Working Group (PCWG2) recommends that PSA change not<br />
be used in isolation to define disease progression or to stop<br />
therapy, particularly during the first several months <strong>of</strong><br />
systemic therapy. Regular PSA evaluation is recommended<br />
in clinical trials, however, so that it can be studied independently<br />
but in the context <strong>of</strong> other disease-related outcomes. 8<br />
Reporting <strong>of</strong> PSA changes in the form <strong>of</strong> descriptive water-<br />
KEY POINTS<br />
● The utility <strong>of</strong> a biomarker depends on the clinical<br />
question and its context <strong>of</strong> use.<br />
● Prostate-specific antigen (PSA), circulating tumor<br />
cell (CTC) count, markers <strong>of</strong> bone turnover, lactate<br />
dehydrogenase, and hemoglobin are traditional biomarkers<br />
that are correlated with overall survival in<br />
castration-resistant prostate cancer (CRPC) but are<br />
not predictive <strong>of</strong> benefit from specific therapies in<br />
CRPC.<br />
● Predictive biomarkers <strong>of</strong> sensitivity to systemic therapies<br />
in CRPC are needed and are likely to be based<br />
on the mechanism <strong>of</strong> action <strong>of</strong> a given agent (i.e.,<br />
androgen receptor activity and MDV3100, microtubule<br />
mutations and docetaxel sensitivity, androgen<br />
precursor levels and abiraterone sensitivity, phosphatase<br />
and tensin homolog loss or phosphoinositide<br />
3-kinase (PI3K) pathway activation and PI3K pathway<br />
inhibitors) and may be determined from tumor<br />
tissue or blood-/plasma-/urine-based assays such as<br />
CTCs or cell-free tumor DNA.<br />
● Improvements in PSA, CTC count, pain, radiographic<br />
tumor burden, and/or quality <strong>of</strong> life and delays in<br />
progression are associated with improved prognosis<br />
post-treatment, and work is ongoing to determine the<br />
association <strong>of</strong> these improvements with overall survival<br />
in patients with CRPC in a range <strong>of</strong> contexts.<br />
● Although there are no current validated surrogates<br />
for overall survival in CRPC, it is imperative that<br />
we incorporate intermediate endpoints and biomarkers<br />
into clinical trial design to develop potential<br />
surrogates <strong>of</strong> activity in order to accelerate drug<br />
development.<br />
Table 1. Prognostic Factors in CRPC a<br />
Baseline Prognostic Factors Post-Treatment Prognostic Factors Predictive Factors<br />
Performance status PSA decline None validated<br />
Visceral metastatic disease Pain improvement See Table 3<br />
Anemia Quality-<strong>of</strong>-life improvement<br />
Alkaline phosphatase and<br />
other bone turnover<br />
markers (e.g., TRAP)<br />
Change in CTC count to � 5<br />
PSA PSA PFS<br />
PSA kinetics Radiographic response PFS<br />
Type <strong>of</strong> progression (bone,<br />
measurable disease, PSA<br />
only)<br />
Induction <strong>of</strong> immunity to tumor<br />
antigens (sipuleucel-T)<br />
CTC count Skeletal-related event<br />
development<br />
LDH Development <strong>of</strong> anemia<br />
Albumin LDH changes<br />
Pain Type <strong>of</strong> progression<br />
Number <strong>of</strong> disease sites Alkaline phosphatase<br />
improvements<br />
Age<br />
VEGF levels<br />
Interleukin-6 levels<br />
Chromogranin-A<br />
C-reactive protein<br />
Bone turnover markers<br />
Gleason sum in primary<br />
Urinary N-telopeptide<br />
Abbreviations: CRPC, castration-resistant prostate cancer; CTC, circulating<br />
tumor cell; LDH, lactate dehydrogenase; PFS, progression-free survival; PSA,<br />
prostate-specific antigen; TRAP, tartrate-resistant acid phosphatase; VEGF,<br />
vascular endothelial growth factor.<br />
a Adapted from Armstrong and colleagues. 35<br />
fall plots provides a visual clue to the effect <strong>of</strong> a systemic<br />
agent on PSA. However, many agents in prostate cancer<br />
have demonstrated benefit without notable early PSA<br />
changes, particularly immunotherapies. 5<br />
PSA doubling time (PSADT) is prognostic for OS in<br />
CRPC. 4 In nonmetastatic CRPC, both the absolute PSA and<br />
the PSADT can identify men at high risk for early disease<br />
progression. In metastatic but asymptomatic CRPC, rapid<br />
PSA kinetics may suggest the need for more aggressive<br />
therapy, whereas reduction in PSADT with chemotherapy is<br />
suggestive <strong>of</strong> a better prognosis. 4,6 However, PSADT may<br />
also change over time without intervention, and changes in<br />
PSA kinetics have not been demonstrated to be surrogates <strong>of</strong><br />
OS broadly, so these findings must be interpreted cautiously.<br />
12<br />
In neuroendocrine or small cell prostate cancer, very little<br />
or no PSA is produced, and therefore PSA changes do not<br />
correlate with disease status. Chromogranin A levels are<br />
prognostic but not predictive in these cases, and thus other<br />
biomarkers such as CTCs will be needed in this emerging<br />
and virulent subset <strong>of</strong> CRPC. 13<br />
CTCs<br />
To establish distant metastases, invasive cancer cells<br />
likely circulate in the bloodstream and home in on their<br />
target <strong>of</strong> choice, which in CRPC is <strong>of</strong>ten bone. CellSearch<br />
(Veridex LLC <strong>of</strong> Raritan, NJ) is the only FDA-cleared<br />
technology for the detection <strong>of</strong> CTCs in patients with cancer.<br />
Using this technology, CTCs are defined as nucleated cells at<br />
least 4 microns in diameter, immunomagnetically captured<br />
from the bloodstream using antibodies against epithelial cell<br />
293
adhesion molecule and further characterized by cytokeratin<br />
expression and lack <strong>of</strong> the leukocyte marker CD45. 14 The<br />
CTC count is a prognostic biomarker in CRPC and in other<br />
solid tumors. Before the start <strong>of</strong> chemotherapy for CRPC,<br />
the detection <strong>of</strong> five or more CTCs is associated with an<br />
inferior OS compared to patients with less than five. Furthermore,<br />
a drop in CTCs to below five during treatment is<br />
associated with an improvement in survival. 15 Whether<br />
CTCs have a greater association with survival than PSA or<br />
radiographic changes over time is an important and unresolved<br />
question. 16<br />
Flares in the CTC count have not been reported, and<br />
changes in the CTC count <strong>of</strong>ten precede changes in PSA. 17<br />
Therefore, the CTC trend may be valuable for therapeutic<br />
decision making in cases where other clinical assessments <strong>of</strong><br />
response are equivocal, though its use in this setting is<br />
purely speculative. Prospective evaluation <strong>of</strong> whether the<br />
CTC count may act as a surrogate for OS is ongoing, but the<br />
initial study assessing its use as a surrogate in CRPC was<br />
promising. 18<br />
The potential use <strong>of</strong> CTCs as a biomarker is not restricted<br />
to the current CTC detection method and definition. One<br />
limitation <strong>of</strong> the CellSearch epithelial-based method is the<br />
lack <strong>of</strong> detection <strong>of</strong> CTCs in many men with progressive,<br />
metastatic CRPC. 17 Recent findings suggest that there is<br />
CTC phenotypic heterogeneity, with some CTCs expressing<br />
not only epithelial proteins, but also mesenchymal and<br />
stemness proteins. 19 Therefore, epithelial-mesenchymal<br />
transitions may explain the relative underdetection <strong>of</strong> CTCs<br />
in patients with advanced malignancy using the standard<br />
epithelial antigen-based technology. A number <strong>of</strong> technologies<br />
that employ nonepithelial targets for CTC capture and<br />
characterization are under development, though the prognostic<br />
and predictive implications <strong>of</strong> these CTC phenotypes<br />
must be independently and prospectively validated.<br />
Because CTCs may be a direct measurement <strong>of</strong> the underlying<br />
tumor biology, enhanced capture <strong>of</strong> CTCs may aid in<br />
the development <strong>of</strong> CTCs as a predictive biomarker. For<br />
example, the identification <strong>of</strong> phosphatase and tensin homolog<br />
(PTEN) loss and androgen receptor amplification in<br />
CTCs suggests that personalization <strong>of</strong> therapy using biomarker<br />
guidance is achievable. 21,22 As additional CTC phenotypes<br />
are discovered, potential therapeutic targets may<br />
emerge. Thus, CTCs may provide prognostic information<br />
through enumeration, but more importantly, they may provide<br />
a noninvasive window into tumor biology and provide<br />
predictive information for therapeutic decision making.<br />
Lactate Dehydrogenase<br />
The enzyme lactate dehydrogenase (LDH) converts pyruvate<br />
to lactate and vice versa as part <strong>of</strong> the normal glycolysis<br />
and gluconeogenesis pathways <strong>of</strong> the cell, and those pathways<br />
are preferentially upregulated in cancer cells as a<br />
result <strong>of</strong> oncogenic signaling. 22 LDH is an independent<br />
prognostic biomarker in CRPC and other tumor types, and<br />
elevations are thought to reflect the underlying tumor burden.<br />
23 Assessments <strong>of</strong> LDH in conjunction with other biomarkers<br />
such as PSA or CTCs may improve on the current<br />
clinical utility <strong>of</strong> LDH for risk stratification and prognostication.<br />
18 Also, although baseline LDH is strongly prognostic<br />
in multivariate models in CRPC, increases in LDH following<br />
therapy carry an unfavorable prognosis and may be useful in<br />
interpreting treatment response. 24 Given the strong associ-<br />
294<br />
ation <strong>of</strong> LDH with OS in CRPC over time, however, serial<br />
measurement and reporting <strong>of</strong> this factor during treatment<br />
and in the context <strong>of</strong> clinical trials are useful and recommended.<br />
Markers <strong>of</strong> Bone Turnover<br />
Prostate cancer commonly metastasizes to the bone, and<br />
this may be mediated by adhesion molecules that target the<br />
bone microenvironment and promote osteomimicry. 25 As<br />
such, agents that impede this tumor-bone stromal interaction<br />
such as zoledronic acid and denosumab delay the<br />
development <strong>of</strong> skeletal-related events in CRPC. 26,27 Bony<br />
metastatic effects can be indirectly measured using bone<br />
turnover markers such as N-telopeptide, tartrate-resistant<br />
acid phosphatase 5b, C-telopeptide, and osteopontin, and<br />
these are being evaluated for use as prognostic and predictive<br />
biomarkers. 28 Bone-derived alkaline phosphatase,<br />
which is a measure <strong>of</strong> osteoblastic activity, is an established<br />
prognostic biomarker in CRPC. 29 The reduction in total<br />
alkaline phosphatase with docetaxel is independently prognostic<br />
in CRPC and may provide evidence <strong>of</strong> a survival<br />
benefit even in the absence <strong>of</strong> decline in PSA or improvement<br />
in imaging. 30 Overall, the measurement <strong>of</strong> total or<br />
bone-derived alkaline phosphatase at baseline and over time<br />
provides prognostic information but does not predict response<br />
to bone-targeted therapy.<br />
Urinary N-telopeptide, which is a breakdown product <strong>of</strong><br />
type 1 collagen, is elevated in men with CRPC and bony<br />
metastases, and high N-telopeptide levels are associated<br />
with an increased risk <strong>of</strong> skeletal-related events, disease<br />
progression, and death. 31 Treatment with the receptor activator<br />
<strong>of</strong> nuclear factor kappa B (RANK)-ligand antagonist<br />
denosumab reduces the levels <strong>of</strong> N-telopeptide in the urine<br />
beyond the reduction seen with zoledronic acid, though<br />
whether this reduction correlates with denosumab’s superiority<br />
in preventing skeletal-related events is unclear. 27 The<br />
utility <strong>of</strong> N-telopeptide and other bone turnover markers as<br />
predictive biomarkers for the use <strong>of</strong> bone-targeted agents is<br />
an area <strong>of</strong> active investigation.<br />
Hemoglobin<br />
BITTING AND ARMSTRONG<br />
Anemia in the setting <strong>of</strong> CRPC is <strong>of</strong>ten multifactorial,<br />
related to androgen deprivation therapy, renal disease,<br />
chemotherapy toxicity, bone marrow infiltration or failure,<br />
or other comorbid conditions. Anemia has long been recognized<br />
as a prognostic factor in CRPC and, in multivariate<br />
analysis, is among the strongest predictors for docetaxelrelated<br />
PSA declines, tumor response, and OS for men with<br />
CRPC. 32 Thus, the measurement <strong>of</strong> hemoglobin over time<br />
has both practical management and prognostic implications<br />
in the treatment <strong>of</strong> men with CRPC.<br />
Post-treatment Prognostic and Potential Surrogate<br />
Endpoints in CRPC<br />
Although imaging measurements are performed routinely<br />
to assess disease status in oncology, the absolute size <strong>of</strong> a<br />
metastatic lesion does not always correlate with clinical<br />
outcome. The use <strong>of</strong> the Response Evaluation Criteria in<br />
Solid Tumors (RECIST) in CRPC is especially problematic,<br />
as RECIST does not account for skeletal lesions, the most<br />
common site <strong>of</strong> metastasis and source <strong>of</strong> morbidity in<br />
CRPC. 33 Bone scintigraphy is used to assess skeletal lesions;
INDIVIDUALIZING TREATMENT FOR MEN WITH CRPC<br />
Table 2. Post-treatment Prognostic and Potential Surrogate Endpoints in CRPC a<br />
Biomarker References Pros Cons<br />
PSA decline 6, 7 Easily measurable Not validated for use with novel agents<br />
Widely available PSA flares may occur<br />
Time � 3 mo Threshold <strong>of</strong> response unclear<br />
Evidence to support use with cytotoxic therapy Not useful with immunotherapy<br />
Subgroups <strong>of</strong> CRPC do not make PSA<br />
CTCs 15, 18, 19 Change before PSA Not present in all CRPC using CellSearch<br />
Tumor-specific Expensive<br />
Strongly prognostic Requires specialized lab<br />
Surrogacy use under evaluation Requires processing within 72 h<br />
May reflect current disease phenotype New CTC definitions and technologies require validation<br />
Bone turnover markers 30 Reflects tumor-stromal interaction Normal in patients without bony metastases<br />
Widely available May be normal if bony metastases<br />
Prognostic <strong>Clinical</strong> relevance <strong>of</strong> partial changes unclear<br />
Quality-<strong>of</strong>-life or pain improvement 6, 36 Direct patient measure Qualitative, requires validated scales<br />
Subjective and variable<br />
May be multifactorial<br />
Inherent biases<br />
Radiographic changes (RECIST 1.1, PCWG2) 8, 16, 34 Well-defined criteria for response Not useful if nonmetastatic disease<br />
Not always measurable (i.e., bone lesions)<br />
Only modestly prognostic<br />
Bone scan flare may occur<br />
Some agents (i.e., immunotherapy) may have benefit<br />
without affecting imaging<br />
PFS (PCWG2) 9, 10 May capture clinical benefit Exact definition is critical<br />
Can measure effect <strong>of</strong> cytostatic agents Not validated as a surrogate for OS<br />
Censorship prevents current surrogate analyses<br />
Abbreviations: CRPC, castration-resistant prostate cancer; CTC, circulating tumor cell; h, hours; mo, months; OS, overall survival; PFS, progression-free survival;<br />
PSA, prostate-specific antigen.<br />
a Adapted from Armstrong and colleagues. 35<br />
however, it is an indirect measurement that detects not only<br />
the cancer but also unassociated inflammation and degeneration.<br />
In fact, effective therapies sometimes cause what<br />
appear to be enlarging bony lesions but actually represent<br />
healing bone surrounding cancerous bone, known as the<br />
flare phenomenon or osteoblastic reactions. 34 To account for<br />
this, bone scan progression per PCWG2 criteria requires the<br />
confirmation <strong>of</strong> bony lesions on a subsequent scan. 8 Overall,<br />
however, there is only a modest correlation between bone<br />
scan changes and survival in CRPC to date. 9<br />
The time from the initiation <strong>of</strong> therapy or study entry to<br />
the documentation <strong>of</strong> disease progression is known as<br />
progression-free survival. Because PFS is generally measurable<br />
sooner than OS, has more independent events than OS,<br />
and is not confounded by subsequent therapeutic interventions,<br />
PFS is a convenient endpoint for use in clinical<br />
trials. 17 The major limitation <strong>of</strong> the PFS endpoint is that<br />
disease progression can be suggested by PSA alone, radiographic<br />
changes based on lesion size or the presence <strong>of</strong> new<br />
lesions, and/or symptomatology, and the clinical significance<br />
<strong>of</strong> each varies. The PCWG2 criteria for disease progression<br />
provides a consensus on the definition, but the relevance <strong>of</strong><br />
this outcome depends on the mechanism <strong>of</strong> action <strong>of</strong> the<br />
therapy. 8 Immunotherapy, for example, prolongs OS without<br />
affecting PFS, therefore, the use <strong>of</strong> PFS as a surrogate<br />
for survival must be considered in the context <strong>of</strong> the therapy<br />
in question and drug mechanism. 5 Given the poor correlation<br />
<strong>of</strong> PFS and OS across phase III trials in CRPC, we are<br />
faced with a conundrum <strong>of</strong> how to approve drugs based only<br />
on PFS benefit without knowing the full effect <strong>of</strong> a novel<br />
agent on OS. 4 Table 2 provides a list <strong>of</strong> potential surrogate<br />
biomarkers in CRPC as well as a brief summary <strong>of</strong> advantages<br />
and disadvantages for each.<br />
Predictive Biomarkers in CRPC<br />
As described above, there are a number <strong>of</strong> prognostic<br />
biomarkers in CRPC. However, there is a paucity <strong>of</strong> biomarkers<br />
under consideration for predictive use, and only<br />
CTCs are being evaluated in clinical trials for surrogacy use.<br />
In an era where costs <strong>of</strong> medical therapies are skyrocketing<br />
Table 3. Predictive Biomarkers under Consideration a<br />
Biomarker Potential CRPC Application<br />
CTC count Potential surrogate for OS with abiraterone,<br />
prognostic predocetaxel, may be mechanismindependent<br />
High urinary N-telopeptide Benefit from denosumab or zoledronic acid<br />
or bone turnover markers<br />
Androgen receptor splice<br />
variants<br />
Levels <strong>of</strong> androgen<br />
precursors<br />
Predict sensitivity or resistance to anti-androgens<br />
(e.g., MDV3100) or to abiraterone<br />
Predict benefit from androgen synthesis inhibitors<br />
(e.g., abiraterone acetate or TAK700)<br />
c-Met/HGF activity Enrich for benefit from c-met inhibitors (e.g.,<br />
cabozantinib)<br />
PTEN loss in CTCs or<br />
metastases<br />
Enrich for benefit from PI3 kinase pathway<br />
inhibitors<br />
Ras/raf mutations Potential benefit from ras pathway inhibitors (e.g.,<br />
sorafenib, vemurafenib)<br />
Tubulin mutations May predict resistance to microtubule-based<br />
therapies (e.g., docetaxel, cabazitaxel)<br />
DNA repair defects Enrich for benefit from PARP inhibitors<br />
Myc amplification Predict sensitivity to cell-cycle inhibitors<br />
(antiproliferation agents)<br />
Abbreviations: CRPC, castration-resistant prostate cancer; CTC, circulating<br />
tumor cell; HGF, hepatocyte growth factor; OS, overall survival; PARP, poly(adenosine<br />
diphosphate [ADP] ribose) polymerase; PTEN, phosphatase and tensin<br />
homolog.<br />
a Adapted from Armstrong and colleagues. 35<br />
295
and treatment options are expanding, the use <strong>of</strong> predictive<br />
biomarkers may allow providers to select or enrich for<br />
patients with CRPC who are most likely to benefit from a<br />
specific therapy. Similar to the benefit seen with HER2targeted<br />
therapy in HER2-amplified breast cancer, one<br />
would hypothesize that patients with CRPC with androgen<br />
receptor amplification are most likely to benefit from an<br />
androgen receptor inhibitor like MDV3100. Likewise, those<br />
with PTEN deletions in their tumor would be expected to<br />
benefit from treatment with a PI3 kinase inhibitor and those<br />
with c-Met overexpression would be expected to benefit from<br />
c-Met inhibition. As prostate cancer subtypes are increasingly<br />
identified, this biomarker-driven paradigm will allow<br />
for the customizing <strong>of</strong> therapeutic interventions. With the<br />
advancing technology for CTC detection and characterization,<br />
this molecular information may eventually be obtainable<br />
from a blood test. Because a biomarker must be<br />
evaluated in a series <strong>of</strong> clinical trials to generate evidence<br />
Authors’ Disclosures <strong>of</strong> Potential Conflicts <strong>of</strong> Interest<br />
for its use, it is imperative that investigators employ trial<br />
designs that allow for appropriate assessment <strong>of</strong> biomarkers.<br />
Table 3 lists predictive biomarkers under consideration<br />
and the evidence supporting their use in CRPC.<br />
Conclusion<br />
All biomarkers currently used clinically in CRPC have<br />
prognostic implications when measured before starting therapy<br />
but have not yet been established as predictive or<br />
surrogate markers in the on- or post-treatment settings.<br />
Ongoing randomized studies <strong>of</strong> active systemic therapies<br />
with prospectively embedded biomarker-based validation<br />
studies are needed before these can be used for definitive<br />
clinical decision making. It is imperative that biomarkers be<br />
studied rigorously in parallel with drug development, given<br />
the potential to maximize benefit and minimize harms and<br />
costs to individual men with CRPC and to society.<br />
Employment or<br />
Leadership Consultant or Stock<br />
Research<br />
Author<br />
Positions Advisory Role Ownership Honoraria Funding<br />
Rhonda L. Bitting Veridex, LLC<br />
Andrew J. Armstrong Amgen; Bristol-<br />
Amgen; Active Biotech;<br />
Myers Squibb;<br />
Dendreon; Bristol-Myers<br />
Novartis; San<strong>of</strong>i<br />
Johnson & Squibb;<br />
Johnson; Pfizer; Dendreon;<br />
San<strong>of</strong>i<br />
ImClone<br />
Systems;<br />
Johnson &<br />
Johnson;<br />
Medivation;<br />
Novartis; Pfizer;<br />
San<strong>of</strong>i<br />
1. Biomarkers Definitions Working Group. Biomarkers and surrogate<br />
endpoints: preferred definitions and conceptual framework. Clin Pharmacol<br />
Ther. 2001;69:89-95.<br />
2. Dancey JE, Dobbin KK, Groshen S, et al. Guidelines for the development<br />
and incorporation <strong>of</strong> biomarker studies in early clinical trials <strong>of</strong> novel agents.<br />
Clin Cancer Res. 2010;16:1745-1755.<br />
3. Prentice RL. Surrogate and mediating endpoints: current status and<br />
future directions. J Natl Cancer Inst. 2009;101:216-217.<br />
4. Armstrong AJ, Garrett-Mayer ES, Yang YC, et al. A contemporary<br />
prognostic nomogram for men with hormone-refractory metastatic prostate<br />
cancer: a TAX327 study analysis. Clin Cancer Res. 2007;13:6396-6403.<br />
5. Kant<strong>of</strong>f PW, Higano CS, Shore ND, et al. Sipuleucel-T immunotherapy<br />
for castration-resistant prostate cancer. N Engl J Med. 2010;363:411-422.<br />
6. Armstrong AJ, Garrett-Mayer E, Ou Yang YC, et al. Prostate-specific<br />
antigen and pain surrogacy analysis in metastatic hormone-refractory prostate<br />
cancer. J Clin Oncol. 2007;25:3965-3970.<br />
7. Petrylak DP, Ankerst DP, Jiang CS, et al. Evaluation <strong>of</strong> prostate-specific<br />
antigen declines for surrogacy in patients treated on SWOG 99-16. J Natl<br />
Cancer Inst. 2006;98:516-521.<br />
8. Scher HI, Halabi S, Tannock I, et al. Design and end points <strong>of</strong> clinical<br />
trials for patients with progressive prostate cancer and castrate levels <strong>of</strong><br />
testosterone: recommendations <strong>of</strong> the Prostate Cancer <strong>Clinical</strong> Trials Working<br />
Group. J Clin Oncol. 2008;26:1148-1159.<br />
9. Halabi S, Vogelzang, NJ, Ou SS, et al. Progression-free survival as a<br />
predictor <strong>of</strong> overall survival in men with castrate-resistant prostate cancer.<br />
J Clin Oncol. 2009;27:2766-2771.<br />
10. Hussain M, Goldman B, Tangen C, et al. Prostate-specific antigen<br />
progression predicts overall survival in patients with metastatic prostate<br />
cancer: data from Southwest <strong>Oncology</strong> Group Trials 9346 (Intergroup Study<br />
0162) and 9916. J Clin Oncol. 2009;27:2450-2456.<br />
11. Thuret R, Massard C, Gross-Goupil M, et al. The postchemotherapy<br />
PSA surge syndrome. Ann Oncol. 2008;19:1308-1311.<br />
12. Smith MR, Manola J, Kaufman DS, et al. Rosiglitazone versus placebo<br />
296<br />
REFERENCES<br />
BITTING AND ARMSTRONG<br />
Expert<br />
Testimony<br />
Other<br />
Remuneration<br />
for men with prostate carcinoma and a rising serum prostate-specific antigen<br />
level after radical prostatectomy and/or radiation therapy. Cancer. 2004;101:<br />
1569-1574.<br />
13. Taplin ME, George DJ, Halabi S, et al. Prognostic significance <strong>of</strong><br />
plasma chromogranin a levels in patients with hormone-refractory prostate<br />
cancer treated in Cancer and Leukemia Group B 9480 study. Urology.<br />
2005;66:386-391.<br />
14. Allard WJ, Matera J, Miller MC, et al. Tumor cells circulate in the<br />
peripheral blood <strong>of</strong> all major carcinomas but not in healthy subjects or<br />
patients with nonmalignant diseases. Clin Cancer Res. 2004;10:6897-6904.<br />
15. de Bono JS, Scher HI, Montgomery RB, et al. Circulating tumor cells<br />
predict survival benefit from treatment in metastatic castration-resistant<br />
prostate cancer. Clin Cancer Res. 2008;14:6302-6309.<br />
16. Sonpavde G, Pond GR, Berry WR, et al. The association between<br />
radiographic response and overall survival in men with metastatic castrationresistant<br />
prostate cancer receiving chemotherapy. Cancer. 2011;117:3963-<br />
3971.<br />
17. Scher HI, Morris MJ, Basch E, et al. End points and outcomes in<br />
castration-resistant prostate cancer: from clinical trials to clinical practice.<br />
J Clin Oncol. 2011;29:3695-3704.<br />
18. Scher HI, Heller G, Molina A, et al. Evaluation <strong>of</strong> circulating tumor cell<br />
(CTC) enumeration as an efficacy response biomarker <strong>of</strong> overall survival (OS)<br />
in metastatic castration-resistant prostate cancer (mCRPC): planned final<br />
analysis (FA) <strong>of</strong> COU-AA-301, a randomized double-blind, placebo-controlled<br />
phase III study <strong>of</strong> abiraterone acetate (AA) plus low-dose prednisone (P) post<br />
docetaxel. J Clin Oncol. 2011;29:293s (suppl; abstr LBA4517ˆ).<br />
19. Armstrong AJ, Marengo MS, Oltean S, et al. Circulating tumor cells<br />
from patients with advanced prostate and breast cancer display both epithelial<br />
and mesenchymal markers. Mol Cancer Res. 2011;9:997-1007.<br />
20. Attard G, Swennenhuis JF, Olmos D, et al. Characterization <strong>of</strong> ERG,<br />
AR and PTEN gene status in circulating tumor cells from patients with<br />
castration-resistant prostate cancer. Cancer Res. 2009;69:2912-2918.<br />
21. Shaffer DR, Leversha MA, Danila DC, et al. Circulating tumor cell
INDIVIDUALIZING TREATMENT FOR MEN WITH CRPC<br />
analysis in patients with progressive castration-resistant prostate cancer.<br />
Clin Cancer Res. 2007;13:2023-2029.<br />
22. Vander Heiden MG, Cantley LC, and Thompson CB. Understanding<br />
the Warburg effect: the metabolic requirements <strong>of</strong> cell proliferation. Science.<br />
2009;324:1029-1033.<br />
23. Halabi S, Small EJ, Kant<strong>of</strong>f PW, et al. Prognostic model for predicting<br />
survival in men with hormone-refractory metastatic prostate cancer. J Clin<br />
Oncol. 2003;21:1232-1237.<br />
24. Halabi S, Ou S, Vogelzang NJ, et al. The prognostic value <strong>of</strong> change in<br />
hemoglobin (HGB), LDH and PSA levels at 3 months from baseline in men<br />
with castrate recurrent prostate cancer. In ASCO Prostate Cancer Symposium<br />
Proceedings. Alexandria, VA: ASCO, 2007 (abstr 217).<br />
25. Chung LW, Huang WC, Sung SY, et al. Stromal-epithelial interaction<br />
in prostate cancer progression. Clin Genitourin Cancer. 2006;5:162-170.<br />
26. Saad F, Gleason DM, Murray R, et al. Long-term efficacy <strong>of</strong> zoledronic<br />
acid for the prevention <strong>of</strong> skeletal complications in patients with metastatic<br />
hormone-refractory prostate cancer. J Natl Cancer Inst. 2004;96:879-882.<br />
27. Fizazi K, Carducci M, Smith M, et al. Denosumab versus zoledronic<br />
acid for treatment <strong>of</strong> bone metastases in men with castration-resistant<br />
prostate cancer: a randomised, double-blind study. Lancet. 2011;377:813-822.<br />
28. Coleman RE, Major P, Lipton A, et al. Predictive value <strong>of</strong> bone<br />
resorption and formation markers in cancer patients with bone metastases<br />
receiving the bisphosphonate zoledronic acid. J Clin Oncol. 2005;23:4925-<br />
4935.<br />
29. Cook RJ, Coleman R, Brown J, et al. Markers <strong>of</strong> bone metabolism and<br />
survival in men with hormone-refractory metastatic prostate cancer. Clin<br />
Cancer Res. 2006;12:3361-3367.<br />
30. Sonpavde G, Pond GR, Berry WR, et al. Serum alkaline phosphatase<br />
changes predict survival independent <strong>of</strong> PSA changes in men with castrationresistant<br />
prostate cancer and bone metastasis receiving chemotherapy. Urol<br />
Oncol. Epub 2010 Sep 29.<br />
31. Brown JE, Cook RJ, Major P, et al. Bone turnover markers as<br />
predictors <strong>of</strong> skeletal complications in prostate cancer, lung cancer, and other<br />
solid tumors. J Natl Cancer Inst. 2005;97:59-69.<br />
32. Armstrong AJ, Tannock IF, de Wit R, et al. The development <strong>of</strong> risk<br />
groups in men with metastatic castration-resistant prostate cancer based on<br />
risk factors for PSA decline and survival. Eur J Cancer. 2010;46:517-525.<br />
33. Eisenhauer EA, Therasse P, Bogaerts J, et al. New response evaluation<br />
criteria in solid tumours: revised RECIST guideline (version 1.1). Eur J<br />
Cancer. 2009;45:228-247.<br />
34. Ryan CJ, Shah S, Efstathiou E, et al. Phase II study <strong>of</strong> abiraterone<br />
acetate in chemotherapy-naive metastatic castration-resistant prostate cancer<br />
displaying bone flare discordant with serologic response. Clin Cancer Res.<br />
2011;17:4854-4861.<br />
35. Armstrong AJ, Eisenberger MA, Halabi S, et al. Biomarkers in the<br />
management and treatment <strong>of</strong> men with metastatic castration-resistant<br />
prostate cancer. Eur Urol. <strong>2012</strong>;61:549-559.<br />
36. Berthold DR, Pond GR, Roessner M, et al. Treatment <strong>of</strong> hormonerefractory<br />
prostate cancer with docetaxel or mitoxantrone: relationships<br />
between prostate-specific antigen, pain, and quality <strong>of</strong> life response and<br />
survival in the TAX-327 study. Clin Cancer Res. 2008;14:2763-2767.<br />
297
KIDNEY CANCER BIOLOGY AND THERAPEUTICS:<br />
VEGFR, MTOR, AND BEYOND<br />
CHAIR<br />
Toni K. Choueiri, MD, MSc<br />
Dana-Farber Cancer Institute<br />
Boston, MA<br />
SPEAKERS<br />
William G. Kaelin, MD<br />
Dana-Farber Cancer Institute<br />
Boston, MA<br />
Daniel Y. C. Heng, MD, MPH<br />
Tom Baker Cancer Centre, University <strong>of</strong> Calgary<br />
Calgary, AB, Canada
The Evolving Landscape <strong>of</strong> Metastatic Renal<br />
Cell Carcinoma<br />
By Daniel Y. C. Heng, MD, MPH, and Toni K. Choueiri, MD, MSc<br />
Overview: The treatment paradigm in metastatic renal cell<br />
carcinoma (mRCC) has evolved over the last 5 years. There are<br />
now seven approved targeted therapies against the vascular<br />
endothelial growth factor (VEGF) and mammalian target <strong>of</strong><br />
rapamycin (mTOR) pathways. The use <strong>of</strong> targeted therapy,<br />
THE ELUCIDATION <strong>of</strong> the von Hippel Lindau (VHL)<br />
tumor suppression gene associated with the hereditary<br />
and sporadic forms <strong>of</strong> clear cell renal cell carcinomas (RCCs)<br />
has sparked a revolution <strong>of</strong> targeted therapy for this disease.<br />
The loss <strong>of</strong> VHL leads to downstream accumulation <strong>of</strong><br />
hypoxia inducible factor (HIF) and subsequent activation <strong>of</strong><br />
tumor promoting pathways including VEGF. 1 In fact, malignancies<br />
<strong>of</strong> the kidney show the greatest range and maximum<br />
expression <strong>of</strong> VEGF, 2 suggesting a rational target in<br />
this disease. mTOR is a serine threonine kinase that is<br />
another important central target in RCC, as it enhances<br />
translation <strong>of</strong> proteins involved in cellular growth, survival,<br />
and angiogenesis.<br />
Inhibitors <strong>of</strong> VEGF and mTOR have dominated the scene<br />
in the treatment <strong>of</strong> mRCC. These drugs have become commonplace<br />
in the treatment algorithm (Table 1) based on the<br />
registration phase III clinical trials.<br />
Choices <strong>of</strong> Targeted Therapy<br />
The list <strong>of</strong> choices for first-line targeted therapy is ever<br />
increasing. Oral VEGF tyrosine kinase inhibitors such as<br />
sunitinib 3 and pazopanib 4 and intravenous antibodies<br />
against VEGF such as bevacizumab 5,6 in combination with<br />
interferon-alpha have demonstrated a convincing<br />
progression-free survival (PFS) benefit in the first-line setting<br />
in patients with favorable or intermediate prognosis.<br />
Patients with poor prognosis benefit from temsirolimus<br />
(mTOR inhibitor), 7 as it is the only therapy to improve<br />
overall survival (OS) in the setting <strong>of</strong> a phase III trial in<br />
patients with poor prognosis.<br />
After failure <strong>of</strong> initial therapy, there are also choices to be<br />
made. On progression after immunotherapy, sorafenib has<br />
demonstrated a PFS benefit compared with placebo. 8 After<br />
failure <strong>of</strong> initial VEGF therapy, both everolimus (mTOR<br />
inhibitor) and axitinib (VEGF inhibitor) have also demonstrated<br />
a PFS benefit. Everolimus was examined in patients<br />
who progressed taking initial sunitinib, sorafenib, or both<br />
drugs and were randomly selected to take everolimus compared<br />
with placebo. The PFS for the everolimus versus<br />
placebo group was 4.9 versus 1.9 months (p � 0.0001),<br />
respectively. 9 Axitinib was recently approved by the FDA<br />
based on the registration phase III trial <strong>of</strong> axitinib compared<br />
with sorafenib in patients previously treated with a broad<br />
range <strong>of</strong> frontline therapies, including mostly prior sunitinib<br />
and cytokines. The response rates and PFS for the axitinib<br />
versus sorafenib group was 19% versus 9% and 6.7 versus<br />
4.7 months (p � 0.0001). 10<br />
Although there is no level I evidence for the drug <strong>of</strong> choice<br />
for patients in whom initial mTOR inhibitors failed, the use<br />
<strong>of</strong> a sequential targeted therapy not previously used remains<br />
a standard practice for these patients. This is also<br />
sequences, combinations, and investigational compounds will<br />
be discussed. Prognostic and predictive tools are detailed,<br />
although much work must be done to find predictive biomarkers<br />
in an effort to individualize therapy for patients.<br />
true for the choice <strong>of</strong> third-line targeted therapy, as there is<br />
no level I evidence to guide us in this setting.<br />
We are now faced with several targeted therapies to<br />
choose from in each treatment setting. As we do not have<br />
any robust, externally validated predictors <strong>of</strong> response to<br />
targeted therapy, the choice <strong>of</strong> targeted therapy is usually<br />
based on physician preference, intravenous versus oral therapy,<br />
reimbursement issues, and toxicity pr<strong>of</strong>ile. For example,<br />
patients with significant lung disease receiving oxygen<br />
or poorly controlled diabetes mellitus may not be optimal<br />
candidates for an mTOR inhibitor, as there is a risk <strong>of</strong><br />
noninfectious pneumonitis and hyperglycemia with this<br />
class <strong>of</strong> agents. Similarly, patients with refractory hypertension<br />
treated with several antihypertensive agents or severe<br />
cardiovascular (CV) morbidities may not initially choose a<br />
VEGF inhibitor, as these are known to increase the risk <strong>of</strong><br />
hypertension and CV events. These examples may be encountered<br />
in daily practice. However, no routine markers to<br />
predict response or toxicity to allow for a more informed<br />
decision about which targeted therapy to choose are available.<br />
Sequences <strong>of</strong> VEGF and mTOR inhibitors are being investigated<br />
in several randomized trials including the RECORD<br />
3 (NCT00903175), 404 study (NCT00474786), and START<br />
(NCT01217931) clinical trials (Table 2). These may help<br />
shed light onto the best sequence <strong>of</strong> drugs to use although<br />
they do not add to the personalized medicine approach.<br />
Different combinations <strong>of</strong> targeted therapy have also been<br />
studied, including the TORAVA trial <strong>of</strong> temsirolimus and<br />
bevacizumab in the frontline setting, which demonstrated<br />
only increased toxicity and no convincing evidence <strong>of</strong> added<br />
benefit. 11 Other combinations such as sunitinib and bevacizumab<br />
have proven to be toxic and have adverse effects such<br />
as thrombotic thrombocytopenic purpura, which is rarely<br />
seen when the agents are used alone. 11 A phase III CALGB<br />
trial investigating second-line use <strong>of</strong> everolimus compared<br />
with everolimus plus bevacizumab is underway with OS as<br />
the primary endpoint (NCT01198158). Unless substantial<br />
differences in durable complete responses or OS are documented,<br />
combination therapy remains investigational.<br />
From the Tom Baker Cancer Center, University <strong>of</strong> Calgary, Calgary, Alberta, Canada;<br />
Dana Farber Cancer Institute and Harvard Medical School, Boston, MA.<br />
Authors’ disclosures <strong>of</strong> potential conflicts <strong>of</strong> interest are found at the end <strong>of</strong> this article.<br />
Address reprint requests to Daniel Y. C. Heng, MD, MPH, Tom Baker Cancer Center,<br />
University <strong>of</strong> Calgary, Calgary, Alberta, Canada, 1331 – 29 th St. NW, Calgary, Alberta,<br />
Canada, T2N 4N2; email: daniel.heng@albertahealthservices.ca.<br />
© <strong>2012</strong> by <strong>American</strong> <strong>Society</strong> <strong>of</strong> <strong>Clinical</strong> <strong>Oncology</strong>.<br />
1092-9118/10/1-10<br />
299
Prognostic Factors in Advanced RCC<br />
Prognostic factors have been developed to assist in patient<br />
counseling, risk-directed treatment, and clinical trial design.<br />
These clinical factors that predict prognosis are a<br />
combination <strong>of</strong> patient factors, indicators <strong>of</strong> tumor burden,<br />
pro-inflammatory markers, and treatment-related factors<br />
including prior cytoreductive nephrectomy. 12<br />
Many multivariable models such as the Memorial Sloan-<br />
Kettering Cancer Center (MSKCC) model 13 have been created<br />
in an effort to stratify groups <strong>of</strong> patients with different<br />
biology. More recently, the International mRCC Database<br />
Consortium developed criteria from a large populationbased<br />
database <strong>of</strong> patients treated with VEGF targeted<br />
therapy. The independent predictors <strong>of</strong> poor OS include<br />
anemia, hypercalcemia, thrombocytosis, neutrophilia, a<br />
Karn<strong>of</strong>sky Performance Status <strong>of</strong> less than 80%, and a<br />
diagnosis-to-treatment interval <strong>of</strong> less than 1 year. Patients<br />
are then segregated into three risk categories: favorable risk<br />
(no risk factors, median OS not reached), intermediate risk<br />
(one to two risk factors, median OS <strong>of</strong> 27 months), and poor<br />
risk (three or more risk factors, median OS <strong>of</strong> 8.8 months). 14<br />
This multivariable model has been externally validated in<br />
an even more modern set <strong>of</strong> patients treated with VEGF<br />
targeted therapy with median OSs at 44 months, 21 months,<br />
and 8 months for the favorable, intermediate, and poor risk<br />
groups, respectively (p � 0.0001). This same model has also<br />
been validated in a set <strong>of</strong> patients who were previously<br />
treated with VEGF inhibitors and received a second line <strong>of</strong><br />
systemic therapy. 15 Data from these models suggest continued<br />
improvements in OS across different risk groups, which<br />
is a testament to the effectiveness <strong>of</strong> modern-day targeted<br />
therapy.<br />
Prediction <strong>of</strong> Response to Targeted Therapy: Steps<br />
Toward Personalized Medicine<br />
Currently there are no clinical factors or biomarkers that<br />
can conclusively predict which targeted therapies patients<br />
will respond to. There are some serum levels <strong>of</strong> proteins in<br />
KEY POINTS<br />
● There are now seven targeted therapies that are<br />
approved by the U.S. Food and Drug Administration<br />
for the treatment <strong>of</strong> metastatic renal cell carcinoma<br />
and are included in the treatment algorithm. The<br />
choice <strong>of</strong> drug is dependent on the context <strong>of</strong> the<br />
corresponding clinical trial, physician experience,<br />
drug availability, and patient preference.<br />
● Currently there are no standard, routinely used baseline<br />
biomarkers to predict response and toxicities.<br />
Recent studies have found that the development <strong>of</strong><br />
hypertension as well as other therapy-related toxicities<br />
and single nucleotide polymorphisms may indicate<br />
improved patient outcomes, but using these<br />
potential biomarkers still requires prospective validation.<br />
● There are new agents on the horizon, including<br />
more targeted therapies and next-generation<br />
immunotherapy.<br />
300<br />
Table 1. A Proposed Treatment Algorithm for Patients<br />
with mRCC<br />
Setting Patients<br />
First-Line Therapy Good or intermediate<br />
risk<br />
Second-Line<br />
Therapy<br />
Third-Line<br />
Therapy<br />
Therapies with<br />
Level 1 Evidence Other Options<br />
Sunitinib<br />
Pazopanib<br />
Bevacizumab<br />
� IFN<br />
High dose IL-2 in<br />
highly select patients<br />
Sorafenib<br />
<strong>Clinical</strong> trial<br />
Observation in select<br />
patients<br />
Poor risk Temsirolimus Other VEGF inhibitors<br />
<strong>Clinical</strong> trial<br />
Prior cytokines Sorafenib Sunitinib<br />
Axitinib<br />
Pazopanib<br />
<strong>Clinical</strong> trial<br />
Prior VEGF Axitinib Targeted therapy not<br />
Everolimus previously used<br />
<strong>Clinical</strong> trial<br />
Prior mTOR No data<br />
available<br />
Any No data<br />
available<br />
HENG AND CHOUEIRI<br />
Targeted therapy not<br />
previously used<br />
<strong>Clinical</strong> trial<br />
Targeted therapy not<br />
previously used<br />
<strong>Clinical</strong> trial<br />
Abbreviations: IFN, interferon; IL, interleukin; mRCC, metastatic renal cell<br />
carcinoma; mTOR, mammalian target <strong>of</strong> rapamycin; VEGF, vascular endothelial<br />
growth factor.<br />
the angiogenesis pathway that may be prognostic <strong>of</strong> OS, but<br />
predictive markers <strong>of</strong> response remain elusive. 12 Single<br />
nucleotide polymorphisms (SNPs) are single-base pair<br />
changes within a gene that may or may not affect gene<br />
function, and many have been explored for their prognostic<br />
or predictive value.<br />
For patients treated with pazopanib, SNPs in two<br />
interleukin-8 and HIF1A loci were associated with a significant<br />
difference in PFS whereas SNPs in HIF1A, NR1/2, and<br />
three VEGFA loci were associated with overall response<br />
rates. 16 For patients treated with sunitinib, 136 patients<br />
with clear cell mRCC were examined to determine a favorable<br />
genetic pr<strong>of</strong>ile, which was found to include an A allele in<br />
the CYP3A5 6986A/G loci, an absent CAT copy in the NR1/3<br />
haplotype, and a TCG copy in the ABCB1 haplotype. Patients<br />
with this favorable pr<strong>of</strong>ile had an improved PFS and<br />
OS compared to those without. 17 A VEGF SNP in a different<br />
loci was found to be associated with sunitinib-induced hypertension<br />
and another VEGF SNP and VEGF Receptor-2<br />
SNP were found to be together associated with OS. 18 Another<br />
study found two VEGF Receptor-3 SNPs to be associated<br />
with PFS when treated with sunitinib. 19 These results<br />
are interesting but are currently restricted to the caucasian<br />
population, as there are substantial racial differences in<br />
SNPs. These SNPs require further prospective evaluation<br />
while ensuring correction for multiple testing to see whether<br />
incorporating them in the decision-making process <strong>of</strong> choosing<br />
targeted therapy for an individual patient improves<br />
outcome.<br />
Recently, studies <strong>of</strong> toxicities due to targeted therapy have<br />
demonstrated better treatment outcomes when toxicity is<br />
encountered. For example, the development <strong>of</strong> hypertension<br />
during the first cycle <strong>of</strong> sunitinib treatment was associated<br />
with a better overall response rate, PFS, and OS. 20 Similar<br />
findings, including fatigue and hand-foot syndrome being
METASTATIC RENAL CELL CARCINOMA<br />
Table 2. Selected Ongoing mRCC <strong>Clinical</strong> Trials Investigating Sequencing, Combinations, and New Drugs<br />
Sequencing RECORD 3 Randomized Phase II<br />
NCT00903175<br />
Trials Population Randomization Primary Endpoint<br />
START Trial Phase III NCT01217931 Treatment naive mRCC with<br />
nephrectomy<br />
associated with better outcomes, have been shown. These<br />
are important associations; however, clinicians should<br />
not discontinue targeted therapy if toxicities do not develop<br />
because the discriminatory value and accuracy in determining<br />
hypertension or other toxicities is unknown.<br />
Additionally, these biomarkers are only helpful once the<br />
administration <strong>of</strong> the drug has already started and<br />
thus do not help the clinician choose which drug to use<br />
initially.<br />
Exploring New Agents and Mechanisms <strong>of</strong> Action<br />
Newer agents are on the horizon to expand the treatment<br />
armamentarium in advanced RCC. They are not yet FDA<br />
approved at this time. A more potent and specific VEGF<br />
tyrosine kinase inhibitor Tivozanib (AV-951) is expected to<br />
report PFS endpoints against sorafenib in the treatment<br />
naive setting (Table 2). Dovitinib (TKI258) is a fibroblast<br />
growth factor (FGF) inhibitor as well as a VEGF inhibitor<br />
and is currently being studied as part <strong>of</strong> the GOLD trial<br />
against sorafenib in third-line therapy after failure <strong>of</strong> one<br />
VEGF and one mTOR inhibitor. The FGF pathway is hypothesized<br />
to be an angiogenic escape mechanism that is<br />
upregulated when the tumor develops resistance against our<br />
current VEGF and mTOR inhibitors. Dovitinib will test the<br />
hypothesis that targeting the FGF angiogenic escape mechanism<br />
will lead to further responses and prolongation <strong>of</strong><br />
survival.<br />
Next-generation immunotherapy is being investigated in<br />
Treatment naive mRCC Sunitinib 3 everolimus vs.<br />
everolimus 3 sunitinib<br />
Pazopanib 3 bevacizumab vs.<br />
pazopanib 3 everolimus vs.<br />
everolimus 3 bevacizumab<br />
vs. everolimus 3 pazopanib<br />
vs. bevacizumab 3<br />
pazopanib vs. bevacizumab<br />
3 everolimus<br />
PFS <strong>of</strong> first-line treatment<br />
(finished accrual)<br />
Time to treatment failure <strong>of</strong> a<br />
sequence<br />
404 Study Phase III NCT00474786 mRCC refractory to sunitinib Temsirolimus vs. sorafenib PFS <strong>of</strong> second-line treatment<br />
(finished accrual)<br />
Combinations BEST Trial Phase III NCT 00378703 Treatment naive clear cell<br />
predominant mRCC<br />
CALGB 90,802 Phase III<br />
NCT01198158<br />
New Drugs ADAPT Trial AGS-003 (Autologous<br />
Dendritic Cell Immunotherapy)<br />
Phase III<br />
BMS936558 (PD-1 inhibitor) Phase<br />
II (phase III trial upcoming)<br />
NCT01354431<br />
Dovitinib (TKI258) (FGF and VEGF<br />
inhibitor) Phase III NCT01030783<br />
Tivozanib (AV951) (potent and<br />
specific VEGF inhibitor) Phase III<br />
NCT01030783<br />
mRCC with one prior VEGF<br />
inhibitor<br />
Bevacizumab vs. temsirolimus vs.<br />
bevacizumab � sorafenib vs.<br />
temsirolimus � sorafenib<br />
Everolimus � bevaciuzmab vs.<br />
everolimus<br />
Treatment naive mRCC AGS-003 with sunitinib/standard<br />
therapy vs. sunitinib/standard<br />
therapy<br />
mRCC with at least one prior<br />
therapy<br />
mRCC with clear cell component<br />
with one prior VEGF inhibitor<br />
and one prior mTOR inhibitor<br />
mRCC with clear cell component<br />
and previous nephrectomy<br />
PFS <strong>of</strong> first-line treatment<br />
BMS936558 0.3 mg/kg q 3 PFS<br />
weekly vs. BMS936558 2<br />
mg/kg q 3 weekly vs.<br />
BMS936558 10 mg/kg<br />
q 3 weekly<br />
Dovitinib vs. sorafenib PFS <strong>of</strong> third-line treatment<br />
Tivozanib vs. sorafenib PFS (data to be presented)<br />
Abbreviations: FGF, fibroblast growth factor; mRCC, metastatic renal cell carcinoma; mTOR, mammalian target <strong>of</strong> rapamycin; OS, overall survival; PD-1, programmed<br />
death � 1; PFS, progression-free survival; VEGF, vascular endothelial growth factor.<br />
mRCC. Programmed death-1 (PD-1) is a member <strong>of</strong> the<br />
immunoglobulin gene family and is expressed after T-cell<br />
activation to inhibit T-cell receptor signaling as a way to<br />
regulate the T-cell response. Higher preoperative soluble<br />
PD-1 ligand levels in patients with RCC were associated<br />
with larger tumors (p � 0.001), tumors <strong>of</strong> advanced stage<br />
(p � 0.017), grade (p � 0.044), and tumors with necrosis (p �<br />
0.003). A doubling <strong>of</strong> these levels was associated with a 41%<br />
increased risk <strong>of</strong> death (p � 0.010). 21 Thus, not only is the<br />
upregulation <strong>of</strong> the PD-1 mechanism a potential prognostic<br />
factor, but it has become a potential target for nextgeneration<br />
immune-based targeted therapy.<br />
BMS 936558 is a PD-1 inhibitor and was studied in a<br />
larger phase I multicenter trial where 126 patients (18 with<br />
RCC) were treated with several dose levels. 22 Across all<br />
doses, the most common adverse events grades 3 to 4 were<br />
fatigue (6.3%) and diarrhea (0.8%). One patient died with<br />
sepsis while being treated for drug-related grade 4 pneumonitis.<br />
There were 18 patients with mRCC, 16 <strong>of</strong> whom were<br />
treated with the 10 mg/kg dose. The objective investigator–<br />
assessed overall response rate was 31.2% (5 out <strong>of</strong> 16). The<br />
median duration <strong>of</strong> treatment was 7.6 months and the<br />
median duration <strong>of</strong> response was 4.0 months. Thus, a phase<br />
II dose-varying trial for mRCC is underway (Table 2), with a<br />
phase III trial anticipated shortly. This drug is promising<br />
but remains investigational at this time.<br />
To further immunotherapy and personalized medicine,<br />
AGS-003 is being investigated. It is an autologous dendritic<br />
OS<br />
OS<br />
301
cell immunotherapy in which a small tumor specimen isolated<br />
from the patient during nephrectomy or metastatectomy<br />
is taken along with a leukopheresis sample <strong>of</strong><br />
monocytes that differentiate into dendritic cells. They are<br />
coelectroporated with the RCC and CD40L RNA and then<br />
eventually injected back into the patient. Recently, AGS-003<br />
was studied along with sunitinib in a phase II trial <strong>of</strong> 21<br />
patients with no grade 3 or 4 adverse events. The overall<br />
response rate was 38% and the median PFS was 11.2<br />
months. 23 Because almost half <strong>of</strong> the patients had poor<br />
prognostic pr<strong>of</strong>iles, 14 this PFS is encouraging and warrants<br />
further study. Thus, a phase III randomized trial <strong>of</strong><br />
Authors’ Disclosures <strong>of</strong> Potential Conflicts <strong>of</strong> Interest<br />
Author<br />
Employment or<br />
Leadership<br />
Positions<br />
Consultant or<br />
Advisory Role<br />
Daniel Y. C. Heng Bayer; Novartis;<br />
Pfizer<br />
Toni K. Choueiri Bayer;<br />
GlaxoSmithKline;<br />
Novartis; Pfizer<br />
1. Rini BI, Small EJ. Biology and clinical development <strong>of</strong> vascular endothelial<br />
growth factor-targeted therapy in renal cell carcinoma. J Clin Oncol.<br />
2005;23:1028-1043.<br />
2. Jubb AM, Pham TQ, Hanby AM, et al. Expression <strong>of</strong> vascular endothelial<br />
growth factor, hypoxia inducible factor 1alpha, and carbonic anhydrase IX in<br />
human tumours. J Clin Pathol. 2004;57:504-512.<br />
3. Motzer RJ, Hutson TE, Tomczak P, et al. Sunitinib versus interferon alfa<br />
in metastatic renal-cell carcinoma. N Engl J Med. 2007;356:115-124.<br />
4. Sternberg CN, Davis ID, Mardiak J, et al. Pazopanib in locally advanced<br />
or metastatic renal cell carcinoma: results <strong>of</strong> a randomized phase III trial.<br />
J Clin Oncol. 2010;28:1061-1068.<br />
5. Rini BI, Halabi S, Rosenberg JE, et al. Phase III trial <strong>of</strong> bevacizumab<br />
plus interferon alfa versus interferon alfa monotherapy in patients with<br />
metastatic renal cell carcinoma: final results <strong>of</strong> CALGB 90206. J Clin Oncol.<br />
2010;28:2137-2143.<br />
6. Escudier BJ, Bellmunt J, Negrier S, et al. Final results <strong>of</strong> the phase III,<br />
randomized, double-blind AVOREN trial <strong>of</strong> first-line bevacizumab (BEV) �<br />
interferon-�2a (IFN) in metastatic renal cell carcinoma (mRCC). J Clin Oncol.<br />
2009;27:239s (suppl; abstr 5020).<br />
7. Hudes G, Carducci M, Tomczak P, et al. Temsirolimus, interferon alfa, or<br />
both for advanced renal-cell carcinoma. N Engl J Med. 2007;356:2271-2281.<br />
8. Escudier B, Eisen T, Stadler WM, et al. Sorafenib in advanced clear-cell<br />
renal-cell carcinoma. N Engl J Med. 2007;356:125-134.<br />
9. Motzer RJ, Escudier B, Oudard S, et al. Efficacy <strong>of</strong> everolimus in<br />
advanced renal cell carcinoma: a double-blind, randomised, placebocontrolled<br />
phase III trial. Lancet. 2008;372:449-456.<br />
10. Rini BI, Escudier B, Tomczak P, et al. Comparative effectiveness <strong>of</strong><br />
axitinib versus sorafenib in advanced renal cell carcinoma (AXIS): a randomised<br />
phase 3 trial. Lancet. 2011;378:1931-1939.<br />
11. Negrier S, Gravis G, Perol D, et al. Temsirolimus and bevacizumab, or<br />
sunitinib, or interferon alfa and bevacizumab for patients with advanced<br />
renal cell carcinoma (TORAVA): a randomised phase 2 trial. Lancet Oncol.<br />
2011;12:673-680.<br />
12. Tang PA, Vickers MM, Heng DY. <strong>Clinical</strong> and molecular prognostic<br />
factors in renal cell carcinoma: what we know so far. Hematol Oncol Clin<br />
North Am. 2011;25:871-891.<br />
13. Motzer RJ, Bacik J, Murphy BA, et al. Interferon-alfa as a comparative<br />
302<br />
AGS-003 with sunitinib/standard therapy compared with<br />
sunitinib/standard therapy is open and enrolling patients.<br />
Conclusion<br />
The treatment <strong>of</strong> mRCC has certainly evolved over the<br />
last 5 years, with more and more treatments available.<br />
There is an urgent need for biomarkers to help clinicians<br />
select which drug is most suitable for each specific patient.<br />
New drugs with different mechanisms <strong>of</strong> action are currently<br />
being investigated, making this an exciting time in<br />
the realm <strong>of</strong> mRCC research.<br />
Stock<br />
Ownership Honoraria<br />
REFERENCES<br />
Research<br />
Funding<br />
Pfizer<br />
Expert<br />
Testimony<br />
HENG AND CHOUEIRI<br />
Other<br />
Remuneration<br />
treatment for clinical trials <strong>of</strong> new therapies against advanced renal cell<br />
carcinoma. J Clin Oncol. 2002;20:289-296.<br />
14. Heng DY, Xie W, Regan MM, et al. Prognostic factors for overall<br />
survival in patients with metastatic renal cell carcinoma treated with<br />
vascular endothelial growth factor-targeted agents: results from a large,<br />
multicenter study. J Clin Oncol. 2009;27:5794-5799.<br />
15. Heng DY, Xie W, Bjarnason GA, et al. A unified prognostic model for<br />
first- and second-line targeted therapy in metastatic renal cell carcinoma<br />
(mRCC): results from a large international study. J Clin Oncol. 2010;28:347s<br />
(suppl; abstr 4523).<br />
16. Xu CF, Bing NX, Ball HA, et al. Pazopanib efficacy in renal cell<br />
carcinoma: evidence for predictive genetic markers in angiogenesis-related<br />
and exposure-related genes. J Clin Oncol. 2011;29:2557-2564.<br />
17. van der Veldt AA, Eechoute K, Gelderblom H, et al. Genetic polymorphisms<br />
associated with a prolonged progression-free survival in patients with<br />
metastatic renal cell cancer treated with sunitinib. Clin Cancer Res. 2011;17:<br />
620-629.<br />
18. Kim JJ, Vaziri SA, Rini BI, et al. Association <strong>of</strong> VEGF and VEGFR2<br />
single nucleotide polymorphisms with hypertension and clinical outcome in<br />
metastatic clear cell renal cell carcinoma patients treated with sunitinib.<br />
Cancer. Epub 2011 Aug 31.<br />
19. Garcia-Donas J, Esteban E, Leandro-Garcia LJ, et al. Single nucleotide<br />
polymorphism associations with response and toxic effects in patients with<br />
advanced renal-cell carcinoma treated with first-line sunitinib: a multicentre,<br />
observational, prospective study. Lancet Oncol. 2011;12:1143-1150.<br />
20. Rini BI, Cohen DP, Lu DR, et al. Hypertension as a biomarker <strong>of</strong><br />
efficacy in patients with metastatic renal cell carcinoma treated with<br />
sunitinib. J Natl Cancer Inst. 2011;103:763-773.<br />
21. Frigola X, Inman BA, Lohse CM, et al. Identification <strong>of</strong> a soluble form<br />
<strong>of</strong> B7-H1 that retains immunosuppressive activity and is associated with<br />
aggressive renal cell carcinoma. Clin Cancer Res. 2011;17:1915-1923.<br />
22. McDermott DF, Drake CG, Sznol M, et al. A phase I study to evaluate<br />
safety and antitumor activity <strong>of</strong> biweekly BMS-936558 (anti-PD-1, MDX-<br />
1106/ONO-4538) in patients with RCC and other advanced refractory malignancies.<br />
J Clin Oncol. 2011;29 (suppl; abstr 331).<br />
23. Figlin RA, Amin A, Dudek A, et al. Phase II study combining personalized<br />
dendritic cell (DC)-based therapy, AGS-003, with sunitinib in metastatic<br />
renal cell carcinoma (mRCC). J Clin Oncol. <strong>2012</strong>;30 (suppl; abstr 348).
RECENT INNOVATIONS IN THE MANAGEMENT OF<br />
GENITOURINARY CANCERS (eQ&A)<br />
CHAIR<br />
Matt D. Galsky, MD<br />
The Tisch Cancer Institute, Mount Sinai School <strong>of</strong> Medicine<br />
New York, NY<br />
SPEAKERS<br />
Elisabeth I. Heath, MD<br />
Karmanos Cancer Institute<br />
Detroit, MI<br />
Robert John Motzer, MD<br />
Memorial Sloan-Kettering Cancer Center<br />
New York, NY
New Developments in Urothelial Cancer<br />
Overview: Urothelial cancer is among the most chemotherapysensitive<br />
neoplasms <strong>of</strong> all the solid tumors. However, for the<br />
majority <strong>of</strong> patients with advanced disease, response durations<br />
with conventional treatment are relatively short. Secondline<br />
systemic treatment regimens are associated with modest<br />
response rates and poor outcomes. Trials in both the first- and<br />
second-line settings have demonstrated that a ceiling in<br />
efficacy has likely been reached with cytotoxic drugs, particularly<br />
in unselected patient populations. Promising areas <strong>of</strong><br />
EACH YEAR, approximately 56,000 patients in the<br />
United States will be diagnosed with urothelial cancer<br />
and more than 10,000 patients will succumb to the disease. 1<br />
Although the majority <strong>of</strong> patients present with clinically<br />
localized disease, a substantial proportion will ultimately<br />
develop metastatic recurrence whereas a smaller percentage<br />
will have metastatic disease at the time <strong>of</strong> initial presentation.<br />
Once metastasis occurs, the median survival for patients<br />
with urothelial cancer is approximately 1 to 2 years.<br />
In an attempt to improve outcomes in advanced disease,<br />
research efforts have traditionally focused on the development<br />
<strong>of</strong> combination cytotoxic regimens, and more recently<br />
have begun to unravel the molecular pathogenesis <strong>of</strong> urothelial<br />
cancer in an effort to target tumor-specific genetic<br />
aberrations.<br />
Chemotherapy in Advanced Urothelial Cancer<br />
The Development <strong>of</strong> MVAC<br />
During the 1980s, multiagent chemotherapeutic regimens<br />
were developed in an attempt to exploit non–cross-resistant<br />
cytotoxics with somewhat nonoverlapping toxicity pr<strong>of</strong>iles.<br />
In 1985, a landmark trial reported the initial experience<br />
with the MVAC (methotrexate, vinblastine, doxorubicin,<br />
and cisplatin) regimen. 2 Remarkably, <strong>of</strong> the 24 patients<br />
enrolled, responses were observed in 71% (95% CI, 53% to<br />
89%), with complete clinical responses in 50% (95% CI, 30%<br />
to 70%). Subsequent studies confirmed the activity <strong>of</strong> MVAC<br />
in larger cohorts, albeit with slightly lower response proportions,<br />
and established MVAC as a standard <strong>of</strong> care. Attempts<br />
to improve on the results with MVAC have included<br />
dose-dense administration with growth factor support. In a<br />
randomized phase III trial <strong>of</strong> dose-dense compared with<br />
conventional administration <strong>of</strong> MVAC, the former regimen<br />
failed to detect a meaningful improvement in median survival,<br />
but did result in an improvement in 5-year survival<br />
rate (21.8% [95% CI, 14.5% to 29.2%] compared with 13.5%<br />
[95% CI, 7.4% to 19.6%; p � 0.042]). 3<br />
Newer-Generation Cytotoxic Agents<br />
During the 1990s, several newer cytotoxics, including the<br />
taxanes and gemcitabine, were explored in urothelial cancer.<br />
Only a few <strong>of</strong> these regimens demonstrated sufficient<br />
activity to advance to randomized phase III trials (Table 1).<br />
These trials demonstrated superior outcomes with MVAC<br />
compared with docetaxel plus cisplatin and uninterpretable<br />
results from a comparison <strong>of</strong> MVAC with paclitaxel plus<br />
carboplatin, as a result <strong>of</strong> early study closure. 4-5 The results<br />
<strong>of</strong> a randomized trial <strong>of</strong> gemcitabine plus cisplatin (GC)<br />
304<br />
By Matthew D. Galsky, MD<br />
investigation include integrating predictive biomarkers to<br />
optimize patient selection for specific therapies, disrupting<br />
driving oncogenomic mutations, and associated signaling<br />
pathways and cotargeting both tumor and the immune system<br />
or tumor stroma. In addition, expanded sources <strong>of</strong> evidence<br />
generation are <strong>of</strong> interest in an effort to refine treatment for<br />
the general population <strong>of</strong> patients with advanced urothelial<br />
cancer, not only those who meet the narrow eligibility criteria<br />
used in most clinical trials.<br />
compared with MVAC, however, did alter the standard <strong>of</strong><br />
care. 6 Although not designed as a noninferiority trial, this<br />
study demonstrated similar antitumor outcomes, yet a favorable<br />
toxicity pr<strong>of</strong>ile, with GC compared with MVAC.<br />
Since the establishment <strong>of</strong> GC as a treatment standard in<br />
advanced urothelial carcinoma, there have only been two<br />
completed phase III trials attempting to improve on this<br />
therapy. The addition <strong>of</strong> paclitaxel to the GC regimen did<br />
not improve median survival in patients with advanced<br />
urothelial carcinoma. 7 Bamias and colleagues performed a<br />
randomized phase III trial <strong>of</strong> dose-dense MVAC (n � 118)<br />
compared with dose-dense GC (gemcitabine 2,500 mg/m 2<br />
plus cisplatin 70 mg/m 2 administered every 2 weeks, n �<br />
57). 8 As a result <strong>of</strong> poor accrual, the MVAC arm was<br />
supplemented with additional patients who received MVAC<br />
<strong>of</strong>f-study, leading to difficulty interpreting the overall study<br />
results. The dose-dense GC arm demonstrated similar outcomes<br />
compared with the dose-dense MVAC arm (overall<br />
response rate, 47.4% compared with 47.4%, p � 0.9; median<br />
survival, 18.4 months compared with 20.7 months, p � 0.7).<br />
However, the methodologic issues with this study have at<br />
least, in part, limited the penetration <strong>of</strong> dose-dense GC into<br />
routine practice.<br />
Second-Line Chemotherapy<br />
There have been only two completed phase III trials<br />
evaluating chemotherapeutic regimens for patients experiencing<br />
disease progression despite first-line treatment. Bellmunt<br />
and colleagues randomly assigned 370 patients with<br />
progressive urothelial cancer after first-line platinum-based<br />
chemotherapy to vinflunine compared with best supportive<br />
care. 9 In the intent-to-treat population, treatment with<br />
vinflunine did not result in a substantial improvement in<br />
survival. However, when only eligible patients were analyzed,<br />
there was a substantially longer median survival with<br />
vinflunine, resulting in marketing authorization by the<br />
Committee for Medicinal Products for Human Use <strong>of</strong> the<br />
European Medicines Agency.<br />
The German Association <strong>of</strong> Urological <strong>Oncology</strong> performed<br />
a randomized phase III trial <strong>of</strong> gemcitabine plus<br />
paclitaxel as second-line therapy in patients with advanced<br />
From the Tisch Cancer Institute, Division <strong>of</strong> Hematology and <strong>Oncology</strong>, Department <strong>of</strong><br />
Medicine, Mount Sinai School <strong>of</strong> Medicine, New York, NY.<br />
Author’s disclosure <strong>of</strong> potential conflicts <strong>of</strong> interest are found at the end <strong>of</strong> this article.<br />
Address reprint requests to Matthew D. Galsky, MD, Mount Sinai School <strong>of</strong> Medicine,<br />
1 Gustave L Levy Place, New York, NY 10029; email: matthew.galsky@mssm.edu.<br />
© <strong>2012</strong> by <strong>American</strong> <strong>Society</strong> <strong>of</strong> <strong>Clinical</strong> <strong>Oncology</strong>.<br />
1092-9118/10/1-10
NOVEL DEVELOPMENTS IN UROTHELIAL CANCER<br />
Table 1. Randomized Phase III Trials <strong>of</strong> Cisplatin-Based<br />
Chemotherapy in Advanced Urothelial Carcinoma<br />
Regimen<br />
urothelial cancer, with the primary goal <strong>of</strong> evaluating the<br />
impact <strong>of</strong> duration <strong>of</strong> therapy. 10 In this trial, 102 patients<br />
were randomly assigned to short-term (six cycles) compared<br />
with prolonged therapy (treatment until disease progression)<br />
with gemcitabine plus paclitaxel. Notably, the<br />
response rate with both treatment arms was approximately<br />
40%. Neither overall survival nor progression-free survival<br />
differed between the treatment arms, although the pro-<br />
KEY POINTS<br />
No. <strong>of</strong><br />
Patients<br />
Response (%)<br />
Overall Complete<br />
Survival<br />
(months) p<br />
MVAC 120 36 13 12.5 � 0.0002<br />
Cisplatin 126 11 3 8.2<br />
MVAC 55 65 35 12.6 � 0.05<br />
CISCA 55 46 25 10<br />
MVAC 86 59 24 12.5 0.17<br />
FAP 83 42 10 12.5<br />
MVAC 129 58 9 14.1 0.122<br />
DD-MVAC 134 72 21 15.5<br />
MVAC 205 46 12 14.8 0.746<br />
Gemcitabine � cisplatin 203 50 12 13.8<br />
MVAC 109 54 23 14.2 0.025<br />
Docetaxel � cisplatin 111 37 13 9.3<br />
MVAC 44 40 13 14.2 0.41<br />
Paclitaxel � carboplatin 41 28 3 13.8<br />
Gemcitabine � cisplatin 314 44 11 12.7 0.075<br />
Gemcitabine � cisplatin<br />
� paclitaxel<br />
312 56 14 15.8<br />
DD-MVAC 118 47 15 18.4 0.7<br />
DD-gemcitabine � cisplatin 57 47 10 20.7<br />
Abbreviations: MVAC, methotrexate, vinblastine, doxorubicin, cisplatin;<br />
CISCA, cyclophosphamide, cisplatin, doxorubicin; CMV, cisplatin, methotrexate,<br />
vinblastine; MV, methotrexate, vinblastine; FAP, fluorouracil, interferon �2b,<br />
cisplatin; DD, dose dense.<br />
● Cisplatin-based combination chemotherapy regimens,<br />
particularly methotrexate, vinblastine, doxorubicin<br />
and cisplatin (MVAC) and gemcitabine plus<br />
cisplatin, have long been the first-line treatment<br />
standards for advanced urothelial cancer.<br />
● There have been no improvements in the efficacy <strong>of</strong><br />
first-line chemotherapy for advanced urothelial cancer<br />
in the last 30 years.<br />
● A large proportion urothelial carcinomas are considered<br />
unfit for cisplatin, and the combination <strong>of</strong> gemcitabine<br />
plus carboplatin is a reasonable standard on<br />
the basis <strong>of</strong> phase III testing.<br />
● Ongoing efforts to improve outcomes focus on integrating<br />
antiangiogenic agents, targeting oncogenes<br />
and associated signaling pathways, stimulating an<br />
antitumor immune response, and refining patient<br />
selection for specific treatments.<br />
● Expansion <strong>of</strong> the evidence base beyond randomized<br />
trials is likely needed to help address many unanswered<br />
questions encountered in the daily care <strong>of</strong><br />
patients with advanced urothelial cancer.<br />
longed schedule was difficult to evaluate because the majority<br />
<strong>of</strong> patients experienced rapid disease progression and/or<br />
toxicity.<br />
In the absence <strong>of</strong> Level 1 evidence, the taxanes or pemetrexed<br />
are commonly used in the second-line setting on the<br />
basis <strong>of</strong> modest activity demonstrated in single-arm phase II<br />
trials. 11-13<br />
Cisplatin-Ineligible Patients<br />
There have not been adequately powered randomized<br />
phase III trials evaluating cisplatin-compared with<br />
carboplatin-based therapy in advanced urothelial cancer to<br />
address a potential survival benefit with the former regimens.<br />
However, a meta-analysis <strong>of</strong> randomized trials revealed<br />
that cisplatin-based chemotherapy was associated<br />
with a significant increase in the likelihood <strong>of</strong> achieving an<br />
objective response (response rate [RR] � 1.33, [95% CI, 1.04<br />
to 1.71], p � 0.025) or a complete response (RR � 3.54 [95%<br />
CI, 1.48 to 8.49], p � 0.004) compared with carboplatinbased<br />
chemotherapy. 14 Although these data support the<br />
preference for cisplatin-based regimens, urothelial cancer is<br />
largely a disease <strong>of</strong> the elderly and as a result <strong>of</strong> multiple<br />
reasons including impairment in renal function and performance<br />
status, approximately 30% to 50% <strong>of</strong> patients are<br />
deemed ineligible for cisplatin-based chemotherapy. 15 Investigators<br />
have long appreciated the disconnect between<br />
the efficacy <strong>of</strong> cisplatin-based therapy and the effectiveness<br />
<strong>of</strong> this treatment when applied to the general population <strong>of</strong><br />
patient with bladder cancer and have designed trials specifically<br />
for cisplatin-ineligible patients.<br />
Two randomized phase III trials have been initiated in the<br />
cisplatin-ineligible population. EORTC 30986 was a randomized<br />
phase II/III trial <strong>of</strong> gemcitabine plus carboplatin<br />
(GCa) compared with methotrexate, vinblastine, and carboplatin<br />
(M-CAVI) in “unfit” patients (WHO performance status<br />
<strong>of</strong> 2 and/or creatinine clearance 30–60 mL/min) with<br />
metastatic urothelial cancer. 16 There was no significant<br />
difference in the progression-free survival (GCa, 5.8 months;<br />
M-CAVI, 4.2 months; hazard ratio [HR], 1.04; 95% CI, 0.8 to<br />
1.35; p � 0.78) or the overall survival (GCa, 9.3 months;<br />
M-CAVI, 8.1 months; HR, 0.94; 95% CI, 0.72 to 1.22; p �<br />
0.64) between the treatment arms. However, GCa was better<br />
tolerated, solidifying the role <strong>of</strong> this regimen as a standard<br />
in the “unfit” population. A phase III trial comparing vinflunine<br />
plus gemcitabine compared with placebo plus gemcitabine<br />
closed early as a result <strong>of</strong> poor accrual.<br />
Trials in the “unfit” patient population have been hampered<br />
by a lack <strong>of</strong> a standard definition <strong>of</strong> what constitutes<br />
cisplatin ineligibility. In an effort to develop a consensus<br />
definition <strong>of</strong> patients with metastatic urothelial cancer “unfit”<br />
for cisplatin-based chemotherapy, a working group assembled<br />
and surveyed genitourinary oncologists with the<br />
goal <strong>of</strong> establishing uniform eligibility criteria for future<br />
clinical trials. 17 According to this consensus definition, patients<br />
meeting at least one <strong>of</strong> the following criteria were<br />
considered unfit for cisplatin: Eastern Cooperative <strong>Oncology</strong><br />
Group (ECOG) performance status <strong>of</strong> 2, creatinine clearance<br />
below 60 mL/min, grade 2 or greater hearing loss, grade 2 or<br />
greater neuropathy, and/or New York Heart Association<br />
Class III heart failure.<br />
305
Novel Approaches in Advanced Urothelial Cancer<br />
Identifying Predictive Biomarkers and Therapeutic Targets<br />
Because urothelial cancer is relatively chemotherapy sensitive,<br />
yet only a fraction <strong>of</strong> patients respond to any given<br />
regimen, there has been much interest in developing tools<br />
to allow more rational use <strong>of</strong> existing and future drugs.<br />
DNA-repair genes, or their protein products, have been <strong>of</strong><br />
particular interest as predictive biomarkers on the basis <strong>of</strong><br />
the mechanism <strong>of</strong> action <strong>of</strong> the conventional cytotoxics used<br />
in the management <strong>of</strong> urothelial cancer. In a retrospective<br />
study, levels <strong>of</strong> the DNA-repair genes ERCC1, RRM1,<br />
BRCA1, and caveolin-1, in tumor tissue from 57 patients<br />
with bladder cancer treated with cisplatin-based combination<br />
chemotherapy, were analyzed. 18 The median survival<br />
<strong>of</strong> patients in this cohort was higher in patients with low<br />
ERCC1 levels (25.4 vs. 15.4 months; p � 0.03). However,<br />
development <strong>of</strong> ERCC1 as a potential prognostic and/or<br />
predictive biomarker has been hampered by difficulties with<br />
analytic validation.<br />
The genomic complexity <strong>of</strong> most solid tumors suggests<br />
that relying on a single gene or protein as a predictive<br />
biomarker is unlikely to yield major improvements in patient<br />
selection. Alternatively, gene “signatures” <strong>of</strong> response<br />
to chemotherapy may be more attractive. The conventional<br />
approach to predictive gene expression model development<br />
is limited, however, because the signature is applicable only<br />
to the particular drug or combination and patient population<br />
for which the signature was developed. To overcome these<br />
limitations, Theodorescu and colleagues have developed a<br />
novel bioinformatics approach known as Coexpression Extrapolation<br />
(COXEN). 19 COXEN uses the publicly available<br />
gene-expression pr<strong>of</strong>iling data and drug sensitivity data<br />
from the NCI-60 cell line panel as a “Rosetta Stone” to<br />
predict chemotherapy sensitivity <strong>of</strong> gene-expression–pr<strong>of</strong>iled<br />
tumor samples. A neoadjuvant study to demonstrate<br />
the clinical utility <strong>of</strong> COXEN in treatment selection for<br />
patients with urothelial cancer is currently being planned.<br />
Pr<strong>of</strong>iling tumors for potentially “actionable” genomic mutations<br />
also <strong>of</strong>fers the potential for personalized application<br />
<strong>of</strong> targeted therapeutics. Sjodahl and colleagues performed<br />
mutation analyses <strong>of</strong> 16 genes (FGFR3, PIK3CA, PIK3R1,<br />
PTEN, AKT1, KRAS, HRAS, NRAS, BRAF, ARAF, RAF1,<br />
TSC1, TSC2, APC, CTNNB1, and TP53) in 145 urothelial<br />
cancer samples. 20 Notably, this study confirmed that FGFR3<br />
and PIK3CA mutations were most commonly found in noninvasive<br />
low-grade tumors, although a smaller proportion <strong>of</strong><br />
muscle-invasive specimens also harbored these mutations.<br />
The potential importance <strong>of</strong> adenomatous polyposis coli<br />
signaling and the mammalian target <strong>of</strong> rapamycin (mTOR)<br />
regulatory tuberous sclerosis complex genes (TSC2 and<br />
TSC2) were also identified in this study. Actionable mutations/aberrations<br />
found in other solid tumors, such EGFR<br />
mutations and EML4-ALK fusions, have been identified<br />
very rarely in urothelial cancer specimens.<br />
Targeting Oncogenes<br />
Although activating mutations in EGFR are not detected<br />
in urothelial cancer, overexpression or amplification <strong>of</strong><br />
EFGR pathway family members are present in a subset <strong>of</strong><br />
tumors, and have been implicated in the pathogenesis <strong>of</strong> the<br />
disease. Multiple phase II trials have explored single-agent<br />
inhibitors <strong>of</strong> the EGFR pathway using different agents,<br />
306<br />
methods <strong>of</strong> patient selection, and trial designs. A study <strong>of</strong><br />
lapatinib as second-line therapy in patients with metastatic<br />
urothelial cancer reported a response rate <strong>of</strong> only 1.7%;<br />
however, stable disease was correlated with EGFR overexpression,<br />
and, to some extent, HER2 overexpression. 21 A<br />
randomized discontinuation trial <strong>of</strong> lapatinib in patients<br />
with tumors harboring HER2 gene amplification, which<br />
included patients with urothelial cancer, demonstrated no<br />
objective responses in the urothelial cancer cohort. 22 Trials<br />
combining chemotherapy with inhibitors <strong>of</strong> the EGFR pathway<br />
have yielded more intriguing results, but the single-arm<br />
nature <strong>of</strong> these trials has precluded an assessment <strong>of</strong> the<br />
contribution <strong>of</strong> the EGFR pathway inhibitor, and few randomized<br />
trials have been initiated.<br />
Although activating mutations in FGFR3 are most commonly<br />
found in non–muscle-invasive urothelial cancer, recent<br />
studies have demonstrated the presence <strong>of</strong> FGFR3<br />
mutations in 10% to 20% <strong>of</strong> muscle-invasive urothelial<br />
cancer specimens. 23 Dovitinib is a small-molecule inhibitor<br />
<strong>of</strong> several tyrosine kinase receptors, including the vascular<br />
endothelial growth factor receptor (VEGFR) and fibroblast<br />
growth factor receptor (FGFR), and has demonstrated inhibition<br />
<strong>of</strong> tumor growth and proliferation in urothelial carcinoma<br />
models. A multicenter, two-stage, open-label phase II<br />
trial is currently evaluating the safety and efficacy <strong>of</strong><br />
dovitinib in patients with advanced urothelial carcinoma<br />
who have experienced disease progression despite prior<br />
systemic therapy, both in cohorts with and without FGFR3<br />
mutations (NCT00790426).<br />
Targeting Angiogenesis<br />
MATTHEW D. GALSKY<br />
Multiple lines <strong>of</strong> evidence support a therapeutic role for<br />
targeting the tumor vasculature in urothelial cancer. The<br />
triplet <strong>of</strong> gemcitabine, cisplatin, and the anti-VEGF antibody<br />
bevacizumab in a phase II <strong>of</strong> 43 chemotherapy-naïve<br />
patients with metastatic urothelial cancer, demonstrated an<br />
overall response rate <strong>of</strong> 73% and an encouraging median<br />
overall survival <strong>of</strong> 19.1 months; it is currently being explored<br />
in a randomized trial by the Cancer and Leukemia<br />
Group B. 24<br />
On the basis <strong>of</strong> preclinical studies demonstrating synergistic<br />
antitumor activity by combining VEGFR and plateletderived<br />
growth factor receptor (PDGFR) kinase inhibitors,<br />
phase II trials <strong>of</strong> sorafenib, sunitinib, and pazopanib, have<br />
been performed in urothelial cancer and demonstrated modest<br />
antitumor activity with the latter drugs. 25-26 Although<br />
supported by preclinical studies showing additive to synergistic<br />
activity, the combination <strong>of</strong> GC plus sunitinib was<br />
poorly tolerated in a phase II trial as a result <strong>of</strong> hematologic<br />
toxicity. 27<br />
Two randomized phase II trials evaluating the impact <strong>of</strong><br />
adding antiangiogenic therapy to cytotoxic regimens in<br />
urothelial carcinoma have been reported. In a placebocontrolled<br />
phase II trial <strong>of</strong> docetaxel with or without vandetanib,<br />
a small-molecule inhibitor <strong>of</strong> the VEGFR and<br />
epithelial growth factor receptor (EGFR) tyrosine kinases, in<br />
the second-line setting, the combination failed to improve<br />
outcomes. 28 Similarly, a trial <strong>of</strong> GC with or without<br />
sorafenib failed to improve outcomes, but closed early as a<br />
result <strong>of</strong> poor accrual limiting the power to detect differences<br />
between the arms. 29
NOVEL DEVELOPMENTS IN UROTHELIAL CANCER<br />
Targeting the Immune System<br />
Preclinical and clinical studies suggest that urothelial<br />
cancer is immunogenic. For example, higher numbers <strong>of</strong><br />
CD8-positive tumor-infiltrating lymphocytes in bladder cancer<br />
specimens have been correlated with substantial improved<br />
disease-free and overall survival. 30 Despite the<br />
immunogenicity <strong>of</strong> urothelial cancer, patients with urothelial<br />
cancer also exhibit tumor-associated immunologic tolerance.<br />
Blocking immune regulatory checkpoints may<br />
overcome tumor-induced immune tolerance in urothelial<br />
cancer. Ipilimumab, a human immunoglobulin G (IgG1)�<br />
anti-CTLA-4 monoclonal antibody, was administered before<br />
cystectomy in a pro<strong>of</strong>-<strong>of</strong>-concept study <strong>of</strong> 12 patients with<br />
clinically localized bladder cancer. 31 Treatment resulted in<br />
perivascular infiltration <strong>of</strong> cells positive for CD3, CD8, CD4,<br />
and granzyme. In addition, patients demonstrated an increase<br />
in CD4-positive T cells with high expression <strong>of</strong><br />
inducible costimulator (ICOS hi ) in tumor tissue and systemic<br />
circulation, and eight <strong>of</strong> 12 patients experienced<br />
downstaging <strong>of</strong> their primary tumor. A phase II trial <strong>of</strong><br />
sequential chemotherapy followed by chemotherapy plus<br />
ipilimumab, in an attempt to “autovaccinate” patients before<br />
introduction <strong>of</strong> immune checkpoint blockade, has recently<br />
been initiated.<br />
Given the high risk <strong>of</strong> recurrence in patients with muscleinvasive<br />
bladder cancer after cystectomy, immune-based<br />
approaches have also been <strong>of</strong> interest in decreasing the<br />
likelihood <strong>of</strong> relapse. Lapuleucel-T is an activated cellular<br />
therapy consisting <strong>of</strong> autologous peripheral blood mononuclear<br />
cells cultured ex vivo with BA7072, a recombinant<br />
fusion antigen consisting <strong>of</strong> portions <strong>of</strong> the intracellular and<br />
extracellular regions <strong>of</strong> HER2/neu linked to granulocytemacrophage<br />
colony-stimulating factor. A placebo-controlled<br />
randomized phase II trial is ongoing exploring the impact <strong>of</strong><br />
Author’s Disclosure <strong>of</strong> Potential Conflicts <strong>of</strong> Interest<br />
Author<br />
Employment or<br />
Leadership<br />
Positions<br />
Consultant or<br />
Advisory Role<br />
Matthew D. Galsky Amgen; AVEO;<br />
Bristol-Myers<br />
Squibb;<br />
GlaxoSmithKline;<br />
Pfizer<br />
1. Jemal A, Bray F, Center MM, et al. Global cancer statistics. CA Cancer<br />
J Clin. 2011;61:69-90.<br />
2. Sternberg CN, Yagoda A, Scher HI, et al. Preliminary results <strong>of</strong> M-VAC<br />
(methotrexate, vinblastine, doxorubicin and cisplatin) for transitional cell<br />
carcinoma <strong>of</strong> the urothelium. J Urol. 1985;133:403-407.<br />
3. Sternberg CN, de Mulder P, Schornagel JH, et al. Seven year update <strong>of</strong><br />
an EORTC phase III trial <strong>of</strong> high-dose intensity M-VAC chemotherapy and<br />
G-CSF versus classic M-VAC in advanced urothelial tract tumours. Eur J<br />
Cancer. 2006;42:50-54.<br />
4. Bamias A, Aravantinos G, Deliveliotis C, et al. Docetaxel and cisplatin<br />
with granulocyte colony-stimulating factor (G-CSF) versus MVAC with<br />
G-CSF in advanced urothelial carcinoma: a multicenter, randomized, phase<br />
III study from the Hellenic Cooperative <strong>Oncology</strong> Group. J Clin Oncol.<br />
2004;22:220-228.<br />
5. Dreicer R, Manola J, Roth BJ, et al. Phase III trial <strong>of</strong> methotrexate,<br />
vinblastine, doxorubicin, and cisplatin versus carboplatin and paclitaxel in<br />
patients with advanced carcinoma <strong>of</strong> the urothelium. Cancer. 2004;100:1639-<br />
1645.<br />
Lapuleucel-T on overall survival in patients with Her-2–<br />
expressing muscle-invasive bladder cancer after cystectomy.<br />
The Rapid Learning Health System and<br />
Urothelial Cancer<br />
Although randomized controlled clinical trials are the<br />
standard for evidence development in medicine, given issues<br />
<strong>of</strong> feasibility and limited patient and economic resources,<br />
there are questions that arise in the daily care <strong>of</strong> patients<br />
with metastatic urothelial cancer that will never be answered<br />
by randomized trials. Furthermore, the pace <strong>of</strong><br />
evidence development through traditional clinical trials is<br />
painstakingly slow. In the meantime, only a small percentage<br />
<strong>of</strong> patients treated for cancer have their clinical data<br />
captured for research purposes. Complementary approaches<br />
to evidence generation must be embraced to support realtime<br />
clinical practice decisions and to enhance the care <strong>of</strong> all<br />
patients with urothelial cancer, not just those who fit the<br />
narrow eligibility criteria used in randomized trials. The<br />
rapid learning health care model envisions the use <strong>of</strong> routinely<br />
collected real-time clinical data, from electronic<br />
medical records or other sources, potentially linked to<br />
“omic” data, to continuously drive both scientific discovery<br />
and patient care through an iterative process. 32 However,<br />
linking data across institutions requires investment,<br />
information-technology infrastructure, and attention to<br />
privacy regulations. Pilot efforts such as the Retrospective<br />
International Study <strong>of</strong> Invasive/Advanced Cancer <strong>of</strong> the<br />
Urothelium (RISC; www.mssm.edu/research/programs/<br />
genitourinary-cancer-research/risc) are currently underway<br />
to link clinical data derived from patients with bladder<br />
cancer from a variety <strong>of</strong> institutions to demonstrate feasibility,<br />
standardize data collection and reporting, establish<br />
infrastructure, and ultimately aid in the design <strong>of</strong> future<br />
projects that will bring rapid learning to the care <strong>of</strong> patients<br />
with bladder cancer.<br />
Stock<br />
Ownership Honoraria<br />
REFERENCES<br />
Research<br />
Funding<br />
Celgene;<br />
Novartis; Pfizer;<br />
Viatar<br />
Expert<br />
Testimony<br />
Other<br />
Remuneration<br />
6. von der Maase H, Hansen SW, Roberts JT, et al. Gemcitabine and<br />
cisplatin versus methotrexate, vinblastine, doxorubicin, and cisplatin in<br />
advanced or metastatic bladder cancer: results <strong>of</strong> a large, randomized,<br />
multinational, multicenter, phase III study. J Clin Oncol. 2000;18:3068-3077.<br />
7. Bellmunt J, Von der Maase H, Mead GM, et al. Randomized phase III<br />
study comparing paclitaxel/cisplatin/gemcitabine and gemcitabine/cisplatin<br />
in patients with locally advanced or metastatic urothelial cancer without<br />
prior systemic therapy; EORTC30987/Intergroup Study. J Clin Oncol.<br />
2007;25 (suppl; abstr 5030).<br />
8. Bamias A, Karadimou A, Lampaki S, et al. Prospective, randomized<br />
phase III study comparing two intensified regimens (methotrexate/vinblastine/doxorubicin<br />
hydrochloride/cisplatin [MVAC] versus gemcitabine/cisplatin)<br />
in patients with inoperable or recurrent urothelial cancer. J Clin Oncol.<br />
2011;29 (suppl; abstr 4510).<br />
9. Bellmunt J, Theodore C, Demkov T, et al. Phase III trial <strong>of</strong> vinflunine<br />
plus best supportive care compared with best supportive care alone after a<br />
platinum-containing regimen in patients with advanced transitional cell<br />
carcinoma <strong>of</strong> the urothelial tract. J Clin Oncol. 2009;27:4454-4461.<br />
307
10. Albers P, Park SI, Niegisch G, et al. Randomized phase III trial <strong>of</strong> 2nd<br />
line gemcitabine and paclitaxel chemotherapy in patients with advanced<br />
bladder cancer: short-term versus prolonged treatment [German Association<br />
<strong>of</strong> Urological <strong>Oncology</strong> (AUO) trial AB 20/99]. Ann Oncol. 2011;22:288-294.<br />
11. Galsky MD, Mironov S, Iasonos A, et al. Phase II trial <strong>of</strong> pemetrexed as<br />
second-line therapy in patients with metastatic urothelial carcinoma. Invest<br />
New Drugs. 2007;25:265-270.<br />
12. McCaffrey JA, Hilton S, Mazumdar M, et al. Phase II trial <strong>of</strong> docetaxel<br />
in patients with advanced or metastatic transitional-cell carcinoma. J Clin<br />
Oncol. 1997;15:1853-1857.<br />
13. Sweeney CJ, Roth BJ, Kabbinavar FF, et al. Phase II study <strong>of</strong><br />
pemetrexed for second-line treatment <strong>of</strong> transitional cell cancer <strong>of</strong> the<br />
urothelium. J Clin Oncol. 2006;24:3451-3457.<br />
14. Galsky MD, Chen GJ, Oh WK, et al. Comparative effectiveness <strong>of</strong><br />
cisplatin-based and carboplatin-based chemotherapy for treatment <strong>of</strong> advanced<br />
urothelial carcinoma. Ann Oncol. <strong>2012</strong>;23:406-410.<br />
15. Dash A, Galsky MD, Vickers AJ, et al. Impact <strong>of</strong> renal impairment on<br />
eligibility for adjuvant cisplatin-based chemotherapy in patients with urothelial<br />
carcinoma <strong>of</strong> the bladder. Cancer. 2006;107:506-513.<br />
16. De Santis M, Bellmunt J, Mead G, et al. Randomized phase II/III trial<br />
comparing gemcitabine/carboplatin (GC) and methotrexate/carboplatin/vinblastine<br />
(M-CAVI) in patients (pts) with advanced urothelial cancer (UC)<br />
unfit for cisplatin-based chemotherapy (CHT): phase III results <strong>of</strong> EORTC<br />
study 30986. J Clin Oncol. 2010;28 (suppl; abstr LBA4519).<br />
17. Galsky MD, Hahn NM, Rosenberg J, et al. Treatment <strong>of</strong> patients with<br />
metastatic urothelial cancer “unfit” for cisplatin-based chemotherapy. J Clin<br />
Oncol. 2011;29:2432-2438.<br />
18. Bellmunt J, Paz-Ares L, Cuello M, et al. Gene expression <strong>of</strong> ERCC1 as<br />
a novel prognostic marker in advanced bladder cancer patients receiving<br />
cisplatin-based chemotherapy. Ann Oncol. 2007;18:522-528.<br />
19. Williams PD, Cheon S, Havaleshko DM, et al. Concordant gene expression<br />
signatures predict clinical outcomes <strong>of</strong> cancer patients undergoing<br />
systemic therapy. Cancer Res. 2009;69:8302-8309.<br />
20. Sjodahl G, Lauss M, Gudjonsson S, et al. A systematic study <strong>of</strong> gene<br />
mutations in urothelial carcinoma; inactivating mutations in TSC2 and<br />
PIK3R1. PLoS One. 2011;6:e18583.<br />
21. Wulfing C, Machiels JP, Richel DJ, et al. A single-arm, multicenter,<br />
open-label phase 2 study <strong>of</strong> lapatinib as the second-line treatment <strong>of</strong> patients<br />
308<br />
MATTHEW D. GALSKY<br />
with locally advanced or metastatic transitional cell carcinoma. Cancer.<br />
2009;115:2881-2890.<br />
22. Galsky MD, Von H<strong>of</strong>f DD, Neubauer M, et al. Target-specific, histologyindependent,<br />
randomized discontinuation study <strong>of</strong> lapatinib in patients with<br />
HER2-amplified solid tumors. Invest New Drugs. Epub 2010 Sep 22.<br />
23. Kompier LC, Lurkin I, van der Aa MN, et al. FGFR3, HRAS, KRAS,<br />
NRAS and PIK3CA mutations in bladder cancer and their potential as<br />
biomarkers for surveillance and therapy. PLoS One. 2010;5:e13821.<br />
24. Hahn NM, Stadler WM, Zon RT, et al. Phase II trial <strong>of</strong> cisplatin,<br />
gemcitabine, and bevacizumab as first-line therapy for metastatic urothelial<br />
carcinoma: Hoosier <strong>Oncology</strong> Group GU 04-75. J Clin Oncol. 2011;29:1525-<br />
1530.<br />
25. Dreicer R, Li H, Stein M, et al. Phase 2 trial <strong>of</strong> sorafenib in patients<br />
with advanced urothelial cancer: a trial <strong>of</strong> the Eastern Cooperative <strong>Oncology</strong><br />
Group. Cancer. 2009;115:4090-4095.<br />
26. Gallagher DJ, Milowsky MI, Gerst SR, et al. Phase II study <strong>of</strong> sunitinib<br />
in patients with metastatic urothelial cancer. J Clin Oncol. 2010;28:1373-<br />
1379.<br />
27. Galsky MD, Sonpavde G, Hellerstedt BA, et al. Phase II study <strong>of</strong><br />
gemcitabine, cisplatin, and sunitinib in patients with advanced urothelial<br />
carcinoma. Presented at the ASCO Genitourinary <strong>Oncology</strong> Symposium,<br />
March 5-7, 2010, San Francisco, CA (abstr 276).<br />
28. Choueiri T, Vaishampayan U, Yu EY, et al. A double-blind randomized<br />
trial <strong>of</strong> docetaxel plus vandetanib versus docetaxel plus placebo in platinumpretreated<br />
advanced urothelial cancer. Presented at the ASCO Genitourinary<br />
<strong>Oncology</strong> Symposium, March 5-7, 2010, San Francisco, CA (abstr LBA239).<br />
29. Krege S, Rexer H, vom Dorp F, et al. Gemcitabine and cisplatin with or<br />
without sorafenib in urothelial carcinoma (AUO-AB 31/05). J Clin Oncol.<br />
2010;28 (suppl; abstr 4574).<br />
30. Sharma P, Shen Y, Wen S, et al. CD8 tumor-infiltrating lymphocytes<br />
are predictive <strong>of</strong> survival in muscle-invasive urothelial carcinoma. Proc Natl<br />
Acad Sci USA.2007;104:3967-3972.<br />
31. Carthon BC, Wolchok JD, Yuan J, et al. Preoperative CTLA-4 blockade:<br />
tolerability and immune monitoring in the setting <strong>of</strong> a presurgical clinical<br />
trial. Clin Cancer Res. 2010;16:2861-2871.<br />
32. Abernethy AP, Etheredge LM, Ganz PA, et al. Rapid-learning system<br />
for cancer care. J Clin Oncol. 2010;28:4268-4274.
New Developments in Prostate Cancer Therapy<br />
By Madhuri Bajaj, MD, and Elisabeth I. Heath, MD<br />
Overview: Prostate cancer is the most common nonskin<br />
malignant neoplasm in men worldwide. In the United States,<br />
241,740 new diagnoses <strong>of</strong> prostate cancer and 28,170 prostate<br />
cancer deaths have been estimated for <strong>2012</strong>, representing<br />
28% <strong>of</strong> new cancer cases and 10% <strong>of</strong> male cancer deaths. 1<br />
Although metastatic prostate cancer remains an incurable<br />
disease, substantial advances have been made in therapeutic<br />
options available for men in the past several years. Development<br />
<strong>of</strong> novel agents that modulate the androgen receptor<br />
pathway, growth factor signaling pathways, and immune function<br />
and bone targeting pathways has been the focus <strong>of</strong><br />
ANDROGENS ARE the key regulators <strong>of</strong> cell growth<br />
and proliferation in prostate cancer. Androgen deprivation<br />
therapy is initially a highly effective therapy because<br />
<strong>of</strong> the induction <strong>of</strong> apoptosis. Persistent androgen receptor<br />
(AR) activation is an important mediator <strong>of</strong> disease progression<br />
in castrate-resistant prostate cancer (CRPC). 2 There<br />
are multiple mechanisms by which this activation happens,<br />
including AR overexpression, AR mutations that increase<br />
androgen sensitivity to or activation by other steroids,<br />
increased local androgen production by prostate cells via<br />
expression <strong>of</strong> steroidogenic enzymes, AR activation via<br />
crosstalk <strong>of</strong> signal transduction pathways (epidermal<br />
growth factor, insulinlike growth factor, interleukin 6),<br />
modulated expression <strong>of</strong> coactivators or corepressors <strong>of</strong> AR,<br />
and proteolytic processing <strong>of</strong> AR to an androgenindependent<br />
is<strong>of</strong>orm. 3 Preclinical research has validated<br />
these concepts and thus has served as the basis for the<br />
translation <strong>of</strong> novel, potent AR-targeted therapies for patients<br />
with prostate cancer who experience relapse after<br />
initial androgen inhibition. Several clinical studies have<br />
demonstrated that CRPC cells continue to be under the<br />
influence <strong>of</strong> androgen signaling as evidenced by the high<br />
number <strong>of</strong> AR expression. 4 As a result, these newer agents<br />
tend to be more specific targets <strong>of</strong> enzymes downstream in<br />
the hormonal cascade.<br />
Abiraterone<br />
Abiraterone acetate is more potent than ketoconazole and<br />
a selective inhibitor <strong>of</strong> the 17�-hydroxylase and the C17,<br />
20-lyase function <strong>of</strong> CYP17A. 5 In a phase II trial, 47 men<br />
with metastatic CRPC deemed refractory to docetaxel-based<br />
chemotherapy were given oral abiraterone at 1,000 mg/d.<br />
The results were prostate-specific antigen (PSA) decreases<br />
<strong>of</strong> 30% or higher, 50% or higher, and 90% or higher seen in<br />
32 <strong>of</strong> 47 patients (68%), 24 <strong>of</strong> 47 patients (51%), and 7 <strong>of</strong> 47<br />
patients (15%), respectively. 6 These results, coupled with a<br />
favorable safety pr<strong>of</strong>ile, have laid the foundation for the<br />
development <strong>of</strong> two randomized, double-blind, placebocontrolled,<br />
phase III clinical trials. Recently, in the first <strong>of</strong><br />
the phase III trials, 1,195 patients with metastatic CRPC<br />
previously treated with docetaxel who were given abiraterone<br />
plus prednisone showed improved overall survival<br />
compared with those given placebo plus prednisone (median<br />
overall survival time, 14.8 vs. 10.9 months; hazard ratio<br />
[HR], 0.65; p � 0.0001). All secondary end points, including<br />
time to PSA progression (10.2 vs. 6.6 months; p � 0.001),<br />
progression-free survival (5.6 vs. 3.6 months; p � 0.001), and<br />
therapeutic strategies because <strong>of</strong> its significance in the biology<br />
<strong>of</strong> prostate cancer progression. Several <strong>of</strong> the agents<br />
have gained U.S. Food and Drug Administration (FDA) approval,<br />
whereas many are in late-stage clinical trials. With the<br />
growth <strong>of</strong> available treatment options, a major challenge as<br />
we move forward will be to determine the best sequence<br />
and/or combination <strong>of</strong> therapy that will result in maximum<br />
clinical efficacy with minimum toxicity. Highlighted in this<br />
publication are several <strong>of</strong> the exciting advances in prostate<br />
cancer therapy for patients with metastatic, castrate-resistant<br />
prostate cancer.<br />
PSA response rate (29% vs. 6%, p � 0.001), favored abiraterone.<br />
7 Common adverse effects with this agent include<br />
hypokalemia, hypertension, and pedal edema. The effects<br />
are explained by a syndrome <strong>of</strong> mineralocorticoid excess. On<br />
the basis <strong>of</strong> these results, the FDA granted approval in April<br />
2011 <strong>of</strong> abiraterone for the treatment <strong>of</strong> patients with<br />
metastatic CRPC whose disease had progressed regardless<br />
<strong>of</strong> their docetaxel-based chemotherapy. The second <strong>of</strong> the<br />
phase III trials is investigating abiraterone in asymptomatic<br />
or mildly symptomatic men with metastatic CRPC who had<br />
not received prior chemotherapy. This trial has completed<br />
accrual, with final results pending (NCT00887198).<br />
The clinical trial <strong>of</strong> abiraterone compared with placebo in<br />
the postdocetaxel population is also important because considerable<br />
progress has been made in the area <strong>of</strong> circulating<br />
tumor cells. Evaluation <strong>of</strong> circulating tumor cells was embedded<br />
in the phase III trial as a potential surrogate<br />
endpoint for overall survival. At the 2011 Annual Meeting <strong>of</strong><br />
the <strong>American</strong> <strong>Society</strong> <strong>of</strong> <strong>Clinical</strong> <strong>Oncology</strong>, results from this<br />
phase III trial confirmed that pretreatment circulating tumor<br />
cells and lactate dehydrogenase, alone and in combination,<br />
served as prognostic biomarkers. 8 Interestingly, PSA<br />
did not. This important trial will set the foundation for<br />
future trials that incorporate biomarkers as surrogate endpoints.<br />
TAK-700<br />
TAK-700 is a novel, selective CYP450c17 inhibitor similar<br />
to abiraterone in its mechanism <strong>of</strong> reducing testosterone and<br />
dehydroepiandrosterone levels. In a phase I/II study <strong>of</strong> this<br />
compound in asymptomatic patients with metastatic CRPC,<br />
the drug was tolerated well at various doses, and there was<br />
a 50% decrease in 12 <strong>of</strong> 15 patients who were treated with<br />
300 mg or more twice daily for 3 months or longer. 9 No<br />
dose-limiting toxicity was seen, and the most common adverse<br />
events were fatigue (62%), nausea (38%), constipation<br />
(35%), and vomiting (30%). The preliminary phase I/II study<br />
results have led to a phase II, ongoing evaluation <strong>of</strong> TAK-<br />
From the Karmanos Cancer Institute/Wayne State University School <strong>of</strong> Medicine, Detroit,<br />
MI.<br />
Authors’ disclosures <strong>of</strong> potential conflicts <strong>of</strong> interest are found at the end <strong>of</strong> this article.<br />
Address reprint requests to Elisabeth I. Heath, MD, Karmanos Cancer Institute/Wayne<br />
State University School <strong>of</strong> Medicine, 4100 John R, Detroit, MI; email: heathe@<br />
karmanos.org.<br />
© <strong>2012</strong> by <strong>American</strong> <strong>Society</strong> <strong>of</strong> <strong>Clinical</strong> <strong>Oncology</strong>.<br />
1092-9118/10/1-10<br />
309
700 at a dose <strong>of</strong> 400 mg twice daily with prednisone in<br />
patients with metastatic CRPC. 8 There are now two phase<br />
III, multicenter, randomized, double-blind trials that are<br />
evaluating TAK-700 plus prednisone compared with placebo<br />
plus prednisone in patients with metastatic CRPC. One trial<br />
is evaluating chemotherapy-naive patients (primary endpoints:<br />
overall survival and radiographic progression-free<br />
survival) 10 (NCT01193244), whereas the other focuses on<br />
the postdocetaxel progression patient population (primary<br />
endpoint: overall survival) 11 (NCT01193257). Both trials are<br />
actively recruiting, with a target accrual <strong>of</strong> 1,000 to 1,400<br />
patients, and results are expected to be available by 2013–<br />
2014. Accrual to the postdocetaxel chemotherapy protocol<br />
may be more challenging in the future because <strong>of</strong> the FDA<br />
approval <strong>of</strong> abiraterone and the positive phase III results <strong>of</strong><br />
MDV3100.<br />
Studies evaluating combination therapy <strong>of</strong> abiraterone or<br />
TAK-700 with docetaxel chemotherapy are under way<br />
(NCT01400555 and NCT01084655, respectively). These results<br />
are eagerly awaited because each <strong>of</strong> these agents<br />
results in a reasonable PSA response, leading to the unanswered<br />
question <strong>of</strong> whether the synergistic combination will<br />
produce an even stronger PSA response.<br />
MDV3100<br />
Another hormonally driven strategy is to target the AR<br />
directly. MDV3100 is an oral AR antagonist that directly<br />
inhibits AR by irreversibly binding to the receptor. This<br />
interaction impairs AR nuclear translocation, DNA binding,<br />
and recruitment <strong>of</strong> coactivators. 12 Preclinical studies have<br />
demonstrated that MDV3100 is a more potent binder to the<br />
AR receptor than bicalutamide, thus leading to complete<br />
suppression <strong>of</strong> the AR pathway. 13 In a phase I/II study,<br />
MDV3100 showed antitumor activity in patients with metastatic<br />
CRPC. 14 In this trial, 56% <strong>of</strong> 140 patients demon-<br />
KEY POINTS<br />
● Currently, multiple new treatment options approved<br />
by the US Food and Drug Administration (FDA) are<br />
available for the treatment <strong>of</strong> metastatic castrateresistant<br />
prostate cancer (CRPC), including cabazitaxel,<br />
sipuleucel-T, and abiraterone.<br />
● MDV3100, an androgen receptor antagonist, and<br />
radium-223, a novel radiopharmaceutical, have also<br />
shown improvement in overall survival when compared<br />
with placebo in men and will likely have a role<br />
in treating patients with CRPC after chemotherapy.<br />
● Denosumab, a monoclonal antibody to RANK ligand,<br />
is now an FDA-approved therapy for the prevention<br />
<strong>of</strong> skeletal-related events from bone metastases secondary<br />
to prostate cancer.<br />
● Additional novel agents that inhibit androgen, angiogenesis,<br />
cell survival and signaling, bone interface,<br />
and immunologic pathways are actively being investigated<br />
in ongoing phase III clinical trials.<br />
● Optimal sequencing and/or combination therapy <strong>of</strong><br />
active agents has yet to be determined in men with<br />
metastatic CRPC.<br />
310<br />
BAJAJ AND HEATH<br />
strated decreases in serum PSA <strong>of</strong> 50% or more, and 61 <strong>of</strong><br />
the 109 patients had stabilized bone disease after treatment.<br />
The Atrial Fibrillation Follow-up Investigation <strong>of</strong> Rhythm<br />
Management (AFFIRM) trial is a randomized, double blind,<br />
placebo-controlled, multinational trial <strong>of</strong> 1,199 men assigned<br />
to 160 mg/d <strong>of</strong> MDV3100 (800 patients) or placebo<br />
(399 patients). MDV3100 was associated with a median<br />
overall survival <strong>of</strong> 18.4 months compared with 13.6 months<br />
for patients assigned to placebo (HR, 0.631). Median<br />
progression-free survival also favored MDV3100 (8.3 vs. 2.9<br />
months). 15 Approximately 30% <strong>of</strong> patients assigned to<br />
MDV3100 had complete or partial response compared with<br />
1.3% in the placebo group. The drug also was associated with<br />
a PSA reduction <strong>of</strong> at least 50% from baseline in 54% <strong>of</strong> the<br />
MDV3100 group compared with 1.5% <strong>of</strong> the placebo group<br />
and at least a 90% reduction from baseline in 25% <strong>of</strong> the<br />
MDV3100 group compared with 1% <strong>of</strong> the placebo group.<br />
Median time to PSA progression was 8.3 months in the<br />
experimental group compared with 3 months in the placebo<br />
group. MDV3100 was well tolerated. The most common<br />
adverse events that occurred more frequently in the<br />
MDV3100 group (� 2%) than in the placebo group included<br />
fatigue, diarrhea, and hot flushes. Five <strong>of</strong> 800 patients<br />
treated with MDV3100 in the study reported seizures compared<br />
with no seizures among the placebo arm. The 0.6%<br />
seizure rate attributed to MDV3100 is below the approximate<br />
1.5% rate seen in an earlier study <strong>of</strong> the drug using<br />
higher doses.<br />
The Prevention <strong>of</strong> VTE after Acute Ischemic Stroke with<br />
LMWH Enoxaparin (PREVAIL) trial is a phase III trial<br />
evaluating patients with chemotherapy-naive CRPC treated<br />
with 160 mg/d <strong>of</strong> MDV3100 with standard <strong>of</strong> care compared<br />
with placebo with standard <strong>of</strong> care (NCT01212991). With a<br />
target accrual goal <strong>of</strong> 1,700 patients nearly complete, the<br />
study has the primary endpoints <strong>of</strong> overall survival and<br />
progression-free survival and secondary endpoints <strong>of</strong> time<br />
to initiation <strong>of</strong> cytotoxic chemotherapy and time to first<br />
skeletal-related event (SRE). 16<br />
ARN-509 is a second-generation antiandrogen that is<br />
currently undergoing clinical evaluation. ARN-509 inhibits<br />
both AR nuclear translocation and AR binding to androgen<br />
response elements in DNA. The compound also does not<br />
exhibit agonist activity in prostate cancer cells that overexpress<br />
AR. In a recent study, ARN-509 was optimized for<br />
inhibition <strong>of</strong> AR transcriptional activity and prostate cancer<br />
cell proliferation, pharmacokinetics, and in vivo efficacy. In<br />
a clinically valid murine xenograft model <strong>of</strong> human CRPC,<br />
ARN-509 showed greater efficacy than MDV3100. Maximal<br />
therapeutic response in this model was achieved at 30 mg/kg<br />
daily <strong>of</strong> ARN-509, whereas the same response required 100<br />
mg/kg daily <strong>of</strong> MDV3100 and higher steady-state plasma<br />
concentrations. 17 Thus, ARN-509 appears to have a higher<br />
therapeutic index than current AR antagonists. ARN-509<br />
seems to be a promising therapy in both castration-sensitive<br />
and castration-resistant forms <strong>of</strong> prostate cancer. Currently,<br />
it is in phase II clinical trials (NCT01171898).<br />
Agents such as abiraterone, TAK-700, and MDV3100 are<br />
appealing options even during clinical trials because patients<br />
are <strong>of</strong>ten inclined to consider treatment with pills<br />
rather than systemic chemotherapy. Furthermore, there is a<br />
familiarity <strong>of</strong> androgen modulators as therapy because patients<br />
undergoing treatment with luteinizing hormonereleasing<br />
hormone or gonadotropin-releasing hormone
DEVELOPMENTS IN PROSTATE CANCER THERAPY<br />
agents have most likely been treated with antiandrogens,<br />
such as flutamide and nilutamide, in the past. Finally,<br />
agents modulating the androgen signaling axis frequently<br />
result in PSA reduction, another familiar signal <strong>of</strong> treatment<br />
effect. However, a major challenge now is how to best<br />
position these oral agents to maximize efficacy. MDV3100<br />
has shown success in patients in whom docetaxel-based<br />
chemotherapy has failed, which is in the same patient group<br />
as those who benefited from abiraterone. Therefore, with<br />
two highly active oral agents, there will be concerns regarding<br />
which agent should be initially administered in patients<br />
in whom chemotherapy has failed. The treatment landscape<br />
will also be affected when the results <strong>of</strong> the clinical trials<br />
conducted in patients before chemotherapy are available.<br />
The sequencing and/or role <strong>of</strong> combination therapy is important<br />
and practical and must be further explored in future<br />
clinical trials.<br />
Cell Signaling Pathways<br />
Angiogenesis, the process <strong>of</strong> new blood vessel formation, is<br />
a crucial step in the propagation <strong>of</strong> malignant tumor growth<br />
and metastasis. Among the multiple proangiogenic factors<br />
that promote the process <strong>of</strong> vessel formation, vascular endothelial<br />
growth factor (VEGF) is one <strong>of</strong> the most important.<br />
Bevacizumab is a humanized monoclonal antibody directed<br />
against VEGF-A and causes potent inhibition <strong>of</strong> VEGF<br />
receptor (VEGFR) signaling and angiogenesis. Bevacizumab<br />
is approved for use in combination with chemotherapy for<br />
patients with metastatic colorectal, breast, and lung cancers.<br />
In prostate cancer, bevacizumab has been evaluated<br />
in several clinical trials, including the Cancer and Leukemia<br />
Group B (CALGB) phase II trial <strong>of</strong> bevacizumab in combination<br />
with docetaxel and estramustine in 79 patients with<br />
metastatic CRPC. 18 A PSA decrease <strong>of</strong> more than 50%<br />
from baseline occurred in 81% <strong>of</strong> patients, the median time<br />
to progression was 9 months, and overall survival was<br />
21 months. These favorable trials led to a recent phase<br />
III randomized placebo-controlled trial <strong>of</strong> docetaxel, prednisone,<br />
and bevacizumab compared with docetaxel and<br />
prednisone in 1,050 patients with chemotherapy-naive metastatic<br />
CRPC with the primary endpoint <strong>of</strong> overall survival.<br />
Final results published in 2011 reported that although there<br />
was median progression-free survival in the bevacizumab<br />
arm <strong>of</strong> 9.9 months compared with the 7.5 months <strong>of</strong> the<br />
control arm (p � 0.0001), the overall survival time was not<br />
statistically significant. 19 Bevacizumab has notable toxicities,<br />
including hypertension, thromboembolism, hemorrhage,<br />
gastrointestinal perforation, and proteinuria. Unfortunately,<br />
a negative study with bevacuzimab poses a challenge for<br />
further development <strong>of</strong> agents modulating the VEGF signaling<br />
axis. Multitargeted tyrosine kinase inhibitors against<br />
VEGFR, such as cediranib and sunitinib, have been evaluated<br />
in phase II and III clinical trials, respectively, but the<br />
response rates have not been overwhelmingly encouraging.<br />
In fact, the phase III sunitinib trial was recently terminated<br />
for futility.<br />
However, a novel VEGFR tyrosine kinase inhibitor that is<br />
showing tremendous response rates in imaging studies,<br />
including bone scans, is cabozantinib (XL184). Cabozantinib<br />
is an inhibitor <strong>of</strong> multiple kinase signaling pathways, including<br />
MET, RET, VEGFR2/KDR, and KIT. MET is a<br />
receptor tyrosine kinase that has roles in oncogenic signaling,<br />
angiogenesis, and metastasis. Androgen deprivation<br />
activates MET signaling in prostate cancer cells. Activated<br />
MET is particularly highly expressed in bone. Preclinical<br />
studies have suggested that MET signaling may promote<br />
survival <strong>of</strong> prostate cancer cells. 20 In a recent phase II study,<br />
cabozantinib showed promising activity in men with bone<br />
metastases, with substantial improvement in bone scans in<br />
most patients. Patients with metastatic CRPC with progressive<br />
measurable disease received cabozantinib at 100 mg/d<br />
orally during 12 weeks, with a primary endpoint <strong>of</strong> objective<br />
response rates. Accrual was halted at 168 patients based on<br />
an observed high rate <strong>of</strong> clinical activity. Of the 100 evaluable<br />
patients with a median age <strong>of</strong> 68, 86% <strong>of</strong> patients had<br />
complete or partial resolution <strong>of</strong> lesions on bone scan as<br />
early as week 6. A total <strong>of</strong> 64% had improved pain, and the<br />
most common related grade 3/4 adverse events were fatigue<br />
(11%), hypertension (7%), and hand-foot syndrome (5%); no<br />
related grade 5 adverse events were reported. The PSA<br />
changes were independent <strong>of</strong> clinical activity, and the overall<br />
week 12 disease control rate was 71%. 21 Current studies<br />
are proceeding with a lower dose <strong>of</strong> cabozantinib to improve<br />
drug tolerability (NCT01347788, NCT01428219). The relationship<br />
between c-met inhibition and metastatic bone disease<br />
has yet to be properly elucidated, but multiple research<br />
efforts are under way to improve our understanding <strong>of</strong> this<br />
novel finding.<br />
Another antiangiogenesis inhibitor in a phase III clinical<br />
trial is tasquinimod. Tasquinimod is an orally active<br />
quinolone-3-carboxamide. Nearly 70% <strong>of</strong> men who took tasquinimod<br />
in a phase II trial did not progress at 6 months<br />
compared with 30% <strong>of</strong> men who took placebo. The phase III<br />
study is now open and expects to enroll 1200 patients<br />
(NCT01234311).<br />
Bone Targeting<br />
Until recently, the only standard <strong>of</strong> care was to administer<br />
vitamin D, calcium, and a bisphosphonate, such as zoledronic<br />
acid, to help minimize bone resorption, which leads to<br />
reduction <strong>of</strong> SREs. Denosumab, a humanized monoclonal<br />
antibody with specificity for the RANK ligand, was shown to<br />
be superior to zoledronic acid in a study <strong>of</strong> 1,901 men with<br />
CRPC. Denosumab delayed or prevented SREs more effectively<br />
than zoledronic acid. 22 The median time to first<br />
on-study SRE, the primary endpoint, was 20.7 months for<br />
denosumab compared with 17.1 months for zoledronic acid<br />
(HR, 0.82; 95% CI, 0.71 to 0.95; p � 0.0002). No differences<br />
were found in PSA time, overall disease progression, or<br />
overall survival. The two treatment groups had a similar<br />
frequency <strong>of</strong> serious toxicities; the cumulative incidence <strong>of</strong><br />
osteonecrosis <strong>of</strong> the jaw was similar in the two groups (2.3%<br />
for denosumab vs. 1.3% for zoledronic acid). These results,<br />
combined with two other pivotal phase III trials <strong>of</strong> the same,<br />
led to the FDA approval <strong>of</strong> denosumab for prevention <strong>of</strong><br />
skeletal complications in patients with bone metastases<br />
from solid tumors, except multiple myeloma.<br />
Denosumab was also found to improve bone metastasis–<br />
free survival (29.5 months) compared with placebo in men<br />
with M0 CRPC disease (25.2 months). However, no survival<br />
benefits were seen with denosumab in the phase III trial<br />
that was reported at the European Multidisciplinary Cancer<br />
Congress and the European <strong>Society</strong> for Medical <strong>Oncology</strong><br />
Meeting 2011. In September 2011, denosumab received FDA<br />
approval for its indication to increase bone mass in patients<br />
at high risk for fracture receiving androgen deprivation<br />
311
therapy for nonmetastatic prostate cancer. Fortunately, for<br />
men with CRPC, there are now two FDA-approved agents to<br />
help improve and strengthen metastatic bones. Zoledronic<br />
acid is administered via intravenous infusion and requires<br />
monitoring <strong>of</strong> renal function. Denosumab is administered<br />
via subcutaneous injection and requires monitoring <strong>of</strong> calcium<br />
and other electrolytes. Both agents have a role in the<br />
treatment <strong>of</strong> prostate cancer.<br />
Another agent that targets the osteoclasts, osteoblasts,<br />
and bone microenvironment is dasatinib. Dasatinib is an<br />
oral tyrosine kinase inhibitor with activity against Src<br />
kinases. A phase I/II trial evaluating dasatinib alone and in<br />
combination with docetaxel chemotherapy in CRPC reported<br />
activity with bone turnover markers with mild PSA response.<br />
23 Currently, a phase III trial with an estimated<br />
enrollment <strong>of</strong> 1,500 patients comparing docetaxel and dasatinib<br />
with docetaxel and placebo is under way<br />
(NCT00744497).<br />
A new drug that targets bone signaling pathways leading<br />
to actual overall survival is radium-223 chloride. Radium-<br />
223 chloride is an intravenously administered radiopharmaceutical<br />
that targets bone metastasis with high-energy,<br />
short-range �-particles. At the European Multidisciplinary<br />
Cancer Congress and the European <strong>Society</strong> for Medical<br />
<strong>Oncology</strong> Meeting 2011, results from a trial <strong>of</strong> nearly 1,000<br />
patients with CRPC with 2:1 randomization to either<br />
radium-223 chloride or placebo revealed a superior median<br />
overall survival <strong>of</strong> 14 months in the radium-223 chloride<br />
arm compared with the 11.2 months in the placebo arm. 24<br />
This 30% reduction in the risk <strong>of</strong> death (HR, 0.695, p �<br />
0.00185) in overall survival is important because the typical<br />
treatment <strong>of</strong> bone targeting agents results in mediating<br />
symptom relief and not necessarily affecting survival. The<br />
study results were promising enough that the data safety<br />
and monitoring committee for this study halted the accrual.<br />
The adverse effects include hematologic effects, such as<br />
anemia, and mild gastrointestinal toxicities. Next steps in<br />
the development <strong>of</strong> radium-223 chloride are eagerly<br />
awaited.<br />
Cytotoxic Therapy<br />
In 2004, our current stand-<strong>of</strong>-care chemotherapy, docetaxel,<br />
in combination with prednisone had shown efficacy<br />
in patients with CRPC. An overall survival benefit <strong>of</strong> almost<br />
t3 months and a PSA response <strong>of</strong> more than 50% were seen<br />
in almost one-half <strong>of</strong> patients compared with mitoxantrone<br />
and prednisone. 9,25 In March 2010, the efficacy results <strong>of</strong> a<br />
new taxane, cabazitaxel, for use in docetaxel-treated pa-<br />
Authors’ Disclosures <strong>of</strong> Potential Conflicts <strong>of</strong> Interest<br />
Author<br />
Employment or<br />
Leadership<br />
Positions<br />
Consultant or<br />
Advisory Role<br />
tients with CRPC were presented. This large, multicenter<br />
study showed an overall survival benefit <strong>of</strong> 3 months when<br />
compared with mitoxantrone. 26,27 On the basis <strong>of</strong> these<br />
results, the FDA-approved cabazitaxel as a second-line<br />
therapy for prostate cancer. As expected, clinical trials<br />
conducted in 2011 are primarily combination therapy evaluating<br />
cabazitaxel with chemotherapy, such as docetaxel<br />
(NCT01308567), abiraterone (NCT01522536), and tasquinimod<br />
(NCT01513733). <strong>Clinical</strong> trials evaluating optimal sequencing<br />
<strong>of</strong> cabazitaxel and abiraterone and other future<br />
agents must be efficiently designed to quickly meet its<br />
objectives.<br />
Conclusion<br />
In the last several years, notable advances have been<br />
made in the field <strong>of</strong> prostate cancer. Treatments emerging<br />
from our knowledge <strong>of</strong> the cell biology, androgen regulation,<br />
immunology, and chemoresistance <strong>of</strong> prostate cancer have<br />
led to the development <strong>of</strong> mechanism-based drug discovery.<br />
This in turn has led to various clinical trials based on a<br />
sound biologic rationale. Several phase III trials testing<br />
rational drug combinations in prostate cancer are ongoing.<br />
However, clinical trials to determine optimal sequence <strong>of</strong><br />
therapy are yet to be conducted.<br />
The role <strong>of</strong> immunotherapy, including the 2010 FDAapproved<br />
sipuleucel-T, is also an important part <strong>of</strong> the<br />
treatment paradigm and one that would benefit from biomarker<br />
identification. Sipuleucel-T is the first therapeutic<br />
cancer vaccine to gain FDA approval for patients with<br />
metastatic CRPC. The landmark phase III study <strong>of</strong><br />
sipuleucel-T showed an overall survival benefit <strong>of</strong> 4.1<br />
months; however, tumor response rates were minimal. 28<br />
The study validated the efficacy <strong>of</strong> immunotherapy in prostate<br />
cancer and has led to an investigation <strong>of</strong> additional<br />
clinical trials <strong>of</strong> sipuleucel-T in prostate cancer. Additional<br />
promising agents in phase III clinical trials include anti-<br />
CTLA-4 (NCT01057810) and ProstVAC-V/F (NCT01322490).<br />
Multiple new drugs have recently been approved for the<br />
treatment <strong>of</strong> prostate cancer, but they all have not been able<br />
to demonstrate cure <strong>of</strong> the disease. In addition to targeting<br />
the different mechanisms <strong>of</strong> action discussed in this publication,<br />
multiple other molecular targets also promise to<br />
provide the next generation <strong>of</strong> advances. We must continue<br />
to maintain a close collaboration between basic and clinical<br />
science so that our knowledge <strong>of</strong> the molecular physiology<br />
can lead to strategic development <strong>of</strong> further viable drug<br />
targets.<br />
Stock<br />
Ownership Honoraria<br />
Research<br />
Funding<br />
Madhuri Bajaj*<br />
Elisabeth I. Heath Amgen;<br />
AstraZeneca;<br />
BiPar Sciences;<br />
Bristol-Myers<br />
Squibb;<br />
GlaxoSmithKline;<br />
Pfizer; Seattle<br />
Genetics;<br />
Zymogenetics<br />
*No relevant relationships to disclose.<br />
312<br />
Expert<br />
Testimony<br />
BAJAJ AND HEATH<br />
Other<br />
Remuneration
DEVELOPMENTS IN PROSTATE CANCER THERAPY<br />
1. Siegel R, Naishadham D, Jemal A. Cancer statistics, <strong>2012</strong>. CA Cancer<br />
J Clin. <strong>2012</strong>;62:10-29.<br />
2. Chen Y, Sawyers CL, Scher HI. Targeting the androgen receptor<br />
pathway in prostate cancer. Curr Opin Pharmacol. 2008;8:440-448.<br />
3. Devlin HL, Mudryj M. Progression <strong>of</strong> prostate cancer: multiple pathways<br />
to androgen independence. Cancer Lett. 2009;274:177-186.<br />
4. Taplin ME, Regan MM, Ko YJ, et al. Phase II study <strong>of</strong> androgen<br />
synthesis inhibition with ketoconazole, hydrocortisone, and dutasteride in<br />
asymptomatic castration-resistant prostate cancer. Clin Cancer Res. 2009;15:<br />
7099-7105.<br />
5. Yap TA, Carden CP, Attard G, et al. Targeting CYP17: established and<br />
novel approaches in prostate cancer. Curr Opin Pharmacol. 2008;8:449-457.<br />
6. Danila DC, Morris MJ, de Bono JS, et al. Phase II multicenter study <strong>of</strong><br />
abiraterone acetate plus prednisone therapy in patients with docetaxeltreated<br />
castration-resistant prostate cancer. J Clin Oncol. 2010;28:1496-<br />
1501.<br />
7. de Bono JS, Logothetis CJ, Molina A, et al. Abiraterone and increased<br />
survival in metastatic prostate cancer. N Engl J Med. 2011;364:1995-2005.<br />
8. Scher HI, Heller G, Molina A, et al. Evaluation <strong>of</strong> circulating tumor cell<br />
(CTC) enumeration as an efficacy response biomarker <strong>of</strong> overall survival (OS)<br />
in metastatic castration-resistant prostate cancer (mCRPC): Planned final<br />
analysis (FA) <strong>of</strong> COU-AA-301, a randomized double-blind, placebo-controlled<br />
phase III study <strong>of</strong> abiraterone acetate (AA) plus low-dose prednisone (P) post<br />
docetaxel. J Clin Oncol. 2011;29 (suppl; abstr LBA4517)<br />
9. Petrylak DP, Tangen CM, Hussain MH, et al. Docetaxel and estramustine<br />
compared with mitoxantrone and prednisone for advanced refractory<br />
prostate cancer. N Engl J Med. 2004;351:1513-1520.<br />
10. U.S. National Institutes <strong>of</strong> Health. Study Comparing Orteronel Plus<br />
Prednisone in Patients With Chemotherapy-Naive Metastatic Castration-<br />
Resistant Prostate Cancer [<strong>Clinical</strong> Trials.gov identifier NCT01193244].<br />
http://clinicaltrials.gov/ct2/show/NCT01193244. Accessed March 20, <strong>2012</strong>.<br />
11. U.S. National Institutes <strong>of</strong> Health. Study Comparing Orteronel Plus<br />
Prednisone in Patients With Metastatic Castration-Resistant Prostate Cancer<br />
[<strong>Clinical</strong> Trials.gov identifier NCT01193257]. http://clinicaltrials.gov/ct2/<br />
show/NCT01193257. Accessed March 20, <strong>2012</strong>.<br />
12. Wu Y, Rosenberg JE, Taplin ME. Novel agents and new therapeutics in<br />
castration-resistant prostate cancer. Curr Opin Oncol. 2011;23:290-296.<br />
13. Tran C, Ouk S, Clegg NJ, et al. Development <strong>of</strong> a second-generation<br />
antiandrogen for treatment <strong>of</strong> advanced prostate cancer. Science. 2009;324:<br />
787-790.<br />
14. Scher HI, Beer TM, Higano CS, et al. Antitumour activity <strong>of</strong> MDV3100<br />
in castration-resistant prostate cancer: a phase 1-2 study. Lancet. 2010;375:<br />
1437-1446.<br />
15. Scher HI, Fizazi K, Saad F, et al. Effect <strong>of</strong> MDV3100, an androgen<br />
receptor signaling inhibitor (ARSI), on overall survival in patients with<br />
prostate cancer postdocetaxel: Results from the phase III AFFIRM study.<br />
J Clin Oncol. <strong>2012</strong>;30 (suppl 5; abstr LBA1).<br />
16. U.S. National Institutes <strong>of</strong> Health. PREVAIL: A Multinational Phase 3<br />
REFERENCES<br />
R, Double-Blind, Placebo-Controlled Efficacy and Safety Study <strong>of</strong> Oral<br />
MDV3100 in Chemotherapy-Naive Patients with Progressive Metastatic<br />
Prostate Cancer Who Have Failed Androgen Deprivation Therapy [<strong>Clinical</strong><br />
Trials.gov identifier NCT01212991]. clinicaltrials.gov/ct2/show/<br />
NCT01212991. Accessed March 20, <strong>2012</strong>.<br />
17. Clegg NJ, Wongvipat J, Tran C, et al. ARN-509: a novel anti-androgen<br />
for prostate cancer treatment. Cancer Res. Epub <strong>2012</strong> Jan 29.<br />
18. Picus J, Halabi S, Kelly WK, et al. A phase 2 study <strong>of</strong> estramustine,<br />
docetaxel, and bevacizumab in men with castrate-resistant prostate cancer:<br />
results from Cancer and Leukemia Group B Study 90006. Cancer. 2011;117:<br />
526-533.<br />
19. Kelly WK, Halabi S, Carducci MA, et al. A randomized, double-blind,<br />
placebo-controlled phase III trial comparing docetaxel, prednisone, and placebo<br />
with docetaxel, prednisone, and bevacizumab in men with metastatic<br />
castration-resistant prostate cancer (mCRPC): Survival results <strong>of</strong> CALGB<br />
90401. J Clin Oncol. 2010;28:18s (suppl; abstr LBA4511).<br />
20. Zhang S, Zhau HE, Osunkoya AO, et al. Vascular endothelial growth<br />
factor regulates myeloid cell leukemia-1 expression through neuropilin-1dependent<br />
activation <strong>of</strong> c-MET signaling in human prostate cancer cells. Mol<br />
Cancer. 2010;9:9.<br />
21. Hussain M, Smith MR, Sweeney C, et al. Cabozantinib (XL184) in<br />
metastatic castration-resistant prostate cancer (mCRPC): results from a<br />
phase II randomized discontinuation trial. J Clin. Oncol. 2011;29 (suppl;<br />
abstr 4516).<br />
22. Fizazi K, Carducci M, Smith M, et al. Denosumab versus zoledronic<br />
acid for treatment <strong>of</strong> bone metastases in men with castration-resistant<br />
prostate cancer: a randomised, double-blind study. Lancet. 2011;377:813-822.<br />
23. Araujo JC, Mathew P, Armstrong AJ, et al. Dasatinib combined with<br />
docetaxel for castration-resistant prostate cancer: results from a phase 1-2<br />
study. Cancer. <strong>2012</strong>;118:63-71.<br />
24. Parker C, Heinrich, D, O’Sullivan JM., et al. Overall urvival benefit <strong>of</strong><br />
adium-223 chloride (Alpharadin) in the treatment <strong>of</strong> atients with Ssymptomatic<br />
bone metastases in castration-resistant prostate cancer (CRPC): a<br />
Pphase III randomized trial (ALSYMPCA). In: European Multidisciplinary<br />
Cancer Congress; 2011 September 24, 2011; Stockholm, Sweden; 2011. p.<br />
1LBA.<br />
25. Tannock IF, de Wit R, Berry WR, et al. Docetaxel plus prednisone or<br />
mitoxantrone plus prednisone for advanced prostate cancer. N Engl J Med.<br />
2004;351:1502-1512.<br />
26. de Bono JS, Oudard S, Ozguroglu M, et al. Prednisone plus cabazitaxel<br />
or mitoxantrone for metastatic castration-resistant prostate cancer progressing<br />
after docetaxel treatment: a randomised open-label trial. Lancet. 2010;<br />
376:1147-1154.<br />
27. Galsky MD, Dritselis A, Kirkpatrick P, et al. Cabazitaxel. Nat Rev Drug<br />
Discov. 2010;9:677-678.<br />
28. Kant<strong>of</strong>f PW, Higano CS, Shore ND, et al. Sipuleucel-T immunotherapy<br />
for castration-resistant prostate cancer. N Engl J Med. 2010;363:411-422.<br />
313
ADJUVANT THERAPY FOR OLDER PATIENTS<br />
CHAIR<br />
Arti Hurria, MD<br />
City <strong>of</strong> Hope<br />
Duarte, CA<br />
SPEAKERS<br />
Cynthia Owusu, MD, MS<br />
Case Western Reserve University<br />
Cleveland, OH<br />
Nadine Jackson McCleary, MD<br />
Dana-Farber Cancer Institute<br />
Boston, MA<br />
Jeffrey Crawford, MD<br />
Duke University Medical Center<br />
Durham, NC
Adjuvant Therapy for Older Women with<br />
Early-Stage Breast Cancer: Treatment<br />
Selection in a Complex Population<br />
By Cynthia Owusu, MD, MS, Arti Hurria, MD, and Hyman Muss, MD<br />
Overview: Breast cancer is a disease <strong>of</strong> aging. However,<br />
older women with breast cancer are less likely to participate in<br />
clinical trials or to receive recommended treatment. This<br />
undertreatment has contributed to a lag in breast cancer<br />
survival outcomes for older women compared with that for<br />
their younger counterparts. The principles that govern recommendations<br />
for adjuvant treatment <strong>of</strong> breast cancer are the<br />
same for younger and older women. Systemic adjuvant treatment<br />
recommendations should be <strong>of</strong>fered on the basis <strong>of</strong><br />
tumor characteristics that divide patients into three distinct<br />
subgroups. These include (1) older women with hormone<br />
receptor (HR)-positive and human epidermal growth factor 2<br />
(HER2)-negative breast cancer who should be <strong>of</strong>fered endocrine<br />
therapy; (2) older women with HR-negative and HER2negative<br />
breast cancer who should be <strong>of</strong>fered adjuvant<br />
BREAST CANCER is the most common cancer in <strong>American</strong><br />
women and the second leading cause <strong>of</strong> cancerrelated<br />
deaths among women. In 2011, approximately<br />
230,480 new cases were diagnosed in the United States,<br />
with an expected 39,520 deaths. 1,2 The most important risk<br />
factor for breast cancer is age. The estimated lifetime risk <strong>of</strong><br />
a new breast cancer is 1 in 15, 1 in 29, 1 in 27 and 1 in 207<br />
for women 70 years or older, 60 to 69, 40 to 59, and 39 or<br />
younger, respectively. 1 The median age at the time <strong>of</strong> breast<br />
cancer diagnosis is currently 61 years and an estimated 45%<br />
<strong>of</strong> women are 65 or older at the time <strong>of</strong> initial diagnosis. 2,3<br />
Recent gains in life expectancy, coupled with aging as a risk<br />
factor for breast cancer, makes breast cancer primarily a<br />
disease <strong>of</strong> older women, with increasing public health importance.<br />
In 1980, persons 65 and older represented 11.3%<br />
<strong>of</strong> the total population, but by 2030 this proportion is<br />
expected to increase to 20%. 3 In addition, by 2030, persons<br />
older than 75 will be expected to account for just under 50%<br />
<strong>of</strong> the total cohort older than 65. 4 Given the nonlinear<br />
age-risk relationship and increasing life expectancy, a substantial<br />
proportion <strong>of</strong> older women are expected to be affected<br />
by breast cancer.<br />
Age-Related Cancer Health Disparities<br />
Evaluation <strong>of</strong> the biology <strong>of</strong> breast cancer by patient age<br />
has shown that hormone receptor-positive, low S-phase, low<br />
tumor grade and HER2-negative tumors are more common<br />
among older than younger women, 5 although these differences<br />
are relatively modest. Despite the favorable tumor<br />
pr<strong>of</strong>ile <strong>of</strong> breast cancer in older women, this has not translated<br />
into any major survival advantage for older women<br />
with breast cancer in comparison with their younger counterparts.<br />
In a study that drew data from National Vital<br />
Statistics Reports and the Surveillance Epidemiology and<br />
End Results database <strong>of</strong> the National Cancer Institute,<br />
Smith and colleagues 6 found that although the rate <strong>of</strong> breast<br />
cancer death in the general population and the adjusted risk<br />
<strong>of</strong> death among women with newly diagnosed disease are<br />
declining among all age groups, the least decline has been<br />
among older women. Relative to 1990, the rate <strong>of</strong> breast<br />
chemotherapy; and (3) older women with HER2-positive disease<br />
who should be <strong>of</strong>fered chemotherapy with trastuzumab.<br />
Exceptions to these guidelines may be made for older women<br />
with small node-negative tumors or frail older women with<br />
limited life expectancy, where close surveillance may be a<br />
reasonable alternative. Addressing the current age-related<br />
disparities in breast cancer survival will require that older<br />
women are <strong>of</strong>fered the same state-<strong>of</strong>-the-art-treatment as<br />
their younger counterparts, with a careful weighing <strong>of</strong> the<br />
risks and benefits <strong>of</strong> each treatment in the context <strong>of</strong> the<br />
individual’s preferences. In addition, older women should be<br />
encouraged to participate in breast cancer clinical trials to<br />
generate additional chemotherapy efficacy, toxicity, and quality<br />
<strong>of</strong> life data.<br />
cancer death in the general population decreased 2.5% per<br />
year for women age 20 to 49, 2.1% per year for those age 50<br />
to 64, 2% per year for those age 65 to 74, but 1.1% per year<br />
for those age 75 years and older. Moreover, among women<br />
with newly diagnosed breast cancer between 1980 and 1997,<br />
the adjusted risk <strong>of</strong> death decreased by 3.6% per year among<br />
women younger than age 75 compared with 1.3% per year<br />
among those age 75 and older (p � 0.01). These differences<br />
were even greater for older black women. The age-related<br />
disparity in survival outcomes was hypothesized to be related<br />
to the undertreatment <strong>of</strong> older women with breast<br />
cancer. In an analysis <strong>of</strong> 9,766 patients enrolled in the<br />
TEAM (Tamoxifen Exemestane Adjuvant Multinational)<br />
randomized controlled trial conducted with postmenopausal<br />
women with hormone receptor-positive breast cancer, increasing<br />
age was associated with a higher disease-specific<br />
mortality. 7 Treatments received and tumor characteristics<br />
did not completely explain the age-related differences in<br />
survival outcomes but older patients in this trial were much<br />
less likely to receive chemotherapy. Together, these data<br />
clearly underscore the fact that breast cancer is an important<br />
disease <strong>of</strong> older women who bear a disproportionate<br />
burden <strong>of</strong> the morbidity and mortality associated with the<br />
disease and who should be <strong>of</strong>fered proven treatments that<br />
improve survival outcomes. Given that treatment differences<br />
do not completely explain the age-related disparities<br />
in survival outcomes, additional population and translational<br />
studies are needed to provide further insight into the<br />
reasons for these disparities.<br />
From the Case Western Reserve University School <strong>of</strong> Medicine, Cleveland, OH; City <strong>of</strong><br />
Hope Medical Center and Beckman Research Institute, Duarte, CA; and, Lineberger<br />
Comprehensive Cancer Center, University <strong>of</strong> North Carolina, Chapel Hill, NC.<br />
Authors’ disclosures <strong>of</strong> potential conflicts <strong>of</strong> interest are found at the end <strong>of</strong> this article.<br />
Address reprint requests to Cynthia Owusu, MD, MS, Case Western Reserve University<br />
School <strong>of</strong> Medicine, Division <strong>of</strong> Hematology/<strong>Oncology</strong> and Case Comprehensive Cancer<br />
Center-BHC 5055, 11100 Euclid Avenue, Cleveland, OH 44106; e-mail: cynthia.owusu@<br />
case.edu.<br />
© <strong>2012</strong> by <strong>American</strong> <strong>Society</strong> <strong>of</strong> <strong>Clinical</strong> <strong>Oncology</strong>.<br />
1092-9118/10/1-10<br />
3
Adjuvant Systemic Therapy<br />
Adjuvant systemic therapy refers to the administration <strong>of</strong><br />
anticancer therapy after primary breast surgery for early<br />
stage breast cancer with the goal <strong>of</strong> eradicating occult<br />
micrometastatic disease thought to be responsible for distant<br />
recurrence. Systemic treatment modalities include<br />
endocrine therapy and chemotherapy with or without trastuzumab.<br />
Treatment decision making regarding adjuvant<br />
systemic therapy for older women should involve consideration<br />
<strong>of</strong> such factors as the risk <strong>of</strong> morbidity and mortality<br />
from breast cancer, life expectancy and treatment tolerance,<br />
and patient preference. A geriatric assessment that includes<br />
evaluation <strong>of</strong> functional, cognitive, nutritional, and psychosocial<br />
status, and review <strong>of</strong> comorbidities and concomitant<br />
medications is particularly helpful in estimating the health<br />
status and life expectancy <strong>of</strong> older adults. Each <strong>of</strong> the<br />
components <strong>of</strong> a geriatric assessment can identify older<br />
adults at increased risk for morbidity and mortality. This<br />
assessment can also identify areas <strong>of</strong> vulnerability that may<br />
affect the patient’s ability to participate in a treatment plan<br />
(for example, ability to take medications on one’s own).<br />
Items in a geriatric assessment are able to identify older<br />
adults at risk <strong>of</strong> chemotherapy toxicity. 8,9 In a multisite<br />
study <strong>of</strong> 500 older adults with cancer, conducted by the<br />
Cancer and Aging Research Group, five geriatric assessment<br />
questions were independent predictors <strong>of</strong> the risk <strong>of</strong> chemotherapy<br />
toxicity, in addition to age, tumor, treatment, and<br />
laboratory variables. The geriatric assessment questions<br />
that were predictive <strong>of</strong> chemotherapy toxicity included hearing<br />
impairment (rated at fair or worse), difficulty in walking<br />
one block, need for assistance with taking medications, one<br />
or more falls in the past 6 months, and decrease in social<br />
activities because <strong>of</strong> either physical or emotional health. 8<br />
This predictive model for chemotherapy toxicity is being<br />
validated specifically in older adults receiving adjuvant<br />
chemotherapy for breast cancer (clinicaltrials.gov<br />
NCT01472094). In another study <strong>of</strong> 518 evaluable older<br />
adults with cancer (Chemotherapy Risk Assessment Scale<br />
for High Age Patients), predictors <strong>of</strong> chemotherapy toxicity<br />
included measures <strong>of</strong> functional status (ability to complete<br />
4<br />
KEY POINTS<br />
● The undertreatment <strong>of</strong> older women with breast<br />
cancer has contributed to poorer survival outcomes<br />
for older women than for younger women.<br />
● The principles that guide breast cancer treatment<br />
recommendations for younger and older women are<br />
fundamentally the same.<br />
● Breast cancer treatment decision making should consider<br />
the risk <strong>of</strong> breast cancer relapse, patient health<br />
status, life expectancy, and patient preference.<br />
● The use <strong>of</strong> systemic therapy in the adjuvant setting<br />
should be based on tumor characteristics and include<br />
endocrine therapy, and/or chemotherapy with or<br />
without trastuzumab.<br />
● Omission <strong>of</strong> postoperative radiation therapy is a<br />
reasonable strategy for older women with small<br />
hormone-receptor positive tumors.<br />
instrumental activities <strong>of</strong> daily living), cognition (Mini-<br />
Mental Status score), and nutrition (Mini-Nutritional Assessment<br />
score). 9 This information can be used in the<br />
treatment decision-making process in order to estimate life<br />
expectancy and risk <strong>of</strong> chemotherapy side effects and to<br />
identify areas <strong>of</strong> intervention to assist the patient during her<br />
treatment course.<br />
Endocrine Therapy<br />
OWUSU, HURRIA, AND MUSS<br />
The majority <strong>of</strong> older patients present with hormone<br />
receptor–positive breast cancer. Endocrine therapy is the<br />
mainstay <strong>of</strong> adjuvant therapy for these patients and results<br />
in proportional reductions in the risk <strong>of</strong> relapse and dying<br />
<strong>of</strong> breast cancer that exceed any available chemotherapy<br />
regimen. Current consensus guidelines recommend adjuvant<br />
systemic endocrine therapy for hormone receptorpositive<br />
breast cancer. The National Comprehensive Center<br />
Network (NCCN) guidelines 10 recommend the use <strong>of</strong> adjuvant<br />
endocrine therapy for women with hormone receptorpositive<br />
breast cancer regardless <strong>of</strong> age, menopausal status,<br />
or HER2 status, with the possible exception <strong>of</strong> women<br />
with lymph node-negative cancers 0.5 cm or less, or 0.6 to<br />
1.0 cm in diameter with favorable prognostic features;<br />
benefit from endocrine therapy is likely to be small for<br />
tumors <strong>of</strong> this size. In contrast, the St. Gallen International<br />
Consensus Panel recommends adjuvant systemic endocrine<br />
therapy for all women with endocrine-responsive disease<br />
with no exceptions and has defined endocrine-responsive<br />
tumors as those with as few as 1% <strong>of</strong> cells staining positive<br />
for hormone receptor proteins. 11 The <strong>Society</strong> <strong>of</strong> International<br />
Geriatric-<strong>Oncology</strong> (SIOG) breast cancer task force<br />
recommended that older women with endocrine-responsive<br />
breast cancer be <strong>of</strong>fered systemic endocrine therapy. However,<br />
for women with minimal risk disease, the decision to<br />
<strong>of</strong>fer endocrine therapy should be based on a risk-benefit<br />
analysis. 12<br />
Tamoxifen is the most firmly established adjuvant endocrine<br />
therapy for both premenopausal and postmenopausal<br />
women. Supporting these recommendations are results<br />
from the Early Breast Cancer Trialists’ Collaborative<br />
Group (EBCTCG) overview analysis, which demonstrated<br />
that over a 15-year period, use <strong>of</strong> adjuvant tamoxifen<br />
therapy for women with known estrogen receptor-positive<br />
disease, compared with no tamoxifen, decreased the 15-year<br />
risk <strong>of</strong> recurrence and death by 39% and 31%, respectively,<br />
regardless <strong>of</strong> age. 13 It is clear from these data that<br />
tamoxifen is <strong>of</strong> benefit for older women. In addition to<br />
decreasing the risk <strong>of</strong> disease relapse and death, there are<br />
also potential nonbreast cancer benefits <strong>of</strong> tamoxifen therapy<br />
in postmenopausal women. Tamoxifen may prevent<br />
osteoporosis 14 and reduce the risk <strong>of</strong> cardiovascular disease.<br />
15 Adjuvant tamoxifen therapy is, however, underutilized<br />
in older women. Women 80 years or older are half as<br />
likely to report having had a discussion about tamoxifen<br />
with their doctor compared with women 65 to 79 years, and<br />
women age 85 to 92 years are 25% less likely to receive a<br />
tamoxifen prescription than those 80 to 84 years. 16 Additionally,<br />
older women are more likely to self-discontinue and<br />
to be nonadherent to tamoxifen before the recommended<br />
treatment period <strong>of</strong> 5 years, 17,18 undercutting the treatment<br />
benefit from tamoxifen. Oncologists should always ask their<br />
patients if they are taking their prescribed medications and
TREATMENT FOR OLDER WOMEN WITH BREAST CANCER<br />
should reinforce the importance <strong>of</strong> adherence to maximize<br />
treatment benefit.<br />
The adjuvant use <strong>of</strong> aromatase inhibitors (anastrozole,<br />
letrozole, exemestane) for postmenopausal women with<br />
early breast cancer has been evaluated in several studies.<br />
These studies have involved the use <strong>of</strong> aromatase inhibitors<br />
either as initial adjuvant therapy, 19 as sequential therapy<br />
after 2 to 3 years <strong>of</strong> tamoxifen, 20 or as extended therapy<br />
after 4.5 to 6 years <strong>of</strong> tamoxifen. 21 The findings <strong>of</strong> these<br />
studies are consistent in demonstrating that the use <strong>of</strong> a<br />
third-generation aromatase inhibitor for postmenopausal<br />
women with hormone receptor–positive breast cancer, regardless<br />
<strong>of</strong> patient age, is superior in decreasing the risk <strong>of</strong><br />
disease recurrence, including ipsilateral and contralateral<br />
breast cancer, and distant metastatic disease compared with<br />
tamoxifen. Additionally, sequential use <strong>of</strong> aromatase inhibitors<br />
20 or extended therapy 21 has been shown to provide an<br />
overall survival advantage compared with tamoxifen use for<br />
5 years. No survival advantage has been demonstrated with<br />
the upfront use <strong>of</strong> aromatase inhibitors for 5 years compared<br />
with tamoxifen.<br />
There are differences in the toxicity pr<strong>of</strong>iles <strong>of</strong> aromatase<br />
inhibitors and tamoxifen. The incidence <strong>of</strong> venous thromboembolic<br />
disease, cerebrovascular events, endometrial cancer,<br />
vaginal bleeding, and hot flashes are less likely to be<br />
associated with aromatase inhibitors than tamoxifen,<br />
whereas the incidence <strong>of</strong> musculoskeletal pain, osteoporosis,<br />
and bone fractures have been found to be higher with<br />
aromatase inhibitors. 19 Emerging data also suggest that<br />
aromatase inhibitors may be associated with a small but<br />
higher risk <strong>of</strong> cardiovascular events compared with tamoxifen,<br />
22,23 but not compared with placebo. 24 In a metaanalysis<br />
<strong>of</strong> seven trials in which aromatase inhibitors were<br />
compared with tamoxifen, longer duration <strong>of</strong> aromatase<br />
inhibitor use or aromatase inhibitor use alone for 5 years<br />
was associated with a higher likelihood <strong>of</strong> cardiovascular<br />
events compared with tamoxifen alone (odds ratio [OR] �<br />
1.26, 95% CI 1.10–1.43). 25 Because studies <strong>of</strong> aromatase<br />
inhibitors have not included extensive follow-up, the full<br />
effect <strong>of</strong> aromatase inhibitors on cardiovascular disease and<br />
coronary heart risk remains to be determined.<br />
Although there is little evidence <strong>of</strong> age-related differences<br />
in the benefits <strong>of</strong> aromatase inhibitors for postmenopausal<br />
women, results <strong>of</strong> studies designed to examine age-related<br />
differences in the pattern <strong>of</strong> toxicity have been mixed. In<br />
general, the incidence <strong>of</strong> grade 3–5 nonfracture-related adverse<br />
events is higher among women 75 and older than<br />
among women less than 75 years. 26 However, a comparison<br />
<strong>of</strong> the quality <strong>of</strong> life and the side effect pr<strong>of</strong>ile for women who<br />
participated in MA-17, a study in which 5 years <strong>of</strong> letrozole<br />
was compared with placebo, showed no age-related differences<br />
in side effects. 24 The long-term consequences and<br />
implications <strong>of</strong> these side effects and any-age-related differences<br />
remain to be well characterized.<br />
Based on the results from recent studies that favor aromatase<br />
inhibitors over tamoxifen, current guidelines recommend<br />
that aromatase inhibitors should be <strong>of</strong>fered to all<br />
postmenopausal women with hormone receptor-positive<br />
early stage breast cancer, either alone, as sequential therapy<br />
after 2–3 years <strong>of</strong> tamoxifen. Given the lack <strong>of</strong> overall<br />
survival advantage associated with aromatase inhibitor use<br />
for 5 years, for women with pre-existing heart disease or<br />
bone loss, use <strong>of</strong> tamoxifen for 5 years or a switching<br />
strategy is a reasonable approach. For women with low<br />
grade, node-negative tumors 1 cm or smaller, endocrine<br />
therapy may be optional and observation acceptable, although<br />
the risks and benefits should be discussed with the<br />
patient.<br />
Adjuvant Chemotherapy<br />
Cytotoxic chemotherapy can be considered for older<br />
women with either node-positive or high-risk node-negative<br />
disease, particularly, triple-negative breast cancer. An<br />
abundance <strong>of</strong> literature has demonstrated the benefit <strong>of</strong><br />
adjuvant chemotherapy for younger women, with benefit<br />
decreasing as age increases. The EBCTCG overview analysis,<br />
13 which has 15 years <strong>of</strong> follow-up data, demonstrated<br />
that adjuvant chemotherapy reduced the annual risk <strong>of</strong><br />
recurrence by 37% and 19%, for women younger than 50 and<br />
50 to 69, respectively. The annual risk <strong>of</strong> death was reduced<br />
by 30% and 12%, for women younger than 50 and 50 to 69,<br />
respectively. The benefit <strong>of</strong> adjuvant chemotherapy for<br />
women with early stage breast cancer over age 70 was<br />
difficult to assess in the EBCTCG overview analysis<br />
because <strong>of</strong> the paucity <strong>of</strong> randomized trials that incorporated<br />
this age group. Of 29,000 women included in 60<br />
adjuvant polychemotherapy trials, 4% were 70 and older.<br />
This paucity <strong>of</strong> data has prevented definitive estimates<br />
regarding the magnitude <strong>of</strong> benefit <strong>of</strong> chemotherapy for<br />
women age 70 and older. In addition, with advancing age,<br />
organ function and performance status decline, and comorbidities<br />
increase, making the risks associated with chemotherapy<br />
even greater. Moreover, the risk reductions for<br />
chemotherapy are lower for postmenopausal women than for<br />
premenopausal women, although the reasons are unclear.<br />
Coupled with the apparent decline in the efficacy <strong>of</strong> chemotherapy<br />
with age is the increased risk <strong>of</strong> death from competing<br />
illnesses (comorbidity), leading to additional decline in<br />
the benefit from chemotherapy. As a result <strong>of</strong> all these<br />
factors, older women with early stage breast cancer receive<br />
adjuvant chemotherapy considerably less frequently than do<br />
younger women.<br />
However, a growing body <strong>of</strong> evidence suggests that adjuvant<br />
chemotherapy leads to improved survival outcomes<br />
for older women with breast cancer, particularly older<br />
women with hormone receptor-negative and node-positive<br />
breast cancer. A retrospective analysis <strong>of</strong> four Cancer and<br />
Leukemia Group B (CALGB) randomized clinical trials<br />
showed superior disease-free and overall survival benefits<br />
with the use <strong>of</strong> more aggressive chemotherapy (compared<br />
with less aggressive chemotherapy), among 6,487 women<br />
with node-positive breast cancer in all age groups, including<br />
70 and older. 27 This benefit, however, came at the cost <strong>of</strong><br />
increased risk <strong>of</strong> toxicity in older women, with older women<br />
more likely to discontinue treatment and to have an increased<br />
risk <strong>of</strong> treatment-related mortality. Additionally,<br />
data from large population studies have demonstrated a<br />
survival benefit from adjuvant chemotherapy for older<br />
women with hormone-negative or node-positive early stage<br />
breast cancer. 28,29<br />
Results from randomized controlled clinical trials that<br />
have specifically focused on older women with breast cancer,<br />
though scant, have helped to fill the gap on chemotherapy<br />
benefit for older women with early stage breast cancer.<br />
In the largest study in this population to date, a CALGB and<br />
Breast Cancer Intergroup study, 33 patients 65 and older<br />
5
with early-stage breast cancer were randomly assigned to<br />
either standard treatment (doxorubicin and cyclophosphamide<br />
[AC] for four cycles or cyclophosphamide, methotrexate,<br />
and 5-fluorouracil [CMF] for six cycles) or to an<br />
experimental arm <strong>of</strong> single-agent capecitabine for six cycles.<br />
30 At a median follow-up <strong>of</strong> 2.4 years, the relapse-free<br />
survival for patients receiving single-agent capecitabine was<br />
inferior to that for women receiving standard therapy (hazard<br />
ratio [HR] � 2.09 (1.38–3.17) p � 0 0.0001), as was<br />
overall survival (HR � 1.85 (1.11–3.08) p � 0.02). An<br />
unplanned subset analysis demonstrated that standard<br />
combination chemotherapy was particularly effective in<br />
patients with hormone receptor-negative disease. Overall,<br />
these results demonstrate that standard combination chemotherapy<br />
in the adjuvant setting, provides an overall<br />
survival benefit for older women with breast cancer, particularly<br />
those with hormone-receptor negative disease.<br />
The optimum chemotherapy regimen for treating breast<br />
cancer in the adjuvant setting remains debatable. Regardless,<br />
anthracycline-based regimens have become the norm,<br />
particularly for high-risk disease. However, the long-term<br />
complications associated with anthracycline use include,<br />
but are not limited to, dose-dependent cardiomyopathy. Age<br />
is a risk factor for cardiac disease, including anthracyclinerelated<br />
cardiomyopathy. Often, this precludes the use <strong>of</strong><br />
anthracycline-based regimens in older women with breast<br />
cancer. Jones and colleagues, 31 in a US <strong>Oncology</strong> trial,<br />
compared an anthracycline-based regimen (AC for four cycles)<br />
with a nonanthracycline taxane-based regimen docetaxel<br />
and cyclophosphamide (TC) for four cycles in 1,016<br />
women with node-negative and node-positive disease. At a<br />
median follow-up <strong>of</strong> 7 years, TC use was associated with<br />
superior disease-free survival (HR � 0.74, 95% CI (0.56–<br />
0.98) p � 0 0.03) and overall survival (HR � 0.69, 95%<br />
(0.50–0.97) p � 0 0.03) compared with AC. Sixteen percent<br />
<strong>of</strong> the study population were 65 and older. Unplanned<br />
subgroup analysis showed that TC was associated with<br />
superior disease-free and overall survival in all age groups,<br />
including older ages. TC therefore is a reasonable treatment<br />
regimen in the adjuvant setting for older women, particularly<br />
those with pre-existing heart disease and other contraindications<br />
to anthracycline-based regimens. A recent study<br />
has shown that this regimen is well tolerated in older<br />
women. 32 In another EBCTCG overview analysis in which<br />
different polychemotherapy regimens for early stage breast<br />
cancer were compared, adding a taxane to an anthracyclinebased<br />
chemotherapy regimen or a higher cumulative-dosage<br />
anthracycline-based regimen reduced breast cancer mortality,<br />
on average, by one-third. This benefit was irrespective<br />
<strong>of</strong> node status, tumor size, tumor grade, or age (�70). 33<br />
No definitive conclusion could be drawn regarding women<br />
70 and older because few women in that age group were<br />
included in the meta-analyses.<br />
For older women with node-negative, hormone-positive<br />
breast cancer, gene-expression pr<strong>of</strong>iling analysis can be used<br />
to identify women with high-risk disease who are likely to<br />
benefit from chemotherapy. The most widely used assay for<br />
this purpose is the 21-gene assay, which quantifies the<br />
likelihood <strong>of</strong> breast cancer recurrence in women with nodenegative,<br />
estrogen receptor-positive breast cancer and predicts<br />
the magnitude <strong>of</strong> endocrine therapy and chemotherapy<br />
benefit. 34 To the extent that the 21-gene assay allows for<br />
6<br />
individualization <strong>of</strong> cancer treatment, it is indeed a useful<br />
tool for the management <strong>of</strong> older patients with breast<br />
cancer, who were well represented in the validation cohorts<br />
(NSABP-14 and NSABP-20) for the assay. Moreover, the<br />
predictive ability <strong>of</strong> the 21-gene assay has been found to be<br />
independent <strong>of</strong> age. 34 Older patients with low scores are not<br />
likely to derive substantial benefit from chemotherapy,<br />
whereas those with high scores may derive great benefit.<br />
The benefit <strong>of</strong> adjuvant chemotherapy among patients with<br />
intermediate scores on the assay is being evaluated in the<br />
Tailor Rx study.<br />
Adjuvant! Online is another tool that can assist in decision<br />
making regarding the benefits <strong>of</strong> adjuvant endocrine therapy<br />
and chemotherapy for an individual patient. This online<br />
tool (available at www.adjuvantonline.com) summarizes the<br />
absolute benefit <strong>of</strong> chemotherapy and endocrine therapy,<br />
taking into account the patient’s age, brief assessment <strong>of</strong><br />
comorbid medical illnesses, and tumor characteristics. 35 To<br />
further inform this discussion, the risks associated with<br />
chemotherapy can be calculated with the predictive models<br />
for chemotherapy toxicity 8,9 as described earlier (See “Adjuvant<br />
Systemic Therapy”).<br />
Adjuvant Trastuzumab for HER2-Positive<br />
Breast Cancer<br />
OWUSU, HURRIA, AND MUSS<br />
Amplification or overexpression <strong>of</strong> HER2 is seen in approximately<br />
10% to 15% <strong>of</strong> invasive breast cancers in older<br />
women, 5 and it is associated with an unfavorable prognosis.<br />
A substantial body <strong>of</strong> literature from phase III trials in<br />
the adjuvant setting has demonstrated considerable benefit<br />
in disease-free and overall survival when trastuzumab is<br />
used either concurrently or sequential to chemotherapy<br />
compared with chemotherapy alone. 36-38 The main adverse<br />
effect associated with trastuzumab use is cardiotoxicity. In<br />
five phase III trials <strong>of</strong> adjuvant trastuzumab, the incidence<br />
<strong>of</strong> severe heart failure (New York Heart Association class III<br />
or IV), ranged from 0 to 3.9% among patients receiving<br />
trastuzumab, compared with 0 to 1.3% among patients who<br />
did not receive trastuzumab. 39 In the Breast Cancer International<br />
Research Group (BCIRG) 006 study, 38 two<br />
trastuzumab-containing regimens (AC plus docetaxel and<br />
trastuzumab and a nonanthracycline regimen <strong>of</strong> docetaxel,<br />
carboplatin, and trastuzumab [TCH]) were compared with<br />
standard chemotherapy alone. This study demonstrated<br />
disease-free and overall survival benefits with the use <strong>of</strong><br />
trastuzumab plus chemotherapy compared with chemotherapy<br />
alone. There was no substantial difference between the<br />
two trastuzumab-containing arms. Moreover, the incidence<br />
<strong>of</strong> cardiotoxicity associated with the nonanthracycline-based<br />
trastuzumab regimen was lower than that associated with<br />
the anthracycline-based trastuzumab regimen.<br />
Consistent with the under-representation <strong>of</strong> older women<br />
in breast cancer clinical trials <strong>of</strong> chemotherapy, older women<br />
have also been under-represented in clinical trials <strong>of</strong> trastuzumab<br />
therapy. With the notable exception <strong>of</strong> cardiac<br />
dysfunction, which was found to be associated with increasing<br />
age (older than 50), limitations in data collection precluded<br />
a determination <strong>of</strong> whether the toxicity pr<strong>of</strong>ile <strong>of</strong><br />
trastuzumab in older patients was different from that in<br />
younger patients. The reported clinical experience was also<br />
not adequate to determine whether the efficacy improvements<br />
(overall and disease-free survival) associated with<br />
trastuzumab in older patients was different from that in
TREATMENT FOR OLDER WOMEN WITH BREAST CANCER<br />
Endocrine-positive,<br />
HER2-negative<br />
Endocrine-negative,<br />
HER2-negative<br />
Table 1. Summary <strong>of</strong> Recommendations for Adjuvant Systemic Therapy in Early Stage Breast Cancer in Older Women<br />
patients younger than 65. Regardless, in the absence <strong>of</strong><br />
contraindications, trastuzumab is currently recommended<br />
for the adjuvant treatment <strong>of</strong> HER2-positive breast cancer,<br />
even for older women. In older women, a nonanthracycline,<br />
trastuzumab-containing regimen <strong>of</strong> TCH is <strong>of</strong>ten used because<br />
<strong>of</strong> the increased risk <strong>of</strong> cardiotoxicity associated with<br />
the anthracycline and trastuzumab regimen.<br />
In summary, the principles that govern recommendations<br />
for systemic adjuvant treatment <strong>of</strong> breast cancer are the<br />
same for younger and older women. Older women should<br />
be <strong>of</strong>fered guideline-recommended therapies (Table 1).<br />
Broadly, these recommendations should be <strong>of</strong>fered along<br />
three clinically distinct subgroups based on tumor characteristics.<br />
(1) Older women with hormone receptor-positive<br />
and HER2-negative breast cancer who should be <strong>of</strong>fered<br />
endocrine therapy regardless <strong>of</strong> node status. Gene expression<br />
pr<strong>of</strong>iling assay may be used to determine the added<br />
benefit <strong>of</strong> chemotherapy for those with node-negative<br />
disease; (2) older women with hormone receptor–negative<br />
and HER2-negative breast cancer (triple-negative breast<br />
cancer) who should be <strong>of</strong>fered adjuvant chemotherapy; and<br />
(3) older women with HER2-positive disease who should<br />
be <strong>of</strong>fered chemotherapy with trastuzumab. In the last<br />
group, women with hormone receptor-positive tumors<br />
should also be <strong>of</strong>fered endocrine therapy. Exceptions to<br />
these guidelines may be made for older women in any <strong>of</strong> the<br />
three subgroups who have node-negative disease and a<br />
tumor less than 1 cm or for frail older women with limited<br />
life expectancy, where close surveillance may be a reasonable<br />
alternative.<br />
Adjuvant Radiation Therapy<br />
Node-negative,<br />
Tumor size � 1cm<br />
No adjuvant therapy or<br />
Consider hormonal therapy if tumor size � 0.6 cm,<br />
grade 2, or other high risk features<br />
No adjuvant therapy or<br />
Consider chemotherapy if tumor size � 0.6 cm plus<br />
other high risk features<br />
HER2-positive No adjuvant therapy or<br />
Consider chemotherapy with trastuzumab if tumor<br />
size � 0.6 cm plus high risk features<br />
Until recently, the guideline recommendation, regardless<br />
<strong>of</strong> age, was for all women to receive radiation therapy<br />
after breast-conserving surgery and for postmastectomy<br />
radiation to be <strong>of</strong>fered to women with a high probability <strong>of</strong><br />
local recurrence. A recent meta-analysis <strong>of</strong> the EBCTCG<br />
supports these recommendations, showing that radiation<br />
therapy decreased the 10-year risk <strong>of</strong> any first recurrence<br />
from 35% to 19% and the 15-year risk <strong>of</strong> breast cancer death<br />
from 25% to 21% among women treated with breastconserving<br />
surgery. 40 Although the proportional reductions<br />
in relapse were similar among all women, the absolute<br />
benefits varied substantially by age, grade, estrogen receptor<br />
status, tamoxifen use, and extent <strong>of</strong> surgery. The authors<br />
concluded that radiation therapy after breast-conserving<br />
surgery halves the rate at which the disease recurs and<br />
reduces the breast cancer death rate by about a sixth.<br />
However, older women with small tumors can be spared<br />
radiation therapy. In a landmark randomized controlled<br />
study by Hughes and colleagues 41,42 636 women 70 or older<br />
who had undergone lumpectomy for stage I hormone<br />
receptor-positive breast cancer were randomly assigned to<br />
receive either radiation therapy and adjuvant tamoxifen<br />
for 5 years or to adjuvant tamoxifen for 5 years alone. The<br />
results demonstrated no substantial differences between<br />
the two groups with regard to mastectomy rates for local<br />
recurrence, distant metastases, or overall survival (median<br />
follow-up <strong>of</strong> 12 years). Of the 49% <strong>of</strong> patients who died<br />
during follow-up, 3% died <strong>of</strong> breast cancer. The only statistically<br />
significant difference was found in the rate <strong>of</strong> local<br />
or regional recurrence at 5 years, (2% among women who<br />
had radiation therapy compared with 9% who did not).<br />
Based on results from this study, one may reasonably<br />
consider lumpectomy (with surgically clear margins)<br />
without radiation therapy for women 70 and older with<br />
clinically negative lymph nodes, a tumor 2 cm or smaller,<br />
and hormone receptor-positive breast cancer who agree<br />
to take endocrine therapy. This strategy is limited by the<br />
high rate <strong>of</strong> nonadherence and early discontinuation <strong>of</strong><br />
adjuvant systemic endocrine therapy among older<br />
women. 43,44 Omission <strong>of</strong> postoperative radiation therapy,<br />
coupled with nonadherence to adjuvant systemic endocrine<br />
therapy, may result in earlier recurrences and, ultimately,<br />
poorer survival outcomes for older women. A favorable<br />
outcome from this approach can be achieved only when<br />
older women are adherent to prescribed oral endocrine<br />
therapies. Adherence to prescribed endocrine therapy<br />
should therefore be discussed and encouraged at follow-up<br />
visits.<br />
Conclusion<br />
Node-negative,<br />
Tumor size � 1 cm Node-positive<br />
Hormonal therapy<br />
Consider chemotherapy based on geneexpression<br />
pr<strong>of</strong>iling results<br />
Hormonal therapy<br />
Chemotherapy<br />
Chemotherapy alone Chemotherapy alone<br />
Chemotherapy with trastuzumab<br />
Add hormonal therapy if hormone<br />
positive<br />
Chemotherapy with trastuzumab<br />
Add hormonal therapy<br />
if hormone positive<br />
Breast cancer is a disease <strong>of</strong> aging. With minor differences,<br />
existing data support similar recommendations for<br />
both younger and older women. However, there are agerelated<br />
differences in treatment patterns, with older women<br />
less likely than younger women to receive standard therapies.<br />
Furthermore, survival outcomes lag behind that <strong>of</strong><br />
younger women. Closing the current gap in age-related<br />
disparities in breast cancer survival will require that<br />
older women are <strong>of</strong>fered the same state-<strong>of</strong>-the-art treatment<br />
as younger women, with a careful weighing <strong>of</strong> the risks<br />
and benefits <strong>of</strong> each treatment in the context <strong>of</strong> the individual’s<br />
preferences. Newer tools that estimate life<br />
expectancy and toxicity as well as the potential benefits <strong>of</strong><br />
therapy should make it easier for oncologists to make better<br />
treatment decisions with older patients. In addition, older<br />
women should be encouraged to participate in breast cancer<br />
7
clinical trials to generate additional efficacy and toxicity<br />
data. Such information will provide further knowledge so<br />
Authors’ Disclosures <strong>of</strong> Potential Conflicts <strong>of</strong> Interest<br />
Author<br />
Employment or<br />
Leadership<br />
Positions<br />
Consultant or<br />
Advisory Role<br />
Cynthia Owusu*<br />
Arti Hurria Amgen;<br />
Genentech; GTx<br />
Hyman Muss Boehringer<br />
Ingelheim; Eisai;<br />
Pfizer; Sandoz<br />
*No relevant relationships to disclose.<br />
1. Siegel R, Ward E, Brawley O, et al. Cancer statistics, 2011: the impact <strong>of</strong><br />
eliminating socioeconomic and racial disparities on premature cancer deaths.<br />
CA Cancer J Clin. 2011;61:212-236.<br />
2. DeSantis C, Siegel R, Bandi P, et al. Breast cancer statistics, 2011. CA<br />
Cancer J Clin. 2011;61:409-418.<br />
3. SEER stat fact sheets. http://seer.cancer.gov/csr/1975_2008/results_<br />
single/sect_01_table.11_2pgs.pdf. Accessed March 8, <strong>2012</strong>.<br />
4. Chu KC, Tarone RE, Kessler LG, et al. Recent trends in U.S. breast<br />
cancer incidence, survival, and mortality rates. J Natl Cancer Inst. 1996;88:<br />
1571-1579.<br />
5. Diab SG, Elledge RM, Clark GM. Tumor characteristics and clinical<br />
outcome <strong>of</strong> elderly women with breast cancer. J Natl Cancer Inst. 2000;92:<br />
550-556.<br />
6. Smith BD, Jiang J, McLaughlin SS, et al. Improvement in breast cancer<br />
outcomes over time: are older women missing out? J Clin Oncol. 2011;29:<br />
4647-4653.<br />
7. van de Water W, Markopoulos C, van de Velde CJ, et al. Association<br />
between age at diagnosis and disease-specific mortality among postmenopausal<br />
women with hormone receptor-positive breast cancer. JAMA. <strong>2012</strong>;<br />
307:590-597.<br />
8. Hurria A, Togawa K, Mohile SG, et al. Predicting chemotherapy toxicity<br />
in older adults with cancer: a prospective multicenter study. J Clin Oncol.<br />
2011;29:3457-3465.<br />
9. Extermann M, Boler I, Reich RR, et al. Predicting the risk <strong>of</strong> chemotherapy<br />
toxicity in older patients: The Chemotherapy Risk Assessment Scale for<br />
High-Age Patients (CRASH) score. Cancer. Epub 2011 Nov 9. doi: 10.1002/<br />
cncr.26646<br />
10. National Comprehensive Cancer Network. <strong>Clinical</strong> Practice Guidelines<br />
in Breast Cancer. Version 1. <strong>2012</strong>. www.nccn.org. Accessed February 28,<br />
<strong>2012</strong>.<br />
11. Goldhirsch A, Wood WC, Gelber RD, et al. Meeting highlights: updated<br />
international expert consensus on the primary therapy <strong>of</strong> early breast cancer.<br />
J Clin Oncol. 2003;21:3357-3365.<br />
12. Wildiers H, Kunkler I, Biganzoli L, et al. Management <strong>of</strong> breast cancer<br />
in elderly individuals: recommendations <strong>of</strong> the International <strong>Society</strong> <strong>of</strong><br />
Geriatric <strong>Oncology</strong>. Lancet Oncol. 2007;8:1101-1115.<br />
13. Effects <strong>of</strong> chemotherapy and hormonal therapy for early breast cancer<br />
on recurrence and 15-year survival: an overview <strong>of</strong> the randomised trials.<br />
Lancet. 2005;365:1687-1717.<br />
14. Love RR, Mazess RB, Barden HS, et al. Effects <strong>of</strong> tamoxifen on bone<br />
mineral density in postmenopausal women with breast cancer. N Engl J Med.<br />
1992;326:852-856.<br />
15. Love RR, Wiebe DA, Feyzi JM, et al. Effects <strong>of</strong> tamoxifen on cardiovascular<br />
risk factors in postmenopausal women after 5 years <strong>of</strong> treatment. J Natl<br />
Cancer Inst. 1994;86:1534-1539.<br />
16. Blackman SB, Lash TL, Fink AK, et al. Advanced age and adjuvant<br />
tamoxifen prescription in early-stage breast carcinoma patients. Cancer.<br />
2002;95:2465-2472.<br />
17. Partridge AH, Wang PS, Winer EP, et al. Nonadherence to adjuvant<br />
tamoxifen therapy in women with primary breast cancer. J Clin Oncol.<br />
2003;21:602–606.<br />
18. Owusu C, Buist DS, Field T, et al. Tamoxifen discontinuance among<br />
older women with estrogen-receptor-positive breast cancer. J Clin Oncol.<br />
2006;24: (suppl; abstr 648).<br />
19. Howell A, Cuzick J, Baum M, et al. Results <strong>of</strong> the ATAC (Arimidex,<br />
8<br />
that oncologists can <strong>of</strong>fer older women the best treatment<br />
options.<br />
Stock<br />
Ownership Honoraria<br />
REFERENCES<br />
Research<br />
Funding<br />
Abraxis<br />
BioScience;<br />
Celgene;<br />
GlaxoSmithKline<br />
OWUSU, HURRIA, AND MUSS<br />
Expert<br />
Testimony<br />
Other<br />
Remuneration<br />
Tamoxifen, Alone or in Combination) trial after completion <strong>of</strong> 5 years’<br />
adjuvant treatment for breast cancer. Lancet. 2005;365:60-62.<br />
20. Coombes RC, Hall E, Gibson LJ, et al. A randomized trial <strong>of</strong> exemestane<br />
after two to three years <strong>of</strong> tamoxifen therapy in postmenopausal women<br />
with primary breast cancer. N Engl J Med. 2004;350:1081-1092.<br />
21. Goss PE, Ingle JN, Martino S, et al. A randomized trial <strong>of</strong> letrozole in<br />
postmenopausal women after five years <strong>of</strong> tamoxifen therapy for early-stage<br />
breast cancer. N Engl J Med. 2003;349:1793-1802.<br />
22. Mouridsen H, Keshaviah A, Coates AS, et al. Cardiovascular adverse<br />
events during adjuvant endocrine therapy for early breast cancer using<br />
letrozole or tamoxifen: safety analysis <strong>of</strong> BIG 1-98 trial. J Clin Oncol.<br />
2007;25:5715-5722.<br />
23. Buzdar A, Howell A, et al. Comprehensive side-effect pr<strong>of</strong>ile <strong>of</strong> anastrozole<br />
and tamoxifen as adjuvant treatment for early-stage breast cancer:<br />
long-term safety analysis <strong>of</strong> the ATAC trial. Lancet Oncol. 2006;7:633-643.<br />
24. Muss HB, Tu D, Ingle JN, et al. Efficacy, toxicity, and quality <strong>of</strong> life in<br />
older women with early-stage breast cancer treated with letrozole or placebo<br />
after 5 years <strong>of</strong> tamoxifen: NCIC CTG intergroup trial MA.17. J Clin Oncol.<br />
2008;26:1956-1964.<br />
25. Amir E, Seruga B, Niraula S, et al. Toxicity <strong>of</strong> adjuvant endocrine<br />
therapy in postmenopausal breast cancer patients: a systematic review and<br />
meta-analysis. J Natl Cancer Inst. 2011;103:1299-1309.<br />
26. Crivellari D, Sun Z, Coates AS, et al. Letrozole compared with tamoxifen<br />
for elderly patients with endocrine-responsive early breast cancer: the<br />
BIG 1-98 trial. J Clin Oncol. 2008;26:1972-1979.<br />
27. Muss HB, Woolf S, Berry D, et al. Adjuvant chemotherapy in older and<br />
younger women with lymph node-positive breast cancer. JAMA. 2005;293:<br />
1073-1081.<br />
28. Du XL, Jones DV, Zhang D. Effectiveness <strong>of</strong> adjuvant chemotherapy for<br />
node-positive operable breast cancer in older women. J Gerontol A Biol Sci<br />
Med Sci. 2005;60:1137-1144.<br />
29. Elkin EB, Hurria A, Mitra N, et al. Adjuvant chemotherapy and<br />
survival in older women with hormone receptor-negative breast cancer:<br />
assessing outcome in a population-based, observational cohort. J Clin Oncol.<br />
2006;24:2757-2764.<br />
30. Muss HB, Berry DA, Cirrincione CT, et al. Adjuvant chemotherapy in<br />
older women with early-stage breast cancer. N Engl J Med. 2009;360:2055-<br />
2065.<br />
31. Jones S, Holmes FA, O’Shaughnessy J, et al. Docetaxel with cyclophosphamide<br />
is associated with an overall survival benefit compared with doxorubicin<br />
and cyclophosphamide: 7-year follow-up <strong>of</strong> US <strong>Oncology</strong> Research<br />
Trial 9735. J Clin Oncol. 2009;27:1177-1183.<br />
32. Freyer G, Campone M, Peron J, et al. Adjuvant docetaxel/cyclophosphamide<br />
in breast cancer patients over the age <strong>of</strong> 70: results <strong>of</strong> an observational<br />
study. Crit Rev Oncol Hematol. 2011;80:466-473. Epub 2011 May 11.<br />
33. Peto R, Davies C, et al. Comparisons between different polychemotherapy<br />
regimens for early breast cancer: meta-analyses <strong>of</strong> long-term<br />
outcome among 100,000 women in 123 randomised trials. Lancet. <strong>2012</strong>;379:<br />
432-444.<br />
34. Paik S, Shak S, Tang G, et al. A multigene assay to predict recurrence<br />
<strong>of</strong> tamoxifen-treated, node-negative breast cancer. N Engl J Med. 2004;351:<br />
2817-2826.<br />
35. Ravdin PM, Simin<strong>of</strong>f LA, Davis GJ, et al. Computer program to assist<br />
in making decisions about adjuvant therapy for women with early breast<br />
cancer. J Clin Oncol. 2001;19:980-991.<br />
36. Piccart-Gebhart MJ, Procter M, Leyland-Jones B, et al. Trastuzumab
TREATMENT FOR OLDER WOMEN WITH BREAST CANCER<br />
after adjuvant chemotherapy in HER2-positive breast cancer. N Engl J Med.<br />
2005;353:1659-1672.<br />
37. Romond EH, Perez EA, Bryant J, et al. Trastuzumab plus adjuvant<br />
chemotherapy for operable HER2-positive breast cancer. N Engl J Med.<br />
2005;353:1673-1684.<br />
38. Slamon D, Eiermann W, Robert N, et al. Adjuvant trastuzumab in<br />
HER2-positive breast cancer. N Engl J Med. 2011;365:1273-1283.<br />
39. Ewer SM, Ewer MS. Cardiotoxicity pr<strong>of</strong>ile <strong>of</strong> trastuzumab. Drug Saf.<br />
2008;31:459-467.<br />
40. Darby S, McGale P, et al. Effect <strong>of</strong> radiotherapy after breast-conserving<br />
surgery on 10-year recurrence and 15-year breast cancer death: meta-analysis<br />
<strong>of</strong> individual patient data for 10,801 women in 17 randomised trials. Lancet.<br />
2011;378:1707-1716.<br />
41. Hughes KS, Schnaper LA, Berry D, et al. Lumpectomy plus tamoxifen<br />
with or without irradiation in women 70 years <strong>of</strong> age or older with early<br />
breast cancer. N Engl J Med. 2004;351:971-977.<br />
42. Hughes KS, Schnaper LA, Cirrincione C, et al. Lumpectomy plus<br />
tamoxifen with or without irradiation in women age 70 or older with early<br />
breast cancer. J Clin Oncol. 2010;28:15s (suppl; abstr 507).<br />
43. Owusu C, Buist DS, Field TS, et al. Predictors <strong>of</strong> tamoxifen discontinuation<br />
among older women with estrogen receptor-positive breast cancer.<br />
J Clin Oncol. 2008;26:549-555.<br />
44. Partridge AH, LaFountain A, Mayer E, et al. Adherence to initial<br />
adjuvant anastrozole therapy among women with early-stage breast cancer.<br />
J Clin Oncol. 2008;26:556-562.<br />
9
Adjuvant Treatment <strong>of</strong> Older Patients with<br />
Lung Cancer<br />
Overview: Although advances in the molecular biology <strong>of</strong><br />
lung cancer have rapidly impacted management <strong>of</strong> patients<br />
with advanced stage non-small cell lung cancer (NSCLC), the<br />
principal treatment in the adjuvant setting <strong>of</strong> early stage<br />
NSCLC remains platinum-based chemotherapy regimens. The<br />
evidence available from clinical trials demonstrates a similar<br />
benefit <strong>of</strong> adjuvant chemotherapy in fit, older patients as well<br />
as younger patients. Observational studies suggest that adjuvant<br />
chemotherapy for older patients provides comparable<br />
survival benefit, along with increased toxicity. The lower use<br />
LUNG CANCER is the leading cause <strong>of</strong> cancer mortality<br />
worldwide. Largely because <strong>of</strong> the cumulative effects <strong>of</strong><br />
smoking and other exposures, the risk <strong>of</strong> lung cancer increases<br />
with age. In the United States, Surveillance, Epidemiology,<br />
and End Results (SEER) data demonstrate that<br />
68% <strong>of</strong> patients diagnosed with NSCLC are age 65 or older. 1<br />
The median age at diagnosis is 71. The incidence <strong>of</strong> lung<br />
cancer peaks in the age 65 to 74 group with an incidence <strong>of</strong><br />
31.9%. However, the rate <strong>of</strong> lung cancer in the age 75 to 84<br />
group is nearly as high at 28.9%. At age 85 and older, the<br />
incidence drops to 7.3%. Whereas this decline in the very old<br />
may partially represent under diagnosis, it also speaks to<br />
the complex interaction between smoking, cumulative life<br />
exposure, and competing risks for mortality, which reduces<br />
the likelihood <strong>of</strong> long-term survival in smokers and former<br />
smokers.<br />
Increasing age is also associated with increasing likelihood<br />
<strong>of</strong> comorbidities, which may limit treatment options for<br />
patients. The intersection <strong>of</strong> age-related functional impairment<br />
and comorbidities affect the patient’s overall performance<br />
status, which has been a powerful predictor <strong>of</strong><br />
outcomes for patients with lung cancer ever since Karn<strong>of</strong>sky’s<br />
initial description. 2 For these reasons, older patients<br />
may be less likely to receive surgery and adjuvant treatment<br />
than their younger counterparts. When this decision is<br />
made on the basis <strong>of</strong> significant functional impairment that<br />
substantially alters the risk/benefit ratio for therapeutic<br />
intervention, this may be quite appropriate. However, withholding<br />
treatment based on age alone is not appropriate<br />
given the substantial body <strong>of</strong> data suggesting that the fit<br />
elderly population can have similar benefits compared to<br />
younger patients.<br />
Further complicating this field is the limited evidence<br />
specific to the elderly population, since the vast majority <strong>of</strong><br />
patients enrolled on clinical trials in lung cancer are<br />
younger than age 70. 3 Therefore, the clinician is <strong>of</strong>ten left<br />
with uncertainty when evaluating the true risk/benefit <strong>of</strong><br />
treatment in a 75- or 80-year-old individual with or without<br />
substantial comorbidities.<br />
Surgery in the Older Patient with Lung Cancer<br />
Traditionally, lung cancer surgery has been reserved for<br />
the healthy elderly population. Even in this population, the<br />
risk <strong>of</strong> operative and postoperative complications is increased,<br />
related, at least in part, to the decline in organ<br />
function with age. It is well known that cardiovascular<br />
By Jeffrey Crawford, MD<br />
<strong>of</strong> chemotherapy in the older population also suggests that the<br />
selection <strong>of</strong> appropriate patients remains an important part <strong>of</strong><br />
the decision process. Carboplatin therapy may be substituted<br />
for cisplatin in selected older patients, and different options<br />
exist for the second cytotoxic chemotherapy agent. As in all<br />
patients, and particularly in this vulnerable population receiving<br />
cytotoxic chemotherapy, supportive care is vital. Increasing<br />
enrollment <strong>of</strong> the older population in clinical trials will be<br />
important to improve the evidence for our decision-making in<br />
the future.<br />
function decreases with age, including a decrease in elasticity<br />
<strong>of</strong> the arterial system, loss <strong>of</strong> myocytes, loss <strong>of</strong> atrial<br />
pacemaker cells, and increased fibrosis <strong>of</strong> the cardiac fibroskeleton.<br />
4 Thus, even the patient with no prior cardiac<br />
history, an increase in arrhythmias and other cardiac events<br />
may occur. Secondly, the age-related decline in renal function<br />
occurs with a decline in renal flow, decreased glomerular<br />
filtration rate, and creatinine clearance. 5<br />
Both renal and cardiac function issues have a substantial<br />
impact on subsequent decisions regarding drug administration,<br />
such as cisplatin-based chemotherapy and related<br />
hydration, and also significant considerations for anesthesia<br />
and other perioperative medications. The decline in hematopoietic<br />
reserve may be associated with increased toxicity <strong>of</strong><br />
chemotherapy but also may substantially affect the need for<br />
hematopoietic support during and following surgical interventions.<br />
6 Particularly important when considering lung<br />
cancer surgery is the decline in lung function that occurs as<br />
a function <strong>of</strong> age. 7 For subjects who have never smoked,<br />
there is a gradual reduction in lung function, with a decline<br />
in FEV 1 <strong>of</strong> 25% between ages 50 and 75, with further decline<br />
over age 75. However, for those who have smoked the<br />
majority <strong>of</strong> their adult lives, there may be a more substantial<br />
reduction in lung function by as much as 75% or greater.<br />
With emphysema and other smoking-related lung diseases,<br />
operative and postoperative risks and long-term quality <strong>of</strong><br />
life may be significantly compromised and/or the patient<br />
may not be a candidate at all, despite having early-stage<br />
disease. Fortunately, for former smokers who quit smoking<br />
before significant lung injury occurs, the subsequent loss <strong>of</strong><br />
lung function is more consistent with the gradual agerelated<br />
decline seen in the nonsmoking population.<br />
Because <strong>of</strong> improvements in surgical techniques, including<br />
less invasive surgery such as video-assisted thoroscopic<br />
approaches, better preoperative evaluation, and improved<br />
perioperative management <strong>of</strong> patients, defining which patients<br />
with early-stage lung cancer are operable has been<br />
evolving. Studies over the last decade have suggested improvement<br />
in outcomes for older patients. 8<br />
From the Duke University Medical Center, Durham, NC.<br />
Author’s disclosure <strong>of</strong> potential conflicts <strong>of</strong> interest are found at the end <strong>of</strong> this article.<br />
Address reprint requests to Jeffrey Crawford, MD, Duke University Medical Center,<br />
2301 Erwin Rd., Durham, NC 27710; email: Crawf006@mc.duke.edu.<br />
© <strong>2012</strong> by <strong>American</strong> <strong>Society</strong> <strong>of</strong> <strong>Clinical</strong> <strong>Oncology</strong>.<br />
1092-9118/10/1-10<br />
315
Table 1. Surgical Management and Outcomes in Elderly<br />
Patients with Early Stage NSCLC<br />
More recent studies provide a current idea <strong>of</strong> the risks and<br />
outcomes for older patients with early stage NSCLC. A<br />
nested case-control study in France between January 2004<br />
and December 2008 matched nearly 2,000 patients age 70 or<br />
older with stage I/II NSCLC with matched controls younger<br />
than age 70 in order to compare surgical treatment and<br />
postoperative outcomes. 9 These results are outlined in Table<br />
1. Although the frequency <strong>of</strong> radical lymph-node dissection<br />
was more common in the younger patient, there was otherwise<br />
no significant difference in the type <strong>of</strong> resection between<br />
younger and older patients. The frequency <strong>of</strong><br />
lobectomy was 77% in the younger patient and 79% in the<br />
older patients, with pneumonectomy in 11% <strong>of</strong> the younger<br />
patients and 8% in the older patients. Sublobar resections<br />
occurred in only 7% <strong>of</strong> the patients younger than age 70 and<br />
6% over the age <strong>of</strong> 70. Postoperative mortality was higher in<br />
the older patient with a 30-day mortality <strong>of</strong> 3.6% compared<br />
with 2.2% in the younger patients (p � 0.01). For the older<br />
population, perhaps an even more important endpoint is the<br />
90-day mortality. This endpoint was 4.7% in those patients<br />
older than age 70 compared with 2.5% in the population<br />
younger than age 70 (p � 0.0002). This result in particular<br />
shows the vulnerability <strong>of</strong> this population after thoracic<br />
surgery. However, the relatively low mortality in this age<br />
group across a wide number <strong>of</strong> institutions in France reassures<br />
us that patients appropriately selected for surgery can<br />
have similar outcomes.<br />
Improved outcomes with higher surgery rates for older<br />
patients with early-stage NSCLC have been recently reported<br />
from a national cohort study <strong>of</strong> over 17,000 Medicare<br />
KEY POINTS<br />
� age 70 � age 70<br />
n � 1,969 n � 1,969<br />
Lobectomy 77% 79% NS<br />
Pneumonectomy 11% 8% NS<br />
30 d mortality 2.2% 3.6% p � 0.01<br />
90 d mortality 2.5% 4.7% p � 0.0002<br />
● The majority <strong>of</strong> lung cancer occurs in the older<br />
population.<br />
● Advances in staging, surgery, and perioperative management<br />
have increased the number <strong>of</strong> older patients<br />
who are candidates for surgical resection for earlystage<br />
lung cancer.<br />
● Although the elderly population is clinically underrepresented<br />
in clinical trials, the data suggest that<br />
older, fit patients may have similar benefit with<br />
adjuvant chemotherapy than younger patients with<br />
a modest increase in toxicities.<br />
● Observational databases suggest that these benefits<br />
extend to the patients outside <strong>of</strong> clinical trials. In the<br />
United States, carboplatin is commonly substituted<br />
for cisplatin in the older patient population and<br />
benefits may be comparable within the limits <strong>of</strong> this<br />
data.<br />
316<br />
patients with stage I or II NSCLC, identified through the<br />
SEER database between 2001 and 2005. 10 In this study,<br />
areas <strong>of</strong> high and low rates <strong>of</strong> curative surgery for earlystage<br />
lung cancer were compared in an attempt to determine<br />
the effectiveness <strong>of</strong> surgery in older and sicker patients. In<br />
the low surgery areas, less than 63% <strong>of</strong> patients underwent<br />
surgery compared with greater than 79% in high surgery<br />
areas. Those high surgery areas also operated on more<br />
patients with advanced age and patients with chronic obstructive<br />
pulmonary disease. Despite this, the overall 1-year<br />
mortality was 18% in the high surgery area compared with<br />
22.8% in the low surgery area (adjusted odds ratio [OR] 0.89;<br />
95% confidence interval [CI] 0.86–0.93). In addition, the<br />
1-year lung cancer specific mortality was lower in the high<br />
surgery area at 12% compared with 16.9% for the low<br />
surgery area (OR 0.86; 95% CI 0.82–0.91). The training<br />
effect and expertise <strong>of</strong> the areas with higher rates <strong>of</strong> surgery<br />
was associated with older and sicker patients undergoing<br />
resection with improved survival. This speaks strongly to<br />
the need for identifying broader expertise and centers <strong>of</strong><br />
excellence for surgical intervention in older patients, both in<br />
the academic and community settings.<br />
Adjuvant Chemotherapy <strong>Clinical</strong> Trials<br />
JEFFREY CRAWFORD<br />
The 5-year survival rate for surgical resection alone in<br />
NSCLC is approximately 55% to 65% for patients with stage<br />
I disease, 40% to 55% for patients with stage II, and 20% to<br />
25% for stage III. Based on the Mountain staging system,<br />
which has served as the basis for the majority <strong>of</strong> adjuvant<br />
chemotherapy trial results currently available, 11 the Lung<br />
Adjuvant Cisplatin Evaluation (LACE) pooled analysis has<br />
reported survival improvement following surgical resection<br />
with the use <strong>of</strong> adjuvant chemotherapy between 4% and<br />
15%. 12 Among these trials, JBR.10 was a study conducted by<br />
the National Cancer Institute <strong>of</strong> Canada, comparing cisplatin<br />
and vinorelbine with observation. A subset analysis has<br />
been reported for the patients older than age 65 in this<br />
trial. 13 Three hundred twenty-seven patients age 65 or<br />
younger were compared to 155 patients older than age 65.<br />
The baseline demographics were similar for the two groups,<br />
with the exception <strong>of</strong> histology with 58% <strong>of</strong> younger patients<br />
having adenocarcinoma compared with 43% <strong>of</strong> the older<br />
patients. Squamous cell cancer was seen in 32% <strong>of</strong> the<br />
younger patients and 49% <strong>of</strong> the older patients (p � 0.001).<br />
Performance status 0 was also more frequent in the younger<br />
population—53% compared with 41% (p � 0.01). Chemotherapy<br />
significantly prolonged overall survival in the older<br />
population with a hazard ratio <strong>of</strong> 0.61 (95% CI, 0.38–0.98).<br />
This benefit was quite similar to the effect seen for the<br />
overall study population. Differences were seen in the mean<br />
dose intensity <strong>of</strong> vinorelbine and cisplatin, with the elderly<br />
patients receiving fewer doses <strong>of</strong> both agents. Fewer elderly<br />
patients completed treatment and more patients refused<br />
treatment. There were no significant differences in toxicities,<br />
hospitalizations, or treatment-related deaths by age<br />
group. There was a higher mortality rate from nonmalignant<br />
causes in the older population—21.1% compared with<br />
11.9% in the younger population, although this did not reach<br />
statistical significance. The authors concluded that, despite<br />
the fact that the older population received less chemotherapy<br />
overall, the improvement in survival <strong>of</strong> adjuvant vinorelbine<br />
and cisplatin was comparable to the result seen in<br />
the younger population and also had acceptable toxicity.
OLDER PATIENTS WITH LUNG CANCER<br />
Fig. 1. Analysis by age <strong>of</strong> survival comparing surgery with chemotherapy<br />
to surgery alone (NSCLC) meta-analysis.<br />
A meta-analysis has also been performed <strong>of</strong> individual<br />
patient data comparing the effects <strong>of</strong> adding adjuvant chemotherapy<br />
to surgery through the NSCLC Meta-analyses<br />
Collaborative Group. 14 The first meta-analysis <strong>of</strong> surgery<br />
plus chemotherapy compared with surgery alone includes 34<br />
trials in 8,447 patients. The overall benefit <strong>of</strong> adding chemotherapy<br />
after surgery was an absolute increase in survival<br />
<strong>of</strong> 4% at 5 years with a hazard ratio <strong>of</strong> 0.86 (95% CI<br />
0.81–0.92; p � 0.0001). Figure 1 demonstrates the hazard<br />
ratio from that study. As shown, in all age subgroups, the<br />
addition <strong>of</strong> chemotherapy to surgery improved survival, with<br />
a trend for the greatest effect in the group older than age 70.<br />
Although this is clearly a subset analysis and must be taken<br />
with caution, it at least suggests that adjuvant chemotherapy<br />
in older patients, who are appropriately selected, as they<br />
would have been in these clinical trials, can provide survival<br />
benefit at least comparable to that <strong>of</strong> younger patients.<br />
Observational Studies <strong>of</strong> Adjuvant Chemotherapy in<br />
the Older Population<br />
How do these results from randomized clinical trials and<br />
meta-analyses <strong>of</strong> these studies apply to clinical practice? An<br />
observational cohort study utilizing the SEER registry and<br />
Medicare database has been reported to address this question.<br />
15 In this study, 3,324 patients older than age 65 were<br />
identified as having surgery for stage II and IIIA NSCLC.<br />
This included cases <strong>of</strong> lung cancer diagnosed up to 2005 with<br />
follow-up data through December <strong>of</strong> 2007. The primary<br />
endpoint was to look at overall survival. In this group <strong>of</strong><br />
patients, 21% received platinum-based chemotherapy.<br />
There was improvement in overall survival for patients who<br />
received chemotherapy, with a hazard ratio <strong>of</strong> 0.78. Beneficial<br />
results were seen in both stage II and stage IIIA<br />
patients. Within age strata, improved survival was seen in<br />
the population younger than age 70 (HR 0.74, CI 0.62–0.88).<br />
There was also improvement in overall survival for the<br />
population age 70 to 79 (HR 0.82, CI 0.71–0.94). However,<br />
no survival benefit was observed in the population older<br />
than age 80 (HR 1.33, CI 0.86–2.06). The use <strong>of</strong> adjuvant<br />
chemotherapy was associated with an increased odds ratio<br />
<strong>of</strong> serious adverse events as determined by hospitalization<br />
(OR 2.0, CI 1.5–2.6). Given the nature <strong>of</strong> this study, there<br />
is no comparison population younger than age 65 and details<br />
<strong>of</strong> exact toxicities are limited. This study extends the observations<br />
from the clinical trial setting to clinical practice<br />
regarding the potential benefit <strong>of</strong> adjuvant chemotherapy in<br />
the elderly. Appropriate caveats also include that the benefit<br />
<strong>of</strong> adjuvant chemotherapy is clearly not established in the<br />
population older than age 80 and the effect <strong>of</strong> adverse events<br />
must be also considered. Overall, there was 3.1% mortality<br />
within 12 weeks <strong>of</strong> treatment for this population.<br />
At the <strong>American</strong> <strong>Society</strong> <strong>of</strong> <strong>Clinical</strong> <strong>Oncology</strong> Annual<br />
Meeting in 2011 (ASCO 2011), Cuffe and colleagues reported<br />
on the patterns <strong>of</strong> use <strong>of</strong> adjuvant chemotherapy among<br />
surgically resected patients with NSCLC in Ontario, with<br />
the focus on the population <strong>of</strong> patients age 70 and older. 16<br />
Although this represents more than 50% <strong>of</strong> the patients with<br />
lung cancer, in JBR.10, only 15% <strong>of</strong> patients were older than<br />
age 70, and the overall LACE analysis included only 9% <strong>of</strong><br />
patients in this age group. This study evaluated the use <strong>of</strong><br />
adjuvant chemotherapy and associated outcomes from 2001<br />
to 2003, before results <strong>of</strong> JBR.10 and other clinical trials<br />
demonstrating the benefit <strong>of</strong> adjuvant chemotherapy were<br />
known. This “preadoption” time period was compared to the<br />
“postadoption” time period <strong>of</strong> 2004 to 2006. The primary<br />
study outcome was overall survival, with a secondary endpoint<br />
<strong>of</strong> rate <strong>of</strong> hospitalization within six months <strong>of</strong> surgery<br />
as a surrogate for toxicity. In this study, 6,570 patients were<br />
identified who underwent surgical resection within 24 weeks<br />
<strong>of</strong> diagnosis. Patients who received neoadjuvant radiation<br />
and/or chemotherapy were excluded, leaving a population <strong>of</strong><br />
6,304 patients. In this group, 3,541 patients were younger<br />
than age 70 with 1,217 patients age 70 to 74, 980 patients<br />
age 75 to 79, and 466 patients older than age 80. Other<br />
variables associated with survival differences included age<br />
and Charlson comorbidity scores. The majority <strong>of</strong> chemotherapy<br />
was delivered in the group younger than age 70.<br />
Very few patients older than age 80 were treated with<br />
chemotherapy. Overall survival by age group clearly favored<br />
the population younger than age 70, with the worse survival<br />
rate in the population older than age 80. However, when the<br />
age groups were compared between the pre- and postadoption<br />
time periods, improvement in survival was seen in both<br />
the population younger than age 70 and older than age 70,<br />
with hazard ratios <strong>of</strong> 0.85 and 0.87 respectively. By comparison,<br />
no difference was seen in the population older than age<br />
80, with a hazard ratio <strong>of</strong> 1.0. Toxicities, defined by hospitalization<br />
within 6 weeks <strong>of</strong> surgery, were higher in the<br />
population older than age 75 (p � 0.001) and ranged between<br />
11% and 18%, reflecting postoperative complications.<br />
By contrast, hospitalization rates between 6 and 24 weeks <strong>of</strong><br />
surgery, when a patient would have received chemotherapy,<br />
were similar across all age groups and varied between 27%<br />
and 32%. The authors suggest that there is an association<br />
between the adoption <strong>of</strong> adjuvant chemotherapy in the older<br />
population and the survival improvement, although the<br />
majority <strong>of</strong> patients did not receive chemotherapy. The<br />
benefit <strong>of</strong> adjuvant chemotherapy in patients older than age<br />
80 was not clarified by this study, since so few patients<br />
received treatment.<br />
An additional study further explored the SEER Medicare<br />
database to form a comparison between carboplatin- and<br />
cisplatin-based regimens. 17 Although cisplatin-based chemotherapy<br />
has been the standard recommended therapy in<br />
the adjuvant setting, the older patient population may have<br />
poor tolerance for this medication or be unable to receive it<br />
317
ecause <strong>of</strong> organ dysfunction, neuropathy, or other comorbidities.<br />
One prospective trial, CALGB 9633, was performed<br />
utilizing carboplatin and paclitaxel in patients with stage IB<br />
disease. 18 However, this trial was underpowered and only<br />
showed a trend toward overall survival. Because <strong>of</strong> this, the<br />
effectiveness <strong>of</strong> carboplatin-based adjuvant chemotherapy<br />
has been questioned. In this particular study, the authors<br />
looked at 3,324 patients age 65 and older from the SEER<br />
Medicare database with resected stage II and IIIA NSCLC<br />
diagnosed between 1992 and 2005. 15 In this study, 636<br />
patients, or 19%, received platinum-based chemotherapy<br />
within 3 months <strong>of</strong> surgery. The overall population had a<br />
significant improvement in survival with a hazard ratio <strong>of</strong><br />
0.79 (95% CI, 0.71–0.89). Among these patients, 105, or<br />
16.5%, received cisplatin-based therapy, and 489 patients,<br />
or 76.9%, received carboplatin-based therapy. The hazard<br />
ratio for cisplatin-based chemotherapy was 0.76 (95% CI,<br />
0.60–0.96). The hazard ratio for carboplatin-based chemotherapy<br />
compared with no adjuvant chemotherapy was 0.76<br />
(95%, CI 0.68–0.86). When cisplatin- and carboplatin-based<br />
regimens were compared directly, there appeared to be no<br />
difference in survival with a hazard ratio for carboplatin <strong>of</strong><br />
0.91 (95% CI, 0.70–1.18). Chemotherapy-related toxicities<br />
were similar between the two platinum-based regimens,<br />
except for a lower rate <strong>of</strong> infection and emesis for the<br />
carboplatin-treated patient and borderline reduction in dehydration,<br />
also favoring carboplatin. This study documented<br />
the common practice <strong>of</strong> substituting carboplatin for cisplatin<br />
in the older patient receiving adjuvant chemotherapy. The<br />
same caveat exists that the healthier patients with better<br />
outcomes may have potentially received chemotherapy.<br />
With that said, the survival appeared similar in those<br />
patients treated with cisplatin- or carboplatin-based therapy.<br />
In this regard, the use <strong>of</strong> a comprehensive geratric<br />
assessment can be helpful in determining both life expectancy<br />
and morbidity from treatment <strong>of</strong> the older patient<br />
with cancer. 22 These tools are being incorporated into prospective<br />
clinical trials <strong>of</strong> older patients with lung cancer and<br />
need to be studied in the surgical population as well so we<br />
can better inform our patients <strong>of</strong> the risks and benefits <strong>of</strong> our<br />
treatments.<br />
Current Status<br />
Although the majority <strong>of</strong> clinical trial data involves<br />
cisplatin-based chemotherapy with vinorelbine, as well as<br />
older regimens such as cisplatin/etoposide, the comparability<br />
<strong>of</strong> cisplatin-based regimens in advanced stage disease<br />
has led to the conclusion that multiple options for platinumbased<br />
chemotherapy can be considered in the adjuvant<br />
setting. The current intergroup study, E1505, allows the<br />
physician a choice <strong>of</strong> cisplatin-based regimens, including<br />
vinorelbine, docetaxel, gemcitabine, or pemetrexed. 19 In<br />
particular, the cisplatin/pemetrexed regimen was added<br />
most recently to this study, based on its efficacy and potential<br />
lower toxicity in advanced stage patients with nonsquamous<br />
NSCLC. 20 A phase II comparison between adjuvant<br />
cisplatin/vinorelbine and cisplatin/pemetrexed was reported<br />
at ASCO 2011. 21 The aim <strong>of</strong> this phase II trial was to<br />
compare dose delivery and clinical feasibility <strong>of</strong> cisplatin/<br />
pemetrexed compared with cisplatin/vinorelbine in the adjuvant<br />
setting. This study was done predominantly in a<br />
younger population with a mean age <strong>of</strong> 59, with 132 patients<br />
randomly selected. However, delivery <strong>of</strong> the intended dose <strong>of</strong><br />
318<br />
JEFFREY CRAWFORD<br />
cisplatin/pemetrexed was higher at 74.6%, compared with<br />
20% for the cisplatin/vinorelbine. The median number <strong>of</strong><br />
cycles <strong>of</strong> treatment was four <strong>of</strong> cisplatin/pemetrexed compared<br />
with three for the cisplatin/vinorelbine arm. The time<br />
to withdrawal from therapy differed significantly, favoring<br />
the cisplatin/pemetrexed arm (p � 0.001). In early follow-up<br />
at 4 months, there had been two deaths on the cisplatin/<br />
vinorelbine arm and one on the cisplatin/pemetrexed arm.<br />
Although this study has a very small number <strong>of</strong> patients, it<br />
does help support the option <strong>of</strong> cisplatin/pemetrexed in the<br />
current ECOG 1505 study that hopefully will lead to more<br />
prospective data in the older population.<br />
In this regard, it is important that we support enrollment<br />
<strong>of</strong> patients on ECOG 1505 across all eligible age populations.<br />
There is currently no large, randomized age-specific adjuvant<br />
trial to provide the robust database we need to better<br />
inform our patients and ourselves <strong>of</strong> the risk/benefit ratio <strong>of</strong><br />
treatment <strong>of</strong> this population. However, by enrolling older<br />
eligible patients in the current adjuvant trial, we will have<br />
the opportunity to develop a significant database <strong>of</strong> toxicities<br />
and long-term outcomes in the older patients compared with<br />
younger patients. An interim report on the demographics<br />
and toxicities <strong>of</strong> E1505 was reported at ASCO 2011. 19<br />
Although the median age is 61 to 62, it is notable that a<br />
substantial percentage <strong>of</strong> patients were in their seventies,<br />
with patients enrolled up to ages 84 and 86. This trial is<br />
comparing standard cisplatin chemotherapy doublets with<br />
or without the addition <strong>of</strong> 1 year <strong>of</strong> bevazucimab therapy.<br />
Differences in the bevazucimab arm in terms <strong>of</strong> toxicity<br />
include higher rates <strong>of</strong> neutropenia, as well as hypertension<br />
and proteinuria. It is also important to recognize that across<br />
all age groups, grade 5 toxicities <strong>of</strong> these regimens occurred<br />
in 2.2% <strong>of</strong> the control arm and 3.3% in the bevazucimab arm.<br />
This speaks to the importance <strong>of</strong> appropriate patient selection,<br />
careful follow-up, and management with state-<strong>of</strong>-theart<br />
supportive care to minimize deaths from toxicity in this<br />
potentially curative setting.<br />
Compared with breast cancer, colon cancer, and other<br />
areas where adjuvant chemotherapy has been a standard<br />
for decades, the incorporation <strong>of</strong> adjuvant chemotherapy is<br />
relatively recent in the therapeutic history <strong>of</strong> NSCLC. The<br />
toxicities <strong>of</strong> our current cytotoxic regimens, along with the<br />
vulnerability <strong>of</strong> our older patients—with and without comorbid<br />
disease—need better definition than is currently available.<br />
Subset analyses <strong>of</strong> randomized clinical trials and<br />
meta-analyses, as well as observational studies, do provide<br />
direction for the practicing oncologist regarding older patients<br />
and the treatment options in the setting <strong>of</strong> early stage<br />
NSCLC. Exciting advances with targeted therapeutic approaches<br />
based on the molecular biology <strong>of</strong> the cancer have<br />
occurred in patients with advanced stage lung cancer. These<br />
same approaches clearly need to be studied prospectively in<br />
patients with early stage lung cancer across all age groups.<br />
The potential benefit <strong>of</strong> reduced toxicity <strong>of</strong> these agents may<br />
be particularly important in the older population. In the<br />
meantime, the principles learned from other populations<br />
treated with adjuvant chemotherapy should be heeded.<br />
There is no evidence currently for nonplatinum-based chemotherapy<br />
in the adjuvant setting, although the substitution<br />
<strong>of</strong> carboplatin for cisplatin may be an appropriate option<br />
for some older patients. Although evidence supports the use<br />
<strong>of</strong> adjuvant chemotherapy in appropriate patients older<br />
than age 70, there is insufficient evidence to recommend
OLDER PATIENTS WITH LUNG CANCER<br />
adjuvant chemotherapy at this point to patients age 80 and<br />
older. The potential importance <strong>of</strong> delivering standard-dose<br />
therapy compared with potential myelotoxicity and compli-<br />
Author’s Disclosure <strong>of</strong> Potential Conflicts <strong>of</strong> Interest<br />
Author<br />
Employment or<br />
Leadership<br />
Positions<br />
Consultant or<br />
Advisory Role<br />
Jeffrey Crawford Agennix; Amgen;<br />
Genentech<br />
1. Ries LAG, Melbert D, Krapcho M, et al. (eds.). SEER Cancer Statistics<br />
Review, 1975–2005. Bethesda, MD: National Cancer Institute; 2005.<br />
2. Karn<strong>of</strong>sky D, Burchenal J. The <strong>Clinical</strong> Evaluation <strong>of</strong> Chemotherapeutic<br />
Agents in Cancer. In MacLeod CM (ed). Evaluation <strong>of</strong> Chemotherapeutic<br />
Agents. New York, NY: Columbia University Press, 1949; 196.<br />
3. Weiss J, Langer C. NSCLC in the elderly—the legacy <strong>of</strong> therapeutic<br />
neglect. Current Treatment Options in <strong>Oncology</strong>. 2009;10:180-194.<br />
4. Cheitlin M. Cardiovascular physiology-changes with aging. Am J Geriatr<br />
Cardiol. 2003;12:9-13.<br />
5. Mühlberg W, Platt D. Age-dependent changes <strong>of</strong> the kidneys: Pharmacological<br />
implications. Gerontology. 1999; 45:243-253.<br />
6. Dees E, O’Reilly SN, Goodman S, et al. A prospective pharmacologic<br />
evaluation <strong>of</strong> age-related toxicity <strong>of</strong> adjuvant chemotherapy in women with<br />
breast cancer. Cancer Invest. 2000;18:521-529.<br />
7. Fletcher C, Peto R. The natural history <strong>of</strong> chronic airflow obstruction.<br />
BMJ. 1977;1:1645-1648.<br />
8. Rajdev L, Keller S. Surgery for Lung Cancer in Elderly Patients. In<br />
Govindan R (ed). ASCO Educational Book. Alexandria, VA: <strong>American</strong> <strong>Society</strong><br />
<strong>of</strong> <strong>Clinical</strong> <strong>Oncology</strong>, 2006; 463-467.<br />
9. Rivera C, Falcoz PE, Bernard A, et al. Surgical management and<br />
outcomes <strong>of</strong> elderly patients with early stage non-small cell lung cancer: a<br />
nested case-control study. Chest. 2011;140:874-880.<br />
10. Gray S, Landrum M, Lamont E. Improved outcomes associated with<br />
higher surgery rates for older patients with early stage non-small cell lung<br />
cancer. Cancer. Epub 2011 Jul 28.<br />
11. Mountain CF. Revisions in the international system for staging lung<br />
cancer. Chest. 1997;111:1710-1717.<br />
12. Pignon J, Tribodet H, Scagliotti G, et al. Lung adjuvant cisplatin<br />
evaluation: A pooled analysis by the LQCE Collaborative Group. J Clin Oncol<br />
2008;26:3552-3559.<br />
13. Pepe C, Hasan B, Winton T, et al. Adjuvant vinorelbine and cisplatin in<br />
elderly patients: National Cancer Institute <strong>of</strong> Canada and Intergroup Study<br />
JBR. 10. J Clin Oncol. 2007;25:1553-1561.<br />
cations in the older postoperative patient with comorbid<br />
disease demands thoughtful and proactive patient management<br />
and supportive care.<br />
Stock<br />
Ownership Honoraria<br />
REFERENCES<br />
Research<br />
Funding<br />
Agennix; Amgen<br />
Expert<br />
Testimony<br />
Other<br />
Remuneration<br />
14. NSCLC Meta-analyses Collaborative Group, Arriagada R, Auperin A,<br />
et al. Adjuvant chemotherapy, with or without postoperative radiotherapy, in<br />
operable non-small-cell lung cancer: two meta-analyses <strong>of</strong> individual patient<br />
data. Lancet. 2010;375:1267-1277.<br />
15. Wisnivesky J, Smith C, Packer S, et al. Survival and risk <strong>of</strong> adverse<br />
events in older patients receiving postoperative adjuvant chemotherapy for<br />
resected stages II-IIIA lung cancer: Observational cohort study. BMJ. 2011;<br />
343:d4013.<br />
16. Cuffe S, Booth C, Peng Y, et al. Adoption <strong>of</strong> adjuvant chemotherapy<br />
(ACT) for non-small cell lung cancer (NSCLC) in the elderly: A populationbased<br />
outcomes study. J Clin Oncol. 2011;29(suppl; abstr 7012)<br />
17. Gu F, Strauss G, Wisnivesky J. Platinum-based adjuvant chemotherapy<br />
(ACT) in elderly patients with non-small cell lung cancer (NSCLC) in the<br />
SEER-Medicare database: Comparison between carboplatin- and cisplatinbased<br />
regimens. J Clin Oncol. 2011;29(suppl; abstr 7014).<br />
18. Strauss G, Herndon J, Maddeus Ma, et al. Adjuvant paclitaxel plus<br />
carboplatin compared with observation in stage IB non-small-cell lung cancer:<br />
CALGB 9633 with the Cancer and Leukemia Group B, Radiation Therapy<br />
<strong>Oncology</strong> Group and North Central Cancer Treatment Group Study Groups.<br />
J Clin Oncol. 2008;26:5043-5051.<br />
19. Wakelee H, Dahlberg S, Keller D, et al. Interim report <strong>of</strong> on-study<br />
demographics and toxicity from E1505, a phase III randomized trial <strong>of</strong><br />
adjuvant (adj) chemotherapy (chemo) with or without bevacizumab (B) for<br />
completely resected early-stage non-small cell lung cancer (NSCLC). J Clin<br />
Oncol. 2011;29(suppl; abstr 7013)<br />
20. Scagliotti GV, Parikh P, von Pawel J, et al. Phase III study comparing<br />
cisplatin plus gemcitabine with cisplatin plus pemetrexed in chemotherapynaïve<br />
patients with advanced stage non-small cell lung cancer. J Clin Oncol.<br />
2008;26:3543-3551.<br />
21. Kreuter M, Vansteenkiste J, Fischer J Randomized phase II trial on<br />
refinement <strong>of</strong> early-stage NSCLC adjuvant chemotherapy with cisplatin and<br />
pemetrexed (CPx) versus cisplatin and vinorelbine (CVb): TREAT. J Clin<br />
Oncol. 2011;29(suppl; abstr 7002)<br />
319
TREATMENT OF OLDER PATIENTS WITH<br />
ADVANCED CANCER: BALANCING EFFICACY<br />
WITH TOXICITY (eQ&A)<br />
CHAIR<br />
Supriya Gupta Mohile, MD, MS<br />
University <strong>of</strong> Rochester Medical Center<br />
Rochester, NY<br />
SPEAKERS<br />
Heidi D. Klepin, MD, MS<br />
Wake Forest University School <strong>of</strong> Medicine<br />
Winston-Salem, NC<br />
Arati V. Rao, MD<br />
Duke University Medical Center<br />
Durham, NC
Considerations and Controversies in the<br />
Management <strong>of</strong> Older Patients with<br />
Advanced Cancer<br />
By Supriya Gupta Mohile, MD, MS, Heidi D. Klepin, MD, MS, and Arati V. Rao, MD<br />
Overview: The incidence <strong>of</strong> cancer increases with age. Oncologists<br />
need to be adept at assessing physiologic and<br />
functional capacity in older patients in order to provide safe<br />
and efficacious cancer treatment. Assessment <strong>of</strong> underlying<br />
health status is especially important for older patients with<br />
advanced cancer, for whom the benefits <strong>of</strong> treatment may be<br />
low and the toxicity <strong>of</strong> treatment high. The comprehensive<br />
geriatric assessment (CGA) is the criterion standard for evaluation<br />
<strong>of</strong> the older patient. The combined data from the CGA<br />
can be used to stratify patients into categories to better<br />
predict risk for chemotherapy toxicity as well as overall<br />
APPROXIMATELY 60% <strong>of</strong> all cancers and 70% <strong>of</strong><br />
cancer-related deaths occur in persons age 65 and<br />
older. 1 It is uncertain whether there is a benefit to initiating<br />
treatment because data are lacking on efficacy and safety <strong>of</strong><br />
the proposed treatment in older patients with health issues<br />
other than cancer. The Comprehensive Geriatric Assessment<br />
(CGA) is a tool used by geriatricians to assess functional<br />
status, comorbidity, cognition, social support system,<br />
nutrition, and medication use. Results from the CGA can aid<br />
oncologists in predicting outcomes and selecting appropriate<br />
treatment regimens for their patients. The CGA can also<br />
help identify and follow-up on symptoms in older patients<br />
that can affect quality <strong>of</strong> life. In this chapter, we discuss the<br />
CGA and its role in the care <strong>of</strong> older patients with advanced<br />
cancer as well as treatment dosing considerations and management<br />
<strong>of</strong> under-recognized cancer-related symptoms in<br />
older patients.<br />
CGA<br />
Although the commonly used Karn<strong>of</strong>sky Performance Status<br />
and Eastern Cooperative <strong>Oncology</strong> Group (ECOG) performance<br />
measures correlate with treatment toxicity, these<br />
tools alone do not predict outcomes as well as the CGA in<br />
the older population. The CGA, a compilation <strong>of</strong> standardized<br />
tools to assess geriatric domains, can help characterize<br />
physiologic age and can detect unsuspected conditions<br />
that may affect cancer treatment in more than 50% <strong>of</strong> older<br />
patients.<br />
Components <strong>of</strong> the CGA<br />
Each domain within the CGA has been shown to predict<br />
morbidity and mortality in community-dwelling older adults<br />
(Table 1). Dependence for Activities <strong>of</strong> Daily Living and<br />
Instrumental Activities <strong>of</strong> Daily Living (ADLs, IADLs) has<br />
been shown to be predictive <strong>of</strong> mortality in geriatric oncology,<br />
2 and the incidence <strong>of</strong> ADL and IADL deficiencies is<br />
higher for older patients with cancer than for age-matched<br />
controls. 3 The geriatrics literature supports the use <strong>of</strong> a<br />
directly observed assessment <strong>of</strong> physical function in order to<br />
assess the risk <strong>of</strong> falls and identify vulnerability in older<br />
patients who may otherwise seem “fit.” The prevalence <strong>of</strong><br />
comorbidity increases with age and can affect survival <strong>of</strong><br />
patients with advanced cancer. In a study <strong>of</strong> 19,268 patients<br />
with newly diagnosed cancer, decreased duration <strong>of</strong> survival<br />
outcomes. The CGA can also be used to identify and follow-up<br />
on possible functional consequences from treatment. A variety<br />
<strong>of</strong> screening tools might be useful in the oncology practice<br />
setting to identify patients who may benefit from further<br />
testing and intervention. In this chapter, we discuss how the<br />
principles <strong>of</strong> geriatrics can help improve the clinical care <strong>of</strong><br />
older adults with advanced cancer. Specifically, we discuss<br />
assessing tolerance for treatment, options for chemotherapy<br />
scheduling and dosing for older patients with advanced cancer,<br />
and management <strong>of</strong> under-recognized symptoms in older<br />
patients with cancer.<br />
was associated with comorbid conditions. 4 In communitydwelling<br />
older individuals, the prevalence <strong>of</strong> polypharmacy<br />
ranges from 15.6% to 94.3%. 5 Polypharmacy can increase<br />
the risk <strong>of</strong> adverse events from chemotherapy. Weight loss<br />
is a marker for declining nutritional status and is <strong>of</strong>ten<br />
observed in the geriatric population, particularly in people<br />
who are frail. 6 In studies <strong>of</strong> community-dwelling older people,<br />
there was a two-fold increased risk <strong>of</strong> death among<br />
people who had lost 5% <strong>of</strong> their body weight. 7 Approximately<br />
20% <strong>of</strong> community-dwelling older adults screen positively<br />
for some degree <strong>of</strong> cognitive disorder. The presence <strong>of</strong> cognitive<br />
disorders, particularly more advanced disease, may<br />
limit life expectancy. Because cognitive issues are common<br />
in older adults, screening for impairment prior to initiating<br />
treatment is necessary in order to appropriately evaluate<br />
whether patients have capacity for informed consent. Additionally,<br />
patients with cognitive disorders may have more<br />
difficulty reporting treatment-related side effects. Depression<br />
and social isolation are important prognostic factors in<br />
older patients undergoing treatment for cancer.<br />
Prediction <strong>of</strong> Outcomes in Older Patients with Advanced Cancer<br />
Similar to the situation with community-dwelling older<br />
adults, several geriatric domains have been shown to influence<br />
survival for older patients with cancer. For example,<br />
in the cancer setting, weight loss, malnutrition, and IADL<br />
deficits before diagnosis have been associated with worse<br />
overall survival rates. 8 Kanesvaran and colleagues evaluated<br />
the effect <strong>of</strong> CGA domains on overall survival for<br />
patients with advanced cancer and developed a prognostic<br />
scoring system for survival for use by clinicians. Factors that<br />
were independently associated with overall survival included<br />
low albumin level, ECOG performance status <strong>of</strong> 2 or<br />
higher, positive geriatric depression screen, advanced stage<br />
<strong>of</strong> disease, malnutrition, and older age.<br />
From the Geriatric <strong>Oncology</strong> Program at the James Wilmot Cancer Center, University <strong>of</strong><br />
Rochester Medical Center, Rochester, NY; Wake Forest School <strong>of</strong> Medicine, Winston-Salem,<br />
NC; Division <strong>of</strong> Geriatrics, Duke University Medical Center, Durham NC.<br />
Authors’ disclosures <strong>of</strong> potential conflicts <strong>of</strong> interest are found at the end <strong>of</strong> this article.<br />
Address reprint requests to Supriya Gupta Mohile, MD, MS, Geriatric <strong>Oncology</strong> Program<br />
at the James Wilmot Cancer Center, University <strong>of</strong> Rochester Medical Center, 601 Elmwood<br />
Ave, Box 704, Rochester, NY 14642; email: supriya_mohile@urmc.rochester.edu.<br />
© <strong>2012</strong> by <strong>American</strong> <strong>Society</strong> <strong>of</strong> <strong>Clinical</strong> <strong>Oncology</strong>.<br />
1092-9118/10/1-10<br />
321
Two important studies have demonstrated the predictive<br />
value <strong>of</strong> CGA domains for estimating the risk <strong>of</strong> severe<br />
toxicity from chemotherapy. One <strong>of</strong> these studies was conducted<br />
by the Cancer and Aging Research Group at seven<br />
institutions. 9 A CGA was completed before chemotherapy<br />
in 500 patients who were 65 or older and had cancer. Grade<br />
3–5 toxicity related to the chemotherapy occurred in 53%<br />
(50% grade 3, 12% grade 4, 2% grade 5). Risk factors for<br />
grade 3–5 toxicity included age <strong>of</strong> 73 or older, cancer type<br />
(gastrointestinal or genitourinary), standard dose, polyche-<br />
KEY POINTS<br />
Table 1. Domains <strong>of</strong> Geriatric Assessment and Measurement Options<br />
Domains Definition Measurement Options<br />
Function/Physical<br />
Performance<br />
- Ability to carry out one’s self care needs to live independently at home Activities <strong>of</strong> Daily Living<br />
- Ability to care for tasks that allow independence in the community Instrumental Activities <strong>of</strong> Daily Living<br />
- Physical performance is an objective evaluation <strong>of</strong> mobility, balance, and fall risk. History <strong>of</strong> falls<br />
Timed Up and Go<br />
Short Physical Performance Battery<br />
Handgrip testing<br />
Comorbidity/Pharmacy - Chronic diseases may influence life expectancy and tolerance to cancer treatment. Charlson Comorbidity Scale<br />
Medications can increase risk <strong>of</strong> adverse events with cancer treatment. Cumulative Illness Scale-Geriatrics<br />
Comorbidity count and severity<br />
Medication Count<br />
Beers Criteria<br />
Cognition Cognitive deficits are common in older patients and may affect decision-making<br />
capacity and interfere with cancer treatment.<br />
Psychological Status - Depression and anxiety are independently associated with adverse outcomes in<br />
patients with cancer.<br />
Nutrition - Weight loss and anorexia affect tolerance to treatment and survival in older<br />
patients with cancers.<br />
Social Support - Adequate social support is necessary for older patients to successfully undergo<br />
treatment.<br />
● Because the incidence <strong>of</strong> cancer increases with age,<br />
oncologists need to be adept at assessing physiologic<br />
and functional capacity in older patients in order to<br />
provide safe and efficacious cancer treatment.<br />
● The combined data from the Comprehensive Geriatric<br />
Assessment can be used to stratify older patients<br />
into categories to better predict risk for chemotherapy<br />
toxicity as well as overall outcomes.<br />
● Older adults frequently receive therapy <strong>of</strong> a lower<br />
dose intensity, in part because <strong>of</strong> concerns regarding<br />
toxicity.<br />
● Modifications <strong>of</strong> chemotherapy dose intensity and<br />
treatment schedules for older adults considered to be<br />
unfit require attention in clinical trials to inform the<br />
risks and benefits <strong>of</strong> therapy.<br />
● Supportive care to prevent cancer-related symptoms<br />
is essential for the care <strong>of</strong> older patients with cancer.<br />
● Cachexia and sarcopenia, chemotherapy-associated<br />
peripheral neuropathy, and cancer-related fatigue<br />
are symptoms that are under-recognized in older<br />
patients and warrant further study.<br />
322<br />
Mini-Mental Status Examination<br />
Blessed-Orientation-Memory Scale<br />
Short Portable Mental Status Questionnaire<br />
Montreal Cognitive Assessment<br />
Geriatric Depression Scale<br />
Hospital Anxiety and Depression Scale<br />
Mini-Nutritional Assessment<br />
Weight loss<br />
Body Mass Index<br />
Needs assessment <strong>of</strong> financial capabilities,<br />
transportation, and caregiver status.<br />
Medical Outcomes Survey Social Support<br />
motherapy, fall within the past 6 months, assistance needed<br />
for IADLs, and decreased social activity. A risk stratification<br />
schema (number <strong>of</strong> risk factors: percentage <strong>of</strong> grade 3–5<br />
toxicity) was developed. Further validation <strong>of</strong> this schema is<br />
in progress.<br />
In the second study, use <strong>of</strong> the Chemotherapy Risk Assessment<br />
Scale for High-Age Patients (CRASH) was evaluated<br />
in more than 500 patients who were age 70 or older. 10<br />
A CGA was completed before chemotherapy was started.<br />
Toxicity <strong>of</strong> the chemotherapy regimen was adjusted using an<br />
index to estimate the average per-patient risk <strong>of</strong> chemotherapy<br />
toxicity (the MAX2 index): severe toxicity occurred in<br />
64% <strong>of</strong> patients. The best model for hematologic toxicity<br />
included IADL score, serum lactate dehydrogenase level,<br />
diastolic blood pressure, and chemotherapy toxicity. The<br />
best predictive model for nonhematologic toxicity included<br />
performance status, Mini-Mental Status score, Mini-<br />
Nutritional Assessment score, and chemotherapy toxicity.<br />
Information from two-thirds <strong>of</strong> the patients was used to<br />
develop the risk stratification scheme, and the tool was<br />
validated in the remaining one-third.<br />
Overall, these predictive risk stratification schemes allow<br />
clinicians to identify patients at highest risk for chemotherapy<br />
toxicity and can be used in future research to apply<br />
interventions to reduce chemotherapy toxicity in vulnerable<br />
older populations.<br />
Models <strong>of</strong> Care <strong>of</strong> the Older Patient with<br />
Advanced Cancer<br />
MOHILE, KLEPIN, AND RAO<br />
Despite recent studies demonstrating the feasibility <strong>of</strong> the<br />
CGA in oncology, adoption as the standard <strong>of</strong> care has been<br />
slow because <strong>of</strong> lack <strong>of</strong> resources, difficulties with interpretation<br />
<strong>of</strong> results, and challenges in implementing targeted<br />
interventions in specialty clinic settings such as oncology. A<br />
short, simple, validated screening procedure could quickly<br />
identify patients who are at risk for morbidity or death in<br />
the oncology clinic. Partnering with geriatricians or a geri-
OLDER PATIENTS WITH ADVANCED CANCER<br />
atric oncology consultative service can help provide geriatric<br />
assessment.<br />
Screening for Impairments in the <strong>Oncology</strong> Clinic<br />
Several screening tools have been evaluated in oncology<br />
clinical care and research. The Vulnerable Elders Survey-13<br />
(VES-13) is a self-administered survey that consists <strong>of</strong> one<br />
question for age and an additional 12 items assessing<br />
self-rated health, functional capacity, and physical performance.<br />
11 In 400 community-dwelling older adults, increasing<br />
VES-13 scores strongly predicted death and functional<br />
decline. 12 In a national cohort, a high proportion (45.8%) <strong>of</strong><br />
elders with a history <strong>of</strong> cancer also scored as “vulnerable”<br />
on the VES-13. 3 Luciani and colleagues conducted a study in<br />
a population <strong>of</strong> 419 patients age 70 and older with a history<br />
<strong>of</strong> any type <strong>of</strong> solid or hematologic cancer; 13 the sensitivity<br />
and specificity <strong>of</strong> VES-13 was 87% and 62%, respectively,<br />
compared with that <strong>of</strong> the CGA.<br />
The Groningen Frailty Indicator (GFI) is a 15-item survey<br />
that includes questions focusing on mobility/physical fitness,<br />
vision/hearing, nutrition, comorbidity, cognition, and psychosocial<br />
health. 14 The GFI score has been shown to correlate<br />
moderately with CGA results. 15 In a study by Aaldriks<br />
and colleagues, the mortality rate after initiation <strong>of</strong> chemotherapy<br />
was increased for patients with higher baseline GFI<br />
scores. 16 The GFI has also been shown to be predictive <strong>of</strong><br />
outcome in older patients with non-small cell lung cancer<br />
treated with platinum-based doublet chemotherapy. 17<br />
The G8 is an eight-item questionnaire designed to assess<br />
domains <strong>of</strong> nutrition, mobility, cognitive deficit, polypharmacy,<br />
age, and self-perceived health status. Soubeyran and<br />
colleagues recommend a score <strong>of</strong> 14 as a predictor <strong>of</strong> CGA<br />
deficits (90% sensitivity and 60% specificity). 18 In one study,<br />
sensitivity <strong>of</strong> the G8 for identifying deficits in older patients<br />
with cancer was found to be superior to that <strong>of</strong> the VES-13. 18<br />
Because these results have not been consistent, screening<br />
tools should not serve as a substitute for a full CGA.<br />
However, if a full CGA cannot be undertaken, screening<br />
tools may serve as a way to capture some important information<br />
within geriatric domains.<br />
Multidisciplinary Approach to Care<br />
A few studies have demonstrated that combining geriatric<br />
and oncologic approaches can affect treatment decision making<br />
for patients with advanced cancer. In the ELCAPA<br />
study, a geriatrician performed an extensive CGA and<br />
proposed a geriatric intervention plan for older patients with<br />
cancer. 19 After the CGA, the initial cancer treatment plan<br />
was modified for 78 (20.8%) <strong>of</strong> 375 patients, usually to<br />
decrease treatment intensity. In another study <strong>of</strong> 161 patients<br />
(more than 50% <strong>of</strong> whom had advanced cancer),<br />
cancer treatment was changed in 49% <strong>of</strong> patients. 20 In a<br />
pilot study, Horgan and colleagues demonstrated that although<br />
most eligible older patients were not referred for<br />
geriatric assessment, when such assessment was done, the<br />
results guided initial decision making. 21 More data are<br />
necessary to determine whether CGA and interventions can<br />
improve outcomes. It is likely that support from multidisciplinary<br />
expertise, including social work, physical therapy,<br />
occupational therapy, and nutrition can help develop CGAguided<br />
interventions for an at-risk older adult with cancer.<br />
For example, review <strong>of</strong> medication usage by a pharmacist<br />
can decrease suboptimal prescribing and potentially lead to<br />
a decrease <strong>of</strong> adverse drug events. 22<br />
Dosing Considerations for Older Patients with<br />
Advanced Cancer<br />
Age-related physiologic changes and comorbid disease can<br />
increase the risk <strong>of</strong> chemotherapy toxicity. The balance<br />
between risk <strong>of</strong> treatment and potential benefit is particularly<br />
challenging when treating with noncurative intent.<br />
With respect to treatment characteristics, both standard<br />
dosing <strong>of</strong> chemotherapy and polychemotherapy have been<br />
associated with increased toxicity among older patients with<br />
cancer. 9 Dose modification, therefore, is one consideration<br />
for treatment planning to minimize the negative consequences<br />
<strong>of</strong> toxic therapies. Given the paucity <strong>of</strong> clinical trial<br />
data specific to older patients, there are more questions than<br />
answers regarding dose-modification strategies, particularly<br />
in the setting <strong>of</strong> first-line treatment for patients with advanced<br />
cancer.<br />
How Common Is Dose Modification for Older Adults<br />
with Advanced Cancer?<br />
Although several large population-based studies have<br />
shown increased age as a risk factor for lower relative-dose<br />
intensity during adjuvant chemotherapy, few data have<br />
been collected in the metastatic setting. In a retrospective<br />
single-insitution study <strong>of</strong> older patients (mean age, 74.6<br />
years, ECOG score, 0–2) with advanced cancer, 44% had a<br />
dose modification either at initiation <strong>of</strong> treatment or because<br />
<strong>of</strong> toxicity. 23 Gajra and colleagues studied factors associated<br />
with primary reduction <strong>of</strong> chemotherapy dosing during the<br />
first course <strong>of</strong> treatment among older adults receiving palliative<br />
chemotherapy in a secondary analysis <strong>of</strong> a multisite<br />
observational study. 24 Almost one-third (29%) <strong>of</strong> the 319<br />
patients had a primary reduction <strong>of</strong> chemotherapy dosing.<br />
Older age, a primary diagnosis <strong>of</strong> lung cancer, and comorbid<br />
conditions were the factors independently associated with a<br />
reduction in this cohort.<br />
Indications for Dose Adjustment Related to<br />
Aging Physiology<br />
Physiologic changes associated with aging have implications<br />
for chemotherapy toxicity among older adults. Aging is<br />
associated with decreased intestinal absorption, changes in<br />
volume <strong>of</strong> distribution, decreased hepatic metabolism, and<br />
impaired renal excretion. The degree to which these changes<br />
have clinical significance can vary greatly within an older<br />
population. Among the changes, a change in renal function<br />
is the most well described and needs to be accounted for<br />
in dosing considerations. Many chemotherapy drugs are<br />
cleared through the kidneys and it is well-documented that<br />
serum creatinine can provide a substantial underestimation<br />
<strong>of</strong> renal function in older adults. A creatinine clearance<br />
should be calculated for all older adults before chemotherapy<br />
is initiated, to inform dose adjustment for drugs cleared<br />
through the kidneys. Multiple guidelines provide specific<br />
recommendations for renal dose adjustment, including a<br />
position paper published by the International <strong>Society</strong> <strong>of</strong><br />
Geriatric <strong>Oncology</strong> (SIOG). 25<br />
323
What Is the Benefit Versus Cost <strong>of</strong> Dose Modification<br />
in the Metastatic Setting? <strong>Clinical</strong> Trial Evidence in<br />
Colorectal Cancer<br />
Depending on the rationale for dose modification, treatment<br />
efficacy may be compromised by an effort to decrease<br />
toxicity. In contrast, quality <strong>of</strong> life may be substantially<br />
improved by avoiding treatment toxicity. In general, guideline<br />
organizations such as the National Comprehensive<br />
Cancer Network (NCCN) do not recommend dose attenuation<br />
during first-line therapy based on chronologic age if<br />
the patient is “fit.” Many patients seen in clinical practice,<br />
however, are less fit than those enrolled on trials, making<br />
the judgment regarding risks and benefits <strong>of</strong> standard<br />
treatment much more complicated.<br />
A recently published trial addresses, in part, the issue <strong>of</strong><br />
dose reduction <strong>of</strong> first-line treatment. This study, by Seymour<br />
and colleagues, involved the use <strong>of</strong> a 2�2 factorial<br />
design to randomly assign 459 frail (considered by their<br />
physician to be unfit for full-dose combination chemotherapy)<br />
older adults with metastatic colorectal cancer to one <strong>of</strong><br />
four first-line systemic therapies using an attenuated starting<br />
dose (80% <strong>of</strong> standard). 26 The study design allowed for<br />
escalation to full-dose therapy after 6 weeks if tolerated.<br />
The four treatments were 48-hour intravenous fluorouracil<br />
(5FU), oxaliplatin plus 5FU, capecitabine alone, or oxaliplatin<br />
plus capecitabine. In addition to a standard quality-<strong>of</strong>life<br />
questionnaire, a composite outcome <strong>of</strong> overall treatment<br />
utility was incorporated in an attempt to capture patient<br />
and physician satisfaction with the outcome <strong>of</strong> each treatment<br />
decision.<br />
The median age <strong>of</strong> the patients was 74, and 13% were<br />
older than 80. The primary reason the patients were considered<br />
unfit for standard therapy was frailty or older chronologic<br />
age. During protocol treatment, few patients had dose<br />
escalation and 49% required additional dose reduction (below<br />
the 80% <strong>of</strong> the standard dose). There was a trend toward<br />
benefit with use <strong>of</strong> oxaliplatin in this attenuated dosing<br />
schedule, although the primary outcome was not significant<br />
(median progression-free survival, 5.8 months vs. 4.5<br />
months; p � 0.07). No substantial increase in grade 3<br />
toxicity was associated with use <strong>of</strong> oxaliplatin, but oxaliplatin<br />
was associated with increased overall treatment utility,<br />
suggesting a palliative benefit. Alternatively, capecitabine<br />
was equivalent in efficacy to 5FU but <strong>of</strong>fered no benefit in<br />
quality <strong>of</strong> life and was also associated with a substantial<br />
increase in toxicity. This study provides evidence for treatment<br />
<strong>of</strong> unfit older adults with metastatic colorectal cancer<br />
with an attenuated chemotherapy regimen and introduces<br />
additional outcome measures to help quantify the palliative<br />
benefit a patient may receive from therapy.<br />
In addition to dose reduction, questions regarding optimal<br />
dosing schedules, including the option for treatment break,<br />
for older patients are also common. Again, few studies have<br />
addressed this issue specific to older patients, yet older<br />
patients are most likely to have debilitating consequences<br />
from ongoing therapy. For colorectal cancer, some data<br />
can be extrapolated from existing studies. For patients fit<br />
enough to receive combination chemotherapy with an<br />
oxaliplatin-based regimen (FOLFOX), data from randomized<br />
trials support de-escalation to 5FU maintenance after<br />
3 months <strong>of</strong> oxaliplatin. Although a fraction <strong>of</strong> patients<br />
enrolled on these studies 27,28 were older than 75, there is<br />
324<br />
MOHILE, KLEPIN, AND RAO<br />
suggestion <strong>of</strong> similar benefit compared with younger patients<br />
when this strategy is used. 29 Additionally, data from<br />
two randomized trials suggest that selected patients may<br />
retain the efficacy <strong>of</strong> first-line combination therapy with<br />
oxaliplatin- or irinotecan-based regimens despite prescribed<br />
chemotherapy-free intervals. 30,31 Although the evidence is<br />
limited, it is reasonable to incorporate these data into<br />
treatment planning for older patients in an attempt to<br />
minimize the negative consequences <strong>of</strong> therapy on quality <strong>of</strong><br />
life.<br />
Supportive Care Interventions to Maximize Treatment<br />
Efficacy in Older Patients with Advanced Cancer<br />
Several organizations including NCCN and ASCO have<br />
developed guidelines to help manage common side effects.<br />
We focus here on three areas that cause substantial distress<br />
for older patients with cancer and require further research:<br />
cachexia and sarcopenia, chemotherapy-associated peripheral<br />
neuropathy and falls, and cancer-related fatigue. 32,33<br />
Cancer Cachexia and Sarcopenia<br />
Sarcopenia is the progressive generalized loss <strong>of</strong> skeletal<br />
muscle mass, strength, and function. Cachexia has no uniform<br />
definition and is a complex metabolic syndrome that<br />
is characterized by weight loss <strong>of</strong> more than 10%, reduced<br />
food intake (�1,500 kcal/d), and systemic inflammation<br />
(C-reactive protein level �10 mg/L). 34 It is estimated that<br />
50% <strong>of</strong> people older than 80 have sarcopenia. Half <strong>of</strong> all<br />
patients with cancer lose some body weight; one third lose<br />
more than 5% <strong>of</strong> body weight, and as many as 20% <strong>of</strong> all<br />
cancer deaths are caused directly by cachexia. 34<br />
Figure 1 illustrates how the pathophysiology <strong>of</strong> cancer<br />
cachexia and sarcopenia are intertwined. 35 These syn-<br />
Fig 1. Interaction and consequences between elderly host-tumortherapy.<br />
Abbreviations: ADLs, activities <strong>of</strong> daily living.
OLDER PATIENTS WITH ADVANCED CANCER<br />
Table 2. Characteristics <strong>of</strong> <strong>Oncology</strong> Agents that Cause Peripheral Neuropathy<br />
Agent Incidence Type and Symptoms Characteristics<br />
Taxanes: paclitaxel,<br />
docetaxel<br />
Platinum agents: cisplatin,<br />
carboplatin, oxaliplatin<br />
Vinca alkaloids: vincristine,<br />
vinblastine, vindesine,<br />
vinorelbine, vinflunine<br />
Nucleoside analogs:<br />
cytarabine, gemcitabine<br />
36% - Axonal Dose dependent; i.e., based on treatment schedule,<br />
- Tingling and numbness within 24 hr <strong>of</strong> infusion<br />
- Motor neuropathy can cause foot drop<br />
- Acute neuropathic pain<br />
infusion duration, and cumulative dose<br />
30% cisplatin - Pure sensory neuropathy - Amount <strong>of</strong> platinum binding to DNA is comparable to or<br />
60–80%: oxaliplatin<br />
20%: carboplatin when<br />
combined with<br />
paclitaxel<br />
- Autonomic symptoms: dizziness, impotence, orthostatic<br />
hypotension<br />
dromes are associated with increased levels <strong>of</strong> proinflammatory<br />
cytokines that have a direct effect on muscle<br />
metabolism and anorexia. One <strong>of</strong> the most prominent mechanisms<br />
is the anabolic resistance <strong>of</strong> older muscles to postprandial<br />
amino acid loading, which leads to a negative<br />
whole-body protein balance. 36 In turn, this negative balance<br />
leads to hypermetabolism with increased resting energy<br />
expenditure. The consequences <strong>of</strong> sarcopenia and cachexia<br />
in older adults include fatigue, depression, decreased mobility,<br />
depression, and falls.<br />
To date, no clinically applied regimen has been completely<br />
successful in reversing cancer-associated muscle or weight<br />
loss. A Cochrane review demonstrated that megestrol acetate<br />
(Megace) improved appetite and weight gain. 37 However,<br />
fat mass rather than lean body mass increased and<br />
there was no benefit in the quality <strong>of</strong> life. Prednisolone and<br />
dexamethasone have been tested in randomized trials and<br />
both have improved appetite and well-being compared with<br />
placebo. 38 Steroids and megestrol acetate are not recommended<br />
for long-term use because <strong>of</strong> such side effectsas<br />
venous thromboembolism and adrenal insufficiency. Randomized<br />
trials <strong>of</strong> nutritional supplements have shown an<br />
improved net food intake and increased physical activity but<br />
no benefit 39 in terms <strong>of</strong> lean body mass or survival. The<br />
authors <strong>of</strong> another Cochrane review concluded that there<br />
was insufficient data to establish whether eicosapentaenoic<br />
acid (EPA) was better than placebo and that there was no<br />
evidence that EPA improves symptoms when combined with<br />
megestrol acetate. 40 In one large study, 41 patients were<br />
randomly assigned to one <strong>of</strong> the following five arms: medroxyprogesterone<br />
acetate, EPA, l-carnitine, thalidomide,<br />
or medroxyprogesterone acetate plus EPA plus l-carnitine<br />
plus thalidomide. Interim analysis <strong>of</strong> data for 125 patients<br />
showed a substantial improvement <strong>of</strong> fatigue in the<br />
l-carnitine arm and the combination treatment arm. Angio-<br />
exceeds levels known to be cytotoxic to tumor cells<br />
- cumulative dose for cisplatin: 400–500 mg/m 2<br />
- High-frequency hearing loss with cisplatin - cisplatin-induced peripheral neuropathy is irreversible in<br />
- Oropharyngeal dysesthesia with oxaliplatin<br />
30–50% <strong>of</strong> patients<br />
31% with vincristine - Sensorimotor neuropathy - Lifelong sequelae in survivors<br />
- Symptoms usually seen within 3 mos <strong>of</strong> therapy - Vincristine and Vindesine cause worse PN than<br />
- Tingling and numbness in hands and feet<br />
- Muscle weakness and leg cramps<br />
- Paralytic ileus and megacolon<br />
Vinblastine and Vinorelbine<br />
- Cap Vincristine dose at 2 mg<br />
1–10% - Demyelinating polyneuropathy - Typically seen with high doses <strong>of</strong> cytarabine (3 gm/m2 )<br />
- Progressive monophasic course within 2–3 wk after<br />
beginning <strong>of</strong> treatment<br />
- Severe motor weakness<br />
- Quadriparesis requiring ventilatory support<br />
which can also cause central cerebellar neuropathy<br />
Bortezomib 30% - Predominantly sensory and distal nerves - Reversible<br />
- Length dependent - Dose and frequency related<br />
- Axonal loss present<br />
Thalidomide 20–40% - Tingling and numbness in hands and feet - Dose dependent<br />
- Muscle cramps - Increases with age<br />
Bevacizumab Case reports - Severe optic neuropathy - Concurrent radiation may have been a risk factor<br />
tensin converting enzyme-inhibitors, cytokine inhibitors<br />
(infliximab or anti-tumor-necrosis factor), myostatin antagonists,<br />
peroxisome-proliferator-activated-receptor-� (PPAR-�)<br />
agonists, selective androgen-receptor modulators (ostarine),<br />
and gherelin agonists are all currently being evaluated<br />
in clinical trials for the management <strong>of</strong> sarcopenia and<br />
cachexia. Resistance exercise training is an effective intervention<br />
that augments muscle mass and strength and attenuates<br />
the development <strong>of</strong> sarcopenia in older patients.<br />
Chemotherapy-Associated Peripheral Neuropathy and Falls<br />
The incidence <strong>of</strong> cancer treatment-related peripheral neuropathy<br />
is variable and ranges from 30% to 40%. The<br />
neuropathy usually affects distal sites first, and as cumulative<br />
doses increase, symptoms progress in severity. Sensory<br />
symptoms and signs typically develop before motor symptoms,<br />
and pain will develop in a subset <strong>of</strong> patients. More<br />
severe and persistent neuropathy may develop in patients<br />
with pre-existing neuropathy (related to diabetes or vitamin<br />
B12 deficiency). Table 2 lists characteristics <strong>of</strong> oncology<br />
agents that can cause treatment-related peripheral neuropathy.<br />
42 Interestingly, one study demonstrated that patients<br />
age 65 or older do not have greater risk <strong>of</strong> peripheral<br />
neuropathy, and advanced age is not associated with increased<br />
severity <strong>of</strong> the disorder. 43 However, peripheral neuropathy<br />
is <strong>of</strong>ten only partially reversible and can thus cause<br />
substantial long-term morbidity in older patients. The risk<br />
<strong>of</strong> falls increases with each cycle <strong>of</strong> chemotherapy and is<br />
highest for patients with worse symptoms from peripheral<br />
neuropathy. 44 These patients had more severe muscle weakness,<br />
loss <strong>of</strong> balance, and a higher interference with walking<br />
or driving. Physical performance measures from the CGA<br />
can help identify patients and aid in the development <strong>of</strong><br />
interventions for patients with cancer treatment-related<br />
peripheral neuropathy who are at most at risk for falling.<br />
325
Table 3. Etiology <strong>of</strong> Cancer-Related Fatigue<br />
Peripheral component: negative energy imbalance<br />
● Cancer itself or therapy for cancer<br />
● Treatment for cancer: chemotherapy, radiation therapy, surgery, hormone<br />
therapy, targeted kinase inhibitors<br />
● Systemic infections<br />
● Hypothyroidism<br />
● Anemia<br />
● Malnutrition<br />
● Metabolic abnormalities<br />
● Sleep disorders<br />
● Psychological factors: depression, anxiety<br />
Central component<br />
● Hyperactivity <strong>of</strong> the hypothalamic-pituitary-adrenal axis from stress (i.e.,<br />
increase in cortisol or corticotrophin-releasing factor<br />
● Decreased gonadotropin levels<br />
● Increase in the numbers <strong>of</strong> circulating T-lymphocytes<br />
● Increased interleukin-1 receptor antagonist<br />
● Increased soluble tumor necrosis factor receptor type II<br />
● Increased neopterin levels<br />
Currently, no proven pharmacologic treatments are available<br />
for established chemotherapy-associated peripheral<br />
neuropathy. Discontinuation <strong>of</strong> the causative agent or dose<br />
reductions may help. Pharmacologic agents that have been<br />
tried include topical amitriptyline with ketamine, topical<br />
lidocaine, selective serotonin norepinephrine uptake inhibitors,<br />
and antiepileptics. Because chemotherapy-associated<br />
peripheral neuropathy is only partially reversible and<br />
maybe permanent, several studies have focused on prevention<br />
<strong>of</strong> the disorder. 45 Calcium and magnesium infusions<br />
and glutathione have been shown to prevent neurotoxicity<br />
from oxaliplatin and cisplatin, respectively. <strong>Clinical</strong> trials<br />
involving patients receiving oxaliplatin or paclitaxel have<br />
demonstrated that glutamine substantially reduces neuropathy<br />
symptoms along with decreased incidence <strong>of</strong> motor<br />
weakness and gait disturbances.<br />
Cancer-Related Fatigue<br />
The NCCN defines cancer-related fatigue as a distressing<br />
persistent subjective sense <strong>of</strong> tiredness or exhaustion that is<br />
not proportional to recent activity and interferes with usual<br />
functioning. Cancer-related fatigue may be an early symptom<br />
<strong>of</strong> malignant disease and is reported by as many as 40%<br />
<strong>of</strong> patients at the time <strong>of</strong> diagnosis. In addition, as many as<br />
90% <strong>of</strong> patients receiving radiation therapy and as many as<br />
80% <strong>of</strong> patients receiving chemotherapy experience fatigue.<br />
One study demonstrated no association between fatigue and<br />
age. 45 A number <strong>of</strong> studies found an association between<br />
hemoglobin levels, pain, and nonpain symptoms with fatigue.<br />
In the Health and Retirement Study, 67% <strong>of</strong> the<br />
patients with cancer were 65 and olderand approximately<br />
50% had one or two coexisting medical conditions. 46 Patients<br />
with cancer had a substantially higher risk for fatigue,<br />
depression, and pain. Table 3 provides information on the<br />
etiology and possible pathophysiology <strong>of</strong> cancer-related fatigue.<br />
47<br />
326<br />
To date, there is no U.S. Food and Drug Administration<br />
(FDA) approved drug for the treatment <strong>of</strong> cancer-related<br />
fatigue. [In addition, there are no studies that have been<br />
conducted to study CRF specifically in elderly patients with<br />
cancer.] If fatigue is secondary to low hemoglobin, erythropoietin<br />
stimulating agents (ESA) may be used to treat the<br />
underlying anemia. However, ESAs must be used with<br />
caution given the risk <strong>of</strong> thromboembolic disease, and they<br />
continue to be contraindicated in patients receiving myelosuppressive<br />
therapy for curative cancers. A randomized<br />
study <strong>of</strong> methylphenidate (target dose, 54 mg/d) in 148 adult<br />
patients with cancer did not significantly improve cancerrelated<br />
fatigue (p � 0.35). 48 A subset analysis suggested<br />
that patients with more severe fatigue and/or with more<br />
advanced disease did have some improvement in fatigue.<br />
Patients in the methylphenidate arm reported more nervousness<br />
and appetite loss. Several studies have shown no<br />
benefit <strong>of</strong> selective serotonin reuptake inhibitors (sertraline<br />
or paroxetine) for the treatment <strong>of</strong> cancer-related fatigue in<br />
patients with cancer. Modafinil, a wakefulness drug, improved<br />
severe cancer-related fatigue in a large randomized<br />
study. 49<br />
Nonpharmacologic interventions have also been studied<br />
for their effect on cancer-related fatigue. 50 A Cochrane<br />
review demonstrated that exercise improves cancer-related<br />
fatigue during and after cancer therapy. 51 Caution in developing<br />
the exercise prescription is recommended for patients<br />
with bone metastases or myelosuppression andpatients who<br />
are febrile. Exercise interventions included moderately intense<br />
(55% to 75% <strong>of</strong> heart rate) aerobic exercise (e.g.,<br />
walking, cycling), ranging from 10–90 minutes 3 to 7 days<br />
per week. Psychological and behavioral interventions 47 that<br />
have been evaluated for cancer-related fatigue include support<br />
groups, yoga, and cognitive behavioral interventions.<br />
These interventions have improved cancer-related fatigue<br />
and vitality in most studies. These interventions also can<br />
be combined with exercise for maximizing benefit, as was<br />
demonstrated in the GROUP-HOPE trial. 52<br />
Conclusion<br />
MOHILE, KLEPIN, AND RAO<br />
The incidence <strong>of</strong> cancer increases with age. An older<br />
patient’s chronologic age does not always reflect his or her<br />
overall health status. Therefore, oncologists need to be adept<br />
at assessing physiologic and functional capacity in older<br />
patients. The CGA is the criterion standard for evaluation <strong>of</strong><br />
an older patient. The combined data from the CGA can be<br />
used to stratify patients into risk categories to better predict<br />
chemotherapy toxicity and survival. More research is<br />
needed on the best ways to incorporate the CGA into<br />
decision making for cancer treatment, implement novel<br />
treatment dosing options, and develop interventions to improve<br />
symptoms <strong>of</strong> sarcopenia, cachexia, chemotherapyassociated<br />
peripheral neuropathy, and cancer-related<br />
fatigue in older patients with cancer.
OLDER PATIENTS WITH ADVANCED CANCER<br />
Authors’ Disclosures <strong>of</strong> Potential Conflicts <strong>of</strong> Interest<br />
Author<br />
Supriya Gupta Mohile*<br />
Heidi D. Klepin*<br />
Arati V. Rao*<br />
*No relevant relationships to disclose.<br />
Employment or<br />
Leadership<br />
Positions<br />
Consultant or<br />
Advisory Role<br />
1. Smith BD, Smith GL, Hurria A. Future <strong>of</strong> Cancer Incidence in the<br />
United States: Burdens Upon an Aging, Changing Nation. J Clin Oncol.<br />
2009;27:2758-2765.<br />
2. Wedding U, Rohrig B, Klippstein A, et al. Age, severe comorbidity and<br />
functional impairment independently contribute to poor survival in cancer<br />
patients. J Cancer Res Clin Oncol. 2007;133:945-950.<br />
3. Mohile SG, Xian Y, Dale W, et al. Association <strong>of</strong> a cancer diagnosis with<br />
vulnerability and frailty in older Medicare beneficiaries. J Natl Cancer Inst.<br />
2009;101:1206-1215.<br />
4. Piccirillo JF, Tierney RM, Costas I, et al. Prognostic importance <strong>of</strong><br />
comorbidity in a hospital-based cancer registry. JAMA. 2004;291:2441-2447.<br />
5. Buck MD, Atreja A, Brunker CP, et al. Potentially inappropriate<br />
medication prescribing in outpatient practices: prevalence and patient characteristics<br />
based on electronic health records. Am J Geriatr Pharmacother.<br />
2009;7:84-92.<br />
6. Fried LP, Kronmal RA, Newman AB, et al. Risk factors for 5-year<br />
mortality in older adults: the Cardiovascular Health Study. JAMA. 1998;279:<br />
585-592.<br />
7. Diehr P, Bild DE, Harris TB, et al. Body mass index and mortality in<br />
nonsmoking older adults. Am J Pub Health. 1998;88:623-629.<br />
8. Kanesvaran R, Li H, Koo KN, Poon D. Analysis <strong>of</strong> prognostic factors <strong>of</strong><br />
comprehensive geriatric assessment and development <strong>of</strong> a clinical scoring<br />
system in elderly Asian patients with cancer. J Clin Oncol. 2011;29:3620-<br />
3627.<br />
9. Hurria A, Togawa K, Mohile SG, et al. Predicting chemotherapy toxicity<br />
in older adults with cancer: a prospective multicenter study. J Clin Oncol.<br />
2011;29:3457-3465.<br />
10. Extermann M, Boler I, Reich RR, et al. Predicting the risk <strong>of</strong> chemotherapy<br />
toxicity in older patients: the Chemotherapy Risk Assessment Scale<br />
for High-Age Patients (CRASH) score. Cancer. Epub 2011 Nov 9.<br />
11. Saliba D, Elliott M, Rubenstein LZ, et al. The Vulnerable Elders<br />
Survey: a tool for identifying vulnerable older people in the community. JAm<br />
Geriatr Soc. 2001;49:1691-1699.<br />
12. Min LC, Elliott MN, Wenger NS, Saliba D. Higher vulnerable elders<br />
survey scores predict death and functional decline in vulnerable older people.<br />
J Am Geriatr Soc. 2006;54:507-511.<br />
13. Luciani A, Ascione G, Bertuzzi C, et al. Detecting disabilities in older<br />
patients with cancer: comparison between comprehensive geriatric assessment<br />
and vulnerable elders survey-13. J Clin Oncol. 2010;28:2046-2050.<br />
14. Steverink NSJ, Schuurmans H, et al. Measuring frailty: developing<br />
and testing the GFI (Groningen Frailty Indicator). The Gerontologist. 2001;<br />
41(Special Issue 1):236.<br />
15. Schrijvers D, Baiter A, Vos MD, et al. Evaluation <strong>of</strong> the Groningen<br />
Frailty Index (GFI) as a screening tool in elderly patients (pts): an interim<br />
analysis. 2010 ESMO; Abstract 566PD.<br />
16. Aaldriks AA, Maartense E, le Cessie S, et al. Predictive value <strong>of</strong><br />
geriatric assessment for patients older than 70 years, treated with chemotherapy.<br />
Crit Rev Oncol Hematol. 2011;79:205-212.<br />
17. Biesma B, Wymenga AN, Vincent A, et al. Quality <strong>of</strong> life, geriatric<br />
assessment and survival in elderly patients with non-small-cell lung cancer<br />
treated with carboplatin-gemcitabine or carboplatin-paclitaxel: NVALT-3 a<br />
phase III study. Ann Oncol. 2011;22:1520-1527.<br />
18. Soubeyran P, Bellera C, Goyard J, et al. Validation <strong>of</strong> the G8 screening<br />
tool in geriatric oncology: The ONCODAGE project. J Clin Oncol. 2011;29<br />
(suppl; abstr 9001).<br />
19. Caillet P, Canoui-Poitrine F, Vouriot J, et al. Comprehensive geriatric<br />
assessment in the decision-making process in elderly patients with cancer:<br />
ELCAPA study. J Clin Oncol. 2011;29:3636-3642.<br />
20. Chaibi P, Magne N, Breton S, et al. Influence <strong>of</strong> geriatric consultation<br />
with comprehensive geriatric assessment on final therapeutic decision in<br />
elderly cancer patients. Crit Rev Oncol Hematol. 2011;79:302-307.<br />
21. Horgan AM, Leighl NB, Coate L, et al. Impact and feasibility <strong>of</strong> a<br />
comprehensive geriatric assessment in the oncology setting: a pilot study.<br />
Am J Clin Oncol. Epub 2011 Mar 17.<br />
Stock<br />
Ownership Honoraria<br />
REFERENCES<br />
Research<br />
Funding<br />
Expert<br />
Testimony<br />
Other<br />
Remuneration<br />
22. Hanlon JT, Weinberger M, Samsa GP, et al. A randomized, controlled<br />
trial <strong>of</strong> a clinical pharmacist intervention to improve inappropriate prescribing<br />
in elderly outpatients with polypharmacy. Am J Med. 1996;100:428-437.<br />
23. Luciani A, Marussi D, Ascione G, et al. Do elderly cancer patients<br />
achieve an adequate dose intensity in common clinical practice? <strong>Oncology</strong>.<br />
2006;71:382-387.<br />
24. Gajra A, Hardt W, Tew W, et al. Primary dose reduction (PDR) <strong>of</strong><br />
chemotherapy (chemo) in patients (Pts) older than age 65 with advanced<br />
cancer (Ca) and toxicity outcomes. J Clin Oncol. 2011:9037.<br />
25. Lichtman SM, Wildiers H, Launay-Vacher V, Steer C, Chatelut E,<br />
Aapro M. International <strong>Society</strong> <strong>of</strong> Geriatric <strong>Oncology</strong> (SIOG) recommendations<br />
for the adjustment <strong>of</strong> dosing in elderly cancer patients with renal<br />
insufficiency. Eur J Cancer. 2007;43:14-34.<br />
26. Seymour MT, Thompson LC, Wasan HS, et al. Chemotherapy options<br />
in elderly and frail patients with metastatic colorectal cancer (MRC<br />
FOCUS2): an open-label, randomised factorial trial. Lancet. 2011;377:1749-<br />
1759.<br />
27. Tournigand C, Cervantes A, Figer A, et al. OPTIMOX1: a randomized<br />
study <strong>of</strong> FOLFOX4 or FOLFOX7 with oxaliplatin in a stop-and-Go fashion in<br />
advanced colorectal cancer–a GERCOR study. J Clin Oncol. 2006;24:394-400.<br />
28. Chibaudel B, Maindrault-Goebel F, Lledo G, et al. Can chemotherapy<br />
be discontinued in unresectable metastatic colorectal cancer? The GERCOR<br />
OPTIMOX2 Study. J Clin Oncol. 2009;27:5727-5733.<br />
29. Figer A, Perez-Staub N, Carola E, et al. FOLFOX in patients aged<br />
between 76 and 80 years with metastatic colorectal cancer: an exploratory<br />
cohort <strong>of</strong> the OPTIMOX1 study. Cancer. 2007;110:2666-2671.<br />
30. Adams RA, Meade AM, Seymour MT, et al. Intermittent versus<br />
continuous oxaliplatin and fluoropyrimidine combination chemotherapy for<br />
first-line treatment <strong>of</strong> advanced colorectal cancer: Results <strong>of</strong> the randomised<br />
phase 3 MRC COIN trial. Lancet Oncol. 2011;12:642-653.<br />
31. Labianca R, Sobrero A, Isa L, et al. Intermittent versus continuous<br />
chemotherapy in advanced colorectal cancer: a randomised ‘GISCAD’ trial.<br />
Ann Oncol. 2011;22:1236-1242.<br />
32. NCCN guidelines for supportive care. Version 2.2011 http://www.<br />
nccn.org/pr<strong>of</strong>essionals/physician_gls/f_guidelines.asp#suportive. Accessed<br />
March 15, <strong>2012</strong>.<br />
33. Basch E, Prestrud AA, Hesketh PJ, et al. Antiemetics: <strong>American</strong><br />
<strong>Society</strong> <strong>of</strong> <strong>Clinical</strong> <strong>Oncology</strong> <strong>Clinical</strong> Practice Guideline Update. J Clin Oncol.<br />
2011;29:4189-4198.<br />
34. Fearon K, Strasser F, Anker SD, et al. Definition and classification <strong>of</strong><br />
cancer cachexia: an international consensus. Lancet Oncol. 2011;12:489-495.<br />
35. Morley JE, Baumgartner RN. Cytokine-related aging process. J Gerontol<br />
A Biol Sci Med Sci. 2004;59:M924-M929.<br />
36. Cuthbertson D, Smith K, Babraj J, et al. Anabolic signaling deficits<br />
underlie amino acid resistance <strong>of</strong> wasting, aging muscle. FASEB J. 2005;19:<br />
422-444.<br />
37. Berenstein EG, Ortiz Z. Megestrol acetate for the treatment <strong>of</strong><br />
anorexia-cachexia syndrome. Cochrane Database Syst Rev. 2005:CD004310.<br />
38. Willox JC, Corr J, Shaw J, et al. Prednisolone as an appetite stimulant<br />
in patients with cancer. Br Med J. 1984;288:27.<br />
39. Fearon KC, Von Meyenfeldt MF, Moses AG, et al. Effect <strong>of</strong> a protein<br />
and energy dense N-3 fatty acid enriched oral supplement on loss <strong>of</strong> weight<br />
and lean tissue in cancer cachexia: a randomised double blind trial. Gut.<br />
2003;52:1479-1486.<br />
40. Dewey A, Baughan C, Dean T, et al. Eicosapentaenoic acid (EPA, an<br />
omega-3 fatty acid from fish oils) for the treatment <strong>of</strong> cancer cachexia.<br />
Cochrane Database Syst Rev. 2007:CD004597.<br />
41. Mantovani G. Randomised phase III clinical trial <strong>of</strong> 5 different arms <strong>of</strong><br />
treatment on 332 patients with cancer cachexia. Eur Rev Med Pharmacol Sci.<br />
2010;14:292-301.<br />
42. Gutiérrez-Gutiérrez G, Sereno M, et al. Chemotherapy-induced peripheral<br />
neuropathy: <strong>Clinical</strong> features, diagnosis, prevention and treatment<br />
strategies. Clin Transl Oncol. 2010;12:81-91.<br />
43. Argyriou AA, Polychronopoulos P, Koutras A, et al. Is advanced age<br />
327
associated with increased incidence and severity <strong>of</strong> chemotherapy-induced<br />
peripheral neuropathy? Support Care Cancer. 2006;14:223-229.<br />
44. T<strong>of</strong>thagen C, Overcash J, Kip K. Falls in persons with chemotherapyinduced<br />
peripheral neuropathy. Support Care Cancer. <strong>2012</strong>;20:583-589.<br />
45. Prue G, Rankin J, Allen J, et al. Cancer-related fatigue: a critical<br />
appraisal. Eur J Cancer. 2006;42:846-863.<br />
46. Reyes-Gibby CC, Aday LA, et al. Pain, depression, and fatigue in<br />
community-dwelling adults with and without a history <strong>of</strong> cancer. J Pain<br />
Symptom Manage. 2006;32:118-128.<br />
47. Rao AV, Cohen HJ. Fatigue in older cancer patients: etiology, assessment,<br />
and treatment. Semin Oncol. 2008;35:633-642.<br />
48. Moraska AR, Sood A, Dakhil SR, et al. Phase III, randomized, doubleblind,<br />
placebo-controlled study <strong>of</strong> long-acting methylphenidate for cancerrelated<br />
fatigue. J Clin Oncol. 2010;28:3673-3679.<br />
328<br />
MOHILE, KLEPIN, AND RAO<br />
49. Jean-Pierre P, Morrow GR, Roscoe JA, et al. A phase 3 randomized,<br />
placebo-controlled, double-blind, clinical trial <strong>of</strong> the effect <strong>of</strong> modafinil on<br />
cancer-related fatigue among 631 patients receiving chemotherapy: a University<br />
<strong>of</strong> Rochester Cancer Center Community <strong>Clinical</strong> <strong>Oncology</strong> Program<br />
Research base study. Cancer. 2010;116:3513-3520.<br />
50. Mustian KM, Morrow GR, Carroll JK, et al. Integrative nonpharmacologic<br />
behavioral interventions for the management <strong>of</strong> cancer-related fatigue.<br />
Oncologist. 2007;12 Suppl 1:52-67.<br />
51. Cramp F, Daniel J Exercise for the management <strong>of</strong> cancer-related<br />
fatigue in adults. Cochrane Database Syst Rev. 2008:CD006145.<br />
52. Courneya KS, Friedenreich CM, Sela RA, et al. The group psychotherapy<br />
and home-based physical exercise (group-hope) trial in cancer survivors:<br />
Physical fitness and quality <strong>of</strong> life outcomes. Psychooncology. 2003;12:357-<br />
374.
RECENT CLINICAL HIGHLIGHTS IN<br />
GYNECOLOGIC ONCOLOGY<br />
CHAIR<br />
Lynette Denny, MD, PhD<br />
University <strong>of</strong> Cape Town<br />
Rondebosch, South Africa<br />
SPEAKERS<br />
D. Scott McMeekin, MD<br />
University <strong>of</strong> Oklahoma Health Sciences Center<br />
Oklahoma City, OK<br />
Nicoletta Colombo, MD<br />
University <strong>of</strong> Milan-Bicocca<br />
Milan, Italy
International Perspective on the Global<br />
Advances in Gynecologic <strong>Oncology</strong><br />
Overview: The treatment <strong>of</strong> gynecologic cancer has evolved<br />
over the years, with greater emphasis on tailored surgery and<br />
reducing morbidity and mortality related to surgery, particularly<br />
in the management <strong>of</strong> vulvar and cervical cancer. The<br />
addition <strong>of</strong> concurrent chemotherapy to radiation regimens<br />
has improved survival <strong>of</strong> patients with cervical cancer in<br />
developed countries; however, most women with cancer in<br />
developing countries have advanced, untreatable disease and<br />
minimal access to anticancer therapies. In the past 15 years<br />
there has been intense research into alternatives to cervical<br />
GYNECOLOGIC ONCOLOGY includes a spectrum <strong>of</strong><br />
cancer that affects the vulva, vagina, cervix, uterus,<br />
tubes, ovaries, peritoneum, and, in some European countries,<br />
breast. Cancers <strong>of</strong> the lower genital tract, most <strong>of</strong><br />
which are etiologically associated with infection with highrisk<br />
types <strong>of</strong> human papillomavirus (HPV), are more commonly<br />
found in developing compared with developed<br />
countries. The inverse is true for cancers <strong>of</strong> the upper genital<br />
tract. Over the years, substantial advances have been made<br />
in the management <strong>of</strong> gynecological cancers. This article<br />
explores those advances.<br />
Vulvar Cancer<br />
Vulvar cancer is the fourth most common gynecologic<br />
cancer, accounting for 4% to 5% <strong>of</strong> all malignant tumors <strong>of</strong><br />
the female genital tract. Age standardized incidence rates<br />
(ASIRs) vary from 0.1 per 100,000 population (Algeria) to<br />
1.6 to 1.7 per 100,000 population (Zimbabwe and United<br />
Kingdom, respectively). 1 Vulvar cancer is usually diagnosed<br />
in older women (mean age, 65 to 70 years), who <strong>of</strong>ten<br />
have lichen sclerosus et atrophicus or differentiated vulvar<br />
intraepithelial neoplasm, which is not related to HPV.<br />
Increasingly, however, vulvar cancer is diagnosed in<br />
younger women, which relates to persistent infection with<br />
HPV and infection with HIV. Toki and colleagues 2 found<br />
the mean age <strong>of</strong> patients with HPV-related vulvar cancer to<br />
be 55 years, whereas that <strong>of</strong> patients with non–HPVassociated<br />
vulvar cancer was 77 years. Eva and colleagues 3<br />
found that 85.7% <strong>of</strong> 70 women who were diagnosed as<br />
having differentiated vulvar intraepithelial neoplasm had<br />
concurrent, previous, or subsequent vulvar squamous cell<br />
carcinoma. However, only 25% <strong>of</strong> women with HPV-related<br />
undifferentiated vulvar intraepithelial neoplasm and 33%<br />
<strong>of</strong> women with lichen sclerosus et atrophicus and squamous<br />
cell hyperplasia had a history <strong>of</strong> developing invasive vulvar<br />
cancer.<br />
Treatment <strong>of</strong> vulvar cancer is primarily surgical. Up to the<br />
late 1980s the most common procedure was a radical vulvectomy<br />
with en bloc resection <strong>of</strong> bilateral groin node lymphadenectomy—the<br />
so-called butterfly incision pioneered by<br />
Antoine Basset in 1912. In the 1940s, Frederick Taussig<br />
adapted the Basset operation by performing separate groin<br />
incisions, which resulted in improved survival rates, albeit<br />
with very high operative morbidity and mortality rates. The<br />
triple incision technique was introduced, and although no<br />
randomized controlled trials have compared this technique<br />
330<br />
By Lynette Denny, MD, PhD<br />
cytologic testing, particularly in low resourced regions but<br />
also in an attempt to improve on cytologic testing in developed<br />
countries. Surgical staging in endometrial cancer has<br />
enabled the use <strong>of</strong> adjuvant radiation to be individualized to<br />
the patient’s particular risk factors for recurrence. The management<br />
<strong>of</strong> ovarian cancer, long stagnant since the introduction<br />
<strong>of</strong> platinum and paclitaxel as chemotherapeutic agents, is<br />
set to change with the onset <strong>of</strong> molecular and genetic pr<strong>of</strong>iling<br />
and the introduction <strong>of</strong> novel therapies.<br />
with the butterfly incision, retrospective studies found no<br />
inferior outcomes and a low risk <strong>of</strong> bridge recurrence<br />
(2.4%). 4,5 Compared with en bloc resection, rates <strong>of</strong> associated<br />
morbidity, such as wound breakdown and lymph drainage<br />
problems, were substantially lower with the triple<br />
incision technique. 6 However, this technique still involved<br />
removal <strong>of</strong> external genitalia.<br />
The introduction <strong>of</strong> radical wide local incision further<br />
reduced operative morbidity, and this approach, which <strong>of</strong>ten<br />
allowed preservation <strong>of</strong> anatomy, was not shown to be<br />
inferior in terms <strong>of</strong> oncologic safety. Current guidelines<br />
suggest that 1 cm <strong>of</strong> disease-free tissue around the invasive<br />
lesion represents a safe margin; however, hard data on the<br />
value <strong>of</strong> disease-free margins are lacking, and some recent<br />
publications suggest that margins are irrelevant if the<br />
entire lesion is excised. 7 Radical wide local excision has<br />
enabled a reasonably accurate definition <strong>of</strong> lateral compared<br />
with central lesions. Central lesion refers to the ability to<br />
remove the invasive lesion with 1 cm <strong>of</strong> disease-free tissue<br />
without damaging or removing central structures (e.g., clitoris,<br />
urethra, and anus). Lateral lesion allows 1 cm <strong>of</strong><br />
disease-free tissue without compromising central structures.<br />
To preserve the central structures, it is possible to<br />
perform primary chemoradiation and to only perform surgery<br />
for residual disease. Moore et al treated 58 women with<br />
T3 or T4 tumors not amenable to surgical resection with<br />
radiation (1.8Gy daily � 32 fractions) plus weekly cisplatinum<br />
(40mg/m 2 ), followed by surgical resection <strong>of</strong> residual<br />
tumor or biopsy to confirm pathological complete response.<br />
Forty <strong>of</strong> the 58 women completed treatment and 37/58 (64%)<br />
women had a complete clinical response. Of these women,<br />
34 underwent biopsy and 29 (78%) <strong>of</strong> these women had a<br />
complete pathological response. 8 Rogers et al reported on a<br />
retrospective study <strong>of</strong> 50 women treated with chemoradiation<br />
for advanced vulvar cancer in Cape Town from 1982 to<br />
2001. Only 14 (28%) <strong>of</strong> the women had a complete response,<br />
29 (58%) had a partial response. Of the women who under-<br />
From the Department <strong>of</strong> Obstetrics & Gynaecology, University <strong>of</strong> Cape Town/Groote<br />
Schuur Hospital, Cape Town, South Africa.<br />
Author’s disclosures <strong>of</strong> potential conflicts <strong>of</strong> interest are found at the end <strong>of</strong> this article.<br />
Address reprint requests to Lynette Denny, MD, PhD, Department <strong>of</strong> Obstetrics &<br />
Gynaecology, University <strong>of</strong> Cape Town/Groote Schuur Hospital, H45, Old Main Building,<br />
Groote Schuur Hospital, Observatory 7925, Cape Town, South Africa; email: lynette.<br />
denny@uct.ac.za.<br />
© <strong>2012</strong> by <strong>American</strong> <strong>Society</strong> <strong>of</strong> <strong>Clinical</strong> <strong>Oncology</strong>.<br />
1092-9118/10/1-10
GLOBAL ADVANCES IN GYNECOLOGIC ONCOLOGY<br />
went post radiation surgery (7 patients), survival was significantly<br />
better. 9<br />
Systematic inguin<strong>of</strong>emoral lymphadenectomy comprises<br />
resection <strong>of</strong> superficial inguinal lymph nodes and deep<br />
femoral nodes. Although current guidelines recommend at<br />
least 6 nodes per groin be removed, systematic lymphadenectomy<br />
is associated with substantial morbidity, such as<br />
leg edema, lymphocysts, wound breakdown, and erysipelas.<br />
To avoid these complications, the use <strong>of</strong> sentinel node<br />
dissection in vulvar cancer has been investigated. Sentinel<br />
nodes were studied in a large, prospective, multicenter study<br />
(Groningen International Study on Sentinel nodes in Vulvar<br />
cancer [GROINSS-V]), which included 403 women with<br />
unifocal vulvar cancer stage I or II, tumor size less than<br />
4 cm, stromal invasion greater than 1 mm, and clinically<br />
negative lymph nodes. 10 Lymphadenectomy was not performed<br />
in women with sentinel node–negative cancer. Groin<br />
recurrences occurred in 2.3% <strong>of</strong> patients, with a median<br />
follow-up <strong>of</strong> 35 months. Overall disease-specific survival was<br />
97% after 3 years, and morbidity was substantially reduced.<br />
These results are equivalent to results <strong>of</strong> systematic inguin<strong>of</strong>emoral<br />
lymphadenectomy from an oncologic point <strong>of</strong><br />
view.<br />
Vaginal Cancer<br />
Primary cancer <strong>of</strong> the vagina is rare and accounts for<br />
approximately 2% <strong>of</strong> all cancers <strong>of</strong> the female genital tract.<br />
Most women are older than 60 years, and only 10% to 15%<br />
are younger than 50 years. The most common type <strong>of</strong> vaginal<br />
cancer is squamous carcinoma (80% to 90%) and adenocarcinoma<br />
(4% to 10%). Infection with HPV-16 is believed to be<br />
an important etiologic factor. Prognosis is dependent on age,<br />
histologic type, and tumor stage. The optimal treatment is<br />
still controversial because <strong>of</strong> the rarity <strong>of</strong> the disease and<br />
lack <strong>of</strong> appropriate clinical trials. Standard therapy is radiation,<br />
either alone or concurrently with platinum-based<br />
chemotherapy. The advantage <strong>of</strong> radiation is the preservation<br />
<strong>of</strong> the vagina, although the vagina does not always<br />
remain fully functional.<br />
KEY POINTS<br />
● The modern management <strong>of</strong> vulvar cancer is to distinguish<br />
between central and lateral lesions and to<br />
adapt surgery and radiation to reduce morbidity and<br />
enhance quality <strong>of</strong> life.<br />
● Management <strong>of</strong> cervical cancer has not changed substantially<br />
in the past 10 to 20 years; however, approaches<br />
to cervical cancer prevention have<br />
undergone a re-evaluation at both the primary and<br />
secondary prevention levels.<br />
● The surgical staging <strong>of</strong> endometrial cancers has enabled<br />
the use <strong>of</strong> adjuvant therapies to be tailored to<br />
the individual patient.<br />
● Current novel therapies undergoing trials bring hope<br />
to the long stagnant progress in the management <strong>of</strong><br />
ovarian cancer.<br />
● Quality <strong>of</strong> life in the treatment <strong>of</strong> patients with<br />
gynecological malignant tumors should inform all<br />
treatment decisions.<br />
Cervical Cancer<br />
The most noteworthy advances in cervical cancer in the<br />
past 15 years have been the exploration <strong>of</strong> alternative<br />
approaches to cytologic testing and colposcopy for the prevention<br />
<strong>of</strong> cervical cancer in developing countries. In addition,<br />
the approach to screening in developed countries has<br />
changed from conventional cytologic testing to the introduction<br />
<strong>of</strong> liquid-based cytologic and HPV DNA testing. In fact,<br />
many molecular tests to increase the sensitivity and specificity<br />
for the detection <strong>of</strong> cervical cancer precursors are<br />
currently undergoing clinical trials.<br />
The potential utility <strong>of</strong> HPV DNA testing was shown in a<br />
randomized trial conducted in South Africa. 11 This trial <strong>of</strong><br />
more than 6,500 unscreened women, aged 35 to 65 years,<br />
showed a substantial reduction <strong>of</strong> high-grade cervical cancer<br />
precursors in women who screened positive with Hybrid<br />
Capture 2 (Qiagen Inc, Gaithersburg, MD) for high-risk<br />
types <strong>of</strong> HPV DNA and who underwent treatment with<br />
cryotherapy, and this reduction was sustained for 36<br />
months. After 36 months the decrease in the cumulative<br />
detection <strong>of</strong> grade 2 or higher cervical intraepithelial neoplasia<br />
in the HPV treatment arm compared with the<br />
control (delayed treatment) arm was 1.5% compared with<br />
5.6% (difference, 4.1%; p � 0.001). The difference, however,<br />
in the visual inspection with acetic acid (VIA) treatment<br />
group compared with the control arm was significantly less<br />
(3.8% vs. 5.6%; difference, 1.8%; p � 0.002). In India,<br />
Sankaranarayanan and colleagues 12 screened 131,746<br />
healthy women who were randomly assigned to one <strong>of</strong><br />
four groups: screening by HPV testing, cytologic testing,<br />
VIA, or no screening (standard <strong>of</strong> care). There was a<br />
reduction in the numbers <strong>of</strong> advanced cancers and cancer<br />
deaths in the HPV treatment group (hazard ratio, 0.47; 95%<br />
CI, 0.32 to 0.69) compared with no reductions in the VIA or<br />
the cytologic testing treatment groups. These studies and<br />
others conducted in developed countries suggest that HPV<br />
DNA testing has the potential to substantially reduce the<br />
incidence <strong>of</strong> and mortality from cervical cancer. More technologically<br />
accessible and affordable tests for HPV DNA<br />
testing are awaited for implementation in low-resource<br />
settings.<br />
In developed countries, HPV DNA testing has been recommended<br />
as a primary screen in women older than 30<br />
years, with cytologic testing as a triage for women with<br />
positive test results. Only women with positive results on<br />
both tests would be referred for colposcopy and treatment. 13<br />
Other methods <strong>of</strong> triage include p16-INK4A overexpression<br />
and testing for HPV-E6/7 messenger RNA, which are still<br />
under study.<br />
The trends in terms <strong>of</strong> cervical cancer treatment include<br />
the following: (1) fertility-sparing surgery in young women<br />
with early-stage disease; (2) laparoscopic radical hysterectomy<br />
and node dissection; (3) use <strong>of</strong> sentinel nodes to<br />
prevent morbidity associated with systematic lymphadenectomy<br />
(still under study) and; (4) use <strong>of</strong> concurrent<br />
cisplatinum-based chemotherapy concurrently with radiation,<br />
which has shown substantial improvement in overall<br />
survival and disease-free survival in women with advanced<br />
disease.<br />
Although cervical cancer is a relatively rare disease in<br />
developed countries with functional screening programs,<br />
cervical cancer remains the most common cancer among<br />
331
women living in developing countries because <strong>of</strong> the lack <strong>of</strong><br />
effective screening programs. Most women with cancer in<br />
developing countries have advanced and untreatable disease.<br />
In addition, palliative care remains out <strong>of</strong> reach <strong>of</strong><br />
most women in developing countries, leaving these women<br />
to die painful and undignified deaths. Cervical cancer is the<br />
one cancer that illustrates the gross inequities <strong>of</strong> health care<br />
between rich and poor countries.<br />
Endometrial Cancer<br />
Worldwide, endometrial cancer is the most common cancer<br />
<strong>of</strong> the female genital tract and the seventh most common<br />
cause <strong>of</strong> death from cancer in women in western Europe.<br />
Yearly, approximately 7,406 cases are registered in the<br />
United Kingdom, 88,068 in the European Union, and 40,102<br />
in the United States. 14 The median age <strong>of</strong> cancer occurrence<br />
is in the sixth decade, and more than 90% <strong>of</strong> cases occur in<br />
women older than 50 years. Approximately 5% <strong>of</strong> endometrial<br />
cancers are associated with Lynch syndrome II (hereditary<br />
nonpolyposis colorectal carcinoma syndrome); women<br />
with this syndrome have a 30% to 60% risk <strong>of</strong> developing<br />
endometrial cancer.<br />
Type 1 endometrial cancer, or endometrioid adenocarcinoma,<br />
represents 80% <strong>of</strong> cancers, with serous carcinomas<br />
being the prototype <strong>of</strong> type II uterine cancers. Clear cell<br />
cancers are rare and comprise approximately 1% <strong>of</strong> endometrial<br />
adenocarcinomas.<br />
Since 1988, the International Federation <strong>of</strong> Gynecology<br />
and Obstetrics (FIGO) has recommended surgical staging <strong>of</strong><br />
endometrial cancer, which includes systematic pelvic and<br />
para-aortic lymphadenectomy. FIGO adopted a new staging<br />
system in 2009 in which the old FIGO stage 1a and 1B were<br />
amalgamated into stage 1A and the old 1C became stage 1B.<br />
In stage II cancer, the distinction between superficial and<br />
deep stromal invasion was merged as stage II in which<br />
stromal invasion is documented.<br />
The role <strong>of</strong> systematic lymphadenectomy has been the<br />
subject <strong>of</strong> numerous studies and much controversy. The<br />
Adjuvant External Beam Radiotherapy in the Treatment <strong>of</strong><br />
Endometrial Cancer (ASTEC) trial 15 randomized 1,408<br />
women from 85 centers with histologically proven endometrial<br />
cancer to standard surgery (704 women with hysterectomy,<br />
bilateral salpingo-oophorectomy, peritoneal washings,<br />
and palpation <strong>of</strong> para-aortic lymph nodes) or to standard<br />
surgery plus lymphadenectomy (704 women). The study<br />
found no evidence <strong>of</strong> benefit in terms <strong>of</strong> either overall or<br />
recurrence-free survival for pelvic lymphadenectomy in<br />
women with early endometrial cancer. This study was,<br />
however, criticized for numerous protocol violations, and<br />
some authors thought that it did not adequately assess the<br />
role <strong>of</strong> effective lymphadenectomy or the role <strong>of</strong> individualized<br />
adjuvant radiotherapy in endometrial cancer. 16<br />
The current recommendation for new FIGO stage 1A<br />
cancer is to perform standard surgery only, but for high-risk<br />
cases (i.e., 50% invasion, grade 3 lesions), systemic lymphadenectomy<br />
should be performed. The value <strong>of</strong> this approach<br />
is that approximately 30% <strong>of</strong> women with high-risk stage I<br />
disease will not require adjuvant whole pelvic irradiation<br />
because their cancer is node negative. Vaginal brachytherapy<br />
would still be indicated for women with high-risk<br />
characteristics to reduce local recurrence. 17<br />
The value <strong>of</strong> adjuvant radiation for women with endometrial<br />
cancer localized to the uterus is still under scrutiny.<br />
332<br />
External beam whole pelvic radiation has been shown to<br />
reduce the rate <strong>of</strong> locoregional recurrence in intermediaterisk<br />
endometrial cancer; however, a number <strong>of</strong> trials have<br />
failed to demonstrate that radiation improves overall or<br />
disease-specific survival. Post-Operative Radiotherapy in<br />
Endometrial Cancer 2 (PORTEC-2), a randomized trial<br />
comparing vaginal brachytherapy with external beam radiation,<br />
also did not find overall improved survival; however,<br />
quality <strong>of</strong> life was better in women treated with vaginal<br />
brachytherapy only. 17<br />
Ovarian Cancer<br />
LYNETTE DENNY<br />
The ASIR <strong>of</strong> ovarian cancer in the European Union in<br />
2008 14 was 13.4 per 100,000 women, with a mortality rate <strong>of</strong><br />
7.6 per 100,000. In the United Kingdom the ASIR was 17.5<br />
per 100,000 women, with a mortality rate <strong>of</strong> 9.6 per 100 000.<br />
In 2011 in the United States, there were 21,990 new cases <strong>of</strong><br />
ovarian cancer diagnosed and 15,640 deaths, making ovarian<br />
cancer the second most lethal cancer among women in<br />
the United States. 18 Approximately 80% <strong>of</strong> advanced stage<br />
ovarian cancers have high-grade histologic features, and<br />
these tumors can arise from the fimbrial end <strong>of</strong> the fallopian<br />
tube, the peritoneal cavity, or the surface epithelium <strong>of</strong> the<br />
ovary. Efforts to improve on the long-term results <strong>of</strong> primary<br />
therapy (a combination <strong>of</strong> “maximum effort” surgery, including<br />
hysterectomy, bilateral salpingo-oophorectomy, infracolic<br />
omentectomy, and maximum debulking <strong>of</strong> all<br />
macroscopic tumor to � 2 cm <strong>of</strong> residual disease, followed by<br />
chemotherapy using a combination <strong>of</strong> carboplatin and paclitaxel)<br />
through the addition <strong>of</strong> additional cytotoxic agents<br />
have not been successful.<br />
New and effective therapies for ovarian cancer have not<br />
been reported since the introduction <strong>of</strong> platinum-based<br />
drugs and paclitaxel. Neither these drugs nor second-line<br />
therapies (e.g., gemcitabine) have produced substantial advances<br />
in overall survival.<br />
Recently, however, two phase III clinical trials <strong>of</strong> bevacizumab,<br />
an angiogenesis inhibitor and vascular endothelial<br />
growth factor–neutralizing monoclonal antibody, have<br />
shown improvements in progression-free survival but not<br />
overall survival. Tumor angiogenesis, a process in which<br />
cancers induce the formation <strong>of</strong> new blood and lymphatic<br />
vasculature to enable proliferation, invasion, and growth <strong>of</strong><br />
metastasis, appears to be essential to disease progression in<br />
women with epithelial ovarian cancer (EOC). Several antiangiogenic<br />
agents have been investigated in clinical trials,<br />
most directed against vascular endothelial growth factor, a<br />
central promoter <strong>of</strong> the initiation <strong>of</strong> angiogenesis. Two trials<br />
(Gynecologic <strong>Oncology</strong> Group 0218 and ICON7) 19,20 were<br />
conducted in a total <strong>of</strong> 3,401 women with newly diagnosed<br />
cancers, and the third (OCEANS) 21 was conducted in 484<br />
women with platinum-sensitive relapse. Although the data<br />
from these studies are yet to mature, the results indicate a<br />
substantial prolongation <strong>of</strong> progression-free survival when<br />
administered with standard platinum-based combined chemotherapy<br />
followed by continuation <strong>of</strong> bevacizumab therapy.<br />
Several vascular endothelial growth factor receptor tyrosine<br />
kinase inhibitors are under evaluation as either<br />
monotherapy or combination therapy in recurrent ovarian<br />
cancer. One <strong>of</strong> this class <strong>of</strong> drugs, cediranib has demonstrated<br />
in a phase II trial a clinical benefit rate <strong>of</strong> 30% and<br />
median progression-free survival rate <strong>of</strong> 5.2 months. 22
GLOBAL ADVANCES IN GYNECOLOGIC ONCOLOGY<br />
A new class <strong>of</strong> anticancer drugs, known as PARP inhibitors,<br />
are also being evaluated. Repair pathways <strong>of</strong> DNA are<br />
critical for the maintenance <strong>of</strong> genome integrity and the<br />
response to DNA-damaging chemotherapy. Naturally<br />
occurring or environmentally induced single-stranded<br />
DNA breaks are generally repaired by the enzyme PARP.<br />
In the presence <strong>of</strong> a PARP inhibitor, single-stranded breaks<br />
may accumulate, leading to double-stranded breaks, usually<br />
repaired through separate molecules involved in the<br />
process known as homologous recombination repair (HRR).<br />
In cells with defective HRR, accumulated double-stranded<br />
breaks generally lead to apoptosis. It has been demonstrated<br />
that EOC in women with hereditary breast-ovarian cancer<br />
syndrome harbor such defects in HRR because <strong>of</strong> the<br />
germline mutations in BRCA1 and BRCA2. Kaye and colleagues<br />
reported on a phase II randomized trial <strong>of</strong> a PARP<br />
inhibitor known as olaparib compared with placebo in<br />
women with recurrent unselected serous EOC who were in<br />
partial or complete response after the last platinumcontaining<br />
chemotherapy regimen. 23 The drug was well<br />
tolerated, and results showed a substantially prolonged<br />
progression-free survival with a 65% reduction in progression<br />
in the olaparib group. Survival data are not yet available.<br />
Other agents in clinical trials include a therapeutic antifolate<br />
receptor antibody (farletuzamab), which may block<br />
folate receptor � signal transduction and promote immunogenicity.<br />
Folate receptor � is relatively absent in normal<br />
tissues and transduces a mitogenic signal on binding to<br />
circulating folate. EC145 is a synthetic agent composed <strong>of</strong><br />
folate covalently linked to a highly toxic vinca alkaloid. Once<br />
Author’s Disclosures <strong>of</strong> Potential Conflicts <strong>of</strong> Interest<br />
Author<br />
Employment or<br />
Leadership<br />
Positions<br />
Consultant or<br />
Advisory Role<br />
internalized into the cell, EC145 binds to folate receptor �<br />
and the vinca alkaloid is activated.<br />
The use <strong>of</strong> PEGylation to alter the pharmacokinetics<br />
and dynamics <strong>of</strong> cytotoxic agents is another advance. A<br />
novel pegylated topoisomerase I inhibitor has demonstrated<br />
single-agent response rates up to 20% for platinum-resistant<br />
EOC in two trials. 24,25 Resistance to platinum agents is a<br />
major obstacle to therapeutic outcomes in many patients.<br />
Recent studies suggest that methylation <strong>of</strong> promotor<br />
regions <strong>of</strong> genes normally required to induce apoptosis may<br />
be one <strong>of</strong> the causes <strong>of</strong> platinum resistance. DNA methyltransferase<br />
inhibitors may reverse this process. Two studies<br />
have produced some preliminary evidence that demethylation<br />
agents can partially restore sensitivity to platinum<br />
therapy. 26,27<br />
Conclusion<br />
The understanding <strong>of</strong> the natural history <strong>of</strong> gynecologic<br />
malignant tumors, combined with some well-conducted<br />
randomized clinical trials, has markedly advanced the<br />
treatment <strong>of</strong> women with gynecologic cancers during the<br />
past 10 to 20 years. The biggest change has occurred in<br />
cervical cancer prevention rather than management <strong>of</strong> cervical<br />
or other gynecologic cancers. Increasingly, targeted<br />
therapies are showing efficacy in ovarian cancers, and for<br />
others, more rational and less mutilating surgery is being<br />
performed. The role <strong>of</strong> adjuvant therapies in endometrial<br />
cancer still needs to be clarified, but in all cancers <strong>of</strong> the<br />
female genital tract, individualization <strong>of</strong> treatment along<br />
with oversight <strong>of</strong> a multidisciplinary team will most likely<br />
improve outcomes.<br />
Stock<br />
Ownership Honoraria<br />
Lynette Denny GlaxoSmithKline;<br />
Merck<br />
1. Parkin DM, Bray F. Chapter 2: the burden <strong>of</strong> HPV-related cancers.<br />
Vaccine. 2006;24 S3:S3/11-S3/25.<br />
2. Toki T, Kurman RJ, Park JS, et al. Probably nonpapillomavirus etiology<br />
<strong>of</strong> squamous cell carcinoma <strong>of</strong> the vulva in older women: a clinicopathologic<br />
study using in situ hybridization and polymerase chain reaction. Int J<br />
Gynecol Pathol. 1991;10:107-125.<br />
3. Eva LJ, Ganesan R, Chan KK, et al. Differentiated-type vulval intraepithelial<br />
neoplasia has a high-risk association with vulval squamous cell<br />
carcinoma. Int J Gynecol Cancer. 2009;19:741-744.<br />
4. Hacker NF, Leuchter RS, Berek JS, et al. Radical vulvectomy and<br />
bilateral inguinal lymphadenectomy through separate groin incisions. Obstet<br />
Gynecol. 1981;58:574-579.<br />
5. Siller BS, Alvarez RD, Conner WD, et al. T2/3 vulva cancer: a casecontrol<br />
study <strong>of</strong> triple incision versus en bloc radical vulvectomy and inguinal<br />
lymphadenectomy. Gynecol Oncol. 1995;57:335-339.<br />
6. Lin JY, DuBeshter B, Angel C, et al. Morbidity and recurrence with<br />
modifications <strong>of</strong> radical vulvectomy and groin dissection. Gynecol Oncol.<br />
1992;47:116-121.<br />
7. Woelber L, Choschzick M, Eulenburg C, et al. Prognostic value <strong>of</strong><br />
pathological resection margin distance in squamous cell cancer <strong>of</strong> the vulva.<br />
Ann Surg Oncol. 2011;18:3811-3818.<br />
8. Moore DH, Ali S, Koh W-J, Michael H, et al. A phase II trial <strong>of</strong> radiation<br />
therapy and weekly cisplatin chemotherapy for the treatment <strong>of</strong> locallyadvanced<br />
squamous cell carcinoma <strong>of</strong> the vulva: a gynecologic oncology group<br />
study. Gynecol Oncol. <strong>2012</strong>;124:529-533.<br />
REFERENCES<br />
Research<br />
Funding<br />
GlaxoSmithKline;<br />
Merck<br />
Expert<br />
Testimony<br />
Other<br />
Remuneration<br />
9. Rogers LJ, Howard B, Van Wijk L, et al. Chemoradiation in Advanced<br />
Vulval Cancer. Int J Gynecol Cancer. 2009;19:745-751.<br />
10. Van der Zee AG, Oonk MH, De Hullu JA, et al. Sentinel node dissection<br />
is safe in the treatment <strong>of</strong> early-stage vulvar cancer. J Clin Oncol. 2008;26:<br />
884-889.<br />
11. Denny L, Kuhn L, Hu C, et al. Human papillomavirus-based cervical<br />
cancer prevention: long-term results <strong>of</strong> a randomized screening trial. J Natl<br />
Cancer Inst. 2010;102:1-11.<br />
12. Sankaranarayanan R, Nene BM, Shastri SS, et al. HPV screening for<br />
cervical cancer in rural India. N Engl J Med. 2009;360:1385-1394.<br />
13. Guglielmo R, Meijer CJLM. HPV screening: available data and recommendations<br />
for clinical practice. Curr Cancer Ther Rev. 2010;6:104-109.<br />
14. International Agency for Research on Cancer. GLOBOCAN 2008.<br />
http://globocan.iarc.fr. Accessed January <strong>2012</strong>.<br />
15. ASTEC study group, Kitchener H, Swart AM, et al. Efficacy <strong>of</strong> systematic<br />
pelvic lymphadenectomy in endometrial cancer (MRC ASTEC trial): a<br />
randomised study. Lancet. 2009;373:125-136.<br />
16. Creasman WT, Mutch DE, Herzog TJ. ASTEC lymphadenectomy and<br />
radiation therapy studies: are conclusions valid? Gynecol Oncol. 2010;116:<br />
293-294.<br />
17. Nout RA, Smit VT, Putter H, et al. Vaginal brachytherapy versus pelvic<br />
external beam radiotherapy for patients with endometrial cancer <strong>of</strong> highintermediate<br />
risk (PORTEC-2): an open-label, non-inferiority, randomised<br />
trial. Lancet. 2010;375:816-823.<br />
18. Siegel R, Ward E, Brawley O, et al. Cancer statistics, 2011; the impact<br />
333
<strong>of</strong> eliminating socioeconomic and racial disparities on premature cancer<br />
deaths. CA Cancer J Clin. 2011;61:212-236.<br />
19. Burger RA, Brady MF, Bookman MA, et al. Phase III trial <strong>of</strong> bevacizumab<br />
(BEV) in the primary treatment <strong>of</strong> advanced epithelial ovarian cancer<br />
(EOC), primary peritoneal cancer (PPC) or fallopian tube cancer (FTC): a<br />
Gynecologic <strong>Oncology</strong> Group study. J Clin Oncol. 2010:28;18s (suppl; abstr<br />
LBA1).<br />
20. Perren TS, Pfisterer J, Ledermann J, et al. ICON7: a phase III<br />
Gynaecologic Cancer Intergroup (GCIG) trial <strong>of</strong> adding bevacizumab to<br />
standard chemotherapy in women with newly diagnosed epithelial ovarian,<br />
primary peritoneal or fallopian tube cancer. Ann Oncol. 2010:21 (suppl 8;<br />
abstr LBA4).<br />
21. Aghajanian C, Finkler NJ, Rutherford T, et al. OCEANS: a randomized<br />
double -blinded, placebo-controlled phase III trial <strong>of</strong> chemotherapy with or<br />
without bevacizumab (BEV) in patients with platinum-sensitive recurrent<br />
epithelial ovarian (EOC), primary peritoneal (PPC), or fallopian tube cancer<br />
(FTC). J Clin Oncol. 2011:29;TK (suppl; abstr LBA5007).<br />
22. Matulonis UA, Berlin S, Ivy P, et al. Cediranib, an oral inhibitor <strong>of</strong><br />
vascular endothelial growth factor receptor kinases, is an active drug in<br />
334<br />
LYNETTE DENNY<br />
recurrent epithelial ovarian, fallopian tube, and peritoneal cancer. J Clin<br />
Oncol. 2009;27:5601-5606.<br />
23. Ledermann JA, Harter P, Gourley C, et al. Phase II randomized<br />
placebo-controlled study <strong>of</strong> olaparib (AZD2281) in patients with platinumsensitive<br />
relapsed serous ovarian cancer. J Clin Oncol. 2011;29; TK (suppl;<br />
abstr 5003).<br />
24. Vergote IB, Micha JP, Pippitt CH Jr, et al. Phase II study <strong>of</strong> NKTR-102<br />
in women with platinum-resistant/refractory ovarian cancer. J Clin Oncol.<br />
2010;28:15s (suppl; abstr 5013).<br />
25. Garcia AA, Vergote IB, Micha JP, et al. The role <strong>of</strong> NKTR-102 in<br />
women with platinum resistant/refractory ovarian cancer and failure on<br />
pegylated liposomal doxorubicin (PLD). J Clin Oncol. 2011;29 (suppl; abstr<br />
5047).<br />
26. Fu S, Hu W, Iyer R, et al. Phase 1b-2a study to reverse platinum<br />
resistance through use <strong>of</strong> a hypomethylating agent, azacitidine, in patients<br />
with platinum-resistant or platinum-refractory epithelial ovarian cancer.<br />
Cancer. 2011;117:1661-1669.<br />
27. Matei D, Shen C, Fang F, et al. A phase II study <strong>of</strong> decitabine and<br />
carboplatin in recurrent platinum (Pt)-resistant ovarian cancer (OC). J Clin<br />
Oncol. 2011;29;TK (suppl; abstr 5011).
The European <strong>Society</strong> <strong>of</strong> Gynaecological<br />
<strong>Oncology</strong>: Update on Objectives and<br />
Educational and Research Activities<br />
By Renata Brantnerova, Ranjit Manchanda, MD, and Nicoletta Colombo, MD<br />
Overview: The European <strong>Society</strong> <strong>of</strong> Gynaecological <strong>Oncology</strong><br />
(ESGO) is the principal European society contributing to the<br />
study, prevention, and treatment <strong>of</strong> gynecologic cancers.<br />
Founded in 1983, ESGO has more than 1,300 members in more<br />
than 40 European countries and worldwide who benefit from<br />
ESGO’s innovative education and research initiatives and<br />
networking opportunities. ESGO objectives have been recently<br />
identified through a strategic planning process and include<br />
education, care, research, collaboration, awareness, and sustainability.<br />
As a leading gynecologic oncology society, ESGO<br />
holds biennial meetings where experts meet to discuss latest<br />
advances in gynecologic treatment and care. The 17th International<br />
Meeting <strong>of</strong> ESGO (ESGO 17) proved to be a resounding<br />
success, with 2,700 delegates and speakers who gathered<br />
from around the world in the cultured city <strong>of</strong> Milan, Italy. The<br />
structure <strong>of</strong> the congress included keynote lectures, debates,<br />
FOUNDED IN 1983 in Italy as a forum for pr<strong>of</strong>essionals<br />
dedicated to the care <strong>of</strong> women with gynecologic cancers,<br />
the European <strong>Society</strong> <strong>of</strong> Gynaecological <strong>Oncology</strong><br />
(ESGO) will soon celebrate 30 years <strong>of</strong> existence.<br />
The mission <strong>of</strong> ESGO is to improve the health and wellbeing<br />
<strong>of</strong> European women with gynecologic (genital and<br />
breast) cancers through prevention, excellence in care, and<br />
high-quality research and education.<br />
Looking back over the last decade, our membership has<br />
increased rapidly from being an exclusive club <strong>of</strong> about 150<br />
members in 2002 into a truly pan-European and international<br />
community <strong>of</strong> multidisciplinary pr<strong>of</strong>essionals: ESGO<br />
had 1,330 members in 2011, with an additional 240 new<br />
members joining in the last quarter <strong>of</strong> the year. ESGO is a<br />
young society: In 2011, 37% <strong>of</strong> members declared to be<br />
younger than age 45 and 25% <strong>of</strong> all members were part <strong>of</strong><br />
the European Network <strong>of</strong> Young Gynaecologic Oncologists<br />
(ENYGO), which is the ESGO’s European platform for trainees<br />
and pr<strong>of</strong>essionals younger than age 40. ESGO cares about<br />
members from all regions <strong>of</strong> Europe: Members from lowerincome<br />
countries form 27% <strong>of</strong> ESGO active membership.<br />
European Challenges in Gynecologic <strong>Oncology</strong><br />
There are approximately 630,000 women diagnosed with<br />
gynecologic cancers in Europe every year (including breast<br />
cancer). Patterns <strong>of</strong> incidence and mortality are different<br />
from region to region, and there are continuing inequalities<br />
both among and within European countries in terms <strong>of</strong><br />
cancer risk, detection, and treatment. The estimated incidence<br />
for cervical cancer is 54,517 new cases/year with<br />
24,874 annual deaths related to this disease. 1 Cervical<br />
cancer is the seventh most common cancer among European<br />
women <strong>of</strong> all ages, and the second most common cancer in<br />
women ages 15 to 39 years. 2 Rates are highest among<br />
women in countries <strong>of</strong> the former Soviet Union, who have<br />
twice the risk <strong>of</strong> dying as a result <strong>of</strong> cervical cancer as<br />
Western European women. Women in Romania are 11 times<br />
more likely than Finnish women to die as a result <strong>of</strong> cervical<br />
cancer. 3<br />
Ovarian cancer is the fifth most common cancer among<br />
state-<strong>of</strong>-the art sessions, and focused sunrise sessions, together<br />
with oral and poster presentations and satellite symposia<br />
sponsored by pharmaceutical companies. For the first<br />
time, during ESGO 17 the <strong>Society</strong> organized a seminar for<br />
European patient groups with an interest in gynecologic<br />
cancers with the aim <strong>of</strong> facilitating different patientrelated<br />
activities across Europe. Moreover, The European Network <strong>of</strong><br />
Young Gynaecologic Oncologists (ENYGO), the European Network<br />
<strong>of</strong> Gynaecologic <strong>Oncology</strong> Trial Groups (ENGOT), and<br />
the European Network <strong>of</strong> Translational Research in Gynaecological<br />
<strong>Oncology</strong> (ENTRIGO) had their own section during<br />
ESGO 17. ESGO also holds numerous workshops throughout<br />
the calendar year and provides clinical and research grants,<br />
online educational materials, webcasts, and numerous networking<br />
opportunities<br />
women in Europe, with more than 65,000 new cases and<br />
41,000 deaths annually. Rates are highest in Eastern and<br />
Northern Europe and lowest in Southern Europe. 1<br />
Uterine cancer is the fourth most common female cancer<br />
in Europe, with 88,000 new cases/year and 22,000 deaths.<br />
The highest incidence is observed in Eastern and Northern<br />
Europe, and lowest in southern regions. 1<br />
Breast cancer is the most common cancer in women in<br />
Europe, with over 425,000 new cases/year. Breast cancer is<br />
responsible for more deaths among European women than<br />
any other cancer, but survival has improved thanks to<br />
screening and better treatments. However, survival is<br />
higher in Western than in Eastern Europe, with further<br />
differences within Western Europe: lower in the United<br />
Kingdom and Denmark than in Finland, Sweden, France,<br />
Italy, and the Netherlands. 2 These statistics outline how<br />
there are unacceptable inequities in access to information,<br />
prevention, screening, treatment and care across Europe.<br />
Patterns <strong>of</strong> incidence and mortality are different from region<br />
to region, <strong>of</strong>ten reflecting a lack <strong>of</strong> awareness and access to<br />
appropriate services. ESGO is committed to help women<br />
with gynecological cancer across Europe to obtain accurate,<br />
reliable and timely information about their disease, to understand<br />
treatment options and to have access to the best<br />
possible care.<br />
ESGO Objectives<br />
Considering the changing external and internal environment,<br />
as well as global and medical trends affecting the field<br />
<strong>of</strong> gynecologic oncology, ESGO conducted strategic planning<br />
From the European <strong>Society</strong> <strong>of</strong> Gynaecological <strong>Oncology</strong>, the UCL Elizabeth Garrett<br />
Anderson Institute for Women’s Health, and the University <strong>of</strong> Milan Bicocca, European<br />
Institute <strong>of</strong> <strong>Oncology</strong>.<br />
Authors’ disclosures <strong>of</strong> potential conflicts <strong>of</strong> interest are found at the end <strong>of</strong> this article.<br />
Address reprint requests to Nicoletta Colombo, MD, University <strong>of</strong> Milan Bicocca,<br />
European Institute <strong>of</strong> <strong>Oncology</strong>, Via Ripamonti 435, 20141 Milan, Italy; email:<br />
Nicoletta.colombo@ieo.it.<br />
© <strong>2012</strong> by <strong>American</strong> <strong>Society</strong> <strong>of</strong> <strong>Clinical</strong> <strong>Oncology</strong>.<br />
1092-9118/10/1-10<br />
335
during 2011. The aims were to analyze the current positioning<br />
<strong>of</strong> the <strong>Society</strong>; define its future as Europe’s leader in<br />
education, research, and care in the field <strong>of</strong> gynecologic<br />
cancers; and ensure ESGO’s continuous development as an<br />
active resource <strong>of</strong> knowledge, information, collaboration and<br />
networking for its members and all pr<strong>of</strong>essionals in the field.<br />
Having in mind the ESGO mission to improve the health<br />
and well-being <strong>of</strong> European women with genital and breast<br />
cancers, core fields <strong>of</strong> interest were defined and general<br />
objectives were established: 1) education—to provide high<br />
quality educational activities and improve training in<br />
gynaecologic oncology in Europe; 2) care—to establish multidisciplinary<br />
standards <strong>of</strong> care and act as the European<br />
authority in the field; 3) research—to support platforms for<br />
collaborative clinical, translational, and basic research in<br />
gynecologic cancers; 4) collaboration—to promote collaboration<br />
among scientific societies, health care pr<strong>of</strong>essionals,<br />
patient organizations, business, industry, and governmental<br />
bodies; 5) awareness—to raise public and governmental<br />
awareness <strong>of</strong> gynecologic cancers and their prevention and<br />
treatment; and 6) sustainability—to ensure ESGO’s sustainability<br />
and positioning as the leading European pr<strong>of</strong>essional<br />
medical society in its field.<br />
Education and Outreach<br />
One <strong>of</strong> ESGO’s primary missions is to provide high-quality<br />
educational activities and improve training in gynecologic<br />
oncology in Europe.<br />
The highlight <strong>of</strong> the educational activities is organizing<br />
the ESGO international bi-annual meeting, which has a<br />
main focus on education, but with an ongoing trend to<br />
present science and research. The ESGO 17 proved to be a<br />
resounding success, with 2,700 delegates and speakers who<br />
gathered from around the world in the cultured city <strong>of</strong><br />
KEY POINTS<br />
● The European <strong>Society</strong> <strong>of</strong> Gynaecological <strong>Oncology</strong><br />
(ESGO) is the principal European society contributing<br />
to the study, prevention, and treatment <strong>of</strong> gynecological<br />
cancers.<br />
● ESGO objectives were recently revisited through a<br />
strategic planning process and include education,<br />
care, research, collaboration, awareness, and sustainability.<br />
● The highlight <strong>of</strong> the educational activities is ESGO’s<br />
international bi-annual meeting, which has a main<br />
focus on education, but with an ongoing trend to<br />
present science and research.<br />
● During ESGO 17 the <strong>Society</strong> organized a seminar for<br />
European patient groups with an interest in gynecologic<br />
cancers with the aim <strong>of</strong> facilitating different<br />
patient-related activities across Europe.<br />
● The European Network <strong>of</strong> Young Gynaecologic Oncologists<br />
(ENYGO) is a network for young physicians<br />
and trainees in gynecologic oncology and related<br />
subspecialties that represents the needs and aspirations<br />
<strong>of</strong> all trainees involved in the study, prevention,<br />
and treatment <strong>of</strong> gynecologic cancers.<br />
336<br />
BRANTNEROVA, MANCHANDA, AND COLOMBO<br />
Milan, Italy. The structure <strong>of</strong> the congress included keynote<br />
lectures, debates, state-<strong>of</strong>-the art sessions, and focused sunrise<br />
sessions, together with oral and poster presentations<br />
and satellite symposia sponsored by pharmaceutical companies.<br />
The keynote lectures were presented by Pr<strong>of</strong>. Umberto<br />
Veronesi and Pr<strong>of</strong>. Rene Bernards. Pr<strong>of</strong>. Veronesi focused on<br />
the progress in breast cancer management, highlighting his<br />
lifetime achievements on the quality <strong>of</strong> life QOL <strong>of</strong> women<br />
with breast cancer. By proposing a new paradigm “from<br />
maximum tolerable treatment to minimum effective treatment,”<br />
he was able to completely modify the care <strong>of</strong> patients<br />
with breast cancer: going from the anatomic concept <strong>of</strong><br />
cancer spread to the biologic one, the therapy for these<br />
patients changed from the mutilated radical mastectomy to<br />
the conservative quadrantectomy, from the complete lymph<br />
node dissection to the sentinel-node biopsy and more recently<br />
from the lengthy external radiotherapy to the intraoperative<br />
radiotherapy. All these achievements led to a<br />
substantial improvement in the QOL <strong>of</strong> women with breast<br />
cancer.<br />
Pr<strong>of</strong>. Bernards highlighted that fact that only 22% <strong>of</strong><br />
patients with cancer derive substantial benefit from their<br />
treatment, mainly because <strong>of</strong> the intra- and intertumor<br />
heterogeneity. To answer the basic questions <strong>of</strong> whom to<br />
treat and in what way, a more sophisticated approach is<br />
needed. Molecular diagnostics has the potential to replace<br />
the microscope in identifying prognostic and predictive<br />
markers that should guide a more personalized cancer<br />
treatment.<br />
One <strong>of</strong> the main highlights <strong>of</strong> the plenary session included<br />
the presentation <strong>of</strong> the results <strong>of</strong> the secondary end point—<br />
QOL—<strong>of</strong> an international study which randomly assigned<br />
patients with recurrent ovarian cancer to receive maintenance<br />
olaparib or placebo. 4 There was no statistically significant<br />
difference in the QOL scales between treatment arms.<br />
The median time to worsening QOL was 2.8 months for<br />
olaparib compared with 3.7 months for placebo, and 70%<br />
<strong>of</strong> patients receiving olaparib experienced nausea compared<br />
with 43% receiveing placebo (hazard ratio [HR] � 2.18).<br />
Results <strong>of</strong> a phase II study combining olaparib with<br />
pegylated liposomal doxorubicin (PLD) in patients with<br />
advanced solid tumors (28 ovarian cancer, 13 breast cancer,<br />
three others) who had received at least three prior chemotherapy<br />
regimens were presented by Del Conte and colleagues<br />
during the ovarian oral session. 5 The maximum<br />
tolerated dose was not reached. Common adverse events<br />
(AEs) <strong>of</strong> any grade included stomatitis (73%), nausea (71%),<br />
asthenia (57%), pyrexia (43%), anorexia (41%), vomiting<br />
(41%), cough (39%), neutropenia (27%) and palmar-plantar<br />
erythrodysesthesia (25%). Serious AEs occurred in 27% <strong>of</strong><br />
patients, with pneumonitis being most common (n � 3). Two<br />
patient deaths were considered “possibly related” to study<br />
therapy. Treatment with olaparib and PLD resulted in two<br />
(5%) complete responses (CRs), 11 (25%) partial responses<br />
(PRs), and 13 (30%) incidents <strong>of</strong> stable disease (SD). Of<br />
the 13 patients who achieved a PR or better, 11 were BRCA<br />
mutation positive and two were BRCA mutation negative or<br />
unknown.<br />
Dr. Fischerova received the award for the best poster<br />
presentation for her work, “The Role <strong>of</strong> Ultrasound in<br />
Planning Fertility Sparing Surgery and Individual Treatment<br />
in Early Stage Cervical Cancer.” She analyzed the
ESGO <strong>EDUCATIONAL</strong> AND RESEARCH ACTIVITIES<br />
accuracy <strong>of</strong> ultrasound in predicting tumor size, parametrial<br />
and lymph node involvement, distance between the tumor<br />
and the internal cervical os, and depth <strong>of</strong> stromal invasion<br />
in 83 patients undergoing radical hysterectomy for stage IB<br />
cervical cancer. Overall sensitivity <strong>of</strong> ultrasound imaging in<br />
the evaluation <strong>of</strong> parameters required for fertility-sparing<br />
surgery was 68.4%. The lower sensitivity mainly resulted<br />
from lack <strong>of</strong> detection <strong>of</strong> lymph-node involvement. 6<br />
The oral translation research award was assigned to<br />
Dr. Lai 7 for her presentation, “Human Ovarian Cancer<br />
Stem Cells Can Be Efficiently Killed by Gamma Delta<br />
T-Lymphocytes.” A subset <strong>of</strong> cancer stem cells (CSCs) from<br />
SKOV3 cell line were cocultured with gamma delta T cells.<br />
Their proliferation rate decreased to 40% after 48 hours <strong>of</strong><br />
exposure. The gamma delta T cells increased the sensitivity<br />
<strong>of</strong> SKOV3 CSCs to chemotherapeutic drugs. Dr. Lai also<br />
found that gamma delta T cells induced G2/M phase cellcycle<br />
arrest and subsequent apoptosis in SKOV3 CSCs.<br />
Moreover, xenograft mouse models demonstrate that<br />
gamma delta T cells dramatically reduced the tumor burden<br />
in vivo.<br />
Finally, the prize for the best oral presentation in the<br />
young pr<strong>of</strong>essionals’ session was awarded to “A Risk Model<br />
for Secondary Cytoreductive Surgery in Recurrent Ovarian<br />
Cancer: An Evidence-Based Proposal for Patient Selection.” 8<br />
Individual data on 1,075 patients with recurrent ovarian<br />
cancer undergoing secondary cytoreductive surgery were<br />
pooled and analyzed. Complete secondary cytoreduction was<br />
associated with six variables: International Federation <strong>of</strong><br />
Gynecology and Obstetrics (FIGO) stage (odds ratio [OR] �<br />
1.32), residual disease after primary cytoreduction (OR �<br />
1.69), progression-free interval (OR � 2.27), Eastern Cooperative<br />
<strong>Oncology</strong> Group (ECOG) performance status (OR �<br />
2.23), cancer antigen 125 (OR � 1.85), and ascites at<br />
recurrence (OR � 2.79). These variables were entered into<br />
the risk model and assigned scores to identify low- and<br />
high-risk categories. In external validation, the sensitivity<br />
and specificity were 83.3% and 57.6%, respectively.<br />
ESGO also provides online educational resources available<br />
on ESGO website: 1) webcasts <strong>of</strong> ESGO meeting sessions<br />
and e-Posters that extend the life <strong>of</strong> ESGO meetings<br />
and bring them to members who could not attend; 2) an<br />
online Video Library that provides a unique opportunity for<br />
visual education; 3) e-Series, which are webcast presentations<br />
by internationally renowned speakers; 4) educational<br />
DVDs; and 5) the Textbook <strong>of</strong> Gynaecological <strong>Oncology</strong>, a<br />
unique publication that compiles articles from world-famous<br />
authors in the field <strong>of</strong> gynecologic oncology, including surgery,<br />
radiotherapy, chemotherapy and imaging specialties.<br />
Care<br />
Committed to the cause <strong>of</strong> excellence in care, ESGO aims<br />
to establish multidisciplinary standards <strong>of</strong> care and acts<br />
as the European authority in the field. Therefore, together<br />
with the European training standards, ESGO set the<br />
multidisciplinary standards <strong>of</strong> care and provides supervision<br />
for certified training. Within the ESGO-European<br />
Board and College <strong>of</strong> Obstetrics and Gynaecology (EBCOG)<br />
Hospital Accreditation Program, ESGO has provided hospital<br />
recognition <strong>of</strong> excellence in training and care while<br />
certifying training centers across Europe. Trainees trained<br />
at ESGO-accredited centers become ESGO-certified gyneco-<br />
logic oncologists. Currently, more than 35 European hospitals<br />
and training centers have been accredited.<br />
Research<br />
One <strong>of</strong> ESGO’s ambitious goals is to become the voice <strong>of</strong><br />
European research in gynecologic cancers and the coordinating<br />
body <strong>of</strong> European clinical and translational research,<br />
with the aim to improve the prediction, prevention, detection,<br />
and treatment <strong>of</strong> women-specific cancers.<br />
In 2007, at ESGO’s congress in Berlin, Germany, the<br />
European Network <strong>of</strong> Gynaecological Oncological Trial<br />
Groups (ENGOT) was founded. Currently, the network<br />
consists <strong>of</strong> 17 European cooperative groups and national and<br />
regional clinical trial units. A consensus statement on requirements<br />
for trials between academic groups and pharmaceutical<br />
companies has been produced and a roadmap for<br />
clinical trials in Europe has been designed. 9<br />
To complement these existing activities and make further<br />
advances, in 2009 ESGO formed the European Network<br />
<strong>of</strong> Translational Research in Gynaecological <strong>Oncology</strong><br />
(ENTRIGO), a platform that brings together scientists,<br />
clinical academics, and patients. In 2011, ESGO organized<br />
the first ENTRIGO Translational Research Workshop,<br />
which brought together the leading European scientists<br />
in the fields <strong>of</strong> epidemiology, molecular carcinogenesis, personalized<br />
medicine and molecular imaging, pathology, and<br />
cancer genetics and epigenetics. In a long-term perspective,<br />
the aim <strong>of</strong> ENGOT is to run large European Union–funded,<br />
trans-European research programs.<br />
Collaboration and Networking<br />
The goal <strong>of</strong> ESGO is to be a platform promoting and<br />
facilitating collaboration among scientific societies, health<br />
care pr<strong>of</strong>essionals, patient organizations, business, industry,<br />
and governmental bodies. Within this aim, ESGO’s<br />
biennial meetings give the best networking opportunities to<br />
meet and discuss women-specific cancers, not only to ESGO<br />
members but also to the wide community <strong>of</strong> medical pr<strong>of</strong>essionals<br />
interested in these cancers.<br />
Networking is rated by members as one <strong>of</strong> the most<br />
important appeals <strong>of</strong> initiating or renewing ESGO membership.<br />
Task forces have been recently established within the<br />
ESGO community to give members opportunities to exchange<br />
ideas and debate a wide range <strong>of</strong> related topics (e.g.,<br />
cancer in pregnancy, fertility preservation, psycho-oncology,<br />
surgery harmonization). Members are welcome to be involved<br />
in the association, use the ESGO newsletter and<br />
website to promote their open projects, and search for<br />
collaboration with their colleagues.<br />
ESGO pays special attention to cooperation with sister<br />
societies, with regional and international societies <strong>of</strong> gynecologic<br />
cancer and mainly with the International <strong>Society</strong> <strong>of</strong><br />
Gynecologic Cancer (IGCS), <strong>Society</strong> <strong>of</strong> Gynecologic <strong>Oncology</strong><br />
(SGO), and Asian <strong>Society</strong> <strong>of</strong> Gynecologic cancer (AGCS).<br />
ESGO is a member and representative <strong>of</strong> gynecologic oncologic<br />
subspecialty <strong>of</strong> EBCOG and a member <strong>of</strong> European<br />
Cancer Organisation (ECCO) which is the umbrella organization<br />
<strong>of</strong> all European cancer-related societies.<br />
Awareness<br />
One <strong>of</strong> ESGO’s objectives is to raise public and governmental<br />
awareness <strong>of</strong> gynecologic cancers in Europe, and <strong>of</strong><br />
their prevention and treatment. The incidence <strong>of</strong> and mor-<br />
337
tality from gynecologic cancers are increasing in Europe,<br />
and all women—especially older women—are at risk. Trustworthy<br />
public and patient information and education are<br />
lacking, especially in non–English–speaking countries.<br />
As the leading European pr<strong>of</strong>essional society related to<br />
cancer, ESGO shares the goal <strong>of</strong> patient groups to help<br />
women living with gynecologic cancers to obtain accurate,<br />
reliable, and timely information about their disease, to<br />
understand treatment options and to have access to the best<br />
possible care. For the first time, during ESCO 17 the <strong>Society</strong><br />
organized a seminar for European patient groups with an<br />
interest in gynecologic cancers. Fifty participants, including<br />
38 representatives <strong>of</strong> patient organizations, attended.<br />
ESGO shares common goals with patient organizations,<br />
and wishes to become a real facilitator <strong>of</strong> different patientrelated<br />
activities across Europe. The seminar at ESGO 17<br />
was only the start <strong>of</strong> the partnership, and ESGO is committed<br />
to provide a continuing platform for patient organizations<br />
to meet and work together. Ten recommendations<br />
coming out <strong>of</strong> this patient seminar have been approved by<br />
the ESGO Council, and a small committee formed to take<br />
these forward in <strong>2012</strong>.<br />
Focus on the Younger Generation: ENYGO<br />
ENYGO is a network for young pr<strong>of</strong>essionals and trainees<br />
in gynecologic oncology and related subspecialties. ENYGO<br />
was created in 2009 and is supported in its activities by<br />
ESGO. ENYGO is the principal European network representing<br />
the needs and aspirations <strong>of</strong> all trainees involved in<br />
the study, prevention, and treatment <strong>of</strong> gynecological cancers.<br />
It serves as a forum for promoting scientific and social<br />
interaction, discussion, debate, and exchange <strong>of</strong> ideas and<br />
views among trainees. ENYGO currently has approximately<br />
400 members from 40 countries across Europe, with each<br />
country having a national representative. ENYGO is represented<br />
on the ESGO council by its president and supported<br />
by a vice president and an executive group.<br />
Authors’ Disclosures <strong>of</strong> Potential Conflicts <strong>of</strong> Interest<br />
Author<br />
Employment or<br />
Leadership<br />
Positions<br />
Consultant or<br />
Advisory Role<br />
ENYGO Activities at ESGO 17<br />
ENYGO had its own session during the 17th ESGO<br />
conference. Details <strong>of</strong> outcomes from this workshop include:<br />
1) an introduction and retrospection about ENYGO by<br />
Dr. Michaela Bossart and past-president Dr. Boris Vranes;<br />
2) presentation <strong>of</strong> the initial results <strong>of</strong> the first Europe-wide<br />
survey on training experiences <strong>of</strong> European gynecologic<br />
oncology trainees by Dr. Ranjit Manchanda, which clearly<br />
highlighted the differences in training across countries and<br />
the importance and need for accredited training for trainees;<br />
3) an update on Teaching the Teachers workshop presented<br />
by Dr. Jurgen Piek, which highlighted the importance <strong>of</strong><br />
training teaching according to modern concepts and mechanisms<br />
such as the CanMEDS framework and specific, measurable,<br />
attainable, rewarding, time-bound (SMART) rules<br />
to facilitate the phenomenon <strong>of</strong> lifelong learning; 4) a presentation<br />
on training systems in Europe by Dr. Murat<br />
Gultekin; and 5) the launch and hard-copy/CD-ROM distribution<br />
<strong>of</strong> the second edition <strong>of</strong> the Textbook <strong>of</strong> Gynaecological<br />
<strong>Oncology</strong>, which was a big success and greatly appreciated<br />
by all trainees.<br />
Subsequently, Dr. Ranjit Manchanda coordinated an Internet<br />
survey, to which 154 members responded. This<br />
showed that surgical anatomy and laparoscopic surgery in<br />
gynecologic oncology were the topics most members would<br />
like to see covered by workshops. On the basis <strong>of</strong> these<br />
results, ENYGO is facilitating the organization <strong>of</strong> a combined<br />
2-day workshop focusing on surgical anatomy and<br />
laparoscopic surgery in London in September <strong>2012</strong>. After<br />
the successful Teaching the Teachers (TTT) workshop in<br />
Amsterdam, the Netherlands, in May 2011, the next TTT<br />
workshop is proposed to be held in Macedonia in June <strong>2012</strong>.<br />
ENYGO will also help facilitate and organize a young<br />
gynecologic oncologist session (YDS—Young Doctors’ Session)<br />
at the upcoming IGCS conference in Vancouver,<br />
Canada.<br />
Stock<br />
Ownership Honoraria<br />
Research<br />
Funding<br />
Renata Brantnerova*<br />
Ranjit Manchanda Abbott<br />
Laboratories<br />
Nicoletta Colombo Cancer<br />
Research UK;<br />
The Eve Appeal<br />
*No relevant relationships to disclose.<br />
1. International Agency for Research on Cancer. Globocan 2008 Fast Stats<br />
Factsheet: Europe. http://globocan.iarc.fr/factsheet.asp. Accessed February<br />
24, <strong>2012</strong>.<br />
2. Ferlay J, Parkin DM, Steliarova-Foucher E. Estimates <strong>of</strong> cancer incidence<br />
and mortality in Europe in 2008. Eur J Cancer. 2010;46:765-781.<br />
3. Arbyn M, Raifu AO, Autier P, et al. Burden <strong>of</strong> cervical cancer in Europe:<br />
estimates for 2004. Ann Oncol. 2007;18:1708-1715.<br />
4. Ledermann J, Harter P, Gourley C, et al. Phase 2randomized placebocontrolled<br />
study <strong>of</strong> olaparib(AZD2281) in patients with platinum-sensitive<br />
relapsed serous ovarian cancer (PSR SOC) Int J Gynecol Cancer. 2011;21:S13.<br />
5. Del Conte G, Sessa C, von Moos R, et al. Antitumor activity <strong>of</strong> olaparib<br />
(AZD2281) and liposomal doxorubicin in previously treated ovarian cancer<br />
patients. Int J Gynecol Cancer. 2011;21:S20.<br />
338<br />
REFERENCES<br />
BRANTNEROVA, MANCHANDA, AND COLOMBO<br />
Expert<br />
Testimony<br />
Other<br />
Remuneration<br />
6. Fischerova D, Freitag P, Zikan M, et al. The role <strong>of</strong> ultrasound in<br />
planning fertility sparing surgery and individual treatment in early stage<br />
cervical cancer. Int J Gynecol Cancer. 2011;21:S107.<br />
7. Lai D, Wang F, Chen Y at al. Human ovarian cancer stem cells can be<br />
efficiently killed by gamma dekta T-lymphocytes. Int J Gynecol Cancer.<br />
2011;21:S45.<br />
8. Tian WJ, Chi DS, Sehouli J, et al. A risk Model for secondary cytoreductive<br />
surgery in recurrent ovarian cancer: an evidence-based proposal for<br />
patient selection. Int J Gynecol Cancer. 2011;21:S55.<br />
9. Vergote I, Pujade-Lauraine E, Pignata S, et al. European Network <strong>of</strong><br />
Gynaecological Oncological Trial Groups’ requirements for trials between<br />
academic groups and pharmaceutical companies. Int J Gynecol Cancer.<br />
2010;20:476-478.
UPFRONT TREATMENT OF OVARIAN CANCER:<br />
INTERNATIONAL CONSENSUS AND VARIATION<br />
CHAIR<br />
Gordon J. S. Rustin, MD, MSc<br />
Mount Vernon Cancer Centre<br />
Middlesex, United Kingdom<br />
SPEAKERS<br />
Deborah K. Armstrong, MD<br />
Sidney Kimmel Comprehensive Cancer Center at Johns Hopkins University<br />
Baltimore, MD<br />
Keiichi Fujiwara, MD, PhD<br />
Saitama Medical University International Medical Center<br />
Hidaka, Japan
Biologicals in the Upfront Treatment <strong>of</strong><br />
Ovarian Cancer: Focus on Bevacizumab and<br />
Poly (ADP-Ribose) Polymerase Inhibitors<br />
By Rene Roux, MBBS, Martin Zweifel, MD, PhD, and Gordon J. S. Rustin, MD, MSc<br />
Overview: Biologicals have made a major impact in the<br />
management <strong>of</strong> several cancers, but have hitherto had a<br />
negligible impact in ovarian cancer. Fortunately, ovarian cancer<br />
has been much more sensitive to cytotoxic chemotherapy<br />
than many cancers, so treatments were still available. However,<br />
improvements are required as more than 80% <strong>of</strong> patients<br />
who present with advanced ovarian cancer eventually will die<br />
as a result <strong>of</strong> their disease.<br />
ANGIOGENESIS IS particularly relevant in ovarian<br />
cancer. 1,2 Vascular endothelial growth factor (VEGF)<br />
is important in both ovarian physiology and pathology.<br />
VEGF-A levels change cyclically with the menstrual cycle<br />
and blockage <strong>of</strong> the pathway can suppress ovulation. In<br />
ovarian cancer, intratumoral VEGF and VEGF receptor<br />
(VEGFR) 2 expression and the carriage <strong>of</strong> VEGF gene<br />
polymorphisms associated with increased VEGF excretion<br />
are independent poor prognostic factors. Increased levels <strong>of</strong><br />
VEGF have been shown to be an independent prognostic<br />
indicator <strong>of</strong> poor survival in early and advanced ovarian<br />
disease. Tumor VEGF overexpression is associated with<br />
tumor angiogenesis, malignant progression and metastasis,<br />
ascites formation, and early recurrence and death as a result<br />
<strong>of</strong> disease. In animal models, blockage <strong>of</strong> VEGF slows tumor<br />
growth and inhibits ascites formation.<br />
The most successful antiangiogenic agent developed to<br />
date is bevacizumab, a humanized monoclonal antibody<br />
which inhibits the binding <strong>of</strong> VEGF-A to its receptors,<br />
VEGFR1 and VEGFR2. This inhibits the formation <strong>of</strong> new<br />
tumor vessels and causes the remaining tumor vasculature<br />
to regress and normalize which prevents tumor growth and<br />
metastasis. In contrast to its activity in most other malignancies,<br />
bevacizumab has shown single-agent activity in<br />
ovarian cancer. In phase II studies it demonstrated response<br />
rates <strong>of</strong> 15.9% to 21%, a 6-month progression-free survival<br />
(PFS) <strong>of</strong> 28% to 40% and an overall survival (OS) <strong>of</strong> between<br />
10.7 and 16.9 months. 3,4<br />
Bevacizumab As First-Line Treatment in<br />
Ovarian Cancer<br />
Two pivotal studies using bevacizumab in the first-line<br />
management <strong>of</strong> ovarian cancer have recently been reported:<br />
ICON7 and GOG-0218, 5,6 with details in Table 1. Both trials<br />
showed similar improvement in PFS, with the ICON7 trial<br />
showing a significant improvement in OS in a high-risk<br />
subgroup (International Federation <strong>of</strong> Gynecology and Obstetrics<br />
[FIGO] stage III suboptimal debulked and FIGO<br />
stage IV), <strong>of</strong> 36.6 compared with 28.8 months. 5 It could be<br />
postulated that an antivascular approach will be more<br />
effective against larger tumors that require an additional<br />
blood supply than microscopic deposits remaining after<br />
optimal surgery. Major differences between the two trials<br />
included a difference in dose and duration <strong>of</strong> bevacizumab.<br />
GOG218 was a three-arm placebo-controlled trial whereas<br />
ICON7 had two arms with no placebo and included some<br />
stage I/II patients. Only in ICON7 will there be a reliable OS<br />
340<br />
The antiangiogenic antibody bevacizumab and the poly<br />
(ADP-ribose) polymerase (PARP) inhibitor olaparib have recently<br />
been shown to improve progression-free survival <strong>of</strong><br />
patients with ovarian cancer with better hazard ratios in<br />
certain groups than have been seen previously.<br />
end point because there was virtually no cross-over <strong>of</strong><br />
patients between the arms.<br />
Several other trials investigating the role <strong>of</strong> bevacizumab<br />
in first-line treatment <strong>of</strong> ovarian cancer are currently recruiting<br />
patients. The phase IV ROSiA trial (NCT01239732)<br />
is investigating the effect <strong>of</strong> adding 3-weekly bevacizumab<br />
15 mg/kg to paclitaxel 175 mg/m 2 3-weekly or 80 mg/m 2<br />
weekly and carboplatin area under the curve (AUC)<br />
3-weekly, followed by bevacizumab maintenance 3-weekly<br />
up to a total <strong>of</strong> 36 cycles in patients with previously untreated<br />
ovarian cancer. The four-arm phase III mEOC trial<br />
(GOG-0241; NCT01081262) investigates the role <strong>of</strong> bevacizumab<br />
15 mg/kg once every 3 weeks added to carboplatin<br />
and paclitaxel or to oxaliplatin and capecitabine in metastatic<br />
mucinous ovarian cancer.<br />
GOG-0252, a phase III three-arm trial (NCT00951496)<br />
investigating intravenous vs. intraperitoneal paclitaxel and<br />
platinum with the addition <strong>of</strong> bevacizumab 15 mg/kg in all<br />
three arms, including maintenance treatment for 16 cycles<br />
in previously untreated, optimally debulked stage III ovarian<br />
cancer. This will be the first trial investigating the role<br />
<strong>of</strong> bevacizumab in combination with intraperitoneal chemotherapy.<br />
Two trials are exploring the addition <strong>of</strong> bevacizumab to<br />
dose-dense chemotherapy but unfortunately no trials are<br />
comparing dose-dense with standard chemotherapy plus<br />
bevacizumab. The improvement in both PFS (hazard ratio<br />
[HR] � 0.71) and OS (HR � 0.75) from dose-dense paclitaxel<br />
plus 3-weekly carboplatin compared with standard 3-weekly<br />
paclitaxel and carboplatin compares favorably with the improvement<br />
seen with addition <strong>of</strong> bevacizumab. 7 OCTAVIA is<br />
a single-arm phase II trial <strong>of</strong> once every 3 weeks cycles <strong>of</strong><br />
treatment with bevacizumab 7.5 mg/kg added to paclitaxel<br />
80 mg/m 2 on days 1, 8, and 15 and carboplatin on day 1 <strong>of</strong><br />
each cycle. Bevacizumab may then continue to be administered<br />
as monotherapy until disease progression. Safety<br />
results from the trial have recently been reported showing<br />
that the addition <strong>of</strong> bevacizumab is well tolerated. 8 GOG-<br />
From the Mount Vernon Cancer Centre, Northwood, Middlesex, United Kingdom.<br />
Authors’ disclosures <strong>of</strong> potential conflicts <strong>of</strong> interest are found at the end <strong>of</strong> this article.<br />
Address reprint requests to Gordon J.S. Rustin, MD, MSc, Department <strong>of</strong> Medical<br />
<strong>Oncology</strong>, Mount Vernon Cancer Centre, Northwood, Middlesex HA6 2RN, UK; email:<br />
grustin@nhs.net.<br />
© <strong>2012</strong> by <strong>American</strong> <strong>Society</strong> <strong>of</strong> <strong>Clinical</strong> <strong>Oncology</strong>.<br />
1092-9118/10/1–10
ANTIANGIOGENICS AND PARP INHIBITORS IN OVARIAN CANCER<br />
Trial<br />
No. <strong>of</strong><br />
Patients Setting Trial Arms<br />
GOG-0218 6<br />
1873 First (CP � 6 � Pl � 5)<br />
line 3 Pl � 17<br />
ICON7 5<br />
OCEANS 9<br />
1528 First<br />
line<br />
0262 is a phase III trial <strong>of</strong> 3-weekly paclitaxel/carboplatin<br />
compared with dose-dense weekly paclitaxel/3-weekly carboplatin<br />
with optional addition <strong>of</strong> bevacizumab 15 mg/kg in<br />
both arms.<br />
Bevacizumab As Second-Line Treatment in<br />
Ovarian Cancer<br />
The OCEANS trial was a randomized, double-blinded,<br />
placebo-controlled phase III trial <strong>of</strong> carboplatin AUC 4<br />
on day 1 and gemcitabine 1000 mg/m 2 on days 1 and 8,<br />
3-weekly with or without bevacizumab 15 mg/kg in patients<br />
KEY POINTS<br />
Table 1. Summarized Results <strong>of</strong> the Three Phase III trials <strong>of</strong> Bevacizumab in Ovarian Cancer<br />
Bevacizumab<br />
Dose<br />
(mg/kg)<br />
● Bevacizumab increases progression-free survival<br />
(PFS) by 2 to 4 months when added to first- or<br />
second-line chemotherapy and continued as maintenance<br />
treatment in ovarian cancer.<br />
● Bevacizumab 7.5 mg/kg in a subgroup <strong>of</strong> patients<br />
with stage III suboptimally debulked and stage IV<br />
disease improved survival by 8 months.<br />
● Bevacizumab should be considered for patients who<br />
are at high risk <strong>of</strong> having a short PFS because the<br />
bevacizumab might delay symptoms and these patients<br />
with platinum-resistant disease are unlikely<br />
to have an opportunity to receive bevacizumab as<br />
part <strong>of</strong> second-line treatment.<br />
● Poly (ADP-ribose) polymerase (PARP) inhibitors have<br />
shown response rates <strong>of</strong> up to 69% in early-phase<br />
trials <strong>of</strong> a patient population with recurrent ovarian<br />
cancer and enriched for patients with BRCA1/2<br />
mutations.<br />
Overall<br />
Response Rate Progression Death<br />
% 95% CI p HR 95% CI p HR 95% CI p<br />
15 Not<br />
reported<br />
CP � 6 � BEV � 5 0.908* 0.795 to<br />
1.040<br />
(CP � 6 �BEV�5)<br />
3 BEV � 17<br />
484 Second (CG�Pl) � 6<br />
line, 3 Pl � 21<br />
platinum or PD<br />
sensitive (CG�BEV) � 6<br />
3 BEV � 21<br />
or PD<br />
CP � 6 7.5 48 0.87*† 0.77 to<br />
0.99<br />
(CP�BEV) � 6<br />
67 11% to � .001 0.73*‡ 0.60 to<br />
3 BEV � 12<br />
(difference<br />
� 19)<br />
28%<br />
0.93<br />
15 57 51% to<br />
64%<br />
78 73% to<br />
84%<br />
with platinum-sensitive recurrent disease. Preliminary data<br />
were presented at ASCO in 2011. 9 PFS was significantly<br />
increased in the bevacizumab arm (12.4 vs. 8.4 months).<br />
The AURELIA (NCT00976911) trial, which reports this<br />
year, evaluates the impact <strong>of</strong> adding bevacizumab (10 mg/kg<br />
every 2 weeks or 15 mg/kg every 3 weeks) to either dosedense<br />
paclitaxel 80 mg/m 2 weekly or topotecan 4 mg/m 2 on<br />
days 1, 8, and 15 <strong>of</strong> each 4-week cycle (or 1.25 mg/m 2 on day<br />
1 through 5 <strong>of</strong> each 3-week cycle) or liposomal doxorubicin<br />
40 mg/m 2 every 4 weeks. Only patients with platinumresistant<br />
ovarian cancer are eligible.<br />
The GOG-0213 (NCT0056585) trial is investigating the<br />
role <strong>of</strong> adding bevacizumab 15 mg/kg to six cycles <strong>of</strong> carboplatin<br />
AUC 5 and paclitaxel 175 mg/m 2 , followed by 3-weekly<br />
bevacizumab until disease progression in platinum-sensitive<br />
recurrent ovarian cancer. Patients are allowed to have<br />
received prior bevacizumab, and this trial is also evaluating<br />
the role <strong>of</strong> secondary cytoreductive surgery. 10<br />
Antiangiogenic Small-Molecule Tyrosine<br />
Kinase Inhibitors<br />
Median Survival<br />
(months)<br />
Progression<br />
Free Overall<br />
10.3 39.3<br />
.16* 1.036 0.827 to<br />
1.297<br />
.76 11.2 38.7<br />
0.717* 0.625 to � .001* 0.915 0.727 to .45 14.1 39.7<br />
0.824<br />
1.152<br />
0.48 0.39 to<br />
0.61<br />
Abbreviations: BEV, bevacizumab; C, carboplatin; G, gemcitabine; HR, hazard ratio; P, paclitaxel; Pl, placebo.<br />
* HR for progression or death.<br />
† All patients.<br />
§ Restricted mean values.<br />
‡ High-risk patients.<br />
.04 0.85 0.69 to<br />
1.04<br />
� .001 0.64 0.48 to<br />
0.85<br />
.11 22.4§<br />
24.1§<br />
.002 14.5§<br />
18.1§<br />
Not yet<br />
reached<br />
28.8<br />
36.6<br />
� .0001 0.751 0.537 to .094 8.4 Not yet<br />
1.052<br />
reached<br />
� .0001 12.4<br />
Randomized trials <strong>of</strong> three antiangiogenic tyrosine kinase<br />
inhibitors in patients with ovarian cancer should present<br />
results within the next 2 years. 2 The AGO-Ovar 12/LUME-<br />
Ovar 1 trial is evaluating intedanib in combination with<br />
3-weekly paclitaxel and carboplatin, followed by intedanib<br />
maintenance up to a total <strong>of</strong> 120 weeks in patients with<br />
stage IIB to IV epithelial ovarian, fallopian tube, or primary<br />
peritoneal cancer after prior tumor-debulking surgery. Pazopanib<br />
is being studied as maintenance therapy for up to 2<br />
years, in a randomized, two-arm, placebo controlled, phase<br />
III study in women with ovarian, fallopian tube, or primary<br />
peritoneal cancer that has not progressed after completing<br />
first-line chemotherapy for advanced ovarian cancer. The<br />
ICON6 trial is investigating carboplatin and paclitaxel che-<br />
341
motherapy with or without concurrent cediranib and also<br />
the effect <strong>of</strong> maintenance cediranib in platinum-sensitive<br />
relapsed ovarian cancer.<br />
Reasons for Adding a New Treatment<br />
There are many unresolved questions as to the optimal<br />
use <strong>of</strong> bevacizumab in patients with ovarian cancer. It is a<br />
very expensive therapy, so it is important to consider the<br />
questions in Table 2 and the issues that follow when deciding<br />
whether to use this new drug.<br />
Improved Survival<br />
Survival curves from the GOG-0218 and ICON7 trials are<br />
still falling, so it seems unlikely that there will be more<br />
long-term survivors or cures than with cytotoxic chemotherapy<br />
alone. Results from the ICON7 trial show that there is<br />
no improved survival for low-risk patients, but an almost<br />
8-months-longer median survival for patients at high risk<br />
for relapse. If the survival benefit from first-line use is<br />
unclear and OCEANS shows improved survival in patients<br />
after first relapse, a case could be made for reserving it for<br />
relapse therapy.<br />
Increased Response Rate<br />
ICON7 and OCEANS both demonstrated an increased<br />
response rate. A tumor response is likely to lead to resolution<br />
<strong>of</strong> cancer-related symptoms. This is <strong>of</strong> most relevance<br />
to patients who have experienced relapse because most<br />
patients receiving first-line treatment will have minimal<br />
symptoms from their cancer after surgery. Patients who<br />
do have cancer-related symptoms are usually candidates<br />
now for neoadjuvant therapy. However, administering bevacizumab<br />
in this setting might raise concerns regarding<br />
impaired wound healing after subsequent surgery.<br />
Delaying Relapse<br />
It is debatable whether the 2- to 4-month delay in PFS<br />
justifies the ongoing treatment with 3-weekly infusions,<br />
especially in view <strong>of</strong> the results from OCEANS showing that<br />
the addition <strong>of</strong> bevacizumab to second-line treatment seems<br />
to also delay PFS.<br />
Less Toxic than Standard Therapy<br />
Table 2. Unresolved Questions Regarding the Use <strong>of</strong> Bevacizumab in Ovarian Cancer<br />
Unresolved Question Option 1/Comment Option 2/Comment<br />
What dose? 15 mg/kg 3 weekly licensed 7.5 mg/kg 3 weekly probably as effective and cheaper<br />
In combination or alone? With chemotherapy as improves response rate Only as maintenance (only addition <strong>of</strong> maintenance improved<br />
PFS in GOG218)<br />
For what duration? Until progression For defined period-22 cycles<br />
Should evidence <strong>of</strong> progression<br />
be sought?<br />
CA-125 every 6 weeks, and or CT every 3 months<br />
will shorten therapy<br />
Bevacizumab has been shown to add to toxicity <strong>of</strong> standard<br />
chemotherapy, especially hypertension, bleeding,<br />
thromboembolic events, and gastric perforations. Administering<br />
bevacizumab as maintenance therapy reduces time<br />
without therapy.<br />
Rationale for PARP Inhibitors<br />
Poly (ADP-ribose) polymerase (PARP), an enzyme discovered<br />
almost 50 years ago, 11 is crucial for the repair <strong>of</strong><br />
single-strand DNA breaks (SSBs) via the base excision<br />
repair (BER) pathway. Although it has no direct effect on<br />
double-strand DNA breaks (DSBs), 12 when a DNA replication<br />
fork comes across a persistent SSB (i.e., if PARP was<br />
inhibited), the fork could collapse or form a DSB. 13 Tumor<br />
cells deficient in BRCA1/2 are unable to repair these DSBs<br />
as a result <strong>of</strong> defects in homologous recombination. The<br />
concept <strong>of</strong> two genetic mutations, that individually have<br />
little effect but when combined are lethal, is known as<br />
synthetic lethality 14 and such an interaction exists between<br />
BRCA1/2 and PARP. PARP inhibitors lead to selective death<br />
<strong>of</strong> tumor cells that have hereditary or acquired BRCA1/2<br />
(homozygous) mutations but have no selective effect on the<br />
normal cells <strong>of</strong> BRCA carriers (heterozygous). Homologous<br />
recombination defects are thought to occur in up to 50% <strong>of</strong><br />
high-grade serous carcinomas, 15 not only through BRCA1/2<br />
mutations but interestingly, via a “BRCAness” phenotype.<br />
Recent microarray studies have identified a BRCAness gene<br />
expression pr<strong>of</strong>ile in patients with sporadic ovarian cancer<br />
that corresponds with responsiveness to platinum-based<br />
chemotherapy and to PARP inhibitors. 16<br />
<strong>Clinical</strong> Trials<br />
No CA-125 or scans will prolong bevacizumab and delay next<br />
line <strong>of</strong> therapy<br />
Which first-line patients? All stage 4 and stage 3 with suboptimal surgery Try to select those likely to have short remission<br />
Which line <strong>of</strong> therapy? First line Relapse<br />
Abbreviations: CA, cancer antigen; CT, computed tomography; GOG, Gynecologic <strong>Oncology</strong> Group; PFS, progression-free survival.<br />
342<br />
ROUX, ZWEIFEL, AND RUSTIN<br />
A phase I trial investigated olaparib (AZD2281) in patients<br />
with solid tumors refractory to conventional chemotherapy<br />
and enriched for patients with BRCA1/2 mutations.<br />
Partial responses were noted in 63% <strong>of</strong> patients (12 <strong>of</strong> 19)<br />
including eight patients with ovarian cancer. 17 Durable<br />
responses (median, 28 weeks; range, 10 to 86) correlating<br />
with an increased platinum-free interval were reported in<br />
an expanded cohort <strong>of</strong> 50 patients from this trial. The overall<br />
response rate (RR) was 69% in platinum-sensitive disease,<br />
50% in platinum-resistant disease, and 27% in platinumrefractory<br />
disease. Dose-limiting toxicity was seen at 400 mg<br />
and 600 mg twice daily, so the dose-expansion cohort received<br />
200 mg twice daily with minimal toxicity—mainly<br />
fatigue and GI symptoms.<br />
These encouraging results led to several phase II trials; a<br />
single-arm, open-label sequential dosing trial in patients<br />
with BRCA1/2 mutations reported clinical efficacy in both<br />
dosing cohorts, but higher response rates were seen in those<br />
receiving olaparib 400 mg twice daily compared with 100 mg<br />
twice daily (RR, 33% and 12.5%, respectively). 18 The toxicities<br />
were fatigue, sickness, and anemia but mostly grades 1<br />
and 2. A randomized phase II trial comparing the efficacy <strong>of</strong><br />
olaparib at two dose levels with pegylated liposomal doxo-
ANTIANGIOGENICS AND PARP INHIBITORS IN OVARIAN CANCER<br />
rubicin (PLD) in patients with BRCA1/2-mutated ovarian<br />
cancer did not show a significant PFS advantage for either<br />
olaparib dosing schedule (200 mg twice daily, 6.5 months;<br />
400 mg twice daily, 8.8 months) compared with PLD (7.1<br />
months; HR � 0.88; p � 0.66). 19 Nonetheless, with comparable<br />
efficacy, replacing PLD with a tablet is an attractive<br />
option.<br />
The concept <strong>of</strong> BRCAness as described earlier herein led<br />
to the investigation <strong>of</strong> whether PARP inhibitors were clinically<br />
active in patients with sporadic serous ovarian cancers.<br />
A phase II trial <strong>of</strong> olaparib 400 mg twice daily in patients<br />
with (n � 17) and without (n � 47) BRCA1/2 mutations<br />
reported a response rate <strong>of</strong> 41% and 24%, respectively. The<br />
better responses were again seen mostly in patients with<br />
platinum-sensitive disease. 20<br />
A phase II randomized placebo-controlled trial <strong>of</strong> olaparib<br />
monotherapy as maintenance treatment for patients with<br />
relapsed platinum-sensitive serous ovarian cancer was reported<br />
at ASCO in 2011. 21 PFS by Response Evaluation<br />
Criteria in Solid Tumors (RECIST) criteria was significantly<br />
longer in the olaparib group (n � 136) compared with the<br />
placebo group (n � 129;8.4 vs. 4.8 months; HR � 0.35; p �<br />
0.00001; Fig. 1). The PFS did not translate into an OS<br />
benefit, which has led AstraZeneca (Wilmington, DE) to stop<br />
pursuing this indication.<br />
Besides those evaluating olaparib, clinical trials are underway<br />
with several other PARP inhibitors as monotherapy<br />
in ovarian cancer and include MK4827 (Merck; Whitehouse<br />
Station, NY), CEP-9722 (Cephalon, Frazer, PA), ABT-888<br />
(Abbott Laboratories, Abbott Park, IL) and rucaparib (previously<br />
PF-01367338 or AG014699). Preliminary phase I<br />
results for MK4827 have shown efficacy comparable with<br />
olaparib 22 and certainly warrant further investigation in<br />
this setting. BMN-673 (BioMarin Pharmaceutical Inc., Novato,<br />
CA) is the most selective and potent PARP inhibitor<br />
reported to date, and phase I data are awaited with interest.<br />
Authors’ Disclosures <strong>of</strong> Potential Conflicts <strong>of</strong> Interest<br />
Employment or<br />
Leadership<br />
Positions<br />
Consultant or<br />
Advisory Role<br />
Author<br />
Rene Roux*<br />
Martin Zweifel*<br />
Gordon J. S. Rustin AstraZeneca;<br />
Roche<br />
*No relevant relationships to disclose.<br />
1. Campos SM, Ghosh S. A current review <strong>of</strong> targeted therapeutics for<br />
ovarian cancer. J Clin Oncol. 2010;28:149362.<br />
2. Mukherjee L, Rustin G. Antivascular therapy in gynaecological cancers.<br />
In Kehoe S, Edmondson RJ, Gore M, et al (eds): Gynaecological Cancers:<br />
Biology and Therapeutics. London: RCOG Press, 2011;121-137.<br />
3. Burger RA, Sill MW, Monk BJ, et al. Phase II trial <strong>of</strong> bevacizumab in<br />
persistent or recurrent epithelial ovarian cancer or primary peritoneal<br />
cancer: a Gynecologic <strong>Oncology</strong> Group Study. J Clin Oncol. 2007;25:5165-<br />
5171.<br />
4. Cannistra SA, Matulonis UA, Penson RT, et al. Phase II study <strong>of</strong><br />
bevacizumab in patients with platinum-resistant ovarian cancer or peritoneal<br />
serous cancer. J Clin Oncol. 2007;25:5180-5186.<br />
5. Perren TJ, Swart AM, Pfisterer J, et al. A phase 3 trial <strong>of</strong> bevacizumab<br />
in ovarian cancer. N Engl J Med. 2011;365:2484-2496.<br />
6. Burger RA, Brady MF, Bookman MA, et al. Incorporation <strong>of</strong> bevacizumab<br />
in the primary treatment <strong>of</strong> ovarian cancer. N Engl J Med. 2011;365:<br />
2473-2483.<br />
Conclusion<br />
Bevacizumab 7.5 mg/kg 3-weekly, should be considered<br />
for patients who are at high risk <strong>of</strong> having a short PFS<br />
because the bevacizumab might delay time to symptomatic<br />
relapse, and it is this group <strong>of</strong> patients who seem to achieve<br />
a substantial survival gain. For patients with optimally<br />
debulked disease, the greatest benefit from bevacizumab is<br />
more likely to be part <strong>of</strong> relapse therapy. The greatest<br />
benefit from PARP inhibitors will be in patients with BRCA<br />
mutations, but they are also likely to have a role in patients<br />
with high-grade serous ovarian cancers. Myelosuppression<br />
is likely to prevent PARP inhibitor administration in combination<br />
with chemotherapy. Because they have shown<br />
single-agent activity and can delay progression, they have<br />
the potential to be used as another line <strong>of</strong> relapse therapy or<br />
as maintenance therapy after chemotherapy for relapse.<br />
Stock<br />
Ownership Honoraria<br />
REFERENCES<br />
Fig 1. Progression-free survival <strong>of</strong> patients in a phase II randomized<br />
placebo-controlled trial <strong>of</strong> olaparib monotherapy as maintenance<br />
treatment for patients with relapsed platinum-sensitive serous<br />
ovarian cancer. Reprinted with permission. © <strong>American</strong> <strong>Society</strong> <strong>of</strong><br />
<strong>Clinical</strong> <strong>Oncology</strong>. All rights reserved. 21<br />
Boehringer<br />
Ingelheim<br />
Research<br />
Funding<br />
Expert<br />
Testimony<br />
Other<br />
Remuneration<br />
7. Katsumata N, Yasuda M, Takahashi F, et al. Dose-dense paclitaxel once<br />
a week in combination with carboplatin every 3 weeks for advanced ovarian<br />
cancer: a phase 3, open-label, randomised controlled trial. Lancet. 2009;374:<br />
1331-1338.<br />
8. Gonzalez-Martin A, Gladieff L, Stroyakovskiy D, et al. Front-line bevacizumab<br />
(BEV) combined with weekly paclitaxel (wPAC) and carboplatin (C)<br />
for ovarian cancer (OC): safety results from the concurrent chemotherapy<br />
(CT) phase <strong>of</strong> the OCTAVIa study. Presented at: European Multidisciplinary<br />
Cancer Congress; 2011; Stockholm, Sweden (abstr 8002).<br />
9. Aghajanian C, Finkler NJ, Rutherford T, et al. OCEANS: A randomized,<br />
double-blinded, placebo-controlled phase III trial <strong>of</strong> chemotherapy with or<br />
without bevacizumab (BEV) in patients with platinum-sensitive recurrent<br />
epithelial ovarian (EOC), primary peritoneal (PPC), or fallopian tube cancer<br />
(FTC). J Clin Oncol. 2011;29:45S (suppl; abstr LBA5007).<br />
10. Coleman RL. Making <strong>of</strong> a phase III study in recurrent ovarian cancer:<br />
the odyssey <strong>of</strong> GOG 213. Clin Ovarian Cancer. 2008;1:78-80.<br />
11. Chambon P, Weill JD, Mandel P. Nicotinamide mononucleotide activa-<br />
343
tion <strong>of</strong> new DNA-dependent polyadenylic acid synthesizing nuclear enzyme.<br />
Biochem Biophys Res Commun. 1963;11:39-43.<br />
12. Noel G, Giocanti N, Fernet M, et al. Poly(ADP-ribose) polymerase<br />
(PARP-1) is not involved in DNA double-strand break recovery. BMC Cell<br />
Biol. 2003;4:7.<br />
13. Haber JE. DNA recombination: the replication connection. Trends<br />
Biochem Sci. 1999;24:271-275.<br />
14. Ashworth A. A synthetic lethal therapeutic approach: poly(ADP) ribose<br />
polymerase inhibitors for the treatment <strong>of</strong> cancers deficient in DNA doublestrand<br />
break repair. J Clin Oncol. 2008;26:3785-3790.<br />
15. Press JZ, De Luca A, Boyd N, et al. Ovarian carcinomas with genetic<br />
and epigenetic BRCA1 loss have distinct molecular abnormalities. BMC<br />
Cancer. 2008;8:17.<br />
16. Konstantinopoulos PA, Spentzos D, Karlan BY, et al. Gene expression<br />
pr<strong>of</strong>ile <strong>of</strong> BRCAness that correlates with responsiveness to chemotherapy and<br />
with outcome in patients with epithelial ovarian cancer. J Clin Oncol.<br />
2010;28:3555-3561.<br />
17. Fong PC, Boss DS, Yap TA, et al. Inhibition <strong>of</strong> poly(ADP-ribose)<br />
polymerase in tumors from BRCA mutation carriers. N Engl J Med. 2009;<br />
361:123-134.<br />
344<br />
ROUX, ZWEIFEL, AND RUSTIN<br />
18. Fong PC, Yap TA, Boss DS, et al. Poly(ADP)-ribose polymerase inhibition:<br />
frequent durable responses in BRCA carrier ovarian cancer correlating<br />
with platinum-free interval. J Clin Oncol. 2010;28:2512-2519.<br />
19. Kaye SB, Lubinski J, Matulonis U, et al. Phase II, open-label, randomized,<br />
multicenter study comparing the efficacy and safety <strong>of</strong> olaparib, a poly<br />
(ADP-ribose) polymerase inhibitor, and pegylated liposomal doxorubicin in<br />
patients with BRCA1 or BRCA2 mutations and recurrent ovarian cancer.<br />
J Clin Oncol. 2011;30:372-379.<br />
20. Gelmon KA, Tischkowitz M, Mackay H, et al. Olaparib in patients with<br />
recurrent high-grade serous or poorly differentiated ovarian carcinoma or<br />
triple-negative breast cancer: a phase 2, multicentre, open-label, nonrandomised<br />
study. Lancet Oncol. 2011;12:852-861.<br />
21. Ledermann JA, Harter P, Gourley C, et al. Phase II randomized<br />
placebo-controlled study <strong>of</strong> olaparib (AZD2281) in patients with platinumsensitive<br />
relapsed serous ovarian cancer (PSR SOC). J Clin Oncol. 2011;29:<br />
15s (suppl; abstr 5003).<br />
22. Sandhu SK, Wenham, RM, Wilding G, et al. First-in-human trial <strong>of</strong> a<br />
poly(ADP-ribose) polymerase (PARP) inhibitor MK-4827 in advanced cancer<br />
patients (pts) with antitumor activity in BRCA-deficient and sporadic ovarian<br />
cancers. J Clin Oncol 2010;28:15s (suppl; abstr 3001).
Intraperitoneal Treatment in Ovarian Cancer:<br />
The Gynecologic <strong>Oncology</strong> Group Perspective<br />
in <strong>2012</strong><br />
By Deborah K. Armstrong, MD, Keiichi Fujiwara, MD, PhD, and Danijela Jelovac, MD<br />
Overview: The peritoneal cavity is the major site <strong>of</strong> disease in<br />
ovarian cancer. The peritoneal predominance <strong>of</strong> disease provides<br />
a rationale for administration <strong>of</strong> chemotherapy within<br />
the peritoneal cavity. Intraperitoneal (IP) chemotherapy for<br />
ovarian cancer has been studied rigorously for more than 30<br />
years and has been reproducibly shown to improve the survival<br />
<strong>of</strong> patients with ovarian cancer. Three large randomized<br />
trials <strong>of</strong> IP compared with intravenous (IV) therapy have<br />
THE PERITONEAL cavity is the major site <strong>of</strong> disease in<br />
ovarian cancer. 1 Whereas ovarian cancer can spread<br />
hematogenously or via the lymphatic system, the bulk <strong>of</strong> the<br />
tumor will be found on peritoneal surfaces. This peritoneal<br />
disease results from shedding <strong>of</strong> ovarian tumor cells into the<br />
peritoneal cavity, circulation <strong>of</strong> these cells throughout the<br />
abdomen and pelvis, and eventual implantation onto peritoneal<br />
surfaces. Viability <strong>of</strong> these cells and successful tumor<br />
growth is further dependent on the development <strong>of</strong> sufficient<br />
neovasculature to support cell survival and tumor growth.<br />
This unique pattern <strong>of</strong> spread within the relatively accessible<br />
peritoneal cavity has led to attempts at surgical cytoreduction<br />
before administration <strong>of</strong> chemotherapy. Dating<br />
back more than 30 years, nearly every study has demonstrated<br />
an inverse correlation between volume <strong>of</strong> tumor<br />
remaining at the completion <strong>of</strong> initial surgery and overall<br />
survival for patients with ovarian cancer. 2 The peritoneal<br />
predominance <strong>of</strong> ovarian cancer also provides a rationale<br />
for administration <strong>of</strong> chemotherapy within the peritoneal<br />
cavity.<br />
When drugs are administered intravenously they are<br />
immediately diluted in the blood. When the same drugs are<br />
administered via the intraperitoneal (IP) route, the peritoneum<br />
can have sustained exposure to higher concentrations<br />
<strong>of</strong> drugs for a more prolonged period <strong>of</strong> time, whereas normal<br />
tissues such as the bone marrow may be relatively spared.<br />
The pharmacologic advantage <strong>of</strong> administering a drug by<br />
the peritoneal route can be quantified by the ratio <strong>of</strong> the<br />
drug concentration (usually measured in area under the<br />
curve [AUC]) in the peritoneal cavity to that in the plasma<br />
after IP compared with IV injection. Table 1 shows this<br />
pharmacologic advantage for some drugs commonly used in<br />
ovarian cancer. 3-5<br />
The rate at which the peritoneal drug concentration decreases<br />
is a function <strong>of</strong> the volume <strong>of</strong> the fluid in the<br />
peritoneal cavity, the surface area through which the drug<br />
diffuses out <strong>of</strong> the cavity, the permeability <strong>of</strong> this surface,<br />
and the difference in free drug concentration between the<br />
cavity and the plasma. The clinical effectiveness <strong>of</strong> a drug<br />
administered IP is affected not only by this pharmacologic<br />
advantage but also by tumor penetration and distribution <strong>of</strong><br />
the drug within the peritoneal cavity. It is thus predicted<br />
that tumor volume and the presence <strong>of</strong> substantial adhesive<br />
disease will influence the efficacy <strong>of</strong> IP therapy. 5<br />
IP chemotherapy was first used in the 1950s for palliation<br />
<strong>of</strong> ascites, primarily from colorectal carcinoma. In the 1970s<br />
IP treatment became more feasible with the development <strong>of</strong><br />
demonstrated statistically significant improvement in clinical<br />
outcome measures. Despite this, the IP approach has not<br />
gained widespread acceptance in the treatment <strong>of</strong> ovarian<br />
cancer. Here, we review reported, recently completed, and<br />
ongoing trials <strong>of</strong> IP therapy in ovarian cancer including attempts<br />
to improve the tolerability and acceptance <strong>of</strong> this<br />
proven approach.<br />
permanent indwelling peritoneal catheters that allowed for<br />
repetitive IP administration without requiring repeated<br />
placement <strong>of</strong> temporary peritoneal catheters. Shortly after<br />
this breakthrough, IP therapy began to be studied in ovarian<br />
cancer. Following is a review <strong>of</strong> data from completed trials <strong>of</strong><br />
IP therapy in ovarian cancer, and a summary <strong>of</strong> ongoing and<br />
recently completed trials and <strong>of</strong> novel and innovative approaches<br />
to IP therapy, with a particular focus on randomized<br />
phase III trials and studies from the Gynecologic<br />
<strong>Oncology</strong> Group (GOG).<br />
Randomized Trials <strong>of</strong> IP Versus IV Therapy in<br />
Ovarian Cancer<br />
Eight published comparative studies <strong>of</strong> IP compared with<br />
IV administration for initial therapy <strong>of</strong> ovarian cancer were<br />
the subject <strong>of</strong> a Cochrane meta-analysis in 2007. 6 This<br />
analysis <strong>of</strong> 1,819 women showed that they were less likely to<br />
die if they received an IP component to the chemotherapy<br />
(hazard ratio [HR] � 0.79) and that the disease-free interval<br />
was also significantly prolonged (HR � 0.79). They did also<br />
note that there may be greater serious toxicity with IP<br />
therapy and that there is a potential for catheter-related<br />
complications in patients receiving IP therapy.<br />
In January 2006, the U.S. National Cancer Institute (NCI)<br />
released a clinical announcement <strong>of</strong> IP therapy for ovarian<br />
cancer. 7 They evaluated data from eight trials and concluded<br />
that IP chemotherapy is beneficial for optimally<br />
debulked stage III ovarian cancer. On average, for the eight<br />
trials, IP therapy was associated with a 21.6% decrease in<br />
the risk <strong>of</strong> death (HR � 0.79) which is estimated to translate<br />
into a 12-month increase in overall median survival. The<br />
three largest trials, responsible for more than 75% <strong>of</strong> the<br />
patients in this analysis, were conducted as U.S. cooperative<br />
group trials and will be further discussed in this section.<br />
The first <strong>of</strong> these trials, led by Southwest <strong>Oncology</strong> Group<br />
(SWOG #8501) with the GOG (GOG #104), used IV cyclophosphamide<br />
(600 mg/m 2 ) with either IP or IV cisplatin (100<br />
mg/m 2 ) administered every 3 weeks for six cycles. 8 This<br />
study showed a survival advantage for the group receiving<br />
From the Johns Hopkins Sidney Kimmel Cancer Center, Baltimore, MD.<br />
Authors’ disclosures <strong>of</strong> potential conflicts <strong>of</strong> interest are found at the end <strong>of</strong> this article.<br />
Address reprint requests to Deborah K. Armstrong, MD, Associate Pr<strong>of</strong>essor <strong>of</strong> <strong>Oncology</strong>,<br />
Associate Pr<strong>of</strong>essor <strong>of</strong> Gynecology & Obstetrics, Johns Hopkins Sidney Kimmel Cancer<br />
Center, 1650 Orleans Street, Room 190, Baltimore, MD 21231; email: darmstro@jhmi.edu.<br />
© <strong>2012</strong> by <strong>American</strong> <strong>Society</strong> <strong>of</strong> <strong>Clinical</strong> <strong>Oncology</strong>.<br />
1092-9118/10/1-10<br />
345
Table 1. Pharmacologic Advantage for Intraperitoneal<br />
Chemotherapy<br />
Drug Molecular Weight<br />
IP cisplatin. In the 546 eligible patients, the estimated<br />
median survival was substantially longer in the IP group (49<br />
months) compared with the IV group (41 months). The HR<br />
for the risk <strong>of</strong> death was 0.76 (p � 0.02) in favor <strong>of</strong> IP<br />
therapy. Although moderate to severe abdominal pain was<br />
more frequent in the IP group, likely related to the catheter<br />
and volume <strong>of</strong> infusate, grade 3/4 granulocytopenia and<br />
tinnitus, clinical hearing loss, and grade 2 to 4 neuromuscular<br />
toxic effects were markedly more frequent in the IV<br />
group. This is the “purest” <strong>of</strong> the IP trials to date as it used<br />
the same drugs and same doses in both arms with only the<br />
route <strong>of</strong> administration being different. Despite the benefits<br />
observed in this trial, IP therapy was not widely embraced<br />
by the gynecologic oncology community at the time <strong>of</strong> this<br />
publication. This may be because <strong>of</strong> the publication in the<br />
same year <strong>of</strong> the benefits <strong>of</strong> paclitaxel in patients with<br />
suboptimally debulked disease. 9<br />
A second IP trial was also conducted by GOG (GOG #114)<br />
and Southwest <strong>Oncology</strong> Group (SWOG #9227). In this trial<br />
426 patients were randomly assigned to receive either IV<br />
paclitaxel 135 mg/m 2 over 24 hours followed by IV cisplatin<br />
75 mg/m 2 every 3 weeks for six cycles or IV carboplatin<br />
(AUC 9) every 28 days for two cycles followed by a regimen<br />
<strong>of</strong> IV paclitaxel 135 mg/m 2 over 24 hours followed by IP<br />
cisplatin at 100 mg/m 2 every 3 weeks for six cycles (eight<br />
total cycles <strong>of</strong> chemotherapy). 10 This study demonstrated an<br />
improved progression-free survival (median, 28 vs. 22<br />
months; HR � 0.78; p � 0.01) and overall survival (median,<br />
63 vs. 52 months; HR � 0.81; p � 0.05) in favor <strong>of</strong> the IP<br />
KEY POINTS<br />
Ratio <strong>of</strong> Drug AUC,<br />
Peritoneal Cavity/Plasma<br />
Cisplatin 300 12<br />
Carboplatin 371 10–18<br />
Topotecan 458 54<br />
Docetaxel 862 181<br />
Gemcitabine 300 759<br />
Paclitaxel 854 1,000<br />
Abbreviation: AUC, area under the curve.<br />
Adapted and modified from Markman 3 and Fujiwara et al. 4<br />
● Current recommendations for intraperitoneal (IP)<br />
therapy are for patients with disease optimally debulked<br />
after initial surgery.<br />
● Three North <strong>American</strong> randomized phase III trials<br />
comparing IP with intravenous therapy have demonstrated<br />
substantial improvement in progression-free<br />
and overall survival.<br />
● Cisplatin is the mainstay <strong>of</strong> IP platinum treatment<br />
but a recently completed trial (GOG 252) is the first to<br />
use IP carboplatin in an arm <strong>of</strong> a randomized phase<br />
III trial.<br />
● IP therapy after neoadjuvant therapy and interval<br />
debulking is currently being tested.<br />
● IP therapy has not been rigorously tested for patients<br />
with recurrent disease.<br />
346<br />
ARMSTRONG, FUJIWARA, AND JELOVAC<br />
group. Toxicities grade 3 or worse, including neutropenia,<br />
thrombocytopenia, and GI and metabolic toxicities, were<br />
markedly more frequent in the IP group. As a result, 18% <strong>of</strong><br />
the patients on the IP arm received fewer than two courses<br />
<strong>of</strong> IP therapy. Despite the substantial survival improvement<br />
in this study, the gynecologic oncology community again did<br />
not accept IP chemotherapy as standard treatment for ovarian<br />
cancer. Many attributed the benefits seen in the experimental<br />
arm to patients receiving eight cycles <strong>of</strong> therapy or to the<br />
intensive carboplatin rather than the IP therapy.<br />
The third trial was conducted by GOG (GOG #172). In this<br />
study 417 eligible patients with optimally debulked stage III<br />
ovarian cancer were randomly assigned to receive IV paclitaxel<br />
(135 mg/m 2 /24 hours) followed by IV cisplatin (75<br />
mg/m 2 ) or IV paclitaxel (135 mg/m 2 /24 hours) followed by IP<br />
cisplatin (100 mg/m 2 ), plus IP paclitaxel (60 mg/m 2 ) on day<br />
8. 11 Treatments were repeated every 21 days for six cycles.<br />
The median progression-free survival was 18.3 months in<br />
the IV group and 23.8 months in the IP group (HR � 0.77,<br />
p � 0.05). The median overall survival was 49.7 in the IV<br />
group and 65.6 months IP group (HR � 0.73, p � 0.03). The<br />
magnitude <strong>of</strong> improvement in median overall survival associated<br />
with IP/IV administration <strong>of</strong> chemotherapy is similar<br />
to that observed with the introduction <strong>of</strong> either cisplatin or<br />
paclitaxel. The 66-month median survival for the IP arm <strong>of</strong><br />
GOG 172 is the longest survival reported to date from a<br />
randomized trial <strong>of</strong> advanced ovarian cancer.<br />
Once again, the data in support <strong>of</strong> IP therapy have not<br />
resulted in widespread acceptance <strong>of</strong> the IP approach. Many<br />
argued that it was the additional drug delivered in the IP<br />
arm <strong>of</strong> GOG 172 that resulted in the improved outcome. This<br />
is somewhat circuitous logic in that those doses <strong>of</strong> drug<br />
cannot be delivered intravenously. The IP arm was designed<br />
to be intentionally intense, exploiting the ability to administer<br />
more drug per unit <strong>of</strong> time IP than can be delivered<br />
with IV therapy.<br />
There were substantially more patients with grade 3 and<br />
4 leukopenia, thrombocytopenia, and GI toxicity, renal toxicity,<br />
neurologic toxicity, fatigue, infection, metabolic toxicity,<br />
and pain toxicity in the IP arm compared with the IV<br />
arm. Because <strong>of</strong> these toxicities and/or catheter problems,<br />
48% <strong>of</strong> patients in the IP arm received three or fewer IP<br />
treatments, and only 42% patients received the planned six<br />
cycles <strong>of</strong> IP therapy. Given that the study results are based<br />
on an intent-to-treat analysis, modifications <strong>of</strong> the regimen<br />
to improve tolerability could allow for improved completion<br />
rates and possibly even better outcomes.<br />
In a separate quality-<strong>of</strong>-life (QOL) analysis, patients who<br />
received IP therapy had a worse QOL during therapy, but<br />
there was no difference 1 year after completion <strong>of</strong> treatment.<br />
12 We and others have shown that the GOG 172 IP<br />
regimen can be administered successfully with a reasonable<br />
toxicity pr<strong>of</strong>ile when patients are appropriately selected and<br />
the therapy is performed at an experienced center with the<br />
assistance <strong>of</strong> a skilled and dedicated support staff. 13<br />
Some have suggested that GOG 172 may overestimate the<br />
benefit <strong>of</strong> IP therapy. Ozols and colleagues have reported on<br />
a preliminary cross-trial analysis comparing the results <strong>of</strong><br />
IP therapy in GOG 172 with IV carboplatin/paclitaxel in<br />
392 similarly staged patients in GOG #158, 14 calculating<br />
that instead <strong>of</strong> the 15.9-month improvement in median<br />
overall survival, the difference may be substantially less (8.2<br />
months) if carboplatin/paclitaxel had been the comparative
INTRAPERITONEAL TREATMENT IN OVARIAN CANCER<br />
Fig. 1. IP compared to IV chemotherapy Phase III<br />
trials. The bars show the % increase in progression-free<br />
survival (PFS, blue bars) and overall survival (OS, orange<br />
bars) for IP compared with IV therapy. 8,10,11 The red bar<br />
shows the percent increase in OS for the IV arm <strong>of</strong> GOG<br />
172 10 with IV paclitaxel and carboplatin (TC) from GOG<br />
protocol 158. 13<br />
arm. Historical, nonrandomized cross-trial comparisons<br />
such as these lack the validity <strong>of</strong> those generated by prospective<br />
randomization and should not be relied upon for<br />
generating credible conclusions. However, in response to the<br />
assertions <strong>of</strong> Ozols et al., we performed a more rigorous<br />
statistical analysis using this cross-trial comparison showed<br />
that there remained a 19% improvement in overall survival<br />
when the IP arm <strong>of</strong> GOG 172 was compared with the IV<br />
paclitaxel/carboplatin arm <strong>of</strong> GOG 158. 15 Figure 1 shows the<br />
percentage improvement in progression-free and in overall<br />
survival from these three trials and includes the overall<br />
survival improvement comparing the IP arm <strong>of</strong> GOG 172<br />
with the IV paclitaxel/carboplatin arm <strong>of</strong> GOG 158.<br />
Recent Trials and New Approaches to IP Therapy<br />
GOG protocol 252 recently closed to accrual. This trial<br />
included an arm <strong>of</strong> IP carboplatin with weekly IV paclitaxel,<br />
an arm with an IP cisplatin regimen that is a modification <strong>of</strong><br />
Authors’ Disclosures <strong>of</strong> Potential Conflicts <strong>of</strong> Interest<br />
Author<br />
Employment or<br />
Leadership<br />
Positions<br />
Consultant or<br />
Advisory Role<br />
Deborah K. Armstrong Eisai (I);<br />
Genentech;<br />
Genzyme;<br />
Oncogenex<br />
Keiichi Fujiwara Amgen;<br />
Boehringer<br />
Ingelheim;<br />
GlaxoSmithKline;<br />
Zeria Pharma<br />
Danijela Jelovac*<br />
*No relevant relationships to disclose.<br />
1. Jelovac D, Armstrong DK. Recent progress in the diagnosis and treatment<br />
<strong>of</strong> ovarian cancer. CA Cancer J Clin. 2011;61:183-203.<br />
2. Bristow RE, Tomacruz RS, Armstrong DK, et al. Survival impact <strong>of</strong><br />
maximum cytoreductive surgery for advanced ovarian carcinoma during the<br />
platinum-era: a meta-analysis <strong>of</strong> 6,885 patients. J Clin Oncol. 2002;20:1248-<br />
1259.<br />
3. Markman M. Intraperitoneal chemotherapy. Semin Oncol. 1991;18:248-<br />
54.<br />
4. Fujiwara K, Armstrong D, Morgan M, et. al. Principles and practice <strong>of</strong><br />
25<br />
20<br />
15<br />
10<br />
5<br />
0<br />
the IP arm <strong>of</strong> GOG 172, and an arm with IV weekly<br />
paclitaxel and IV carboplatin. This is the first randomized<br />
phase III trial to rigorously evaluate IP carboplatin. All<br />
patients in this trial received bevacizumab. In addition, the<br />
protocol enrolled a limited number <strong>of</strong> patients with suboptimal<br />
residual disease, which will allow for a preliminary<br />
evaluation <strong>of</strong> the role <strong>of</strong> IP therapy in that setting.<br />
Currently, the National Cancer Institute <strong>of</strong> Canada <strong>Clinical</strong><br />
Trials Group, in conjunction with the Gynecologic Cancer<br />
Intergroup, in protocol OV.21 is examining the use <strong>of</strong> IP<br />
platinum-taxane-based chemotherapy after neoadjuvant<br />
chemotherapy and subsequent interval surgical debulking.<br />
To address whether the less systemically toxic carboplatin<br />
can be substituted for cisplatin IP, the first phase <strong>of</strong> the<br />
study will have three arms: one IV only, and two IP-containing<br />
regimens, one with carboplatin and one with cisplatin. At<br />
completion <strong>of</strong> the first stage, one <strong>of</strong> the IP regimens will<br />
proceed into a phase III comparison with the IV arm.<br />
Stock<br />
Ownership Honoraria<br />
REFERENCES<br />
PFS: % Increase OS: % Increase<br />
GOG 104 GOG 114 GOG 172<br />
Research<br />
Funding<br />
Eisai (I);<br />
Exelixis (I);<br />
Morphotek<br />
Bristol-Myers<br />
Squibb;<br />
Chugai Pharma;<br />
Janssen<br />
<strong>Oncology</strong>;<br />
Nihonkayaku;<br />
San<strong>of</strong>i; Taiho<br />
Pharmaceutical<br />
Expert<br />
Testimony<br />
G158 TC c/w<br />
G 172 IP OS<br />
Other<br />
Remuneration<br />
intraperitoneal chemotherapy for ovarian cancer. Int J Gynecol Cancer.<br />
2007;17:1-20.<br />
5. Howell SB. Pharmacologic principles <strong>of</strong> intraperitoneal chemotherapy<br />
for the treatment <strong>of</strong> ovarian cancer. Int J Gynecol Cancer. 2008;18:20-25<br />
(suppl 1).<br />
6. Jaaback K, Johnson N. Intraperitoneal chemotherapy for the initial<br />
management <strong>of</strong> primary epithelial ovarian cancer. The Cochrane Database <strong>of</strong><br />
Systematic Reviews. 2006;1:CD005340.pub2.<br />
7. NCI clinical announcement. http://ctep.cancer.gov/highlights/docs/clin_<br />
347
annc_010506.pdf#search�“intraperitonealchemotherapy”. Accessed March 23,<br />
<strong>2012</strong>.<br />
8. Alberts DS, Liu PY, Hannigan EV, et al. Intraperitoneal cisplatin plus<br />
intravenous cyclophosphamide versus intravenous cisplatin plus intravenous<br />
cyclophosphamide for stage III ovarian cancer. N Engl J Med. 1996;335:1950-<br />
1955.<br />
9. McGuire WP, Hoskins WJ, Brady MF, et al. Cyclophosphamide and<br />
cisplatin compared with paclitaxel and cisplatin in patients with stage III and<br />
stage IV ovarian cancer. N Engl J Med. 1996;334:1-6.<br />
10. Markman M, Bundy BN, Alberts DS, et al. Phase III trial <strong>of</strong> standarddose<br />
intravenous cisplatin plus paclitaxel versus moderately high-dose carboplatin<br />
followed by intravenous paclitaxel and intraperitoneal cisplatin in<br />
small-volume stage III ovarian carcinoma: an intergroup study <strong>of</strong> the Gynecologic<br />
<strong>Oncology</strong> Group, Southwestern <strong>Oncology</strong> Group, and Eastern Cooperative<br />
<strong>Oncology</strong> Group. J Clin Oncol. 2001;19:1001-1007.<br />
348<br />
ARMSTRONG, FUJIWARA, AND JELOVAC<br />
11. Armstrong DK, Bundy B, Wenzel L, et. al. Intraperitoneal cisplatin and<br />
paclitaxel in ovarian cancer. N Engl J Med. 2006;354:34-43.<br />
12. Wenzel L, Huang HQ, Armstrong DK, et. al. Health-related quality <strong>of</strong><br />
life during and after intraperitoneal versus intravenous chemotherapy for<br />
optimally debulked ovarian cancer: a Gynecologic <strong>Oncology</strong> Group study.<br />
J Clin Oncol. 2007;25:437-443.<br />
13. Guile MW, Horne AL, Thompson SD, et al. Intraperitoneal chemotherapy<br />
for stage III ovarian cancer using the Gynecologic <strong>Oncology</strong> Group<br />
protocol 172 intraperitoneal regimen: Effect <strong>of</strong> supportive care using aprepitant<br />
and pegfilgrastim on treatment completion rate. Clin Ovarian Cancer.<br />
2008;1:68-71.<br />
14. Ozols RF, Bookman MA, Young RC. Intraperitoneal chemotherapy for<br />
ovarian cancer. N Engl J Med. 2006;354:1641-1643.<br />
15. Armstrong DK, Brady MF. Intraperitoneal therapy for ovarian cancer:<br />
a treatment ready for prime time. J Clin Oncol. 2006;24:4531-4533.
Dose-Dense Chemotherapy and Neoadjuvant<br />
Chemotherapy for Ovarian Cancer<br />
By Keiichi Fujiwara, MD, PhD, Noriyuki Katsumata, MD, PhD, and<br />
Takashi Onda, MD, PhD<br />
Overview: Two <strong>of</strong> the innovative chemotherapeutic approaches<br />
to ovarian cancer treatment, dose-dense chemotherapy<br />
and neoadjuvant chemotherapy, will be discussed<br />
herein. The primary concept <strong>of</strong> dose-dense chemotherapy is to<br />
administer the same cumulative dose <strong>of</strong> chemotherapy over a<br />
shorter period. Increased dose density is achieved by reducing<br />
the interval between each dose <strong>of</strong> chemotherapy. The<br />
Japanese Gynecologic <strong>Oncology</strong> Group (JGOG) first demonstrated<br />
the survival advantage <strong>of</strong> dose-dense weekly administration<br />
<strong>of</strong> paclitaxel in 2009. However, there are unanswered<br />
questions, such as the question <strong>of</strong> dose-dense carboplatin<br />
versus less dose-intensive regimens. Clear cell or mucinous<br />
carcinomas seem to need other strategies, such as targeted<br />
agents. The aim <strong>of</strong> neoadjuvant chemotherapy is to reduce<br />
OVARIAN CANCER is the most lethal disease among<br />
gynecologic malignancies because this disease remains<br />
most commonly diagnosed in advanced stages. 1<br />
Treatment <strong>of</strong> ovarian cancer has been investigated for<br />
more than 30 years, since cisplatin was introduced in the<br />
1970s. The clinical outcome <strong>of</strong> ovarian cancer was improved<br />
because <strong>of</strong> the development <strong>of</strong> new anticancer agents and<br />
advancements in surgical technique, equipment, and anesthesia.<br />
Consequently, current standard therapy for advanced<br />
ovarian cancer became a combination <strong>of</strong> maximum<br />
surgical effort to remove bulky abdominal disease followed<br />
by chemotherapy with paclitaxel plus carboplatin. However,<br />
this has not changed since 1999, despite great efforts to find<br />
new therapeutic strategy.<br />
The first approach to improve survival was by finding<br />
other effective drugs or treatment modalities. The most<br />
thrilling area at this time is the development <strong>of</strong> new anticancer<br />
drugs, especially targeted agents. It is unfortunate,<br />
however, that most pharmaceutical companies have not paid<br />
great attention to gynecologic cancer. Therefore, we had to<br />
wait until 2010 before the first targeted agent, bevacizumab,<br />
showed positive results in ovarian cancer chemotherapy. 2,3<br />
In the meantime, investigators have conducted academic<br />
trials to find a way to improve the prognosis <strong>of</strong> patients with<br />
ovarian cancer. Those trials include intraperitoneal (IP)<br />
chemotherapy, maintenance chemotherapy, and dose-dense<br />
weekly chemotherapy.<br />
On the basis <strong>of</strong> the results <strong>of</strong> these large-scale randomized<br />
trials, the 4 th Ovarian Cancer Consensus Conference in<br />
2010 4 yielded the following statement: “It was agreed unanimously<br />
that the basic minimum comparator in a phase III<br />
trial <strong>of</strong> advanced ovarian carcinoma must contain a taxane<br />
and a platinum agent given for 6 cycles. Acceptable alternatives<br />
must be supported by at least 1 clinical trial demonstrating<br />
noninferiority or superiority to a standard taxane/<br />
platinum regimen. Acceptable alternatives at present<br />
include IP delivery to patients with small-volume residual<br />
disease, weekly paclitaxel in combination with carboplatin<br />
every 3 weeks, bevacizumab given concurrently with paclitaxel/carboplatin<br />
followed by bevacizumab maintenance,<br />
and 12 months <strong>of</strong> monthly paclitaxel maintenance given to<br />
tumor volume or spread before main treatment. This could<br />
then make the main procedures easier or less invasive, just<br />
like breast-conserving surgery after neoadjuvant chemotherapy.<br />
In advanced ovarian cancer, standard procedure is maximum<br />
primary debulking surgery followed by chemotherapy.<br />
Recently, a prospective randomized trial demonstrated that<br />
neoadjuvant chemotherapy followed by interval debulking<br />
surgery was not inferior to the standard procedure. However,<br />
there are several questions that remain unanswered, such as<br />
the suitable number <strong>of</strong> chemotherapy cycles before interval<br />
debulking surgery. Some <strong>of</strong> those questions regarding dosedense<br />
chemotherapy or neoadjuvant chemotherapy may be<br />
resolved by ongoing or future prospective trials.<br />
patients who achieve a clinical complete response with 6<br />
cycles <strong>of</strong> standard paclitaxel/carboplatin.” Dose-dense paclitaxel<br />
was the one regimen that dramatically improved the<br />
survival <strong>of</strong> patients with ovarian cancer.<br />
A second approach is to attempt to reduce the patient’s<br />
tumor burden, if the clinical outcome is the same regardless<br />
<strong>of</strong> the intensity or aggressiveness <strong>of</strong> the treatment. These<br />
less-invasive treatments will improve the patients’ quality <strong>of</strong><br />
life compared with more intensive therapies. Neoadjuvant<br />
chemotherapy is one <strong>of</strong> these approaches, although it is<br />
controversial. In the 4 th Ovarian Consensus Conference<br />
statement, 4 it was concluded that “Delayed primary surgery<br />
following neoadjuvant chemotherapy is an option for selected<br />
patients with stage IIIC and IV ovarian cancer as<br />
included in EORTC 55971, 5 ” although this was the only<br />
issue among the consensus conference in which the total<br />
consensus was not reached.<br />
Dose-Dense Chemotherapy<br />
The basic concept <strong>of</strong> dose-dense therapy is to administer<br />
the same cumulative dose <strong>of</strong> chemotherapy over a shorter<br />
period. Increased dose density is achieved by reducing the<br />
interval between each dose <strong>of</strong> chemotherapy. The theoretical<br />
basis for this dose-dense chemotherapy strategy is derived<br />
from the Gompertzian model, which is based on Norton-<br />
Simon’s hypothesis. 6 In the Gompertzian model, smaller<br />
tumors grow faster and so tumor regrowth between treatment<br />
cycles is more rapid when cell kill is greatest. The<br />
Norton-Simon model suggests that increasing the dose density<br />
<strong>of</strong> chemotherapy will increase efficacy by minimizing<br />
From the Department <strong>of</strong> Gynecologic <strong>Oncology</strong>, Saitama Medical School International<br />
Medical Center, Saitama, Japan; Department <strong>of</strong> Medical <strong>Oncology</strong>, Nippon Medical<br />
School, Musashikosugi Hospital, Kawasaki-City, Japan; Department <strong>of</strong> Gynecology, Kitasato<br />
University School <strong>of</strong> Medicine, Sagamihara-City, Kanagawa, Japan.<br />
Authors’ disclosures <strong>of</strong> potential conflicts <strong>of</strong> interest are found at the end <strong>of</strong> this article.<br />
Address reprint requests to Keiichi Fujiwara, MD, PhD, Department <strong>of</strong> Gynecologic<br />
<strong>Oncology</strong>, Saitama Medical School, International Medical Center,1397-1 Yamane, Hidaka-<br />
City, Saitama, 350-1298, Japan; email: fujiwara@saitama-med.ac.jp.<br />
© <strong>2012</strong> by <strong>American</strong> <strong>Society</strong> <strong>of</strong> <strong>Clinical</strong> <strong>Oncology</strong>.<br />
1092-9118/10/1-10<br />
349
the opportunity for regrowth <strong>of</strong> tumor cells between cycles <strong>of</strong><br />
chemotherapy.<br />
<strong>Clinical</strong>ly, the benefit <strong>of</strong> dose-dense chemotherapy was<br />
first proven in breast cancer. Two randomized phase III<br />
trials on breast cancer showed improved survival benefit by<br />
administering paclitaxel with weekly compared with every-<br />
3-weeks administration. 7,8<br />
In ovarian cancer, the Japanese Gynecologic <strong>Oncology</strong><br />
Group (JGOG) first demonstrated the survival advantage<br />
<strong>of</strong> dose-dense weekly administration <strong>of</strong> paclitaxel. 9 In<br />
JGOG3016, 637 patients were randomly assigned to receive<br />
six cycles <strong>of</strong> either paclitaxel (180 mg/m 2 ; 3-hour intravenous<br />
[IV] infusion) plus carboplatin (area under the curve<br />
[AUC] 6 mg/mL/min), administered on day 1 <strong>of</strong> a 21-day<br />
cycle (conventional regimen; n � 320), or dose-dense paclitaxel<br />
(80 mg/m 2 ; 1-hour IV infusion) administered on days 1,<br />
8, and 15 plus carboplatin administered on day 1 <strong>of</strong> a 21-day<br />
cycle (dose-dense regimen; n � 317). The primary end point<br />
was progression-free survival, and secondary end points<br />
were overall survival and toxicity. A total <strong>of</strong> 631 patients<br />
were eligible (dose-dense regimen, n � 312; conventional<br />
regimen, n � 319). Median progression-free survival was<br />
significantly longer in the dose-dense treatment group (28.0<br />
months; 95% CI, 22.3 to 35.4 months) than in the conventional<br />
treatment group (17.2 months; 95% CI, 15.7 to 21.1<br />
months; hazard ratio [HR] � 0.71; 95% CI, 0.58 to 0.88; p �<br />
0.0015). Overall survival at 3 years was higher in the<br />
dose-dense regimen group (72.1%) than in the conventional<br />
treatment group (65.1%; HR � 0.75: 95% CI, 0.57 to 0.98;<br />
p � 0.03). Early discontinuation <strong>of</strong> the treatment was more<br />
frequent in dose-dense group (n � 165) than in the conventional<br />
group (n � 117) primarily because <strong>of</strong> the toxicities<br />
(n � 113 vs. n � 69). The most common adverse event was<br />
neutropenia (dose-dense regimen, 286 [92%] <strong>of</strong> 312 patients;<br />
conventional regimen, 276 [88%] <strong>of</strong> 314 patients), but not<br />
statistically different. The frequency <strong>of</strong> grade 3 and 4 anemia<br />
was significantly higher in the dose-dense group (n �<br />
214 [69%]) than in the conventional treatment group (n �<br />
137 [44%]; p � 0.0001). The frequencies <strong>of</strong> other toxicities<br />
including peripheral neuropathy were similar between<br />
groups. On the basis <strong>of</strong> these results, it was concluded that<br />
dose-dense weekly paclitaxel plus carboplatin improved sur-<br />
KEY POINTS<br />
● One trial <strong>of</strong> dose-dense chemotherapy (Japanese Gynecologic<br />
<strong>Oncology</strong> Group [JGOG]-3016) has shown<br />
significant improvement in progression-free survival<br />
and overall survival in patients with advanced ovarian<br />
cancer. Three confirmatory studies are ongoing<br />
worldwide.<br />
● One trial <strong>of</strong> neoadjuvant chemotherapy (European<br />
Organisation for Research and Treatment <strong>of</strong> Cancer<br />
[EORTC]) has shown that there was no substantial<br />
difference in OS in patients with stage III/IV ovarian<br />
cancer. Two confirmatory studies are ongoing worldwide.<br />
● The role <strong>of</strong> dose-dense chemotherapy and neoadjuvant<br />
chemotherapy will be clarified with solid highlevel<br />
evidence in the near future.<br />
350<br />
FUJIWARA, KATSUMATA, AND ONDA<br />
vival compared with the conventional triweekly administration<br />
<strong>of</strong> paclitaxel with the cost <strong>of</strong> modest increase in<br />
toxicities.<br />
The result <strong>of</strong> this trial has markedly influenced the<br />
designs <strong>of</strong> clinical trials, which were planned at the time the<br />
JGOG3016 result was presented. The GOG252 trial was<br />
scheduled to compare IV administration <strong>of</strong> paclitaxel (day 1)<br />
plus intravenous carboplatin (day 1; arm 1) versus IV<br />
paclitaxel (day 1) plus IP carboplatin (day 1; arm 2) versus<br />
modified GOG172 trial winner regimen, which was IV paclitaxel<br />
(day 1) plus IP cisplatin (day 2) plus IP paclitaxel<br />
(day 8; arm 3) (see http://clinicaltrials.gov/ct2/show/<br />
NCT01167712?term�GOG0262&rank�1 for trial information).<br />
All three arms incorporated concurrent and<br />
maintenance bevacizumab. Only arm 3 had administration<br />
<strong>of</strong> paclitaxel on day 8. This was greatly criticized because<br />
there might be a possibility that day-8 administration <strong>of</strong><br />
paclitaxel contributed to the improvement <strong>of</strong> overall survival,<br />
not to the result <strong>of</strong> IP chemotherapy. Therefore, as<br />
Bookman described in the Commentary, 10 “it should be<br />
proven that the net contribution <strong>of</strong> weekly paclitaxel to the<br />
overall survival advantage associated with intraperitoneal<br />
therapy will hopefully be addressed in future studies.” In<br />
fact, the GOG decided to incorporate weekly dose-dense<br />
administration <strong>of</strong> paclitaxel into the two carboplatin arms.<br />
Another scheduled IP trial was OV-20, a National Cancer<br />
Institute <strong>of</strong> Canada/Gynecologic Cancer Intergroup (GCIG)<br />
trial. The trial design <strong>of</strong> OV-20 was similar to that <strong>of</strong><br />
GOG252 trial, except that one <strong>of</strong> two IP arms would be<br />
chosen by randomized phase II fashion, and they did not<br />
combine bevacizumab. This trial design was amended to<br />
incorporate day-8 administration <strong>of</strong> paclitaxel (not dosedense<br />
weekly) for IV and IP carboplatin arms (arm 1 and<br />
arm 2).<br />
Another great movement among gynecologic oncology clinical<br />
trials was to conduct confirmatory trials <strong>of</strong> dose-dense<br />
chemotherapy. In addition to the confirmation <strong>of</strong> the same<br />
dose-dense weekly regimen <strong>of</strong> paclitaxel, these trials try to<br />
answer the following questions. The first is whether weekly<br />
administration <strong>of</strong> carboplatin also contributes to improve<br />
survival. The second question is whether simple division <strong>of</strong><br />
total dose <strong>of</strong> paclitaxel will demonstrate similar efficacy with<br />
less toxicity.<br />
At this time, three randomized trials are ongoing. The<br />
GOG262 trial applied exactly the same trial arms—conventional<br />
triweekly regimen <strong>of</strong> paclitaxel plus carboplatin—<br />
although it only included stage III/IV patients. Use <strong>of</strong><br />
bevacizumab is patient choice (Fig. 1). This trial was closed<br />
in early <strong>2012</strong>. The second trial is MITO-7 study, led by the<br />
Multicenter Italian Trials in Ovarian Cancer Group (Fig. 2).<br />
This study compares the efficacy and toxicity <strong>of</strong> weekly<br />
administration <strong>of</strong> carboplatin in addition to weekly administration<br />
<strong>of</strong> paclitaxel. The dose <strong>of</strong> paclitaxel in the experimental<br />
arm is 60 mg/m 2 instead <strong>of</strong> 80 mg/m 2 in the<br />
JGOG3016 study. The dose <strong>of</strong> carboplatin in the experimental<br />
arm is AUC 2 administered every week. This trial was<br />
opened in 2008, and accrual <strong>of</strong> 800 patients was completed<br />
recently. The third trial is an ambitious study conducted by<br />
Integrated Community <strong>Oncology</strong> Network (ICON) and European<br />
Network <strong>of</strong> Gynaecological Oncological Trial Groups<br />
(ENGOT; Fig. 3). ICON8-ENGOT OV-13 is a three-arm trial<br />
with conventional triweekly administration <strong>of</strong> paclitaxel<br />
plus carboplatin as a comparator (arm 1), and two experi-
CHEMOTHERAPY FOR OVARIAN CANCER<br />
Fig. 1. Study design <strong>of</strong> the GOG0262<br />
trial.<br />
Abbreviations: AUC, area under the<br />
curve; PFS, progression-free survival.<br />
mental arms <strong>of</strong> dose-dense weekly paclitaxel plus triweekly<br />
carboplatin (arm 2) and dose-dense weekly administration <strong>of</strong><br />
both paclitaxel and carboplatin (arm 3). They allowed the<br />
use <strong>of</strong> neoadjuvant chemotherapy as an investigator’s<br />
choice. This trial has just opened in 2011.<br />
Another important unresolved question is the fact that<br />
JGOG3016 failed to show the survival benefit in clear cell<br />
and mucinous cancers. A meta-analysis <strong>of</strong> breast cancer<br />
trials indicated that the dose-dense strategy only contributed<br />
to prolonging survival in the hormone receptor–negative<br />
patients. 11 The patient population who truly benefit<br />
from the dose-dense strategy must be clarified by future<br />
studies.<br />
In conclusion, dose-dense administration <strong>of</strong> paclitaxel<br />
showed a substantial improvement in the survival <strong>of</strong> patients<br />
with ovarian cancer. Further studies ongoing world-<br />
Fig. 2. Study design <strong>of</strong> MITO-7 trial.<br />
Abbreviations: AUC, area under the<br />
curve; PFS, progression-free survival.<br />
wide and meta-analysis will prove the optimal use <strong>of</strong> dosedense<br />
strategy in ovarian cancer.<br />
Neoadjuvant Chemotherapy<br />
The general concept <strong>of</strong> neoadjuvant chemotherapy is to<br />
administer chemotherapy before main treatment such as<br />
surgery or radiation therapy. The purpose <strong>of</strong> this strategy is<br />
to reduce the tumor size or extent <strong>of</strong> cancer spread before<br />
applying the radical main treatment, thus making the<br />
procedure easier or less invasive. It also provides the chance<br />
to know whether the chemotherapy is effective, which is not<br />
possible when the tumor is completely removed.<br />
Neoadjuvant chemotherapy has been studied in several<br />
types <strong>of</strong> cancer, such as breast, prostate, cervix, colorectal,<br />
lung, and esophageal cancers. Among them, breast, prostate,<br />
and cervical cancers have been investigated more<br />
351
than other tumors in which consensus conference or metaanalysis<br />
have already been conducted. 12-14 In the cervical<br />
cancer, neoadjuvant chemotherapy followed by radical hysterectomy<br />
improved survival compared with radiation alone.<br />
But neoadjuvant chemotherapy followed by radiation therapy<br />
negatively affected survival compared with radiation<br />
alone, and the benefit <strong>of</strong> neoadjuvant chemotherapy followed<br />
by radical hysterectomy has not been concluded in comparison<br />
with radical hysterectomy. It has not been compared<br />
with current standard, chemoradiotherapy. In prostate cancer,<br />
neoadjuvant hormone therapy was proven to be effective<br />
only when radiation therapy was applied afterward, but not<br />
beneficial if surgery was conducted after antiandrogenic<br />
therapy. 13 In breast cancer, neoadjuvant therapy was first<br />
used, in the 1980s, typically for patients with inoperable<br />
locally advanced or inflammatory breast cancer, and the<br />
breast-conserving surgery rate dramatically increased. The<br />
next step for the neoadjuvant therapy was to use it as an in<br />
vivo test for chemosensitivity by assessing pathologic complete<br />
response. Currently, by using pathologic response and<br />
other biomarkers as intermediate end points, results from<br />
trials <strong>of</strong> new regimens and therapies that use neoadjuvant<br />
therapy are aimed to proceed and anticipate the results from<br />
larger adjuvant trials. 12<br />
In ovarian cancer, primary debulking surgery followed<br />
by adjuvant chemotherapy is a gold standard procedure.<br />
Although some investigators reported their favorable experience<br />
<strong>of</strong> neoadjuvant chemotherapy followed by interval<br />
debulking surgery, meta-analysis suggested that neoadjuvant<br />
chemotherapy was associated with poorer outcome.<br />
12,15,16 However, there had been no randomized trial<br />
that prospectively demonstrated that primary debulking<br />
surgery is better than neoadjuvant chemotherapy followed<br />
by less-invasive interval debulking surgery. EORTC55971 is<br />
the first prospective randomized study <strong>of</strong> advanced (stage<br />
IIIC or IV) ovarian carcinoma, fallopian tube carcinoma, or<br />
primary peritoneal carcinoma to compare overall survival<br />
between patients who received standard primary debulking<br />
surgery followed by chemotherapy and those who received<br />
352<br />
FUJIWARA, KATSUMATA, AND ONDA<br />
Fig. 3. Study design <strong>of</strong> ICON8-<br />
ENGOT OV-13 trial.<br />
Abbreviations: AUC, area under the<br />
curve; DPS, delayed primary surgery;<br />
EOC, epithelial ovarian cancer; FTC,<br />
fallopian tube cancer; IPS, immediate<br />
primary surgery; PFS, progression-free<br />
survival; PPC, primary peritoneal cancer.<br />
neoadjuvant chemotherapy plus interval debulking surgery.<br />
5 The majority <strong>of</strong> patients who entered this trial (n �<br />
670) had extensive stage IIIC or IV disease at the treatment.<br />
The largest residual tumor 1 cm or less in diameter was<br />
achieved in 41.6% <strong>of</strong> patients after primary debulking and in<br />
80.6% <strong>of</strong> patients after interval debulking. The HR for death<br />
in the neoadjuvant chemotherapy group compared with the<br />
primary debulking surgery group was 0.98 (90% CI, 0.84 to<br />
1.13; p � 0.01 for noninferiority), and the HR for progressive<br />
disease was 1.01 (90% CI, 0.89 to 1.15). Complete resection<br />
<strong>of</strong> all macroscopic disease (at primary or interval surgery)<br />
was the strongest independent variable in predicting overall<br />
survival. Postoperative rates <strong>of</strong> adverse events and mortality<br />
were higher after primary debulking than after interval<br />
debulking.<br />
This study raised considerable controversies 13,17-19 ; thus,<br />
additional phase III trial(s) are necessary to clarify the<br />
benefit <strong>of</strong> the neoadjuvant strategy. Fortunately, two prospective<br />
randomized trials have already completed accrual,<br />
one from the United Kingdom and another from Japan.<br />
The United Kingdom trial CHORUS (Chemotherapy or<br />
Upfront Surgery) is a randomized trial to determine the<br />
impact <strong>of</strong> timing <strong>of</strong> surgery and chemotherapy in patients<br />
with newly diagnosed stage III/IV ovarian, primary peritoneal,<br />
or fallopian tube carcinoma. Study design is similar to<br />
that <strong>of</strong> EORTC55971. The patients were randomly assigned<br />
to receive either immediate primary debulking surgery followed<br />
by six cycles <strong>of</strong> chemotherapy, or neoadjuvant chemotherapy<br />
for three cycles followed by interval debulking<br />
surgery, and then an additional three cycles <strong>of</strong> chemotherapy.<br />
For patients assigned to receive neoadjuvant chemotherapy,<br />
however, histologic or cytologic confirmation <strong>of</strong><br />
target diseases was necessary before starting treatment.<br />
The target accrual was 550 and accrual has already been<br />
accomplished. The data will be combined with EORTC55971<br />
to reliably exclude a 5% to 6% difference in 3-year overall<br />
survival. This trial accomplished enrollment and is waiting<br />
for analysis.<br />
JCOG conducted a randomized trial (JCOG0602) 20 in
CHEMOTHERAPY FOR OVARIAN CANCER<br />
which patients with stage III/IV ovarian, tubal, or primary<br />
peritoneal cancer, were allocated to receive either primary<br />
debulking surgery followed by eight cycles <strong>of</strong> chemotherapy<br />
or to receive neoadjuvant chemotherapy with four cycles <strong>of</strong><br />
chemotherapy followed by interval debulking surgery plus<br />
an additional four cycles <strong>of</strong> chemotherapy. This study is<br />
designed as a noninferiority trial with 300 patients in total.<br />
This trial completed accrual in 2011.<br />
The ultimate goal <strong>of</strong> neoadjuvant chemotherapy is to<br />
minimize the invasiveness <strong>of</strong> surgical treatment without<br />
jeopardizing its efficacy. The most successful example is<br />
breast-conserving surgery after neoadjuvant chemotherapy.<br />
In the EORTC55791 trial, the survival was same between<br />
primary debulking surgery group and neoadjuvant chemotherapy<br />
group, but it was not fully analyzed how they could<br />
minimize the invasiveness <strong>of</strong> surgery between primary and<br />
interval surgery.<br />
The remaining questions on neoadjuvant chemotherapy<br />
include: 1) accurate selection <strong>of</strong> candidate patients, 2) duration<br />
<strong>of</strong> chemotherapy before interval debulking surgery, 3)<br />
selection and duration <strong>of</strong> adjuvant therapy after interval<br />
debulking surgery, and finally 4) whether the failure to<br />
achieve optimal cytoreductive surgery is to the result <strong>of</strong> the<br />
surgeon’s skill or biology <strong>of</strong> the cancer.<br />
For the first question, JCOG investigators conducted a<br />
prospective feasibility study on the selection <strong>of</strong> the patients<br />
who would be suitable for neoadjuvant chemotherapy<br />
Authors’ Disclosures <strong>of</strong> Potential Conflicts <strong>of</strong> Interest<br />
Author<br />
Employment or<br />
Leadership<br />
Positions<br />
Consultant or<br />
Advisory Role<br />
Keiichi Fujiwara Amgen;<br />
Boehringer<br />
Ingelheim;<br />
GlaxoSmithKline;<br />
Zeria Pharma<br />
Noriyuki Katsumata*<br />
Takashi Onda*<br />
*No relevant relationships to disclose.<br />
1. Stuart GC, Kitchener H, Bacon M, et al. 2010 Gynecologic Cancer<br />
InterGroup (GCIG) consensus statement on clinical trials in ovarian cancer:<br />
report from the Fourth Ovarian Cancer Consensus Conference. Int J Gynecol<br />
Cancer. 2011;21:750-755.<br />
2. Burger RA, Brady MF, Bookman MA, et al. Incorporation <strong>of</strong> bevacizumab<br />
in the primary treatment <strong>of</strong> ovarian cancer. N Engl J Med. 2011;365:<br />
2473-2483.<br />
3. Perren TJ, Swart AM, Pfisterer J, et al. A phase 3 trial <strong>of</strong> bevacizumab<br />
in ovarian cancer. N Engl J Med. 2011;365:2484-2496.<br />
4. Thigpen T, duBois A, McAlpine J, et al. First-line therapy in ovarian<br />
cancer trials. Int J Gynecol Cancer. 2011;21:756-1762.<br />
5. Vergote I, Trope CG, Amant F, et al. Neoadjuvant chemotherapy or<br />
primary surgery in stage IIIC or IV ovarian cancer. N Engl J Med. 2010;363:<br />
943-953.<br />
6. Norton L. Theoretical concepts and the emerging role <strong>of</strong> taxanes in<br />
adjuvant therapy. Oncologist. 2001; 3:30-35 (suppl).<br />
7. Seidman AD, Berry D, Cirrincione C, et al. Randomized phase III trial <strong>of</strong><br />
weekly compared with every-3-weeks paclitaxel for metastatic breast cancer,<br />
with trastuzumab for all HER-2 overexpressors and random assignment to<br />
trastuzumab or not in HER-2 nonoverexpressors: final results <strong>of</strong> Cancer and<br />
Leukemia Group B protocol 9840. J Clin Oncol. 2008;26:1642-1649.<br />
8. Sparano JA, Wang M, Martino S, et al. Weekly paclitaxel in the adjuvant<br />
treatment <strong>of</strong> breast cancer. N Engl J Med. 2008;358:1663-1671.<br />
(JCOG0206). 21 They assessed the accuracy <strong>of</strong> clinical diagnosis<br />
on the basis <strong>of</strong> imaging tests, cytology from ascites,<br />
pleural effusion or tumor, and tumor markers (cancer antigen<br />
[CA]-125 � 200 U/mL and Carcinoembryonic antigen<br />
[CEA] � 20 ng/mL). The diagnosis was confirmed by diagnostic<br />
laparoscopy and these results showed that patients<br />
can be correctly diagnosed as having ovarian, fallopian tube,<br />
or primary peritoneal carcinoma with greater than 90%<br />
accuracy by clinical diagnoses on the basis <strong>of</strong> findings<br />
including cytology, according to Bayesian statistical methods.<br />
Other questions, however, have not been investigated<br />
prospectively. We can hypothesize that the greater the<br />
number <strong>of</strong> cycles <strong>of</strong> neoadjuvant chemotherapy, the less<br />
invasive the interval surgery would be because the disseminated<br />
tumor could be eliminated by the chemotherapy.<br />
However, meta-analysis <strong>of</strong> retrospective studies reported<br />
that increasing the number <strong>of</strong> chemotherapy cycles before<br />
interval surgery would negatively affect survival. 16,22 It is<br />
hoped that these important questions will be answered<br />
prospectively in the near future.<br />
Conclusion<br />
The important concept <strong>of</strong> dose-dense chemotherapy and<br />
neoadjuvant chemotherapy in the treatment <strong>of</strong> ovarian<br />
cancer has been tested in phase III trials 5,9 and is a great<br />
influence on clinical practice.<br />
Stock<br />
Ownership Honoraria<br />
REFERENCES<br />
Bristol-Myers<br />
Squibb; Chugai<br />
Pharma;<br />
Janssen<br />
<strong>Oncology</strong>;<br />
Nihonkayaku;<br />
San<strong>of</strong>i; Taiho<br />
Pharmaceutical<br />
Research<br />
Funding<br />
Expert<br />
Testimony<br />
Other<br />
Remuneration<br />
9. Katsumata N, Yasuda M, Takahashi F, et al. Dose-dense paclitaxel once<br />
a week in combination with carboplatin every 3 weeks for advanced ovarian<br />
cancer: a phase 3, open-label, randomised controlled trial. Lancet. 2009;374:<br />
1331-1338.<br />
10. Bookman MA. Dose-dense chemotherapy in advanced ovarian cancer.<br />
Lancet. 2009;374:1303-1305.<br />
11. Bonilla L, Ben-Aharon I, Vidal L, et al. Dose-dense chemotherapy in<br />
nonmetastatic breast cancer: a systematic review and meta-analysis <strong>of</strong><br />
randomized controlled trials. J Natl Cancer Inst. 2010;102:1845-1854.<br />
12. Kaufmann M, von Minckwitz G, Mamounas EP, et al. Recommendations<br />
from an International Consensus Conference on the current status and<br />
future <strong>of</strong> neoadjuvant systemic therapy in primary breast Cancer. Ann Surg<br />
Oncol. Epub 2011 Dec 23.<br />
13. Shelley MD, Kumar S, Wilt T, et al. A systematic review and metaanalysis<br />
<strong>of</strong> randomised trials <strong>of</strong> neo-adjuvant hormone therapy for localised<br />
and locally advanced prostate carcinoma. Cancer Treat Rev. 2009;35:9-17.<br />
14. Neoadjuvant chemotherapy for locally advanced cervical cancer: a<br />
systematic review and meta-analysis <strong>of</strong> individual patient data from 21<br />
randomised trials. Eur J Cancer. 2003;39:2470-2486.<br />
15. Onda T, Yoshikawa H. Neoadjuvant chemotherapy for advanced ovarian<br />
cancer: overview <strong>of</strong> outcomes and unanswered questions. Expert Rev<br />
Anticancer Ther. 2011;11:1053-1067.<br />
16. Bristow RE, Chi DS. Platinum-based neoadjuvant chemotherapy and<br />
353
interval surgical cytoreduction for advanced ovarian cancer: a meta-analysis.<br />
Gynecol Oncol. 2006;103:1070-1076.<br />
17. Chi DS, Bristow RE, Armstrong DK, et al. Is the easier way ever the<br />
better way? J Clin Oncol. 2011;29:4073-4075.<br />
18. Chi DS, Musa F, Dao F, et al. An analysis <strong>of</strong> patients with bulky<br />
advanced stage ovarian, tubal, and peritoneal carcinoma treated with primary<br />
debulking surgery (PDS) during an identical time period as the<br />
randomized EORTC-NCIC trial <strong>of</strong> PDS vs neoadjuvant chemotherapy<br />
(NACT). Gynecol Oncol. <strong>2012</strong>;124:10-14.<br />
19. Vergote I, Tropé CG, Amant F, et al. Neoadjuvant chemotherapy is the<br />
better treatment option in some patients with stage IIIc to IV ovarian cancer.<br />
J Clin Oncol. 2011;29:4076-4078.<br />
354<br />
FUJIWARA, KATSUMATA, AND ONDA<br />
20. Onda T, Matsumoto K, Shibata T, et al. Phase III trial <strong>of</strong> upfront<br />
debulking surgery versus neoadjuvant chemotherapy for stage III/IV ovarian,<br />
tubal and peritoneal cancers: Japan <strong>Clinical</strong> <strong>Oncology</strong> Group Study<br />
JCOG0602. Jpn J Clin Oncol. 2008;38:74-77.<br />
21. Onda T, Kobayashi H, Nakanishi T, et al. Feasibility study <strong>of</strong> neoadjuvant<br />
chemotherapy followed by interval debulking surgery for stage III/IV<br />
ovarian, tubal, and peritoneal cancers: Japan <strong>Clinical</strong> <strong>Oncology</strong> Group Study<br />
JCOG0206. Gynecol Oncol. 2009;113:57-62.<br />
22. Bristow RE, Eisenhauer EL, Santillan A, et al. Delaying the primary<br />
surgical effort for advanced ovarian cancer: A systematic review <strong>of</strong> neoadjuvant<br />
chemotherapy and interval cytoreduction. Gynecol Oncol. 2007;104:480-<br />
490.
UTERINE SARCOMA: CHALLENGING CASES FOR<br />
THE INTERDISCIPLINARY TEAM<br />
CHAIR<br />
Martee L. Hensley, MD<br />
Memorial Sloan-Kettering Cancer Center<br />
New York, NY<br />
SPEAKERS<br />
Anuja Jhingran, MD<br />
University <strong>of</strong> Texas M. D. Anderson Cancer Center<br />
Houston, TX<br />
John P. Curtin, MD<br />
New York University School <strong>of</strong> Medicine<br />
New York, NY
Uterine Sarcomas: Histology and Its<br />
Implications on Therapy<br />
Overview: Uterine sarcomas are rare cancers, they comprise<br />
only 5% <strong>of</strong> all uterine malignancies. There are about 2,000<br />
cases <strong>of</strong> uterine sarcoma diagnosed annually in the United<br />
States. Uterine sarcomas may be categorized as either<br />
favorable-risk, low-grade malignancies with a relatively good<br />
prognosis or as poor-risk, high-grade cancers that carry a<br />
high risk for tumor recurrence and disease progression.<br />
Expert histologic review is critical for appropriate diagnosis<br />
and management. Uterine sarcoma histologies considered to<br />
carry a more favorable prognosis include low-grade endometrial<br />
stromal sarcomas and adenosarcomas. The high-grade<br />
sarcomas include high-grade leiomyosarcomas, high-grade<br />
undifferentiated endometrial sarcomas, and adenosarcomas<br />
with sarcomatous overgrowth.<br />
Low-Grade Endometrial Stromal Sarcomas<br />
ENDOMETRIAL STROMAL sarcomas (ESS) are, by<br />
definition, low-grade malignancies. Histologically they<br />
have bland appearance, with few mitotic figures. Immunohistochemistry<br />
(IHC) stains for desmin, CD10, estrogen<br />
receptor (ER), and progesterone receptor (PR) are typically<br />
positive. Smooth muscle markers (h-caldesmon, smooth<br />
muscle actin) are generally negative in ESS. 1,2 A chromosomal<br />
translocation, (t(7;17)(p15;q21), which fuses two zinc<br />
finger genes, JAZF1/JJAZ, has been described in the majority<br />
<strong>of</strong> ESS and may be useful for distinguishing ESS from<br />
high-grade, undifferentiated endometrial sarcoma (HGUS)<br />
and from leiomyosarcoma (LMS). 3,4<br />
Fifteen percent to 30% <strong>of</strong> patients with ESS may have<br />
evidence <strong>of</strong> metastatic disease at the time <strong>of</strong> diagnosis, with<br />
lung being the most common site for metastatic disease.<br />
However, reflecting the low-grade, favorable behavior <strong>of</strong> this<br />
tumor, five-year survival rates are 60% to 90% across all<br />
stages <strong>of</strong> disease. 5,6 There are no randomized trials assessing<br />
the influence <strong>of</strong> bilateral salpingo-oophorectomy (BSO)<br />
on recurrence and survival in ESS. Some retrospective<br />
studies have shown higher recurrence rates among patients<br />
with retained ovaries. 7 Surveillance, Epidemiology and End<br />
Results (SEER) retrospective data did not show worse overall<br />
survival for women who did not undergo BSO, but this<br />
study did not address recurrence rates, and the number <strong>of</strong><br />
patients with retained ovaries was very small. 8 Lymph node<br />
involvement has been reported to be found in zero to onethird<br />
<strong>of</strong> patients, and whether routine lymph node dissection<br />
<strong>of</strong> normal-appearing nodes in ESS is necessary remains<br />
controversial. 8,9<br />
For patients with uterus-limited, completely resected<br />
ESS, there are no data to support routine adjuvant therapy.<br />
Retrospective SEER data showed poorer overall survival<br />
among patients who received adjuvant pelvic radiation<br />
(80.1%) than among those who had surgery alone (90.7%). 10<br />
There is no role for adjuvant cytotoxic therapy in ESS, and<br />
adjuvant hormonal treatment has not been prospectively<br />
studied. It is reasonable to avoid estrogen replacement<br />
therapy in patients with as diagnosis <strong>of</strong> ESS, although there<br />
are no prospective randomized trials addressing this issue.<br />
356<br />
By Martee L. Hensley, MD<br />
The favorable histology, low-grade uterine sarcomas may be<br />
cured with surgical resection <strong>of</strong> uterus-limited disease. These<br />
tumors are <strong>of</strong>ten hormone-sensitive, and treatment with<br />
hormonal therapies may be efficacious for patients with advanced,<br />
unresectable disease. High-grade uterine leiomyosarcomas<br />
and undifferentiated endometrial sarcomas carry a<br />
high risk for recurrence, even after complete resection <strong>of</strong><br />
uterus-limited disease. No adjuvant intervention has been<br />
shown to improve survival outcomes. Advanced, metastatic<br />
disease is generally treated with systemic cytotoxic therapies,<br />
which may result in objective response but is not curative.<br />
Selected patients with isolated metastatic disease and a long<br />
disease-free interval may benefit from metastatectomy.<br />
Responses to cytotoxic chemotherapy for advanced disease<br />
would be expected to be low in ESS, due to its indolent<br />
growth rate. Since endometrial stromal sarcomas frequently<br />
express ER and PR, objective responses <strong>of</strong> advanced disease<br />
to hormonal interventions, such as treatment with aromatase<br />
inhibitors, have been documented. 11,12<br />
Adenosarcomas<br />
Uterine adenosarcomas are low-grade malignancies that<br />
arise most commonly in the uterine fundus. Histologically<br />
they are characterized by a mixed histologic appearance that<br />
contains benign-appearing glandular epithelial components<br />
and low-grade endometrial stromal sarcoma. 13 Experienced<br />
histologic review is required to exclude the presence <strong>of</strong><br />
“sarcomatous overgrowth” (see below), the presence <strong>of</strong> which<br />
portends a poor prognosis.<br />
In the absence <strong>of</strong> sarcomatous overgrowth, adenosarcomas<br />
have a favorable prognosis, with 5-year survival rates<br />
exceeding 90%. 14 Adenosarcomas are rarer than endometrial<br />
stromal sarcomas, thus data regarding treatment are<br />
very limited. Since the malignant portion <strong>of</strong> adenosarcomas<br />
resembles endometrial stromal sarcoma, management recommendations<br />
are sometimes extrapolated from ESS data.<br />
Like ESS, adenosarcomas commonly express ER, PR, and<br />
CD10 15 ; and, as with ESS, it is reasonable to perform<br />
bilateral oophorectomy, and to avoid hormone replacement<br />
therapy. Table 1 provides a summary <strong>of</strong> histologic features,<br />
prognosis, and management issues for the favorable risk,<br />
low-grade uterine sarcomas.<br />
From the Memorial Sloan-Kettering Cancer Center, Weill Cornell Medical College, New<br />
York, NY.<br />
Author’s disclosures <strong>of</strong> potential conflicts <strong>of</strong> interest are found at the end <strong>of</strong> this article.<br />
Address reprint requests to Author and Session Chair contact information: Martee L.<br />
Hensley, MD, Associate Attending, Gynecologic Medical <strong>Oncology</strong>, Memorial Sloan-<br />
Kettering Cancer Center, Associate Pr<strong>of</strong>essor <strong>of</strong> Medicine, Weill Cornell Medical College,<br />
300 E. 66 th Street, Suite 1355, New York, NY 10065; email: hensleym@mskcc.org.<br />
© <strong>2012</strong> by <strong>American</strong> <strong>Society</strong> <strong>of</strong> <strong>Clinical</strong> <strong>Oncology</strong>.<br />
1092-9118/10/1-10
HISTOLOGIC FEATURES AND MANAGEMENT OF LOW- AND HIGH-GRADE UTERINE SARCOMAS<br />
Endometrial Stromal<br />
Sarcoma<br />
Histologic Features and Management <strong>of</strong> Poor-Risk,<br />
High-Grade Uterine Sarcomas<br />
Leiomyosarcomas<br />
Leiomyosarcomas are generally considered high-grade<br />
cancers. The grading <strong>of</strong> LMS remains controversial, however,<br />
and careful review <strong>of</strong> the histology with an expert<br />
pathologist is recommended. <strong>Clinical</strong>ly, there are some leiomyosarcomas<br />
that may exhibit a more indolent disease<br />
course, although it is not clear yet which histologic features<br />
can definitively classify this subgroup. 16 Most LMS is char-<br />
KEY POINTS<br />
Table 1. Histologic Features and Management Summary for Favorable Histology, Low-Grade Uterine Sarcomas<br />
Histologic Features Surgical Issues Prognosis Systemic Management Issues<br />
Low mitotic rate, IHC positive<br />
for desmin, CD10, ER, PR<br />
JAZF1/JJAZ translocation<br />
Adenosarcoma Benign glandular epithelium<br />
� malignant stromal tissue<br />
that resembles ESS, IHC<br />
positive for ER, PR, CD10<br />
Hysterectomy � BSO, resection<br />
<strong>of</strong> suspicious lymph nodes;<br />
lymph node dissection <strong>of</strong><br />
normal-appearing nodes<br />
remains controversial<br />
Hysterectomy � BSO; resection<br />
<strong>of</strong> suspicious lymph nodes;<br />
lymph node dissection <strong>of</strong><br />
normal-appearing nodes<br />
remains controversial<br />
● Low-grade endometrial stromal sarcomas and adenosarcomas<br />
are hormone-sensitive tumors with a favorable<br />
prognosis.<br />
● High-grade uterine leiomyosarcomas have a high risk<br />
for recurrence after resection <strong>of</strong> uterus-limited disease—approximately<br />
50% at 3 years.<br />
● No adjuvant treatment has been shown to improve<br />
outcomes following resection <strong>of</strong> uterus-limited highgrade<br />
sarcoma—the standard approach following resection<br />
remains observation.<br />
● Cytotoxic treatments with efficacy in high-grade uterine<br />
leiomyosarcoma include doxorubicin with or without<br />
ifosfamide, ifosfamide, gemcitabine, and fixed<br />
dose-rate gemcitabine plus docetaxel.<br />
● Data regarding management <strong>of</strong> high-grade, undifferentiated<br />
endometrial sarcomas are very limited.<br />
Treatment <strong>of</strong> advanced disease is currently extrapolated<br />
from experience with leiomyosarcoma and other<br />
s<strong>of</strong>t tissue sarcomas.<br />
25%–30% <strong>of</strong> patients with uteruslimited<br />
disease may recur.<br />
5-year survival 60%–90% across<br />
all disease stages<br />
5-year overall survival<br />
approximately 90%<br />
Observation after resection <strong>of</strong><br />
uterus-limited disease<br />
Limited data available<br />
regarding adjuvant radiation<br />
or hormonal treatment<br />
Advanced, measurable disease<br />
may respond to hormone<br />
blockade<br />
Expert histologic review to<br />
exclude presence <strong>of</strong><br />
sarcomatous overgrowth<br />
Limited data available<br />
regarding adjuvant radiation<br />
or hormonal treatment<br />
Management strategies are<br />
sometime extrapolated from<br />
ESS<br />
Abbreviations: IHC, immunohistochemistry; ER, estrogen receptor; PR, progesterone receptor; ESS, endometrial stromal sarcoma; BSO, bilateral salpingooophorectomy.<br />
acterized by the histologic presence <strong>of</strong> cytologic atypia,<br />
coagulative necrosis, and a moderate to high mitotic rate. 17<br />
Smooth muscle markers such as h-caldesmon and smooth<br />
muscle actin are typically positive by immunohistochemistry;<br />
p16, p53, and Ki-67 have been reported as useful for<br />
distinguishing LMS from benign smooth muscle tumors. 18<br />
Estimates for the risk <strong>of</strong> recurrence after complete resection<br />
<strong>of</strong> uterus-limited (the International Federation <strong>of</strong> Gynecology<br />
and Obstetrics [FIGO] stage I-limited to the<br />
uterine fundus or stage II-uterine fundus and cervix) uterine<br />
leiomyosarcoma are variable, ranging from 60% to 70% at 2<br />
years 19,20 in retrospective studies, to approximately 50% at<br />
3 years in a prospective study. 21 The aggressive nature <strong>of</strong><br />
this cancer is reflected in the poor overall survival rates. In<br />
one large study the 5-year survival rate was 51% among<br />
patients with FIGO stage I LMS, and only 25% among those<br />
with FIGO stage II LMS. 14 FIGO and the <strong>American</strong> Joint<br />
Committee on Cancer (AJCC) staging systems have been<br />
shown to perform poorly in terms <strong>of</strong> providing accurate<br />
estimates <strong>of</strong> overall survival. 22 A nomogram that includes<br />
age, tumor size, histologic grade, mitotic index, extrauterine<br />
spread, and distant metastases provides better survival<br />
estimates. 23<br />
No form <strong>of</strong> adjuvant treatment has yet been shown to<br />
improve overall survival or progression-free survival. Retrospective<br />
studies suggested that patients who received pelvic<br />
radiation may have fewer local recurrences, however, there<br />
was no improvement in overall survival. 24,25 Furthermore, a<br />
prospective randomized trial <strong>of</strong> adjuvant pelvic radiation<br />
compared to observation did not show benefit for patients<br />
with uterus-limited LMS in terms <strong>of</strong> local recurrence or<br />
survival. 21 One prospective trial <strong>of</strong> adjuvant doxorubicin for<br />
patients with a variety <strong>of</strong> uterine sarcoma histologies did not<br />
show benefit (recurrence rate 41% among patients assigned<br />
to doxorubicin compared with 53% among patients assigned<br />
to observation). 26 Adjuvant chemotherapy using fixed dose-<br />
357
Table 2. Chemotherapy Agents with Activity in Advanced Uterine Leiomyosarcoma<br />
Reference Study Design Agent Response Rate<br />
Sutton, 1992 Prospective phase II Ifosfamide 17%<br />
Omura, 1983 Prospective phase III Doxorubicin 25%<br />
Look, 2003 Prospective phase II, second line Gemcitabine 20%<br />
Sutton, 1996 Prospective Doxorubicin � ifosfamide 30%<br />
Hensley, 2008 Prospective phase II, first line Fixed dose-rate gemcitabine � docetaxel 36%<br />
Hensley, 2008 Prospective phase II, second line 27%<br />
Talbot, 2003 Prospective phase II, s<strong>of</strong>t tissue sarcomas Temozolomide 4%–20%<br />
Ferris, 2010<br />
Anderson, 2005 Retrospective, small series<br />
Monk, <strong>2012</strong> Prospective phase II, first line, uterine LMS Trabectedin 10%<br />
rate gemcitabine plus docetaxel, followed by doxorubicin<br />
was tested in a single-arm phase II trial for patients with<br />
uterus-limited, high-grade LMS. Although 78% <strong>of</strong> patients<br />
were disease free at 2 years, only 52% remained disease free<br />
at 3 years. 27 An international, prospective phase III trial<br />
comparing adjuvant chemotherapy to observation for<br />
women with uterus-limited, high-grade LMS is under development.<br />
Median survival after a diagnosis <strong>of</strong> metastatic or recurrent<br />
leiomyosarcoma is less than 12 months. Patients with<br />
isolated, resectable disease with a long disease-free interval<br />
may be candidates for metastatectomy. Reported outcomes<br />
represent highly selected patients. 28,29 Treatment <strong>of</strong> multisite,<br />
unresectable disease is with systemic cytotoxic therapy.<br />
Response rates for single agents in LMS include ifosfamide<br />
(17%), 30 doxorubicin (25%), 31 gemcitabine (20%), 32 temozolomide<br />
(approximately 4% to 20%, based on limited<br />
data) 33,34,35,36 and trabectedin (8% among patients with<br />
prior treatment, 10% in first-line treatment <strong>of</strong> uterine LMS,<br />
up to 17% in some subsets <strong>of</strong> patients). 37,38,39 Commonly<br />
used combination chemotherapy regimens with activity include<br />
fixed dose-rate gemcitabine plus docetaxel (response<br />
rate 36% as first-line therapy, 40 27% as second-line therapy<br />
in uterine LMS 41 ) and doxorubicin plus ifosfamide (response<br />
rate 30%). 42 Table 2 provides a summary <strong>of</strong> agents with<br />
activity in uterine LMS.<br />
The role <strong>of</strong> targeted therapies in uterine LMS remains to<br />
be elucidated. Retrospective data representing a cohort <strong>of</strong><br />
selected patients with small volume, ER- and/or PR-positive<br />
uterine LMS showed response to treatment with aromatase<br />
inhibition in 9% <strong>of</strong> patients. 43 Data supporting the use <strong>of</strong><br />
vascular endothelial targeted agents are limited. Use <strong>of</strong><br />
such agents is discouraged outside <strong>of</strong> a clinical trial. A<br />
prospective phase II study <strong>of</strong> doxorubicin plus bevacizumab<br />
for patients with previously untreated, metastatic s<strong>of</strong>t tissue<br />
sarcoma (40% <strong>of</strong> whom had uterine LMS) resulted in a<br />
lower-than-expected response rate <strong>of</strong> only 12%, and a high<br />
rate <strong>of</strong> cardiac dysfunction (35% with grade 2 or worse<br />
cardiac toxicity). 44 The multitargeted kinase inhibitor<br />
sunitinib failed to achieve objective responses or disease<br />
stabilization as second- or third-line therapy in uterine<br />
LMS. 45 Sorafenib was similarly disappointing in LMS <strong>of</strong><br />
uterine or nonuterine origin. 46 Whether the vasculartargeted<br />
agent bevacizumab can augment the response rate<br />
and time to disease progression in advanced uterine LMS<br />
when used in combination with fixed dose-rate gemcitabine<br />
plus docetaxel is being investigated in a phase III, placebocontrolled<br />
trial. 47<br />
358<br />
High-Grade, Undifferentiated Endometrial Sarcomas<br />
High-grade, undifferentiated endometrial sarcomas were<br />
previously called high-grade ESS in order to distinguish<br />
them from low-grade ESS. However, since these tumors lack<br />
histologic features that resemble endometrial stroma, the<br />
preferred term is now high-grade, undifferentiated endometrial<br />
sarcoma (HGUS). These tumors have a pleomorphic,<br />
undifferentiated appearance, do not express epithelial or<br />
smooth muscle markers, and are typically ER and PR<br />
negative by immunohistochemistry. A recently reported<br />
t(10;17) genomic rearrangement which yields an oncoprotein<br />
is present in the majority <strong>of</strong> HGUS, a finding that may have<br />
diagnostic and perhaps therapeutic potential. 48 Since lymph<br />
node involvement at time <strong>of</strong> diagnosis is common, lymph<br />
node dissection is <strong>of</strong>ten performed, however, it is not known<br />
whether this leads to improved outcomes given the overall<br />
aggressive behavior <strong>of</strong> this cancer.<br />
High-grade undifferentiated endometrial sarcomas have a<br />
poor prognosis, with 5 year survival rates <strong>of</strong> approximately<br />
25%, and a high risk for disease recurrence after resection <strong>of</strong><br />
uterus-limited disease. 49,50 While retrospective data suggest<br />
that adjuvant pelvic radiation decreases local recurrence<br />
rates, the number <strong>of</strong> patients in such retrospective reports is<br />
small. 51,52 There are no randomized trials assessing the role<br />
<strong>of</strong> pelvic radiation, and no data showing improvements in<br />
progression-free or overall survival.<br />
There have been no prospective trials <strong>of</strong> systemic therapy<br />
specifically for high grade undifferentiated endometrial sarcomas.<br />
Retrospective data, and subsets <strong>of</strong> HGUS patients<br />
enrolled in s<strong>of</strong>t tissue sarcoma studies suggest that HGUS<br />
may respond to doxorubicin or ifosfamide-based regimens.<br />
53,54 Efforts should be made to enroll these patients on<br />
clinical trials testing new agents for sarcomas.<br />
Adenosarcomas with Sarcomatous Overgrowth<br />
MARTEE L. HENSLEY<br />
Uterine adenosarcomas with sarcomatous overgrowth<br />
(ASSO) represent a small subset <strong>of</strong> adenosarcomas. In<br />
contradistinction from adenosarcomas, in which the malignant<br />
portion <strong>of</strong> the tumor has the bland appearance <strong>of</strong> low<br />
grade ESS, the sarcomatous portion <strong>of</strong> ASSO has cellular<br />
atypia, a higher mitotic rate, higher ki-67 staining, and may<br />
have the appearance <strong>of</strong> nonuterine, high-grade sarcomas<br />
such as rhabdomyosarcoma or chondrosarcoma or others.<br />
The sarcomatous overgrowth portion <strong>of</strong> these tumors is<br />
generally CD10 and ER-negative, although PR may be<br />
positive.<br />
ASSO are more likely than adenosarcomas without sarcomatous<br />
overgrowth to have evidence <strong>of</strong> metastatic disease at
HISTOLOGIC FEATURES AND MANAGEMENT OF LOW- AND HIGH-GRADE UTERINE SARCOMAS<br />
the time <strong>of</strong> diagnosis. The disease course is dictated by the<br />
high-grade sarcoma portion <strong>of</strong> the tumor. Survival rates for<br />
patients with metastatic disease are poor: median survival<br />
was 13 months in one study, and the 5-year overall survival<br />
rate was 43% in another. 14,55<br />
There are no prospective data to guide management<br />
decisions for patients with ASSO. Patients are at risk for<br />
both local and distant recurrences, but it is not known<br />
whether any adjuvant strategy can diminish that risk. For<br />
patients with recurrent and/or measurable metastatic disease,<br />
systemic treatment approaches are reasonable, although<br />
the choice <strong>of</strong> agents to use must be extrapolated from<br />
s<strong>of</strong>t tissue sarcoma data. Enrollment in clinical trials for s<strong>of</strong>t<br />
tissue sarcoma should be encouraged. Table 3 provides a<br />
summary <strong>of</strong> histologic features, prognosis, and management<br />
issues for poor-risk, high-grade uterine sarcomas.<br />
Carcinosarcomas<br />
Table 3. Histologic Features and Management Summary for Poor-Risk, High-Grade Uterine Sarcomas<br />
Leiomyosarcoma Cytologic atypia, coagulative<br />
necrosis, high mitotic rate;<br />
High-grade, undifferentiated<br />
endometrial sarcoma<br />
Adenosarcoma with<br />
sarcomatous overgrowth<br />
Carcinosarcomas are uterine tumors containing both a<br />
malignant epithelial component (i.e., adenocarcinoma) and a<br />
malignant mesenchymal component (which may resemble<br />
uterine tissue, in which case it is called homologous; or<br />
non-uterine tissue, in which case it is called heterologous).<br />
Author’s Disclosures <strong>of</strong> Potential Conflicts <strong>of</strong> Interest<br />
Employment or<br />
Leadership<br />
Author<br />
Positions<br />
Martee L. Hensley San<strong>of</strong>i (I)<br />
Histologic Features Surgery for Early-stage Disease Prognosis Systemic Management Issues<br />
IHC positive for smooth muscle<br />
markers; may be positive<br />
for ER, PR<br />
Pleomorphic, undifferentiated<br />
cells; IHC negative for<br />
smooth muscle markers,<br />
negative for ER, PR; t(10;<br />
17) genomic rearrangement<br />
(YWHAE/FAM22)<br />
Sarcomatous portion has<br />
atypia, heterologous<br />
elements resembling<br />
rhabdomyosarcoma,<br />
chondrosarcoma, or others;<br />
CD10 negative; ER negative<br />
Consultant or<br />
Advisory Role<br />
Hysterectomy � BSO; lymph<br />
node dissection not necessary<br />
for normal-appearing lymph<br />
nodes<br />
Hysterectomy � BSO; may<br />
include lymph node dissection<br />
(limited data)<br />
Hysterectomy � BSO; may<br />
include lymph node dissection<br />
(limited data)<br />
These tumors are sometimes considered a subtype <strong>of</strong> uterine<br />
sarcomas, and sometimes considered to be high grade endometrial<br />
cancers. The constraints <strong>of</strong> this publication preclude<br />
a complete discussion <strong>of</strong> these malignancies in this manuscript.<br />
For further information, see other publications. 56<br />
Critical Research Questions<br />
Although uterine sarcomas are rare, adequately powered,<br />
well-designed studies are needed in these diseases. An<br />
international, randomized phase III trial is under development<br />
to address whether adjuvant chemotherapy can improve<br />
survival outcomes among women with completely<br />
resected uterus-limited LMS. Large collaborative trials are<br />
needed to define the role <strong>of</strong> hormonal treatment for ESS,<br />
both as adjuvant therapy and for advanced disease. New<br />
agents are needed for advanced leiomyosarcoma, and prospective<br />
data are needed to define whether current agents<br />
have activity in high grade undifferentiated endometrial<br />
sarcomas. Correlative studies are needed to identify molecular<br />
drivers <strong>of</strong> these tumors, and clinical trials will be<br />
needed to determine whether these drivers may serve as<br />
effective targets for treatment.<br />
Stock<br />
Ownership Honoraria<br />
For uterus-limited LMS:<br />
approximately 50%<br />
recurrence by 3 year;<br />
5-year overall survival<br />
25%–50%; median<br />
survival from diagnosis<br />
<strong>of</strong> metastatic disease is<br />
1 year, but range may<br />
be wide<br />
Limited prospective data;<br />
risk for recurrence<br />
greater than 50%;<br />
survival with metastatic<br />
disease is poor<br />
Approximately 30% have<br />
metastatic disease at<br />
diagnosis; median<br />
survival approximately<br />
13 months<br />
Research<br />
Funding<br />
For uterus-limited disease:<br />
neither pelvic radiation nor<br />
adjuvant chemotherapy has<br />
been shown to improve PFS or<br />
OS<br />
For advanced, measurable<br />
disease: cytotoxic therapy may<br />
achieve objective responses in<br />
approximately 30% <strong>of</strong> patients<br />
No prospective data regarding<br />
adjuvant therapy for completely<br />
resected disease<br />
No prospective data regarding<br />
systemic treatment specifically<br />
for HGUS; enrollment in<br />
clinical trials for s<strong>of</strong>t tissue<br />
sarcoma encouraged<br />
No prospective data regarding<br />
adjuvant therapy for completely<br />
resected disease<br />
No prospective data for<br />
treatment <strong>of</strong> advanced ASSO;<br />
enrollment on clinical trials for<br />
s<strong>of</strong>t tissue sarcoma encouraged<br />
Abbreviations: IHC, immunohistochemistry; ER, estrogen receptor; PR, progesterone receptor; BSO, bilateral salpingo-oophorectomy; PFS, progression-free survival;<br />
OS, overall survival; ASSO, adenosarcoma with sarcomatous overgrowth.<br />
Expert<br />
Testimony<br />
Other<br />
Remuneration<br />
359
1. McCluggage WG, Sumathi VP, Maxwell P. CD10 is a sensitive and<br />
diagnostically useful immunohistochemical marker <strong>of</strong> normal endometrial<br />
stroma and <strong>of</strong> endometrial stromal neoplasms. Histopathology. 2001;39:273-278.<br />
2. Nucci MR, O’Connell JT, Huettner PC, et al. h-Caldesmon expression<br />
effectively distinguishes endometrial stromal tumors from uterine smooth<br />
muscle tumors. Am J Surg Pathol. 2001;25:455-463.<br />
3. Koontz JI, Soreng AL, Nucci M, et al. Frequent fusion <strong>of</strong> the JAZF1 and<br />
JJAZ1 genes in endometrial stromal tumors. Proc Natl Acad Sci U S A.<br />
2001;98:6348-6353.<br />
4. Micci F, Walter CU, Teixeira MR, Panagopoulos I, Bjerkehagen B,<br />
Saeter G, Heim S. Cytogenetic and molecular genetic analyses <strong>of</strong> endometrial<br />
stromal sarcoma: nonrandom involvement <strong>of</strong> chromosome arms 6p and 7p and<br />
confirmation <strong>of</strong> JAZF1/JJAZ1 gene fusion in t(7;17). Cancer Genet Cytogenet.<br />
2003;144:119-124.<br />
5. Bodner K, Bodner-Adler B, Obermair A, Windbichler G, Petru E,<br />
Mayerh<strong>of</strong>er S, Czerwenka K, Leodolter S, Kainz C, Mayerh<strong>of</strong>er K. Prognostic<br />
parameters in endometrial stromal sarcoma: a clinicopathologic study in 31<br />
patients. Gynecol Oncol. 2001;81:160-165.<br />
6. Gadducci A, Sartori E, Landoni F, et al. Endometrial stromal sarcoma:<br />
analysis <strong>of</strong> treatment failures and survival. Gynecol Oncol. 1996;63:247-253.<br />
7. Berchuck A, Rubin SC, Hoskins WJ, et al. Treatment <strong>of</strong> endometrial<br />
stromal tumors. Gynecol Oncol. 1990;36:60-65.<br />
8. Shah JP, Bryant CS, Kumar S, et al. Lymphadenectomy and ovarian<br />
preservation in low-grade endometrial stromal sarcoma. Obstet Gynecol.<br />
2008;112:1102-1108.<br />
9. Dos Santos LA, Garg K, Diaz JP, et al. Incidence <strong>of</strong> lymph node and<br />
adnexal metastasis in endometrial stromal sarcoma. Gynecol Oncol. 2011;<br />
121:319-322.<br />
10. Barney B, Tward JD, Skidmore T, et al. Does radiotherapy or lymphadenectomy<br />
improve survival in endometrial stromal sarcoma? Int J Gynecol<br />
Cancer. 2009;19:1232-1238.<br />
11. Maluf FC, Sabbatini P, Schwartz L, et al. Endometrial stromal sarcoma:<br />
objective response to letrozole. Gynecol Oncol. 2001;82:384-388.<br />
12. Chu MC, Mor G, Lim C, et al. Low-grade endometrial stromal sarcoma:<br />
hormonal aspects. Gynecol Oncol. 2003;90:170-176.<br />
13. Verschraegen CF, Vasuratna A, Edwards C, et al. Clinicopathologic<br />
analysis <strong>of</strong> mullerian adenosarcoma: The M. D. Anderson Cancer Center<br />
experience. Oncol Rep. 1998;5:939-944.<br />
14. Abeler VM, Røyne O, Thoresen S, et al. Uterine sarcomas in Norway. A<br />
histopathological and prognostic survey <strong>of</strong> a total population from 1970 to<br />
2000 including 419 patients. Histopathology. 2009;54:355-364.<br />
15. Soslow RA, Ali A, Oliva E. Mullerian adenosarcomas: an immunophenotypic<br />
analysis <strong>of</strong> 35 cases. Am J Surg Pathol. 2008;32:1013-1021.<br />
16. Veras E, Zivanovic O, Jacks L, et al. “Low-grade leiomyosarcoma” and<br />
late-recurring smooth muscle tumors <strong>of</strong> the uterus: a heterogenous collection<br />
<strong>of</strong> frequently misdiagnosed tumors associated with an overall favorable<br />
prognosis relative to conventional uterine leiomyosarcomas. Am J Surg<br />
Pathol. 2011;35:1626-1637.<br />
17. Bell SW, Kempson RL, Hendrickson MR. Problematic uterine smooth<br />
muscle neoplasms. A clinicopathologic study <strong>of</strong> 213 cases. Am J Surg Pathol.<br />
1994;18:535-558.<br />
18. Chen L, Yang B. Immunohistochemical analysis <strong>of</strong> p16, p53, and Ki-67<br />
expression in uterine smooth muscle tumors. Int J Gynecol Pathol. 2008;27:<br />
326-332.<br />
19. Major FJ, Blessing JA, Silverberg SG, et al. Prognostic factors in<br />
early-stage uterine sarcoma. A Gynecologic <strong>Oncology</strong> Group study. Cancer.<br />
1993;71:1702-1709.<br />
20. Nordal RR, Kristense GB, Kaern J, et al. The prognostic significance <strong>of</strong><br />
stage, tumor size, cellular atypia and DNA ploidy in uterine leiomyosarcoma.<br />
Acta Oncol. 1995;34:797-802.<br />
21. Reed NS, Mangioni C, Malmström H, et al. Phase III randomised study<br />
to evaluate the role <strong>of</strong> adjuvant pelvic radiotherapy in the treatment <strong>of</strong><br />
uterine sarcomas stages I and II: an European Organisation for Research and<br />
Treatment <strong>of</strong> Cancer Gynaecological Cancer Group Study (protocol 55874).<br />
Eur J Cancer. 2008;44:808-818<br />
22. Zivanovic O, Leitao MM, Iasonos A, et al. Stage-specific outcomes <strong>of</strong><br />
patients with uterine leiomyosarcoma: a comparison <strong>of</strong> the international<br />
Federation <strong>of</strong> gynecology and obstetrics and american joint committee on<br />
cancer staging systems. J Clin Oncol. 2009;27:2066-2072.<br />
23. Zivanovic O, Jacks LM, Iasonos A, et al. A nomogram to predict<br />
postresection 5-year overall survival for patients with uterine leiomyosarcoma.<br />
Cancer. <strong>2012</strong>;118:660-669.<br />
24. Giuntoli RL, Metzinger DS, DiMarco CS, et al. Retrospective review <strong>of</strong><br />
208 patients with leiomyosarcoma <strong>of</strong> the uterus: Prognostic indicators,<br />
360<br />
REFERENCES<br />
MARTEE L. HENSLEY<br />
surgical management, and adjuvant therapy. Gynecol Oncol. 2003;89:460-<br />
469.<br />
25. Hornback NB, Omura G, Major FJ. Observations on the use <strong>of</strong> adjuvant<br />
radiation therapy in patients with stage I and II uterine sarcoma. Int J<br />
Radiat Oncol Biol Phys. 1986;12:2127-2130.<br />
26. Omura GA, Blessing JA, Major F, et al. A randomized clinical trial <strong>of</strong><br />
adjuvant adriamycin in uterine sarcomas: a Gynecologic <strong>Oncology</strong> Group<br />
Study. J Clin Oncol. 1985;3:1240-1245.<br />
27. Hensley ML, Wathen K, Maki RG, et al. 3-year follow-up <strong>of</strong> SARC 005:<br />
adjuvant treatment <strong>of</strong> high risk primary uterine leiomyosarcoma with gemcitabine/docetaxel<br />
(GT) followed by doxorubicin (D). Connective Tissue <strong>Oncology</strong><br />
<strong>Society</strong>, Abstract ID: 1130375, 2011.<br />
28. Leitao MM, Brennan MF, Hensley M, et al. Surgical resection <strong>of</strong><br />
pulmonary and extrapulmonary recurrences <strong>of</strong> uterine leiomyosarcoma. Gynecol<br />
Oncol. 2002;87:287-294.<br />
29. Clavero JM, Deschamps C, Cassivi SD, et al. Gynecologic cancers:<br />
factors affecting survival after pulmonary metastasectomy. Ann Thorac Surg.<br />
2006;81:2004-2007.<br />
30. Sutton GP, Blessing JA, Barrett RJ, et al. Phase II trial <strong>of</strong> ifosfamide<br />
and mesna in leiomyosarcoma <strong>of</strong> the uterus: a Gynecologic <strong>Oncology</strong> Group<br />
study. Am J Obstet Gynecol. 1992;166:556-559.<br />
31. Omura GA, Major FJ, Blessing JA, et al. A randomized study <strong>of</strong><br />
adriamycin with and without dimethyl triazenoimidazole carboxamide in<br />
advanced uterine sarcomas. Cancer. 1983;52:626-632.<br />
32. Look KY, Sandler A, Blessing JA, et al. Phase II trial <strong>of</strong> gemcitabine as<br />
second-line chemotherapy <strong>of</strong> uterine leiomyosarcoma: a Gynecologic <strong>Oncology</strong><br />
Group (GOG) study. Gynecol Oncol. 2004;92:644-647.<br />
33. Talbot SM, Keohan ML, Hesdorffer M, et al. A phase II trial <strong>of</strong><br />
temozolomide in patients with unresectable or metastatic s<strong>of</strong>t tissue sarcoma.<br />
Cancer. 2003;98:1942-1946.<br />
34. Boyar M, Hesdorffer M, Keohan M, et al. Phase II study <strong>of</strong> temozolomide<br />
and thalidomide in patients with unresectable or metastatic leiomyosarcoma.<br />
J Clin Oncol. 2005;23 (abstract 9029).<br />
35. Ferriss JS, Atkins KA, Lachance JA, et al. Temozolomide in advanced<br />
and recurrent uterine leiomyosarcoma and correlation with o6methylguanine<br />
DNA methyltransferase expression: A case series. Int J<br />
Gynecol Cancer. 2010;20:120-125.<br />
36. Anderson S, Aghajanian C. Temozolomide in uterine leiomyosarcomas.<br />
Gynecol Oncol. 2005;98:99-103.<br />
37. Garcia-Carbonero R, Supko JG, Maki RG, et al. Ecteinascidin-743<br />
(ET-743) for chemotherapy-naive patients with advanced s<strong>of</strong>t tissue sarcomas:<br />
multicenter phase II and pharmacokinetic study. J Clin Oncol. 2004;24:<br />
5484-5492.<br />
38. Delaloge S, Yovine A, Taamma A, et al. Ecteinascidin-743: a marinederived<br />
compound in advanced, pretreated sarcoma patients-preliminary<br />
evidence <strong>of</strong> activity. J Clin Oncol. 2001;19:1248-1255.<br />
39. Monk BJ, Blessing JA, Street DG, et al. A phase II evaluation <strong>of</strong><br />
trabectedin in the treatment <strong>of</strong> advanced, persistent, or recurrent uterine<br />
leiomyosarcoma: a gynecologic oncology group study. Gynecol Oncol. <strong>2012</strong>;<br />
124:48-52.<br />
40. Hensley ML, Blessing JA, Mannel R, et al. Fixed-dose rate gemcitabine<br />
plus docetaxel as first-line therapy for metastatic uterine leiomyosarcoma: a<br />
Gynecologic <strong>Oncology</strong> Group phase II trial. Gynecol Oncol. 2008;109:329-334.<br />
41. Hensley ML, Blessing JA, DeGeest K, et al. Fixed-dose rate gemcitabine<br />
plus docetaxel as second-line therapy for metastatic uterine leiomyosarcoma:<br />
a Gynecologic <strong>Oncology</strong> Group phase II study. Gynecol Oncol.<br />
2008;109:323-328.<br />
42. Sutton G, Blessing JA, Malfetano JH. Ifosfamide and doxorubicin in the<br />
treatment <strong>of</strong> advanced leiomyosarcomas <strong>of</strong> the uterus: a Gynecologic <strong>Oncology</strong><br />
Group study. Gynecol Oncol. 1996;62:226-229.<br />
43. O’Cearbhaill R, Zhou Q, Iasonos A, et al. Treatment <strong>of</strong> Advanced Uterine<br />
Leiomyosarcoma with Aromatase Inhibitors. Gynecol Oncol. 2010;116:424-429.<br />
44. D’Adamo DR, Anderson SE, Albritton K, et al. Phase II study <strong>of</strong><br />
doxorubicin and bevacizumab for patients with metastatic s<strong>of</strong>t-tissue sarcomas.<br />
J Clin Oncol. 2005;23:7135-7142.<br />
45. Hensley ML, Sill MW, Scribner DR Jr, et al. Sunitinib malate in the<br />
treatment <strong>of</strong> recurrent or persistent uterine leiomyosarcoma: a Gynecologic<br />
<strong>Oncology</strong> Group phase II study. Gynecol Oncol. 2009;115:460-465.<br />
46. Maki RG, D’Adamo DR, Keohan ML, et al. Phase II study <strong>of</strong> sorafenib<br />
in patients with metastatic or recurrent sarcomas. J Clin Oncol. 2009;27:<br />
3133-3140.<br />
47. Hensley ML. Principal Investigator. (GOG 0250, clinical trial identifier<br />
NCT01012297).
HISTOLOGIC FEATURES AND MANAGEMENT OF LOW- AND HIGH-GRADE UTERINE SARCOMAS<br />
48. Lee CH, Ou WB, Mariño-Enriquez A, et al. 14-3-3 fusion oncogenes in<br />
high-grade endometrial stromal sarcoma. Proc Natl Acad SciUSA. <strong>2012</strong> Jan<br />
5. [Epub ahead <strong>of</strong> print]<br />
49. Nordal RR, Kristensen GB, Kaern J, et al. The prognostic significance<br />
<strong>of</strong> surgery, tumor size, malignancy grade, menopausal status, and DNA ploidy<br />
in endometrial stromal sarcoma. Gynecol Oncol. 1996;62:254-259.<br />
50. Gadducci A, Sartori E, Landoni F, et al. Endometrial stromal sarcoma:<br />
analysis <strong>of</strong> treatment failures and survival. Gynecol Oncol. 1996;63:247-253.<br />
51. Berchuck A, Rubin SC, Hoskins WJ, et al. Treatment <strong>of</strong> endometrial<br />
stromal tumors. Gynecol Onco. 1990;36:60-65.<br />
52. Piver MS, Rutledge FN, Copeland L, et al. Uterine endolymphastic<br />
stromal myosis: A collaborative study. Obstet Gynecol 1995;64:173.<br />
53. Sutton G, Blessing JA, Park R, et al. Ifosfamide treatment <strong>of</strong> recurrent<br />
or metastatic endometrial stromal sarcomas previously unexposed to chemotherapy:<br />
a study <strong>of</strong> the Gynecologic <strong>Oncology</strong> Group. Obstet Gynecol. 1996;<br />
87:747-750.<br />
54. Yamawaki T, Shimizu Y, Hasumi K. Treatment <strong>of</strong> stage IV “highgrade”<br />
endometrial stromal sarcoma with ifosfamide, adriamycin, and cisplatin.<br />
Gynecol Oncol. 1997;64:265-269.<br />
55. Krivak TC, Seidman JD, McBroom JW, et al. Uterine adenosarcoma<br />
with sarcomatous overgrowth versus uterine carcinosarcoma: comparison <strong>of</strong><br />
treatment and survival. Gynecol Oncol. 2001;83:89-94.<br />
56. Acharya S, Hensley ML, Montag AC, et al. Rare Uterine Cancers.<br />
Lancet <strong>Oncology</strong>. 2005;6:961-971.<br />
361
Surgical Options for Recurrent<br />
Uterine Sarcomas<br />
By Sharmilee B. Korets, MD, and John P. Curtin, MD<br />
Overview: Leiomyosarcoma, the most frequent pure uterine<br />
sarcoma, is an aggressive tumor with a tendency toward early<br />
relapse. Survival for patients with recurrent disease is poor.<br />
In contrast, endometrial stromal sarcoma, the second most<br />
common uterine sarcoma, is a more indolent malignancy with<br />
a tendency toward recurrence after a long latency period. The<br />
relative infrequency <strong>of</strong> both diseases makes the study and<br />
standardization <strong>of</strong> treatment for recurrent disease challenging.<br />
Treatment <strong>of</strong> recurrence with cytotoxic chemotherapy,<br />
radiation therapy, or hormone therapy produces modest to<br />
poor response rates. Surgical resection is one treatment<br />
modality <strong>of</strong>fering the potential for cure and perhaps a more<br />
durable response than is seen with medical management.<br />
Although initial studies focused on pulmonary metastasec-<br />
Case History: An otherwise healthy, middle-age woman was<br />
diagnosed with leiomyosarcoma at the time <strong>of</strong> a myomectomy<br />
for symptomatic fibroids in 2000. Six weeks later, she underwent<br />
a total abdominal hysterectomy with bilateral salpingooophorectomy<br />
for leiomyosarcoma and was diagnosed with<br />
stage IB disease. Surgery was followed by three cycles <strong>of</strong><br />
adjuvant chemotherapy. She had an initial disease-free interval<br />
<strong>of</strong> 7.5 years. However, in fall 2008, she was diagnosed<br />
with a large left upper lobe thoracic metastasis and a<br />
synchronous left acetabular lesion. She underwent a videoassisted<br />
thoracoscopic left upper lobectomy and mediastinal<br />
lymph node dissection for a 5 cm mass. All pulmonary<br />
disease was completely resected and lymph nodes were negative<br />
for disease. Three months later, she had intermittent<br />
abdominal discomfort and vague bowel symptoms and was<br />
noted to have a jejunal mass that was suspected to be<br />
recurrent disease. The mass was completely removed by<br />
small bowel resection with reanastomosis. At the time <strong>of</strong><br />
surgery, there was no evidence <strong>of</strong> residual intra-abdominal<br />
disease. She had resection <strong>of</strong> the acetabular lesion in spring<br />
2009. At that time, she was thought to be free <strong>of</strong> disease.<br />
However within 1 month, she a calcaneal metastasis was<br />
found, and after a course <strong>of</strong> radiation therapy, systemic<br />
chemotherapy was initiated. She was treated with multiple<br />
chemotherapy regimens over the course <strong>of</strong> the next 16 months.<br />
She then chose to pursue palliative treatment and died<br />
approximately 10 years after the initial diagnosis and 30<br />
months after the initial surgical resection for recurrence.<br />
This case illustrates several key points in our review <strong>of</strong><br />
surgical management <strong>of</strong> recurrent uterine sarcoma and provides<br />
illustrative radiographic images <strong>of</strong> resectable metastases<br />
in uterine sarcoma (Fig. 1).<br />
MALIGNANT MESENCHYMAL tumors <strong>of</strong> the uterus<br />
are rare, accounting for less than 3% <strong>of</strong> all uterine<br />
malignancies. The annual incidence <strong>of</strong> uterine sarcoma<br />
approaches two per 100,000 women. 1 Recent changes in<br />
terminology and classification now exclude carcinosarcomas<br />
from this group, as the biology and clinical behavior <strong>of</strong> those<br />
tumors point toward an epithelial origin. This review focuses<br />
on surgical management <strong>of</strong> recurrence in the most<br />
common pure uterine sarcomas, listed in order <strong>of</strong> incidence:<br />
362<br />
tomy in recurrent s<strong>of</strong>t tissue sarcoma, an increasingly large<br />
body <strong>of</strong> data specifically evaluating outcomes after both<br />
thoracic and extrathoracic metastasectomy in patients with<br />
recurrent uterine sarcoma is now available. Though no prospective<br />
trials have been conducted, retrospective comparisons<br />
<strong>of</strong> chemotherapy or radiation therapy with surgery for<br />
recurrent uterine sarcoma suggest improvement in diseasespecific<br />
survival for the surgery group. Clearly defined factors<br />
are associated with better prognosis after surgical resection<br />
<strong>of</strong> recurrence, including a prolonged disease-free interval<br />
and complete resection <strong>of</strong> disease. In properly selected<br />
women, surgery and even repeated metastasectomy for<br />
recurrent disease may improve survival and should be considered.<br />
leiomyosarcoma, which accounts for the majority <strong>of</strong> uterine<br />
sarcomas, followed by endometrial stromal sarcoma (ESS,<br />
previously called low-grade endometrial stromal sarcoma),<br />
undifferentiated endometrial sarcoma (previously called<br />
high-grade endometrial stromal sarcoma), and adenosarcoma.<br />
Even early-stage uterine sarcomas demonstrate aggressive<br />
clinical behavior and confer a poor prognosis. Leiomyosarcomas<br />
have a propensity toward hematogenous spread<br />
and early recurrence. Five-year survival rates range from<br />
30% to 48%, and relapse rates approach 60%, with 42% <strong>of</strong><br />
relapsed disease occurring outside the pelvis. 2 Most extrapelvic<br />
relapses occur in the lung. As with other s<strong>of</strong>t tissue<br />
sarcomas, leiomyosarcomas are relatively chemoresistant, 3<br />
making treatment <strong>of</strong> recurrent disease challenging. In contrast,<br />
ESS is more indolent, with a tendency toward local or<br />
distant relapse after a long latency period. Five-year survival<br />
rates are between 80% and 100%, 1 the median time to<br />
recurrence in women with stage I disease is 65 months, and<br />
the rate <strong>of</strong> relapse ranges from 36% to 56%. 4 Recurrences<br />
are primarily pelvic, intra-abdominal, or pulmonary;<br />
however, intravascular, cardiac, and central nervous<br />
system metastases have also been reported. Because <strong>of</strong><br />
their slow growth, repeated resection may be warranted,<br />
and even secondary and tertiary cytoreductions likely improve<br />
prognosis. 4 Adenosarcomas are similarly indolent,<br />
with excellent survival when disease is at an early stage at<br />
the time <strong>of</strong> diagnosis and there is a long latency period<br />
before relapse.<br />
The relative rarity <strong>of</strong> uterine sarcomas makes clinical<br />
investigation difficult and prospective randomized trials<br />
nearly impossible, especially for evaluating the management<br />
<strong>of</strong> recurrent disease. Most studies are retrospective,<br />
and various primary disease sites and histologic subtypes<br />
From the Division <strong>of</strong> Gynecologic <strong>Oncology</strong>, New York University School <strong>of</strong> Medicine, New<br />
York, NY.<br />
Authors’ disclosures <strong>of</strong> potential conflicts <strong>of</strong> interest are found at the end <strong>of</strong> this article.<br />
Address reprint requests to John P. Curtin, MD, NYU <strong>Clinical</strong> Cancer Center, 160 E. 34th<br />
St., 4th Floor, New York, NY 10016; email: john.curtin@med.nyu.edu.<br />
© <strong>2012</strong> by <strong>American</strong> <strong>Society</strong> <strong>of</strong> <strong>Clinical</strong> <strong>Oncology</strong>.<br />
1092-9118/10/1-10
SURGICAL OPTIONS FOR UTERINE SARCOMAS<br />
have been combined in order to attain sufficient power.<br />
Thus, optimal management strategies for recurrent disease<br />
and definition <strong>of</strong> favorable clinicopathologic characteristics<br />
for secondary cytoreduction or metastasectomy are not well<br />
established. The purpose <strong>of</strong> this chapter is to better characterize<br />
candidates for resection <strong>of</strong> recurrent disease and to<br />
discuss factors associated with better prognosis.<br />
Treatment Options for Recurrent Disease<br />
Survival for patients with recurrent disease is poor. Systemic<br />
therapy is <strong>of</strong>ten recommended; in ESS, hormone<br />
therapy is frequently used, whereas in leiomyosarcoma,<br />
cytotoxic chemotherapy may be the treatment <strong>of</strong> choice. In<br />
uterine leiomyosarcoma, doxorubicin is the most active<br />
single agent, with a response rate <strong>of</strong> 25%. 5 Combination<br />
therapies containing doxorubicin and ifosfamide and gemcitabine<br />
and docetaxel also have demonstrated activity in<br />
metastatic or recurrent disease. 6,7<br />
External-beam radiation therapy has been used in the<br />
adjuvant setting for uterine sarcoma, and most studies have<br />
demonstrated an improved local control rate without a<br />
substantial effect on overall survival. Radiation therapy<br />
may have a role in the palliative treatment <strong>of</strong> distant<br />
metastases. 1 However, <strong>of</strong> all possible treatments for recurrent<br />
disease, including cytotoxic chemotherapy, hormone<br />
therapy, radiation therapy, or surgery, only surgery is<br />
associated with a high cure rate and prolonged survival. 8<br />
Surgical Resection <strong>of</strong> Metastatic Disease<br />
In a single-institution retrospective cohort study <strong>of</strong> 33<br />
patients with recurrent or metastatic uterine leiomyosarcoma,<br />
survival was improved for patients who had surgical<br />
resection <strong>of</strong> metastases compared with patients who<br />
did not have surgery, with a median overall survival 45<br />
compared with 31 months, respectively. 9 In a larger study<br />
(128 patients), survival after secondary cytoreduction for<br />
recurrent uterine leiomyosarcoma was compared with survival<br />
after chemotherapy and/or radiation therapy for recur-<br />
KEY POINTS<br />
● Recurrent uterine sarcoma is an aggressive disease in<br />
which nonsurgical treatment modalities <strong>of</strong>fer limited<br />
benefit.<br />
● Due to the relative rarity <strong>of</strong> uterine sarcoma, few<br />
prospective data on optimal management <strong>of</strong> recurrent<br />
disease are available.<br />
● Retrospective data demonstrate a survival benefit for<br />
well-selected women in whom surgery is performed<br />
for the management <strong>of</strong> recurrent disease.<br />
● Although initial data suggested that surgery for metastastic<br />
disease should be limited to pulmonary<br />
metastasectomy only, additional studies have shown<br />
that extrathoracic metastasectomy <strong>of</strong>fers a similar<br />
survival advantage.<br />
● Prolonged disease-free interval and complete resection<br />
<strong>of</strong> tumor correlate with improved outcome for<br />
patients undergoing surgery for metastatic uterine<br />
sarcoma.<br />
rent disease. No distinction was made between abdominal,<br />
pelvic, or thoracic procedures for resection. Disease-specific<br />
survival as well as overall survival were substantially improved<br />
for the surgery group compared with the medical<br />
management group, with a mean overall survival <strong>of</strong> 2.0<br />
compared with 1.1 years from the time <strong>of</strong> recurrence. 10<br />
Due to the rarity <strong>of</strong> uterine sarcoma, much <strong>of</strong> the data<br />
regarding surgical management <strong>of</strong> recurrence come from<br />
studies in which heterogeneous histologic subtypes and<br />
primary disease sites are pooled. Because the lung is a<br />
preferred site <strong>of</strong> recurrence for s<strong>of</strong>t tissue sarcoma, a large<br />
body <strong>of</strong> literature focuses on indications for pulmonary metastasectomy<br />
and factors affecting survival after resection.<br />
Pulmonary Metastasectomy<br />
Indications for pulmonary metastasectomy in recurrent<br />
uterine sarcoma can be extrapolated from studies <strong>of</strong> the<br />
procedure for patients with all types <strong>of</strong> s<strong>of</strong>t tissue sarcomas.<br />
In the surgical and thoracic oncology literature, indications<br />
for pulmonary metastasectomy include medical suitability<br />
for surgery, sufficient pulmonary reserve to tolerate loss <strong>of</strong><br />
lung capacity, control <strong>of</strong> disease at the primary site, no<br />
evidence <strong>of</strong> extrapulmonary disease, and no better therapy<br />
available. 11 Pulmonary metastasectomy with curative intent<br />
is widely accepted for well-selected patients with sarcomatous<br />
lung metastases. 3,12<br />
Thoracic procedures for metastasectomy range from thoracoscopic<br />
wedge resection to open pneumonectomy or bilobectomy<br />
via median sternotomy. Wedge resection is<br />
performed in most women undergoing surgery for sarcomatous<br />
lung metastases. 11,13,14 In a series from Brigham and<br />
Women’s hospital, approximately 75% <strong>of</strong> patients who had<br />
pulmonary metastasectomy for recurrent leiomyosarcoma<br />
were treated with wedge resection. In that study, it was also<br />
demonstrated that disease-specific and overall survival were<br />
similar for patients who underwent video-assisted thoracic<br />
surgery (VATS) compared with thoracotomy or sternotomy.<br />
Because recurrent lung metastases will develop in many<br />
women, procedures that preserve the ability to tolerate<br />
repeated resections, such as VATS and wedge resection, are<br />
preferred. 13<br />
In multiple series, 34 to 48% <strong>of</strong> women who underwent<br />
initial pulmonary metastasectomy developed recurrent lung<br />
metastasis and had repeat resection. 13-15 Sixteen to thirtyseven<br />
percent <strong>of</strong> those patients had a tertiary resection. 14,15<br />
Repeated metastasectomies correlate with improved survival,<br />
likely because tumor biology is more favorable in<br />
patients who survive long enough for more than one metastasectomy.<br />
16<br />
Extrathoracic Metastasectomy<br />
More recently, outcomes after extrathoracic metastasectomy<br />
have been evaluated in studies <strong>of</strong> recurrent sarcoma.<br />
13,16,17 For patients undergoing resection <strong>of</strong> recurrent<br />
uterine leiomyosarcoma, survival associated with thoracic<br />
metastasectomy is similar to that associated with nonthoracic<br />
procedures for recurrent disease, with a median<br />
disease-specific survival <strong>of</strong> nearly 4 years from the time <strong>of</strong><br />
first metastasectomy. 17 A long-term survival benefit may be<br />
conferred by metastasectomy for patients with extrapulmonary<br />
disease in the case <strong>of</strong> complete resection. These data<br />
suggest that the traditional criteria for pulmonary metasta-<br />
363
Fig. 1. PET CT <strong>of</strong> synchronous lung and acetabular metastatic<br />
lesions, managed surgically after an initial 7.5-year disease-free<br />
interval.<br />
Abbreviation: PET CT, positron emission tomography/computerized<br />
tomography.<br />
sectomy, listed earlier, should be expanded to encourage<br />
consideration <strong>of</strong> pulmonary metastasectomy for patients<br />
who have either synchronous or prior resectable metastases<br />
outside <strong>of</strong> the chest. 16 This point is illustrated by our case<br />
history. Figure 1 is a positron emission tomography/computerized<br />
tomography (CT) demonstrating the two sites <strong>of</strong><br />
recurrent disease at the time <strong>of</strong> the initial metastasectomy.<br />
Liver Metastasectomy<br />
Resection is frequently performed for liver metastases<br />
resulting from colorectal cancer. However, the role <strong>of</strong> surgery<br />
is less clear in the case <strong>of</strong> sarcomatous liver metastases.<br />
Due to the paucity <strong>of</strong> cases, few studies specifically address<br />
liver metastasectomy in uterine sarcoma. However, several<br />
small studies <strong>of</strong> heterogeneous patient populations with<br />
metastatic leiomyosarcoma have demonstrated that in appropriately<br />
chosen patients, liver resection for metastatic<br />
disease can prolong survival. 18,19 In one small series <strong>of</strong> 66<br />
patients who underwent resection, resection with radi<strong>of</strong>requency<br />
ablation, or radi<strong>of</strong>requency ablation alone, the median<br />
overall survival after the procedure was 47 months.<br />
Longer survival was associated with metastases 3 cm or<br />
smaller and with resection alone compared with radi<strong>of</strong>requency<br />
ablation with or without surgical resection. 19 A<br />
study <strong>of</strong> 11 patients demonstrated a median survival <strong>of</strong> 39<br />
months after resection and improved survival associated<br />
with complete resection <strong>of</strong> metastatic disease. 18<br />
Inferior Vena Cava Resection or Intracardiac<br />
Metastasectomy for Recurrent ESS<br />
Vascular extension is a common characteristic <strong>of</strong> ESS.<br />
Inferior vena cava tumor thrombus likely begins as tumor<br />
growth within the uterine or ovarian veins. 20 Multiple case<br />
364<br />
reports and small series have documented extensive resections<br />
for recurrent ESS with inferior vena cava and intracardiac<br />
extension. 21,22 Preoperatively, imaging studies such<br />
as CT, magnetic resonance imaging, and transesophageal<br />
ultrasound or echocardiography can delineate the extent <strong>of</strong><br />
disease. Depending on the findings <strong>of</strong> these studies, the<br />
surgery may be done by laparotomy, thoracotomy, or combined<br />
sternolaparotomy. In one series <strong>of</strong> 19 patients, cardiopulmonary<br />
bypass was required during seven procedures.<br />
When reconstruction <strong>of</strong> major vascular structures such as<br />
the inferior vena cava was necessary, xenopericardium and<br />
graft replacements were used. In this series, a radical<br />
resection resulting in complete tumor removal was possible<br />
in 10 patients, some <strong>of</strong> whom required concurrent surgical<br />
procedures for synchronous metastases, including pulmonary<br />
metastasectomy or pelvic extenteration. The median<br />
survival was 2 years, with a range <strong>of</strong> 0.3 to 4.5 years. 21<br />
Because <strong>of</strong> the overall excellent prognosis in completely<br />
resected ESS and the likelihood <strong>of</strong> imminent heart failure or<br />
pulmonary tumor embolism in women with intracaval or<br />
intracardiac extension, surgical excision is considered appropriate<br />
for well-selected patients. 22,23<br />
Resection <strong>of</strong> Recurrent Adenosarcoma<br />
Few data are available on resection for the treatment <strong>of</strong><br />
recurrent adenosarcoma. In one small study <strong>of</strong> 23 women, 24<br />
17 had resection <strong>of</strong> recurrence in the vagina, pelvis, or<br />
abdomen. Of these women, eight had a durable and possibly<br />
curative response, with disease-free intervals <strong>of</strong> 5 to 12<br />
years.<br />
Survival after Surgery for Recurrent Disease<br />
Survival after resection for recurrence <strong>of</strong> uterine sarcoma<br />
varies among studies. In studies that include patients with<br />
all types <strong>of</strong> s<strong>of</strong>t tissue sarcoma, 15,25 the median overall<br />
survival as well as disease-specific survival after metastasectomy<br />
is poorer than in studies <strong>of</strong> more homogenous<br />
populations <strong>of</strong> patients with gynecologic sarcomas. 14,17,26<br />
This finding may reflect more aggressive tumor biology in<br />
patients with nongynecologic sarcomatous metastases.<br />
In studies <strong>of</strong> patients with metastatic gynecologic sarcoma,<br />
the median survival after resection <strong>of</strong> metastatic<br />
disease ranges from 24 to 31 months, similar to the survival<br />
for the patient described in our case history. 9,10 In one study,<br />
disease-free survival after surgical resection <strong>of</strong> first recurrence<br />
was nearly 4 years. 17 Five and 10-year survival ranges<br />
from 38% to 47% and 34% to 35%, respectively. 12,14,26<br />
Studies addressing outcomes after surgical resection <strong>of</strong> recurrent<br />
uterine sarcoma are summarized in Table 1. Given<br />
these favorable survival data, it is clear that in the appropriately<br />
selected patient, surgical resection <strong>of</strong> metastatic<br />
disease may prolong disease-free and overall survival.<br />
Characteristics Correlating with Improved Outcome<br />
after Metastasectomy<br />
Characteristics associated with improved survival after<br />
resection <strong>of</strong> recurrent disease have been retrospectively<br />
assessed in multiple studies. Factors that have been assessed<br />
for correlation with improved outcome after metastasectomy<br />
include complete response after upfront<br />
treatment, disease-free interval between primary diagnosis<br />
and surgery for first recurrence, presence <strong>of</strong> residual disease
SURGICAL OPTIONS FOR UTERINE SARCOMAS<br />
Study<br />
after metastasectomy, number <strong>of</strong> metastases, lesion size,<br />
bilaterality <strong>of</strong> pulmonary lesions, and tumor doubling time.<br />
In one study, data on patients who underwent thoracic<br />
procedures for recurrent disease were compared with data<br />
on patients undergoing extrathoracic surgery, and no survival<br />
difference was noted between the two groups. 17<br />
A disease-free interval longer than 6 to 12 months between<br />
initial diagnosis and recurrence may indicate less<br />
aggressive tumor biology and correlates with improved survival<br />
after metastasectomy. 4,10-13,16,17,26 One study demonstrated<br />
a 5-year survival <strong>of</strong> 60% compared with 37% with a<br />
disease-free interval longer than 12 months. Another study<br />
resulted in a marked difference <strong>of</strong> 5.1 compared with 1.5<br />
years mean survival after surgery for recurrence in women<br />
with a disease-free interval <strong>of</strong> more than 12 months. 17<br />
Another study showed a demonstrable survival improvement<br />
with each month increase in disease-free interval. 16<br />
Other factors that have been associated with survival<br />
after surgery for recurrent disease include lesion size, history<br />
<strong>of</strong> prior metastasectomy, number <strong>of</strong> metastases, and<br />
bilaterality <strong>of</strong> lesions. 13,14 The utility <strong>of</strong> these prognostic<br />
indicators may be as markers for resectability <strong>of</strong> the recurrent<br />
tumor. 10 Presence <strong>of</strong> residual disease after resection for<br />
recurrence has been shown by multiple investigators to be<br />
associated with a considerably poorer prognosis and shorter<br />
disease-free survival. 10,15 One study demonstrated a<br />
marked difference in median survival (3.9 vs. 0.7 years after<br />
surgery) according to the completeness <strong>of</strong> resection for<br />
recurrence. 17<br />
In a study <strong>of</strong> patients undergoing repeat pulmonary me-<br />
Authors’ Disclosures <strong>of</strong> Potential Conflicts <strong>of</strong> Interest<br />
Author<br />
John P. Curtin*<br />
Sharmilee B. Korets*<br />
*No relevant relationships to disclose.<br />
Employment or<br />
Leadership<br />
Positions<br />
Table 1. Outcomes <strong>of</strong> Surgery for Recurrent Uterine Sarcoma<br />
Total No. <strong>of</strong> Pts.<br />
(No. with Uterine<br />
Sarcoma) Histology Recurrence Site Outcome<br />
Anderson et al (2001) 11 19 (12) LMS, ESS, nonsarcomatous uterine<br />
malignancies<br />
Anraku et al (2004) 12 133 (11) LMS, cervical cancer, uterine<br />
choriocarcinoma, uterine<br />
adenocarcinoma<br />
Bernstein-Molho et al (2010) 9 33 (33) LMS Lung, pelvis, abdomen, bone,<br />
retroperitoneum, brain,<br />
liver, adrenal glands<br />
Clavero et al (2006) 26 70 (41) LMS, ESS, adenocarcinoma,<br />
“other” sarcoma, SCCA,<br />
choriocarcinoma<br />
Consultant or<br />
Advisory Role<br />
tastasectomy for all s<strong>of</strong>t tissue sarcoma, multivariate analysis<br />
showed that incomplete resection, resection <strong>of</strong> more<br />
than two nodules, resection <strong>of</strong> nodules larger than 2 cm, and<br />
high grade <strong>of</strong> primary tumor were meaningful factors associated<br />
with poorer survival. The 5-year survival rate was<br />
100% for patients who had none <strong>of</strong> these poor prognostic<br />
factors and was 10% for patients who had three factors. 15<br />
Of note, all but one <strong>of</strong> these four characteristics would be<br />
known preoperatively, allowing for careful assessment <strong>of</strong><br />
the potential benefit <strong>of</strong> surgical resection in the case <strong>of</strong><br />
recurrent disease.<br />
In two studies in which the use <strong>of</strong> adjuvant therapy after<br />
surgery for recurrent uterine sarcoma was evaluated, postoperative<br />
chemotherapy or radiation therapy did not improve<br />
outcomes. 14,17<br />
Conclusion<br />
Women with recurrent uterine sarcoma have an overall<br />
poor prognosis. Surgical resection <strong>of</strong> metastases, in a wellselected<br />
patient, <strong>of</strong>fers the possibility <strong>of</strong> cure and improved<br />
disease-specific survival compared with other treatment<br />
modalities. Prolonged disease-free interval and resectability<br />
<strong>of</strong> tumor have been identified as favorable prognostic characteristics<br />
in multiple studies. Many characteristics correlating<br />
with optimal tumor resection, and therefore improved<br />
outcome, are evaluable preoperatively. Thus, with appropriate<br />
preoperative assessment and risk stratification, patients<br />
with recurrent uterine sarcoma who are medically stable<br />
may be considered candidates for surgical resection for the<br />
management <strong>of</strong> recurrent disease.<br />
Stock<br />
Ownership Honoraria<br />
Lung Median survival, 25 mo for patients<br />
with recurrent uterine LMS<br />
Lung 5-yr survival, 37.9% for patients with<br />
recurrent LMS<br />
Research<br />
Funding<br />
Median PFS, 7.9 mo; OS, 45 mo<br />
Lung 5-yr survival, 46.8%; 10-yr survival,<br />
34.3% in all histologic subgroups<br />
Giuntoli et al (2007) 10 128 (128) LMS Pelvis, abdomen, thoracic cavity Median survival, 2.0 yr vs. 1.1 yr in<br />
metastasectomy vs. chemotherapy<br />
and/or RT group<br />
Leitao et al (2002) 17 41 (41) LMS Pelvis, lung, abdomen, bone,<br />
multiple synchronous sites<br />
Median DSS, 3.9 yr; 2-yr survival, 71%<br />
Levenback et al (1992) 14 45 (45) LMS, ESS, carcinosarcoma (7%) Lung 5-yr survival 43%; 10-yr survival, 35%<br />
Abbreviations: LMS, leiomyosarcoma; ESS, endometrial stromal sarcoma; PFS, progression-free survival; OS, overall survival; SCCA, squamous cell carcinoma; RT,<br />
radiation therapy; DSS, disease-specific survival.<br />
Expert<br />
Testimony<br />
Other<br />
Remuneration<br />
365
1. Gadducci A, Cosio S, Romanini A, et al. The management <strong>of</strong> patients<br />
with uterine sarcoma: a debated clinical challenge. Crit Rev Oncol Hematol.<br />
2008;65:129-142.<br />
2. Barakat RR, Markman M, Randall M. Principles and Practice <strong>of</strong> Gynecologic<br />
<strong>Oncology</strong>. 5th ed. Philadelphia: Wolters Kluwer Health/Lippincott<br />
Williams & Wilkins; 2009.<br />
3. Seddon BM, Davda R. Uterine sarcomas-recent progress and future<br />
challenges. Eur J Radiol. 2011;78:30-40.<br />
4. Amant F, Coosemans A, Debiec-Rychter M, et al. <strong>Clinical</strong> management<br />
<strong>of</strong> uterine sarcomas. Lancet Oncol. 2009;10:1188-1198.<br />
5. Omura GA, Major FJ, Blessing JA, et al. A randomized study <strong>of</strong><br />
adriamycin with and without dimethyl triazenoimidazole carboxamide in<br />
advanced uterine sarcomas. Cancer. 1983;52:626-632.<br />
6. Sutton G, Blessing JA, Malfetano JH. Ifosfamide and doxorubicin in the<br />
treatment <strong>of</strong> advanced leiomyosarcomas <strong>of</strong> the uterus: a Gynecologic <strong>Oncology</strong><br />
Group study. Gynecol Oncol. 1996;62:226-229.<br />
7. Hensley ML, Maki R, Venkatraman E, et al. Gemcitabine and docetaxel<br />
in patients with unresectable leiomyosarcoma: results <strong>of</strong> a phase II trial.<br />
J Clin Oncol. 2002;20:2824-2831.<br />
8. Nam JH. Surgical treatment <strong>of</strong> uterine sarcoma. Best Pract Res Clin<br />
Obstet Gynaecol. 2011;25:751-760.<br />
9. Bernstein-Molho R, Grisaro D, Soyfer V, et al. Metastatic uterine<br />
leiomyosarcomas: a single-institution experience. Int J Gynecol Cancer.<br />
2010;20:255-260.<br />
10. Giuntoli RL 2nd, Garrett-Mayer E, Bristow RE, et al. Secondary<br />
cytoreduction in the management <strong>of</strong> recurrent uterine leiomyosarcoma.<br />
Gynecol Oncol. 2007;106:82-88.<br />
11. Anderson TM, McMahon JJ, Nwogu CE, et al. Pulmonary resection in<br />
metastatic uterine and cervical malignancies. Gynecol Oncol. 2001;83:472-<br />
476.<br />
12. Anraku M, Yokoi K, Nakagawa K, et al. Pulmonary metastases from<br />
uterine malignancies: results <strong>of</strong> surgical resection in 133 patients. J Thorac<br />
Cardiovasc Surg. 2004;127:1107-1112.<br />
13. Burt BM, Ocejo S, Mery CM, et al. Repeated and aggressive pulmonary<br />
resections for leiomyosarcoma metastases extends survival. Ann Thorac<br />
Surg. 2011;92:1202-1207.<br />
14. Levenback C, Rubin SC, McCormack PM, et al. Resection <strong>of</strong> pulmonary<br />
metastases from uterine sarcomas. Gynecol Oncol. 1992;45:202-205.<br />
366<br />
REFERENCES<br />
15. Weiser MR, Downey RJ, Leung DH, et al. Repeat resection <strong>of</strong> pulmonary<br />
metastases in patients with s<strong>of</strong>t-tissue sarcoma. J Am Coll Surg.<br />
2000;191:184-190, discussion 190-181.<br />
16. Blackmon SH, Shah N, Roth JA, et al. Resection <strong>of</strong> pulmonary and<br />
extrapulmonary sarcomatous metastases is associated with long-term survival.<br />
Ann Thorac Surg. 2009;88:877-884, discussion 884-875.<br />
17. Leitao MM, Brennan MF, Hensley M, et al. Surgical resection <strong>of</strong><br />
pulmonary and extrapulmonary recurrences <strong>of</strong> uterine leiomyosarcoma.<br />
Gynecol Oncol. 2002;87:287-294.<br />
18. Chen H, Pruitt A, Nicol TL, et al. Complete hepatic resection <strong>of</strong><br />
metastases from leiomyosarcoma prolongs survival. J Gastrointest Surg.<br />
1998;2:151-155.<br />
19. Pawlik TM, Vauthey JN, Abdalla EK, et al. Results <strong>of</strong> a single-center<br />
experience with resection and ablation for sarcoma metastatic to the liver.<br />
Arch Surg. 2006;141:537-543, discussion 543-534.<br />
20. Montag TW, Manart FD. Endolymphatic stromal myosis: surgical and<br />
hormonal therapy for extensive venous recurrence. Gynecol Oncol. 1989;33:<br />
255-260.<br />
21. Renzulli P, Weimann R, Barras JP, et al. Low-grade endometrial<br />
stromal sarcoma with inferior vena cava tumor thrombus and intracardiac<br />
extension: radical resection may improve recurrence free survival. Surg<br />
Oncol. 2009;18:57-64.<br />
22. Veroux P, Veroux M, Nicosia A, et al. Thrombectomy <strong>of</strong> the inferior<br />
vena cava from recurrent low-grade endometrial stromal sarcoma: case report<br />
and review <strong>of</strong> the literature. J Surg Oncol. 2000;74:45-48.<br />
23. Yokoyama Y, Ono Y, Sakamoto T, et al. Asymptomatic intracardiac<br />
metastasis from a low-grade endometrial stromal sarcoma with successful<br />
surgical resection. Gynecol Oncol. 2004;92:999-1001.<br />
24. Clement PB, Scully RE. Mullerian adenosarcoma <strong>of</strong> the uterus: a<br />
clinicopathologic analysis <strong>of</strong> 100 cases with a review <strong>of</strong> the literature. Hum<br />
Pathol. 1990;21:363-381.<br />
25. Smith R, Pak Y, Kraybill W, et al. Factors associated with actual<br />
long-term survival following s<strong>of</strong>t tissue sarcoma pulmonary metastasectomy.<br />
Eur J Surg Oncol. 2009;35:356-361.<br />
26. Clavero JM, Deschamps C, Cassivi SD, et al. Gynecologic cancers:<br />
factors affecting survival after pulmonary metastasectomy. Ann Thorac Surg.<br />
2006;81:2004-2007.
PATIENTS WITH HPV-POSITIVE OROPHARYNX<br />
TUMORS: COMMUNICATING DIAGNOSIS TO<br />
FAMILY AND TREATMENT OPTIONS<br />
CHAIR<br />
David Brizel, MD<br />
Duke University Medical Center<br />
Durham, NC<br />
SPEAKERS<br />
William H. Westra, MD<br />
Johns Hopkins School <strong>of</strong> Medicine<br />
Baltimore, MD<br />
Maura L. Gillison, MD, PhD<br />
The Ohio State University<br />
Columbus, OH
Management <strong>of</strong> Human Papillomavirus–Induced<br />
Oropharynx Cancer<br />
Overview: Oropharynx cancer (OPC) constitutes the most<br />
common location for squamous-cell head and neck cancer,<br />
and most OPC is caused by the human papilloma virus (HPV).<br />
Early-stage (<strong>American</strong> Joint Committee on Cancer [AJCC]<br />
stage I and II) disease should be treated with single modality<br />
surgery or radiotherapy whenever possible. More advanced<br />
presentations generally require combined-modality therapy<br />
with various combinations <strong>of</strong> surgery, radiotherapy, and che-<br />
THE VAST majority (� 90%) <strong>of</strong> squamous-cell head and<br />
neck cancers (HNCs) that are induced by the human<br />
papillomavirus (HPV) originate within the oropharynx. This<br />
anatomic region comprises the tonsils, s<strong>of</strong>t palate, and base<br />
<strong>of</strong> tongue. Understanding the general management principles<br />
<strong>of</strong> oropharynx cancer (OPC) is central to comprehending<br />
the development <strong>of</strong> new therapeutic options for HPVpositive<br />
OPC. The first step in this process is to recognize<br />
that OPC can be clinically divided into early-stage and<br />
advanced-stage presentations. Early-stage presentations<br />
typically include T1N0 and T2N0 presentations (stage I and<br />
II), whereas advanced-stage disease is more heterogeneous,<br />
with presentations ranging from T1N1 to T4N3 (stage III to<br />
IVB).<br />
The fundamental management tenet for early-stage disease<br />
is to adopt a strategy that maximizes the likelihood<br />
that the disease can be treated with single-modality therapy—either<br />
surgery or radiotherapy (RT) alone. These two<br />
modalities have equivalent efficacy in the early-stage setting.<br />
The rationale for the use <strong>of</strong> a single modality is to<br />
minimize morbidity, which invariably increases in step with<br />
the use <strong>of</strong> multiple therapeutic modalities. Surgical management<br />
consists <strong>of</strong> resection <strong>of</strong> the primary site, most <strong>of</strong>ten<br />
with an ipsilateral neck dissection, whereas definitive RT<br />
treats the same primary tumor site and ipsilateral lymph<br />
nodes as surgery. The advent <strong>of</strong> minimally invasive surgical<br />
techniques including transoral robotic resection and transoral<br />
laser excision and the increasing adoption <strong>of</strong> intensitymodulated<br />
RT (IMRT) have all led to substantial reductions<br />
in the long-term functional morbidity <strong>of</strong> treating early-stage<br />
OPC. The National Comprehensive Cancer Center Network<br />
(NCCN) Head and Neck Cancer Guidelines provide additional<br />
detail regarding the choices for radiation fractionation<br />
schemes (www.nccn.org). Postoperative irradiation for<br />
early-stage disease should be used only on the basis <strong>of</strong><br />
adverse histopathology including perineural invasion, positive<br />
margins, the presence <strong>of</strong> multiply involved lymph nodes,<br />
or the presence <strong>of</strong> extracapsular nodal extension. Conversely,<br />
when RT is used as the primary modality, surgery<br />
should be used only for the presence <strong>of</strong> persistent disease or<br />
for the salvage <strong>of</strong> recurrent disease. Primary RT is usually<br />
preferred for tumors originating in the base <strong>of</strong> tongue or<br />
extending from the tonsil onto the s<strong>of</strong>t palate close to the<br />
midline because <strong>of</strong> the higher risk <strong>of</strong> occult contralateral<br />
lymph node involvement, which necessitates treatment <strong>of</strong><br />
both sides <strong>of</strong> the neck. These regions can usually be treated<br />
with less long-term functional morbidity by using RT as<br />
opposed to bilateral neck dissection.<br />
368<br />
By David Brizel, MD<br />
motherapy or molecularly targeted therapy. All <strong>of</strong> these approaches<br />
expose patients to a substantial risk <strong>of</strong> serious<br />
long-term functional morbidity. HPV-induced OPC has a very<br />
favorable prognosis compared with its HPV-negative counterpart<br />
irrespective <strong>of</strong> the treatment platform that is used.<br />
Current clinical trials are investigating the concept <strong>of</strong> therapeutic<br />
deintensification with the dual objectives <strong>of</strong> decreasing<br />
toxicity and maintaining efficacy.<br />
T1N1 and T2N1 presentations are classified as stage III<br />
and technically advanced disease, but the management<br />
principles for stage I/II presentations are still generally<br />
applicable, namely treatment with primary surgery or RT as<br />
single modality with an understanding, however, that the<br />
larger burden <strong>of</strong> disease increases the probability that the<br />
addition <strong>of</strong> adjuvant neck dissection or postoperative irradiation<br />
will be necessary. Traditionally, primary nonsurgical<br />
management has assumed the most prominent role for the<br />
management <strong>of</strong> more extensive stage III and IV presentations,<br />
with surgery being used more in an adjuvant setting<br />
(neck dissection or salvage role for residual/recurrent disease<br />
at the primary site). Three different strategies can<br />
be considered as standards <strong>of</strong> care for management <strong>of</strong><br />
advanced-stage disease: RT alone using modified fractionation,<br />
RT plus concurrent epidermal growth factor receptor<br />
(EGFR) blockade, and RT and concurrent chemotherapy.<br />
Randomized trials have confirmed improvements in locoregional<br />
control with the use <strong>of</strong> both hyperfractionation<br />
and accelerated fractionation by means <strong>of</strong> improvement in<br />
locoregional disease control. 1-3 A recent meta-analysis from<br />
Bourhis and colleagues evaluated updated individual patient’s<br />
data from 6,515 patients in 15 randomized trials <strong>of</strong><br />
patients with stage III/IV oropharynx and larynx carcinoma.<br />
4 It showed an 8.2% absolute difference in 5-year<br />
survival for patients who received altered fractionation RT<br />
compared with conventional once-daily RT (36.7% vs. 28.5%,<br />
respectively).<br />
EGFR is overexpressed in the majority <strong>of</strong> patients with<br />
head and neck squamous-cell carcinoma. Bonner and colleagues<br />
conducted a prospective randomized trial that<br />
tested the value <strong>of</strong> adding EGFR blockade to a course <strong>of</strong><br />
RT. 5,6 Four hundred twenty-four patients were randomly<br />
assigned to receive either RT alone or RT plus weekly<br />
cetuximab, a chimeric monoclonal antibody to the EGFR<br />
receptor. The majority <strong>of</strong> these patients had oropharynx<br />
primary tumors, and 75% <strong>of</strong> them received treatment with<br />
accelerated or hyperfractionated irradiation. The RT/cetuximab<br />
patients had significant improvements in median survival<br />
(49 months vs. 29 months; hazard ratio [HR] � 0.73;<br />
p � 0.02), and 5-year overall survival (46% vs. 36%). Subset<br />
From the Duke University Cancer Institute, Durham, NC.<br />
Author’s disclosure <strong>of</strong> potential conflicts <strong>of</strong> interest are found at the end <strong>of</strong> this article.<br />
Address reprint requests to David Brizel, MD, Duke University Cancer Institute, Box 3085<br />
DUMC, Durham, NC 27710; email: david.brizel@duke.edu.<br />
© <strong>2012</strong> by <strong>American</strong> <strong>Society</strong> <strong>of</strong> <strong>Clinical</strong> <strong>Oncology</strong>.<br />
1092-9118/10/1-10
HPV-INDUCED OROPHARYNX CANCER<br />
analysis demonstrated the most pronounced therapeutic<br />
benefit in patients with oropharynx primaries, N� presentations,<br />
and treatment with accelerated fractionation.<br />
Concurrent chemoradiation represents the largest experience<br />
in the nonsurgical management <strong>of</strong> advanced HNC.<br />
Pignon and colleagues performed an extensive metaanalysis<br />
<strong>of</strong> individual patient data from 17,346 patients<br />
enrolled on 93 randomized controlled trials comparing RT<br />
alone against RT and chemotherapy. 7,8 These trials were<br />
published from 1965 to 2000, and it is important to recognize<br />
that this era was before the use <strong>of</strong> taxane-based chemotherapy<br />
in HNC. Their analysis demonstrated a 21% reduction<br />
in the risk <strong>of</strong> recurrence or death as a result <strong>of</strong> the addition<br />
<strong>of</strong> concurrent chemotherapy to RT. This risk reduction<br />
corresponded to a 6.5% absolute improvement in 5-year<br />
survival (33.7 vs. 27.2%). The majority <strong>of</strong> this improvement<br />
was attributable to a 13% absolute improvement in locoregional<br />
disease control. A more modest 2.9% absolute<br />
improvement in distant failure was attributable to concurrent<br />
therapy. The analysis also showed that the use <strong>of</strong><br />
concurrent cisplatin was most important. Induction chemotherapy<br />
did not provide an improvement in locoregional<br />
disease control.<br />
The role <strong>of</strong> induction chemotherapy continues to be investigated.<br />
Several phase III trials have confirmed that the<br />
addition <strong>of</strong> a taxane to a platinum and flurorouracil (FU)–<br />
based induction regimen yields superior survival compared<br />
with induction with platinum and FU only. Most <strong>of</strong> these<br />
trials administered substandard locoregional therapy, however.<br />
9-11 Whether taxane-based induction followed by state<strong>of</strong>-the-art<br />
concurrent chemoradiation is superior to concurrent<br />
chemoradiation alone, however, remains unknown.<br />
Toxicity considerations must be incorporated into the<br />
treatment decision-making process for HNC in general and<br />
OPC in particular. The standard investigational treatment<br />
paradigm has been one <strong>of</strong> therapeutic intensification via the<br />
combination <strong>of</strong> multiple modalities with increasing disease<br />
stage and tumor burden. Concurrent chemoradiation programs<br />
do improve overall efficacy, but at the price <strong>of</strong> a<br />
considerable increase in toxicity. The incidence <strong>of</strong> grade 3<br />
confluent mucositis approximately doubles with the addi-<br />
KEY POINTS<br />
● The incidence <strong>of</strong> oropharynx cancer is increasing even<br />
as the incidence <strong>of</strong> head and neck cancer (HNC) in<br />
other locations is decreasing.<br />
● This increase is attributable to the human papillomavirus<br />
(HPV), which causes the majority <strong>of</strong> cases <strong>of</strong><br />
oropharynx cancer.<br />
● The prognosis <strong>of</strong> HPV-associated HNC is much more<br />
favorable than that <strong>of</strong> HPV-negative HNC.<br />
● The favorable prognostic effect <strong>of</strong> HPV positivity is<br />
independent <strong>of</strong> treatment platform, although a history<br />
<strong>of</strong> cigarette smoking appears to negate some <strong>of</strong><br />
the benefit.<br />
● Trials under development for HPV-positive HNC will<br />
explore the concept <strong>of</strong> therapeutic de-escalation designed<br />
to reduce treatment-associated morbidity with<br />
preservation <strong>of</strong> thereapeutic efficacy.<br />
tion <strong>of</strong> concurrent chemotherapy to RT-alone regimens. The<br />
percentage <strong>of</strong> patients experiencing serious late toxicities<br />
including cervical fibrosis, dental disease, and neuropathy<br />
increases by approximately 50%. 12 The addition <strong>of</strong> neck<br />
dissection after concurrent chemoradiation also substantially<br />
increases the risk <strong>of</strong> late toxicity after concurrent<br />
chemoradiation. 13 Recent developments in the use <strong>of</strong> postchemoradiation<br />
computed tomography and positron emission<br />
tomography scanning, however, have substantially<br />
improved the ability to separate those patients who do<br />
require adjuvant neck dissection from those who do not. 14<br />
Bentzen and Trotti analyzed the Radiation Therapy <strong>Oncology</strong><br />
Group (RTOG) portfolio <strong>of</strong> randomized trials in HNC<br />
that have either compared different RT schedules against<br />
one another or compared RT and chemotherapy against RT<br />
alone. 15 They showed that both induction chemotherapy and<br />
concurrent chemotherapy approaches increased the overall<br />
burden <strong>of</strong> severe toxicity on the patient by five-fold compared<br />
with conventionally fractionated irradiation. Moreover,<br />
their data clearly demonstrated that a single patient<br />
can have more than one severe toxicity event, something<br />
that is not routinely captured in conventional toxicity scoring<br />
systems.<br />
The toxicity <strong>of</strong> a given treatment strategy can also be<br />
judged by the probability that a patient will be able to<br />
complete the entire prescribed course <strong>of</strong> treatment. Sequential<br />
chemoradiation regimens (induction chemotherapy followed<br />
by concurrent chemoradiotherapy) sound a cautionary<br />
note in this regard. Compliance rates with an entire course<br />
<strong>of</strong> treatment in the published clinical trials range from<br />
only 50% to 73%. 10,16 Sequential therapy regimens that use<br />
state-<strong>of</strong>-the-art concurrent chemoradiation schemes appear<br />
to be particularly problematic from a compliance standpoint.<br />
9,17<br />
A recent evaluation <strong>of</strong> tissue obtained through the Surveillance,<br />
Epidemiology and End Results (SEER) registries<br />
<strong>of</strong> patients with OPC receiving treatment from 1984 to 2004<br />
shows an increasing incidence <strong>of</strong> OPC even as the overall<br />
incidence <strong>of</strong> HNC is declining. 18 OPC now constitutes the<br />
most prevalent subtype <strong>of</strong> HNC within the United States.<br />
Moreover, the majority (approximately 70%) <strong>of</strong> these OPCs<br />
are HPV positive. Current estimations are that by the year<br />
2020 the annual number <strong>of</strong> HPV-positive OPCs will exceed<br />
the annual number <strong>of</strong> HPV-positive uterine cervix carcinomas.<br />
Additional projections are that there will be more than<br />
15,000 cases <strong>of</strong> HPV-positive OPC per year by 2030.<br />
The SEER analysis included patients who received various<br />
types <strong>of</strong> treatment. The overall 5- and 10-year survivals<br />
were approximately two times better for those patients with<br />
HPV-positive disease. HPV-positive disease appears to convey<br />
this more favorable prognosis in a platform-independent<br />
fashion. RTOG-0129 was a phase III trial that compared<br />
standard fractionation RT against accelerated fractionation<br />
RT, with both arms receiving concurrent cisplatin chemotherapy.<br />
19 The accelerated fractionation arm received two<br />
cycles, and the standard fractionation arm received three<br />
cycles. There was no difference in overall outcome. Therefore,<br />
the patients from the two arms were pooled, and<br />
outcomes were reanalyzed as a function <strong>of</strong> HPV positivity as<br />
determined by p16 immunohistochemistry. The probability<br />
<strong>of</strong> long-term survival was more than three times greater for<br />
HPV-positive patients than for their HPV-negative counterparts<br />
(HR � 0.29; p � 0.001). A history <strong>of</strong> cigarette smoking<br />
369
counteracted the favorable effect <strong>of</strong> HPV positivity, however.<br />
HPV-positive patients who received sequential chemoradiation<br />
on the Tax-324 trial also had an approximate three-fold<br />
improvement in overall survival at 5 years compared with<br />
the HPV-negative patients. 20 Patients who are treated<br />
with RT alone also have a two- to threefold more favorable<br />
prognosis than similarly staged patients with HPV-negative<br />
disease. 21 A single-arm trial <strong>of</strong> transoral laser excision <strong>of</strong><br />
OPC has also demonstrated a more favorable prognosis for<br />
HPV-positive patients. 22<br />
The vastly different prognosis <strong>of</strong> OPC according to HPV<br />
status is now guiding the development <strong>of</strong> new therapeutic<br />
strategies. Long-term survival for HPV-negative patients<br />
stubbornly remains in the 30% to 40% range, which is<br />
clearly suboptimal. Therapeutic intensification with the<br />
objective <strong>of</strong> improving efficacy is still appropriate for this<br />
group <strong>of</strong> patients.<br />
The one-size-fits-all intensification approach, with its associated<br />
acute and long-term morbidity, is no longer appropriate<br />
for HPV-positive patients because long-term survival<br />
for these patients is in the 70% to 85% range. Most investigational<br />
efforts for this group are now aimed toward therapeutic<br />
de-escalation with dual objectives <strong>of</strong> reducing toxicity<br />
and maintaining overall therapeutic efficacy. The favorable<br />
effect <strong>of</strong> HPV positivity is equal for multiple treatment<br />
platforms from a relative standpoint, but it is unknown<br />
which, if any, <strong>of</strong> these platforms is superior to the others in<br />
absolute terms.<br />
Several rational strategies exist for treatment de-escalation.<br />
These include the use <strong>of</strong> RT alone instead <strong>of</strong> concurrent<br />
chemoradiation, the use <strong>of</strong> RT and EGFR inhibition<br />
instead <strong>of</strong> concurrent chemoradiation, the use <strong>of</strong> induction<br />
chemotherapy followed by reduced dose (chemo)RT for favorable<br />
responses, and primary surgery instead <strong>of</strong> RT or<br />
chemoradiation.<br />
The Eastern Cooperative <strong>Oncology</strong> Group recently completed<br />
enrollment onto a phase II trial (ECOG 1308; www.<br />
clinicaltrials.gov; NCT01084083) in which 83 patients with<br />
stage III-IVB HPV-positive OPC received three cycles <strong>of</strong><br />
induction chemotherapy consisting <strong>of</strong> paclitaxel, cisplatin,<br />
and cetuximab. Those patients who had a clinical and<br />
radiographic complete response (CR) then received reduceddose<br />
IMRT to 54 Gy via conventional fractionation and<br />
weekly cetxuimab. Those who had less than CR received<br />
accelerated fractionation IMRT (69.3 Gy/33 fractions). Results<br />
<strong>of</strong> this study are pending.<br />
RTOG is currently conducting a phase III trial in patients<br />
with HPV-positive (determined by p16 immunohistochemis-<br />
Author’s Disclosure <strong>of</strong> Potential Conflicts <strong>of</strong> Interest<br />
Author<br />
Employment or<br />
Leadership<br />
Positions<br />
Consultant or<br />
Advisory Role<br />
David Brizel NCI (L) Bristol-Myers<br />
Squibb; Siemens<br />
Molecular<br />
<strong>Oncology</strong><br />
1. Horiot JC, Le Fur R, N�Guyen T, et al. Hyperfractionation versus<br />
conventional fractionation in oropharyngeal carcinoma: final analysis <strong>of</strong> a<br />
randomized trial <strong>of</strong> the EORTC cooperative group <strong>of</strong> radiotherapy. Radiother<br />
Oncol. 1992;25:231-241.<br />
2. Fu KK, Pajak TF, Trotti A, et al. A Radiation Therapy <strong>Oncology</strong> Group<br />
370<br />
try) locally advanced OPC (RTOG 1016; www.clinicaltrials.<br />
gov; NCT01302834). Randomization is stratified by low<br />
and high T stage, low and high N stage, and smoking history<br />
(� or � 10 pack years). Both arms <strong>of</strong> this trial will use<br />
accelerated fractionation IMRT (70 Gy in 6 weeks). The<br />
control arm will receive two cycles <strong>of</strong> cisplatin, and the<br />
experimental arm will receive weekly cetuximab. This trial<br />
will enroll 700 patients overall.<br />
The National Cancer Institute (NCI) Head and Neck<br />
Cancer Steering Committee conducted a clinical trials planning<br />
meeting in November 2011 to formulate a concept that<br />
would be developed into a clinical trial designed to assess<br />
the role <strong>of</strong> transoral resection as a primary modality in the<br />
management <strong>of</strong> HPV-positive OPC. This concept is in the<br />
final stages <strong>of</strong> development, and the clinical trial will be<br />
conducted through the NCI Cooperative Group mechanism.<br />
Most likely, a risk-based approach will be tested. Specifically,<br />
patients with I-IIB tonsil cancer will undergo transoral<br />
resection with lymph node dissection. Those patients<br />
who are at low risk for recurrence on the basis <strong>of</strong> surgical<br />
pathology will be observed. Those patients who are at high<br />
risk for locoregional recurrence (positive margins and/or<br />
positive extracapsular extension [ECE]) will receive standard<br />
postoperative irradiation and concurrent cisplatin chemotherapy.<br />
Those patients with intermediate risk features<br />
such as positive nodes without ECE, perineural invasion, or<br />
angiolymphatic invasion, or high pathologic T stage will<br />
undergo a phase II random assignment between lowercompared<br />
with higher-dose postoperative irradiation. Prospective<br />
and standardized assessments <strong>of</strong> quality <strong>of</strong> life and<br />
patient outcomes will be an important part <strong>of</strong> the evaluation<br />
<strong>of</strong> the transoral resection strategy in addition to the standard<br />
end points <strong>of</strong> locoregional disease control, progressionfree<br />
survival, and overall survival. Furthermore, this<br />
strategy will allow for tissue banking to be performed for all<br />
patients for future evaluation <strong>of</strong> new biologic targets in the<br />
treatment <strong>of</strong> this disease.<br />
A note <strong>of</strong> caution is in order. Therapeutic deintensification<br />
for HPV-positive OPC carries the very real risk <strong>of</strong> a reduction<br />
in efficacy. Conversely, intensification potentially exposes<br />
HPV-negative patients to the risk <strong>of</strong> even more<br />
toxicity without a commensurate improvement in efficacy.<br />
Both de-escalation and escalation strategies should be undertaken<br />
only within the context <strong>of</strong> rigorously designed<br />
clinical trials that will be able to detect adverse outcomes<br />
should they exist. Outside <strong>of</strong> a clinical trial, both HPVpositive<br />
and HPV-negative disease should be managed according<br />
to best practices that are based on existing evidence.<br />
Stock<br />
Ownership Honoraria<br />
REFERENCES<br />
Research<br />
Funding<br />
Expert<br />
Testimony<br />
DAVID BRIZEL<br />
Other<br />
Remuneration<br />
(RTOG) phase III randomized study to compare hyperfractionation and two<br />
variants <strong>of</strong> accelerated fractionation to standard fractionation radiotherapy<br />
for head and neck squamous cell carcinomas: first report <strong>of</strong> RTOG 9003. Int J<br />
Radiat Oncol Biol Phys. 2000;48:7-16.<br />
3. Overgaard J, Hansen HS, Specht L, et al. Five compared with six
HPV-INDUCED OROPHARYNX CANCER<br />
fractions per week <strong>of</strong> conventional radiotherapy <strong>of</strong> squamous-cell carcinoma<br />
<strong>of</strong> head and neck: DAHANCA 6 and 7 randomised controlled trial. Lancet.<br />
2003;362:933-940.<br />
4. Bourhis J, Overgaard J, Audry H, et al. Hyperfractionated or accelerated<br />
radiotherapy in head and neck cancer: a meta-analysis. Lancet. 2006;368:843-<br />
854.<br />
5. Bonner JA, Harari PM, Giralt J, et al. Radiotherapy plus cetuximab for<br />
squamous-cell carcinoma <strong>of</strong> the head and neck. N Engl J Med. 2006;354:567-<br />
578.<br />
6. Bonner JA, Harari PM, Giralt J, et al. Radiotherapy plus cetuximab for<br />
locoregionally advanced head and neck cancer: 5-year survival data from a<br />
phase 3 randomised trial, and relation between cetuximab-induced rash and<br />
survival. Lancet Oncol. 2010;11:21-28.<br />
7. Pignon JP, Bourhis J, Domenge C, Designe L. Chemotherapy added to<br />
locoregional treatment for head and neck squamous-cell carcinoma: Three<br />
meta-analyses <strong>of</strong> updated individual data: MACH-NC Collaborative Group<br />
Meta-Analysis <strong>of</strong> Chemotherapy on Head and Neck Cancer. Lancet. 2000;355:<br />
949-955.<br />
8. Pignon JP, le Maître A, Maillard E, Bourhis J. Meta-analysis <strong>of</strong> chemotherapy<br />
in head and neck cancer (MACH-NC): an update on 93 randomised<br />
trials and 17,346 patients. Radiother Oncol. 2009;92:4-14.<br />
9. Hitt R, López-Pousa A, Martínez-Trufero J, et al. Phase III study<br />
comparing cisplatin plus fluorouracil to paclitaxel, cisplatin, and fluorouracil<br />
induction chemotherapy followed by chemoradiotherapy in locally advanced<br />
head and neck cancer. J Clin Oncol. 2005;23:8636-8645.<br />
10. Posner MR, Hershock DM, Blajman CR, et al. Cisplatin and fluorouracil<br />
alone or with docetaxel in head and neck cancer. New Engl J Med..<br />
2007;357:1705-1715.<br />
11. Vermorken JB, Remenar E, van Herpen C, et al. Standard cisplatin/<br />
infusional 5-fluorouracil (PF) vs docetaxel (T) plus PF (TPF) as neoadjuvant<br />
chemotherapy for nonresectable locally advanced squamous cell carcinoma <strong>of</strong><br />
the head andneck (LA-SCCHN): a phase III trial <strong>of</strong> the EORTC Head and<br />
Neck Cancer Group (EORTC #24971). J Clin Oncol. 2004;22:490s (suppl;<br />
abstr 5508).<br />
12. Denis F, Garaud P, Bardet E, et al. Late toxicity results <strong>of</strong> the GORTEC<br />
94-01 randomized trial comparing radiotherapy with concomitant radiochemotherapy<br />
for advanced-stage oropharynx carcinoma: comparison <strong>of</strong> LENT/<br />
SOMA, RTOG/EORTC, and NCI-CTC scoring systems. Int J Radiat Oncol<br />
Biol Phys. 2003;55:93-98.<br />
13. Machtay M, Moughan J, Trotti A, et al. Factors associated with severe<br />
late toxicity after concurrent chemoradiation for locally advanced head and<br />
neck cancer: an RTOG analysis. J Clin Oncol. 2008;26:3582-3589.<br />
14. Porceddu SV, Pryor DI, Burmeister E, et al. Results <strong>of</strong> a prospective<br />
study <strong>of</strong> positron emission tomography-directed management <strong>of</strong> residual<br />
nodal abnormalities in node-positive head and neck cancer after definitive<br />
radiotherapy with or without systemic therapy. Head Neck. 2011;33:1675-<br />
1682.<br />
15. Bentzen SM, Trotti A. Evaluation <strong>of</strong> early and late toxicities in<br />
chemoradiation trials. J Clin Oncol. 2007;25:4096-4103.<br />
16. Vermorken JB, Remenar E, van Herpen C, et al. Cisplatin, fluorouracil,<br />
and docetaxel in unresectable head and neck cancer. New Engl J Med.<br />
2007;357:1695-1704.<br />
17. Adelstein DJ, Moon J, Hanna E, et al. Docetaxel, cisplatin, and<br />
fluorouracil induction chemotherapy followed by accelerated fractionation/<br />
concomitant boost radiation and concurrent cisplatin in patients with advanced<br />
squamous cell head and neck cancer: A Southwest <strong>Oncology</strong> Group<br />
phase II trial (S0216). Head Neck. 2010;32:221-228.<br />
18. Chaturvedi AK, Engels EA, Pfeiffer RM, et al. Human papillomavirus<br />
and rising oropharyngeal cancer incidence in the United States. J Clin Oncol.<br />
2011;29:4294-4301.<br />
19. Ang KK, Zhang RH, Wheeler DI. A phase III trial (RTOG 0129) <strong>of</strong> two<br />
radiation-cisplatin regimens for head and neck carcinomas (HNC): impact <strong>of</strong><br />
radiation and cisplatin intensity on outcome. J Clin Oncol. 2010;28:15s(suppl;<br />
abstr 5507).<br />
20. Posner MR, Lorch JH, Goloubeva O, et al. Survival and human<br />
papillomavirus in oropharynx cancer in TAX 324: a subset analysis from an<br />
international phase III trial. Ann Oncol. 2011;22:1071-1077.<br />
21. Lassen P, Eriksen JG, Hamilton-Dutoit S, et al. Effect <strong>of</strong> HPVassociated<br />
p16INK4A expression on response to radiotherapy and survival in<br />
squamous cell carcinoma <strong>of</strong> the head and neck. J Clin Oncol 2009;27:1992-<br />
1998.<br />
22. Rich JT, Milov S, Lewis JS Jr.et al. Transoral laser microsurgery (TLM)<br />
�/� adjuvant therapy for advanced stage oropharyngeal cancer: outcomes<br />
and prognostic factors. Laryngoscope. 2009;119:1709-1719.<br />
371
TRANSLATING BIOLOGIC DISCOVERIES INTO<br />
CLINICAL TRIALS<br />
CHAIR<br />
Lisa F. Licitra, MD<br />
Fondazione IRCCS Istituto Nazionale dei Tumori<br />
Milan, Italy<br />
SPEAKERS<br />
Christine H. Chung, MD<br />
Johns Hopkins University<br />
Baltimore, MD<br />
Lillian L. Siu, MD<br />
Princess Margaret Hospital<br />
Toronto, ON, Canada
Important Early Advances in Squamous Cell<br />
Carcinoma <strong>of</strong> the Head and Neck<br />
By Eric Bissada, MD, DMD, Irene Brana, MD, and Lillian L. Siu, MD<br />
Overview: Therapeutic advances in squamous cell carcinomas<br />
<strong>of</strong> the head and neck (SCCHN) are attained by improvement<br />
in locoregional and/or distant disease control, as well as<br />
by reduction in treatment-related morbidity—especially longterm<br />
complications affecting normal organ functions and<br />
quality <strong>of</strong> life. New technological innovations in surgical<br />
management, such as transoral robotic surgery (TORS), and in<br />
radiotherapy (RT), such as proton and carbon ion therapy,<br />
bring promises <strong>of</strong> equal or superior efficacy outcomes coupled<br />
with the potential to minimize normal tissue toxicity. Scientific<br />
THE DELIVERY <strong>of</strong> precision medicine is relevant in the<br />
management <strong>of</strong> both early and advanced stages <strong>of</strong><br />
SCCHN, not only to improve functional outcomes and disease<br />
control, but also to reduce the burden <strong>of</strong> acute and<br />
long-term treatment-related morbidity. Technical precision<br />
may be enhanced surgically with the emergence <strong>of</strong> TORS,<br />
accompanied by the benefit <strong>of</strong> reduced manipulation <strong>of</strong><br />
surrounding normal structures. The use <strong>of</strong> energy particles<br />
such as protons or carbon ions <strong>of</strong>fers precise RT dose<br />
distributions, thus limiting undesirable scatter to nearby<br />
normal organs. In the systemic treatment <strong>of</strong> SCCHN, precision<br />
medicine is based on the principles that “druggable”<br />
molecular drivers <strong>of</strong> sensitivity or resistance exist in cancers,<br />
and that matching <strong>of</strong> pharmaceutical agents to specific<br />
oncogenic aberrations may improve therapeutic outcome.<br />
The evaluation <strong>of</strong> these multidisciplinary advances in SCCH<br />
through the conduct <strong>of</strong> properly designed clinical trials is<br />
necessary to support their application in practice.<br />
Novel Advances in Surgical Management <strong>of</strong> SCCHN<br />
The Emergence <strong>of</strong> TORS and Its Potential Benefits in SCCHN<br />
The use <strong>of</strong> robotic surgery is gaining popularity and is<br />
being applied to several fields including urology, orthopaedic<br />
surgery and cardiac surgery. The latest use <strong>of</strong> this technology<br />
is in otolaryngology in which some surgeons have found<br />
it helpful for the ablation <strong>of</strong> difficult-to-access tumors <strong>of</strong> the<br />
upper aerodigestive tract, particularly oropharyngeal cancers.<br />
TORS is a topic <strong>of</strong> great interest in head and neck<br />
surgery, and some surgeons believe its popularity will only<br />
increase in the coming years. In TORS, the surgeon sits at a<br />
remote console and controls micromanipulators that in turn<br />
move the arms <strong>of</strong> a robot placed at the patient’s bedside. A<br />
highly magnified three-dimensional view <strong>of</strong> the surgical field<br />
is procured with precise, scaled, and filtered motions to the<br />
operating arms. The proponents <strong>of</strong> TORS state several<br />
benefits over the open approach. The avoidance <strong>of</strong> a mandibulotomy<br />
and its associated morbidity is clearly the main<br />
advantage this technology <strong>of</strong>fers. Decreased manipulation<br />
and dissection <strong>of</strong> healthy tissue, improved cosmetic outcome,<br />
decreased need for tracheotomies, early return to oral intake,<br />
and shortened hospital stay are but a few other<br />
potential suggested benefits. 1,2<br />
At present, there are no clear and convincing data to<br />
suggest the superiority <strong>of</strong> either primary surgery or RT in<br />
terms <strong>of</strong> disease control and survival benefit for early-stage<br />
oropharyngeal cancer. 3 Functional outcomes and predicted<br />
insights in the systemic treatment <strong>of</strong> SCCHN, such as novel<br />
approaches to overcome epidermal growth factor receptor<br />
(EGFR) resistance, may enable more effective molecular<br />
targeting in SCCHN beyond the current armamentarium <strong>of</strong><br />
available agents. An overarching theme <strong>of</strong> these early multidisciplinary<br />
advances is to enable the delivery <strong>of</strong> precisionbased<br />
therapeutic regimens from both the technical and<br />
scientific perspectives. Rigorous clinical trial evaluations are<br />
necessary to help define their roles in practice.<br />
quality <strong>of</strong> life after treatment have therefore largely been<br />
the determining factors in the choice <strong>of</strong> therapy <strong>of</strong>fered to<br />
such patients. RT with or without concurrent chemotherapy<br />
is currently the most frequently advocated primary treatment<br />
modality for all stages <strong>of</strong> oropharyngeal cancer,<br />
whereas surgery is usually reserved for local and/or regional<br />
failures. This stems from the belief that the surgical morbidity<br />
can be more substantial than the anticipated morbidity<br />
from nonsurgical approaches. Speech and swallowing can<br />
be substantially compromised by an open surgical approach<br />
to the oropharynx. This coupled with the fact that patients<br />
may ultimately require adjuvant RT or concurrent chemoradiotherapy<br />
(CRT) has led most head and neck oncologists<br />
to agree that primary RT should serve as the definitive<br />
treatment for these patients.<br />
In contrast to advanced-stage oropharyngeal cancers,<br />
speech and eating quality-<strong>of</strong>-life outcomes have not shown<br />
any advantage to RT with or without chemotherapy over<br />
surgery followed by adjuvant RT in early-stage T1/T2 disease.<br />
4 With the advent <strong>of</strong> TORS, surgeons are given a new<br />
and potentially less morbid way <strong>of</strong> approaching the oropharynx<br />
directly, leading many to reconsider surgery as the<br />
primary treatment in selected cases <strong>of</strong> oropharyngeal cancers.<br />
Leaving the surrounding structures untouched has led<br />
many surgeons to believe that TORS is less morbid when<br />
compared with conventional open surgical approaches, and<br />
to nonsurgical treatment regimens. Proponents <strong>of</strong> primary<br />
surgery argue that adjuvant treatment should be based on<br />
operative and pathologic findings tailoring such therapy to<br />
the patients’ needs. Surgically treated early stage T1/T2,<br />
N0/N1 carcinomas <strong>of</strong> the oropharynx avoided the need for<br />
adjuvant RT in up to 38% <strong>of</strong> cases without sacrificing disease<br />
control. 5-7 Primary surgery to such tumors enabled a reduction<br />
<strong>of</strong> the median RT dose to target volumes and prevented<br />
the use <strong>of</strong> concurrent chemotherapy in many cases. 5,7 TORS<br />
may also decrease target volumes to undissected tissues that<br />
From the Department <strong>of</strong> Otolaryngology—Head and Neck Surgery, Division <strong>of</strong> Medical<br />
<strong>Oncology</strong> and Hematology, Princess Margaret Hospital, University Health Network, Toronto,<br />
Ontario, Canada.<br />
Authors’ disclosures <strong>of</strong> potential conflicts <strong>of</strong> interest are found at the end <strong>of</strong> this article.<br />
Address reprint requests to Lillian L. Siu, MD, Division <strong>of</strong> Medical <strong>Oncology</strong> and<br />
Hematology, Drug Development Program, Princess Margaret Hospital, 610 University<br />
Avenue, Suite 5-718, Toronto, ON, M5G 2M9, Canada; email: lillian.siu@uhn.on.ca.<br />
© <strong>2012</strong> by <strong>American</strong> <strong>Society</strong> <strong>of</strong> <strong>Clinical</strong> <strong>Oncology</strong>.<br />
1092-9118/10/1-10<br />
373
would otherwise have been included with more traditional<br />
open approaches because <strong>of</strong> potential tumor seeding. 7<br />
Treatment De-Intensification in Favorable-Risk Patient Groups<br />
De-intensified adjuvant treatment has the potential benefit<br />
<strong>of</strong> lessening the reported long-term adverse effects <strong>of</strong><br />
primary CRT. The demographic shift <strong>of</strong> patients with human<br />
papillomavirus (HPV)-positive oropharyngeal cancers<br />
may present a population that may benefit from the avoidance<br />
<strong>of</strong> the long-term sequelae inherent to RT and CRT.<br />
These are important considerations, particularly with respect<br />
to ongoing efforts to decrease long-term adverse effects<br />
in potentially more favorable groups such as in HPV-positive<br />
patients who are diagnosed younger and who will live<br />
longer. TORS is one <strong>of</strong> several potential de-intensification<br />
strategies that needs to be studied. The ideal indications and<br />
applications <strong>of</strong> its use have not yet been completely elucidated.<br />
Patient selection and adequate preoperative assessment<br />
are essential. This would ideally exclude those<br />
patients who would clearly benefit from adjuvant CRT, such<br />
as those in whom margin clearance would be difficult or<br />
whose disease shows evidence <strong>of</strong> extracapsular nodal extension<br />
and invasion on preoperative imaging. Long-term results<br />
<strong>of</strong> randomized trials comparing primary surgery with<br />
RT are needed to clarify oncologic outcomes and quality-<strong>of</strong>life<br />
issues in early-stage orophayrngeal cancer. This would<br />
help support the belief that dose de-escalation or avoidance<br />
<strong>of</strong> CRT results in reduced late toxicity and improved quality<br />
<strong>of</strong> life.<br />
KEY POINTS<br />
● Examples <strong>of</strong> early advances in surgical, radiation,<br />
and medical oncology have emerged in the treatment<br />
<strong>of</strong> squamous cell carcinomas <strong>of</strong> the head and neck<br />
(SCCHN), including transoral robotic surgery, particle<br />
therapy, and molecular strategies to overcome<br />
epidermal growth factor receptor (EGFR) resistance.<br />
● Transoral robotic surgery provides better access to<br />
challenging anatomic sites while avoiding morbidity<br />
associated with open surgery. Its role alone or in<br />
combination with other treatments, especially to enable<br />
de-intensification <strong>of</strong> favorable risk oropharyngeal<br />
cancers, warrants clinical evaluation.<br />
● Particle therapy such as the use <strong>of</strong> protons or carbon<br />
ions <strong>of</strong>fers dose distribution that may be intensified<br />
or escalated for tumor tissues while scatter to adjacent<br />
critical healthy structures is minimized; definitive<br />
evaluations <strong>of</strong> comparative effectiveness <strong>of</strong><br />
particle therapy compared with photon-based radiation<br />
therapy are awaited.<br />
● Multiple molecular strategies to overcome EGFR resistance<br />
are under clinical development, including<br />
targeting other members <strong>of</strong> the erbB/HER family<br />
besides EGFR, other nodes along the EGFR–mitogenactivated<br />
protein kinase (MAPK) pathway, or oncogenic<br />
pathways other than EGFR, such as the phosphoinositide<br />
3-kinase (PI3K), insulin-like growth factor 1<br />
receptor (IGF-1R) and MET pathways.<br />
374<br />
Novel Advances in Radiotherapy in SCCHN<br />
BISSADA, BRANA, AND SIU<br />
The ability to optimize the therapeutic ratio by delivering<br />
sufficient radiation dose to tumor tissues while sparing<br />
adjacent critical organs is highly relevant in the treatment<br />
<strong>of</strong> SCCHN, and it relies on the dose distribution <strong>of</strong> the<br />
radiation energy. Particle therapy using protons or carbon<br />
ion has gained increasing interest in the treatment <strong>of</strong><br />
SCCHN, largely as a result <strong>of</strong> their physical characteristics<br />
to travel only finite distances through tissues and release<br />
most <strong>of</strong> their energy immediately before they come to rest, in<br />
the so-called Bragg Peak. The prospect <strong>of</strong> delivering less<br />
radiation to nontargeted healthy tissues is appealing, as is<br />
the feasibility to intensify or escalate doses to tumor areas.<br />
In a recent systematic review and meta-analysis <strong>of</strong> RT in<br />
various head and neck cancers comparing photons, protons<br />
and carbon ions, which included 86 observational studies<br />
and eight comparative in silico studies, 5-year survival was<br />
markedly higher after carbon ion therapy compared with<br />
conventional photon therapy for mucosal malignant melanomas.<br />
Additionally, 5-year local control after proton therapy<br />
was markedly higher for paranasal and sinonasal cancer<br />
compared with intensity-modulated photon therapy. No<br />
other substantial differences were observed. However, the<br />
overall quality and quantity <strong>of</strong> data are limited, precluding<br />
definitive comparisons <strong>of</strong> these modalities. 8 Randomized<br />
clinical trials in SCCHN evaluating the comparative effectiveness<br />
<strong>of</strong> different RT techniques and particle therapy are<br />
awaited.<br />
Novel Advances in Systemic Management <strong>of</strong> SCCHN<br />
The systemic management <strong>of</strong> SCCHN continues to represent<br />
an unmet medical need because cetuximab, the monoclonal<br />
antibody against the epidermal growth factor receptor<br />
(EGFR), remains the only molecularly targeted agents approved<br />
by the U.S. Food and Drug Administration (FDA) for<br />
the treatment <strong>of</strong> SCCHN.<br />
Targeting EGFR Resistance in SCCHN<br />
Multiple mechanisms <strong>of</strong> resistance to EGFR inhibition<br />
have been proposed in SCCHN, including: 1) presence <strong>of</strong> the<br />
mutant type III variant <strong>of</strong> EGFR (EGFRvIII) as the result <strong>of</strong><br />
an in-frame deletion mutation <strong>of</strong> exons 2 to 7 spanning the<br />
extracellular ligand-binding domain and 2) extensive crosstalk<br />
among the ErbB/HER family receptors and with other<br />
signaling pathways. EGFRvIII prevents the binding <strong>of</strong> EGF<br />
and other ligands to the receptor and leads to constitutive<br />
activation <strong>of</strong> the EGFR tyrosine kinase domain and its<br />
downstream effectors. 9 With crosstalk, blockade <strong>of</strong> a single<br />
receptor such as EGFR can lead to compensatory upregulation<br />
<strong>of</strong> escape mechanisms along the same EGFR receptor<br />
axis or its downstream effectors, or alternatively, via rescue<br />
from another pathway. Figure 1 depicts several strategies<br />
that are currently being evaluated in clinical trials to<br />
overcome EGFR resistance in cancer, including SCCHN.<br />
Targeting Members <strong>of</strong> the ErbB/HER Family Other Than EGFR<br />
Currently, agents that target multiple members <strong>of</strong> the<br />
ErbB/HER family other than EGFR are actively being developed<br />
in the clinic, with the rationale to overcome de novo<br />
and/or acquired resistance to EGFR inhibition.<br />
Pan-HER inhibitors. Afatinib and dacomitinib are both<br />
orally available, small-molecule, irreversible Pan-HER in-
EARLY ADVANCES IN SCCHN<br />
Fig 1. Targeting epidermal growth<br />
factor resistance in squamous cell carcinoma<br />
<strong>of</strong> the head and neck.<br />
Abbreviations: ERK, extracellular signalregulated<br />
kinase; mTOR, mammalian<br />
target <strong>of</strong> rapamycin; PI3K, phosphoinositide<br />
3-kinase; PTEN, phosphatase and<br />
tensin homolog.<br />
Targeting other<br />
receptor<br />
tyrosine kinases<br />
hibitors with specificities against EGFR, HER2, and HER4<br />
tyrosine kinases. In addition to targeting ErbB/HER family<br />
receptor crosstalk, both <strong>of</strong> these agents have reported activity<br />
against EGFRvIII in preclinical models. In a randomized,<br />
open-label phase II study <strong>of</strong> afatinib compared with cetxuimab<br />
in 124 platinum-refractory patients with recurrent or<br />
metastatic SCCHN, the objective response rate (ORR) favored<br />
afatinib compared with cetuximab (22% vs. 13%,<br />
respectively), as did the median progression-free survival<br />
(PFS; 16 weeks compared with 10 weeks, respectively). The<br />
most common adverse events related to afatinib included<br />
diarrhea and skin toxicity. 10 Afatinib is currently being<br />
evaluated in two phase III trials, including one in the<br />
recurrent or metastatic setting against methotrexate after<br />
progression with platinum-based chemotherapy (NCT-<br />
01345682), and the other in the locally advanced setting<br />
against placebo as maintenance therapy after completion <strong>of</strong><br />
definitive CRT; this latter trial (NCT01345669) is limited to<br />
only those patients whose tumors are HPV negative. Dacomitinib<br />
has undergone single-arm, open-label phase II<br />
evaluation as a first-line therapy in 69 patients with recurrent<br />
or metastatic SCCHN and reported an ORR <strong>of</strong> 11%,<br />
median PFS <strong>of</strong> 2.8 months, and median overall survival (OS)<br />
<strong>of</strong> 7.6 months. Similar to afatinib, diarrhea and skin toxicity<br />
were the most frequent treatment-related adverse events. 11<br />
HER3-targeting agents. HER3 has emerged to become a<br />
key target in the therapeutic strategy to overcome EGFR<br />
resistance. Despite its catalytically inactive tyrosine kinase<br />
domain, HER3 is capable <strong>of</strong> allosterically activating the<br />
kinase domain <strong>of</strong> its heterodimerized partners, preferably<br />
HER2. 12 In addition, phosphorylation <strong>of</strong> the C-terminal tail<br />
<strong>of</strong> HER3 leads to direct recruitment <strong>of</strong> the p85 subunit <strong>of</strong><br />
phosphoinositide 3-kinase (PI3K), rendering it a potent<br />
activator <strong>of</strong> the PI3K-AKT survival pathway. 13,14 Several<br />
HER3 targeting agents are undergoing or completing phase<br />
I development, including MEHD7945A (Genentech, Inc.,<br />
South San Francisco, CA), a humanized, dual-specific, immunoglobulin<br />
(Ig) G1 antibody capable <strong>of</strong> binding to both<br />
PI3K<br />
IRS<br />
Shc<br />
Grb2 Sos<br />
Targeting receptor<br />
tyrosine kinases e.g.<br />
HER family<br />
receptors<br />
TSC2 AKT<br />
PTEN Ras<br />
Targeting<br />
TSC1<br />
multiple<br />
nodes <strong>of</strong> the<br />
Targeting<br />
Raf<br />
same<br />
RHEB<br />
mTOR<br />
multiple<br />
oncogenic<br />
pathways<br />
(horizontal<br />
blockade)<br />
MEK<br />
1/2<br />
ERK<br />
1/2<br />
oncogenic<br />
pathway<br />
(vertical<br />
blockade)<br />
S6K 4E-BP<br />
Transcription<br />
Translation<br />
Growth, Proliferation, Metastasis, Survival<br />
EGFR and HER3 (NCT01207323) 15 ; MM-121/SAR256212<br />
(NCT00734305; Merrimack Pharmaceuticals Inc., Cambridge,<br />
MA and San<strong>of</strong>i <strong>Oncology</strong>, Paris, France) and U3-<br />
1287/AMG 888 (NCT00730470; U3 Pharma, Munich,<br />
Germany and Amgen, Inc., Thousand Oaks, CA), both fully<br />
human monoclonal antibodies directed against HER3;<br />
among others. In a recent report using a mouse monoclonal<br />
antibody (RTJ.2, Santa Cruz Biotechnology, Santa Cruz,<br />
CA; dilution 1:750) against HER3, immunohistochemical<br />
evaluation <strong>of</strong> 387 primary SCCHN specimens showed positive<br />
membranous and cytoplasmic HER3 expression in 34<br />
(8.8%) and 300 (77.5%) cases, respectively. Membranous<br />
HER3 overexpression was significantly associated with<br />
worse OS (p � 0.027) and was an independent prognostic<br />
factor in multivariate analysis. 16 In addition, Wilson and<br />
colleagues demonstrated that a substantial proportion <strong>of</strong><br />
primary SCCHN samples and cell lines have increased<br />
mRNA expression <strong>of</strong> the ligand heregulin, which in turn can<br />
lead to autocrine activation <strong>of</strong> HER3. 17 As such, heregulindriven<br />
tumors with coexpression and activation <strong>of</strong> HER3<br />
may represent a target SCCHN patient population for<br />
HER3-directed therapies.<br />
Targeting EGFR–Mitogen-Activated Protein Kinase and<br />
Other Oncogenic Pathways Concurrently<br />
In many advanced solid tumors, even among molecularly<br />
pr<strong>of</strong>iled patients whose tumors harbor activating mutations,<br />
the activity <strong>of</strong> selective pathway inhibitors administered as<br />
monotherapy appears limited, or in cases <strong>of</strong> initial response,<br />
the duration <strong>of</strong> response is typically short lived. Therapeutic<br />
resistance may occur as a result <strong>of</strong> incomplete pathway<br />
blockade with a single inhibitor, crosstalk and/or upregulation<br />
<strong>of</strong> compensatory escape signaling pathways. A logical<br />
approach to evade such mechanisms <strong>of</strong> resistance is by using<br />
targeted combinations with two agents either interrogating<br />
the same oncogenic signaling pathway (vertical blockade) or<br />
via concurrent modulation <strong>of</strong> two separate signaling pathways<br />
(horizontal blockade). Targeting other receptor ty-<br />
375
osine kinases that can lead to activation <strong>of</strong> both the<br />
mitogen-activated protein kinase (MAPK) and PI3K pathways,<br />
such as the insulin-like growth factor 1 receptor<br />
(IGF-1R) and MET pathways, represents an attractive strategy<br />
to reverse EGFR resistance and is actively being investigated<br />
in the clinical setting. 18,19 In addition, phase I trials<br />
combining inhibitors <strong>of</strong> the MAPK pathway (such as RAF or<br />
MEK) and PI3K pathway (such as PI3K, AKT, or mammalian<br />
target <strong>of</strong> rapamycin [mTOR]) are near completion and<br />
would be <strong>of</strong> interest in SCCHN, especially among patients<br />
whose tumors harbor molecular aberrations in these key<br />
signaling cascades.<br />
Authors’ Disclosures <strong>of</strong> Potential Conflicts <strong>of</strong> Interest<br />
Author<br />
Employment or<br />
Leadership<br />
Positions<br />
Consultant or<br />
Advisory Role<br />
Conclusion<br />
Emerging innovations in surgical, radiation, and medical<br />
oncology aim to improve technical precision and molecular<br />
targeting in disease management for patients with SCCHN.<br />
<strong>Clinical</strong> trial evaluations are needed to translate these early<br />
advances into practice. Ultimately, the goal <strong>of</strong> these<br />
precision-based strategies is to optimize disease control with<br />
increased technical and scientific targeting while sparing<br />
healthy tissue toxicity and late effects, such that translation<br />
to tangible clinical benefits in patients with SCCHN is<br />
achieved.<br />
Stock<br />
Ownership Honoraria<br />
Research<br />
Funding<br />
Eric Bissada*<br />
Irene Brana*<br />
Lillian L. Siu Amgen EntreMed (I) Abraxis<br />
BioScience;<br />
Bristol-Myers<br />
Squibb;<br />
Genentech;<br />
GlaxoSmithKline;<br />
Merck;<br />
Millennium;<br />
Novartis; Pfizer;<br />
Regeneron;<br />
Roche<br />
*No relevant relationships to disclose.<br />
1. Iseli TA, Kulbersh BD, Iseli CE, et al. Functional outcomes after<br />
transoral robotic surgery for head and neck cancer. Otolaryngol Head Neck<br />
Surg. 2009;141:166-171.<br />
2. Genden EM, Desai S, Sung CK. Transoral robotic surgery for the<br />
management <strong>of</strong> head and neck cancer: a preliminary experience. Head Neck.<br />
2009;31:283-289.<br />
3. Cosmidis A, Rame JP, Dassonville O, et al. T1-T2 NO oropharyngeal<br />
cancers treated with surgery alone: a GETTEC study. Eur Arch Otorhinolaryngol.<br />
2004;261:276-281.<br />
4. Allal AS, Nicoucar K, Mach N, et al. Quality <strong>of</strong> life in patients with<br />
oropharynx carcinomas: assessment after accelerated radiotherapy with or<br />
without chemotherapy versus radical surgery and postoperative radiotherapy.<br />
Head Neck 2003;25:833-839.<br />
5. Moncrieff M, Sandilla J, Clark J, et al. Outcomes <strong>of</strong> primary surgical<br />
treatment <strong>of</strong> T1 and T2 carcinomas <strong>of</strong> the oropharynx. Laryngoscope. 2009;<br />
119:307-311.<br />
6. Weinstein GS, O’Malley BW Jr, Snyder W, et al. Transoral robotic<br />
surgery: radical tonsillectomy. Arch Otolaryngol Head Neck Surg. 2007;133:<br />
1220-1226.<br />
7. Weinstein GS, Quon H, O’Malley BW Jr, et al. Selective neck dissection<br />
and deintensified postoperative radiation and chemotherapy for oropharyngeal<br />
cancer: a subset analysis <strong>of</strong> the University <strong>of</strong> Pennsylvania transoral<br />
robotic surgery trial. Laryngoscope 2010;120:1749-1755.<br />
8. Ramaekers BL, Pijls-Johannesma M, Joore MA, et al. Systematic review<br />
and meta-analysis <strong>of</strong> radiotherapy in various head and neck cancers: comparing<br />
photons, carbon-ions and protons. Cancer Treat Rev. 2011;37:185-201.<br />
9. Sok JC, Coppelli FM, Thomas SM, et al. Mutant epidermal growth factor<br />
receptor (EGFRvIII) contributes to head and neck cancer growth and resistance<br />
to EGFR targeting. Clin Cancer Res. 2006;12:5064-5073.<br />
10. Seiwert TY, Fayette J, Del Campo JM, et al. Updated results <strong>of</strong> a<br />
randomized, open-label phase II study exploring BIBW 2992 versus cetux-<br />
376<br />
REFERENCES<br />
BISSADA, BRANA, AND SIU<br />
Expert<br />
Testimony<br />
Other<br />
Remuneration<br />
imab in patients with platinum-refractory metastatic/recurrent head and<br />
neck cancer (SCCHN). Ann Oncol 2010;21:viii316 (abstr 1010PD).<br />
11. Siu LL, Hotte SJ, Laurie SA, et al. Phase II trial <strong>of</strong> the irreversible oral<br />
pan-human EGF receptor (HER) inhibitor PF-00299804 (PF) as first-line<br />
treatment in recurrent and/or metastatic (RM) squamous cell carcinoma <strong>of</strong><br />
the head and neck (SCCHN). J Clin Oncol. 2011;29:45S (suppl; abstr 5561).<br />
12. Jura N, Endres NF, Engel K, et al. Mechanism for activation <strong>of</strong> the EGF<br />
receptor catalytic domain by the juxtamembrane segment. Cell. 2009;137:<br />
1293-1307.<br />
13. Solt<strong>of</strong>f SP, Carraway KL III, Prigent SA, et al. ErbB3 is involved in<br />
activation <strong>of</strong> phosphatidylinositol 3-kinase by epidermal growth factor. Mol<br />
Cell Biol. 1994;14:3550-3558.<br />
14. Prigent SA, Gullick WJ. Identification <strong>of</strong> c-erbB-3 binding sites for<br />
phosphatidylinositol 3�-kinase and SHC using an EGF receptor/c-erbB-3<br />
chimera. EMBO J. 1994;13:2831-2841.<br />
15. Schaefer G, Haber L, Crocker LM, et al. A two-in-one antibody against<br />
HER3 and EGFR has superior inhibitory activity compared with monospecific<br />
antibodies. Cancer Cell. 2011;20:472-486.<br />
16. Takikita M, Xie R, Chung JY, et al. Membranous expression <strong>of</strong> Her3 is<br />
associated with a decreased survival in head and neck squamous cell<br />
carcinoma. J Transl Med. 2011;9:126.<br />
17. Wilson TR, Lee DY, Berry L, et al. Neuregulin-1-mediated autocrine<br />
signaling underlies sensitivity to HER2 kinase inhibitors in a subset <strong>of</strong><br />
human cancers. Cancer Cell. 2011;20:158-172.<br />
18. Wheeler DL, Huang S, Kruser TJ, et al. Mechanisms <strong>of</strong> acquired<br />
resistance to cetuximab: role <strong>of</strong> HER (ErbB) family members. Oncogene.<br />
2008;27:3944-3956.<br />
19. Jameson MJ, Beckler AD, Taniguchi LE, et al. Activation <strong>of</strong> the<br />
insulin-like growth factor-1 receptor induces resistance to epidermal growth<br />
factor receptor antagonism in head and neck squamous carcinoma cells. Mol<br />
Cancer Ther. 2011;10:2124-2134.
Application <strong>of</strong> Genomic and Proteomic<br />
Technologies in Biomarker Discovery<br />
By Elana J. Fertig, PhD, Robbert Slebos, PhD, and Christine H. Chung, MD<br />
Overview: Sequencing <strong>of</strong> the human genome was completed<br />
in 2001. Building on the technology and experience <strong>of</strong> wholeexome<br />
sequencing, numerous cancer genomes have been<br />
sequenced, including head and neck squamous cell carcinoma<br />
(HNSCC) in 2011. Although DNA sequencing data reveals a<br />
complex genome with numerous mutations, the biologic interaction<br />
and clinical significance <strong>of</strong> the overall genetic aberrations<br />
are largely unknown. Comprehensive analyses <strong>of</strong> the<br />
tumors using genomics and proteomics beyond sequencing<br />
THE SEQUENCING <strong>of</strong> the human genome was completed<br />
in 2001 and the whole exome sequencing <strong>of</strong> head<br />
and neck squamous cell carcinoma (HNSCC) was completed<br />
in 2011. 1-4 Although these DNA sequencing data reveal a<br />
complex genome with numerous mutations, the biologic<br />
interaction and clinical significance <strong>of</strong> the overall genetic<br />
aberrations are largely unknown. With rapid development <strong>of</strong><br />
genomic and proteomic technologies, comprehensive analyses<br />
<strong>of</strong> the tumors reflective <strong>of</strong> individual genetic context<br />
are feasible, and these data are expected to increase novel<br />
biomarkers that may have a significant potential in the<br />
clinical management <strong>of</strong> HNSCC. We will summarize the<br />
current genomic and proteomic technologies and approaches<br />
to data analyses. Although the experimental platforms vary,<br />
the analysis techniques for these experiments for cancer<br />
remain platform independent; therefore, we will review<br />
experimental platforms individually but describe analysis<br />
techniques largely without regard for the technology. Furthermore,<br />
we will discuss the general biomarker-discovery<br />
paradigm using the technology and present a few examples<br />
<strong>of</strong> the biomarker discovery and development with clinical<br />
implications and their limitations.<br />
Genomic Technology<br />
In recent years, genomic technologies have expanded from<br />
measuring global gene expression to measuring diverse<br />
genomic features, including exon-level gene expression,<br />
DNA binding, single nucleotide polymorphisms (SNPs), and<br />
DNA methylation patterns (Table 1). Regardless <strong>of</strong> the<br />
biomaterial they quantify (e.g., RNA, DNA or DNA methylation),<br />
the genomic platforms are generally based on either<br />
microarray- or sequencing-based technologies. The first set<br />
<strong>of</strong> microarray technologies focused on the transcriptome,<br />
typically measuring differential expression between case<br />
and control samples for a multitude <strong>of</strong> genes simultaneously.<br />
5 Current microarrays consist <strong>of</strong> sequence-specific<br />
probes designed to hybridize with complementary portions<br />
<strong>of</strong> the fluorescently labeled sample RNA or single-stranded<br />
DNA under analysis. The intensity <strong>of</strong> fluorescence for each<br />
probe increases with the amount <strong>of</strong> sample bound to that<br />
probe, providing a quantitative measure <strong>of</strong> DNA or RNA<br />
abundance, as previously described. 6<br />
However, microarray-based technologies for genomic measurements<br />
are being rapidly replaced by next-generation<br />
sequencing (NGS). 7 The NGS technologies simultaneously<br />
sequence short, single-strand DNA fragments bound to a<br />
solid surface. The sequencers then obtain fluorescence sig-<br />
data can potentially accelerate the rate and number <strong>of</strong> biomarker<br />
discoveries to improve biology-driven classification <strong>of</strong><br />
tumors for prognosis and patient selection for a specific<br />
therapy. In this review, we will summarize the current genomic<br />
and proteomic technologies, general biomarker-discovery<br />
paradigms using the technology and published data in<br />
HNSCC—including potential clinical applications and limitations.<br />
nals for each <strong>of</strong> the sequence fragments based on complimentary<br />
binding <strong>of</strong> fluorescently tagged bases to each base<br />
in the fragment sequentially. 8 The fluorescent signals are<br />
converted to sequences for each fragment and then aligned<br />
to a full-length reference genome to match a sequencing<br />
signal to genomic location. 9 Counting the number <strong>of</strong> sequencing<br />
reads that align to a given genomic location is<br />
analogous to microarray intensities for a probe with a<br />
specific sequence. Similar to microarrays, sequencing technologies<br />
have subsequently evolved from measuring DNA<br />
sequence to also measuring gene expression, DNA methylation<br />
and binding (Table 1). Sequencing technologies can<br />
further identify variation between samples by identifying<br />
genomic locations whereas reads that do not perfectly match<br />
the reference genome may indicate individual genetic variation<br />
or a sequencing error.<br />
Proteomic Technology<br />
Proteomics refers to high-throughput studies <strong>of</strong> proteins<br />
for detection, identification, quantification, and characterization.<br />
Global analysis <strong>of</strong> proteins has long been elusive,<br />
but recent advances in mass-spectrometry (MS) have<br />
made comprehensive and unbiased analysis <strong>of</strong> proteins in<br />
complex mixtures possible. These advances include improved<br />
separation techniques, such as high performance<br />
liquid chromatography (HPLC), 2-dimensional difference gel<br />
electrophoresis (2D-DIGE), new labeling techniques (multicolor<br />
fluorophores, stable isotopes), and new detection and<br />
identification strategies based on MS. Strategies to detect<br />
whole proteins include tissue microarrays (TMAs), protein<br />
microarrays, 2D-DIGE- and MS-based approaches. 6,10<br />
Tissue microarrays (TMAs) allow the detection <strong>of</strong> a single<br />
protein in hundreds <strong>of</strong> separate tissue samples by wellestablished<br />
immunohistochemistry (IHC) using antibodies<br />
specific to the protein <strong>of</strong> interest. The advantage <strong>of</strong> TMAs is<br />
the preservation <strong>of</strong> histologic information, but the quality <strong>of</strong><br />
detection is dependent on biochemical characteristics <strong>of</strong> the<br />
antibody, such as specificity and sensitivity. The IHC has a<br />
From the Department <strong>of</strong> <strong>Oncology</strong>, Johns Hopkins University School <strong>of</strong> Medicine,<br />
Baltimore, MD; Department <strong>of</strong> Cancer Biology, Vanderbilt University School <strong>of</strong> Medicine,<br />
Nashville, TN.<br />
Authors’ disclosures <strong>of</strong> potential conflicts <strong>of</strong> interest are found at the end <strong>of</strong> this article.<br />
Address reprint requests to Christine H. Chung, MD, Department <strong>of</strong> <strong>Oncology</strong>, Johns<br />
Hopkins University School <strong>of</strong> Medicine, 1650 Orleans Street, CRB-1 Room 344, Baltimore,<br />
MD 21231; email: cchung11@jhmi.edu.<br />
© <strong>2012</strong> by <strong>American</strong> <strong>Society</strong> <strong>of</strong> <strong>Clinical</strong> <strong>Oncology</strong>.<br />
1092-9118/10/1-10<br />
377
Table 1. Summary <strong>of</strong> Genomics Measurement Technologies and<br />
Normalization Techniques.<br />
Genomic Measurement Technology Platforms<br />
DNA sequence Sequencing NGS<br />
Gene expression microarray Affymetrix GeneChip<br />
Agilent microarray<br />
Illumina BeadChip<br />
Gene expression RNA-seq NGS<br />
ChIP-chip Roche NibleGen<br />
DNA binding ChIP-seq NGS<br />
Methylation array Illumina BeadChip<br />
Bisulfite sequencing NGS<br />
Methylation<br />
Binding assays (MBDSeq, MeDiP) NGS or microarray<br />
SNP-chip Affymetrix SNP array<br />
Illumina SNP array<br />
SNP<br />
Sequencing NGS<br />
Abbreviations: NGS, next generation sequencing (performed sequencers, including<br />
Illumina Solexa, Applied Biosystems SOLiD and Roche 454); ChIP,<br />
chromatin immunoprecipitation; MBDSeq, methyl-CpG binding domain protein<br />
sequencing; MeDiP, methylated DNA immunoprecipitation; SNP, single nucleotide<br />
polymorphism.<br />
low dynamic range for quantification, and high-quality antibodies<br />
may not be available for all targets <strong>of</strong> interest.<br />
Reverse-phase protein microarrays (RPPM) are similar to<br />
TMAs in that the protein mixture <strong>of</strong> interest is spotted onto<br />
a glass slide and used for hybridization to an antibody. The<br />
spots used for RPPM can be smaller than for TMA, and,<br />
thus, a larger number <strong>of</strong> specimens can be surveyed simultaneously.<br />
Again, the quality <strong>of</strong> the RPPM is dependent on<br />
the availability and biochemical properties <strong>of</strong> the antibody.<br />
Two-dimensional protein separation using 2D-DIGE is<br />
achieved by isoelectric focusing and electrophoresis in a<br />
polyacrylamide gel. Newly developed multicolor fluorophores<br />
allow separate labeling <strong>of</strong> different protein mixtures,<br />
which are then combined and analyzed together by 2D-<br />
DIGE. However, reproducibility and sensitivity <strong>of</strong> the 2D<br />
gels can be problematic, and it may miss some <strong>of</strong> the<br />
smallest and largest proteins in the analysis.<br />
KEY POINTS<br />
● Although DNA sequencing data reveal a complex<br />
genome with numerous mutations, the biologic and<br />
clinical significance <strong>of</strong> genetic aberrations are largely<br />
unknown.<br />
● Microarray-based technologies for gene expression<br />
analysis are being rapidly replaced by next generation–sequencing<br />
technology.<br />
● New advances in proteomic technologies have made<br />
unbiased and quantitative analysis <strong>of</strong> thousands <strong>of</strong><br />
proteins possible, leading to novel insights into tumor<br />
biology and clinical features.<br />
● Although they are powerful discovery tools, each<br />
finding must be vigorously validated before clinical<br />
application.<br />
● New genetic and proteomic analysis technology allows<br />
comprehensive analyses <strong>of</strong> pathways rather<br />
than individual genes or proteins, and they are promising<br />
tools to identify biomarkers and potential therapeutic<br />
targets.<br />
378<br />
FERTIG, SLEBOS, AND CHUNG<br />
Mass spectrometry-based applications have revolutionized<br />
protein detection and quantification, and the advances<br />
are now starting to be used for clinical applications. All MS<br />
instruments work by first ionizing biomolecules into a gasphase,<br />
separating these ions by their mass/charge properties<br />
and detecting them after separation in the instrument. One<br />
<strong>of</strong> the first MS applications uses was Matrix-Assisted Laser<br />
Desorption Ionization (MALDI), where low molecularweight<br />
proteins are ionized from a solid matrix and separated<br />
by mass and charge to yield a detection pr<strong>of</strong>ile unique<br />
for a given mixture <strong>of</strong> proteins. MALDI-MS has the advantage<br />
that it requires minimal sample processing and that<br />
it can analyze a large number <strong>of</strong> samples in a short time.<br />
However, it is limited to the smallest proteins, does not<br />
provide protein identification, and is very prone to experimental<br />
variation.<br />
Since MS instruments are most sensitive and accurate<br />
with small molecules, detection strategies that include digestion<br />
<strong>of</strong> proteins to peptides have been highly successful in<br />
recent years. Here, a complex mixture <strong>of</strong> proteins is first<br />
digested into peptides by trypsin, which cleaves proteins at<br />
lysine (K) and arginine (R) residues. These peptides are then<br />
ionized using a solid matrix (MALDI) or, more commonly, by<br />
electrospray, where a solution <strong>of</strong> peptides is forced through<br />
a thin needle to form a continuous spray that creates peptide<br />
ions for analysis. Additional biochemical separation can be<br />
applied before MS analysis using isoelectric focusing- or<br />
strong cation exchange columns for a multidimensional<br />
peptide separation. After separation and ionization, peptides<br />
may be analyzed and detected directly by MS, creating<br />
a “fingerprint” <strong>of</strong> the original proteins; however, this only<br />
works for relatively simple protein mixtures. For true identification<br />
purposes, peptide ions collide with an inert gas to<br />
form fragment ions that are analyzed in a second MS phase<br />
(tandem-MS or MS/MS) to create fragment MS spectra that<br />
are unique for each peptide. Modern instruments can collide<br />
and analyze thousands <strong>of</strong> peptides per minute, and these<br />
analyses can be performed on a large scale and typically<br />
yield thousands <strong>of</strong> protein identifications with quantitative<br />
information. Identification is achieved by creating a list <strong>of</strong><br />
spectra predicted from genome sequencing information to<br />
include all possible human peptides that could potentially<br />
match the obtained MS spectra. These theoretical spectra<br />
are then matched to the observed MS spectra through<br />
database-search algorithms (Sequest, X!Tandem, Myrimatch).<br />
Knowledge <strong>of</strong> the full sequences <strong>of</strong> all proteins is<br />
then used to reassemble the identified peptides into a list <strong>of</strong><br />
proteins, similar to the process used to create DNA sequencing<br />
alignments from known genomic data. The number <strong>of</strong><br />
times spectra were observed can function as a measure <strong>of</strong><br />
quantity for each protein.<br />
When only a limited number <strong>of</strong> proteins are <strong>of</strong> interest, a<br />
more targeted approach is possible in which the MS instrument<br />
is restricted to a narrow range <strong>of</strong> peptide masses,<br />
which can then be measured with high sensitivity and over<br />
a large dynamic range (selected or multiple reaction monitoring<br />
[SRM or MRM]). These assays can be multiplexed to<br />
create highly specific assays that are also capable <strong>of</strong> measuring<br />
specific protein is<strong>of</strong>orms or modified protein forms.<br />
Proteins can be detected even in lysates generated from<br />
formalin-fixed paraffin-embedded tissues, since trypsin digestion<br />
can free up sufficient amounts <strong>of</strong> peptides for SRM or<br />
MRM detection.
GENOMIC AND PROTEOMIC TECHNOLOGIES AND BIOMARKERS<br />
Fig 1. Analysis steps. Genomics and proteomics analyses from all measurement platforms should follow steps (a)–(f) to obtain clinically<br />
relevant genomic signatures. (a) Sample experimental design accounting for technical artifacts in batches (e.g., processing date, lab technician,<br />
etc). (b) Sample raw output from Affymetrix microarray. (c) Comparing the distribution <strong>of</strong> genomic data in each sample before and after<br />
normalization, at which point measurements for each sample should be on the same scale. (d) Dendrogram for clustering <strong>of</strong> all data colored by<br />
batch to identify artifacts in the data. (e) Heatmap <strong>of</strong> genes that the analysis identified as distinguishing tumor and normal samples (green<br />
represents low measurement values and red high). (f) State predicted based on gene signature from the analysis in (e) for a new test set <strong>of</strong> tumor<br />
and normal samples.<br />
Approaches to Genomic and Proteomic Data Analyses<br />
All genomic and proteomic analyses generally follow the<br />
stages outlined in Figure 1. We will discuss these stages<br />
using genomic analyses since they are better developed for<br />
genomic data, although the various approaches apply to<br />
proteomic datasets as well. First, careful experimental design<br />
is required to ensure that the measurements collected<br />
can quantify the hypothesized genomic state or differences<br />
in tumors and have large enough number <strong>of</strong> samples to<br />
provide sufficient statistical power. Moreover, technical elements<br />
<strong>of</strong> the experimental design, including processing date,<br />
lab technician, and processing center, can introduce significant<br />
artifacts known as batch effects into the data. 11 These<br />
batch effects are present in all genomics data, including<br />
sequencing technologies. Therefore, the experimental design<br />
should ensure that samples for each <strong>of</strong> the biologic conditions<br />
are processed in each <strong>of</strong> the technical conditions (Fig.<br />
1A). If such representation is impossible because <strong>of</strong> the size<br />
or scope <strong>of</strong> the experiment, the experiments should adopt a<br />
randomized block design. Once collected, the raw data (Fig.<br />
1B) must be preprocessed to convert the raw output <strong>of</strong> the<br />
platforms to identifiable genomic or proteomic measurements<br />
that are comparable across the experiments (Fig. 1C),<br />
including statistical removal <strong>of</strong> or control for batch effects<br />
(Fig. 1D). Only after that can computational algorithms be<br />
applied to infer gene-expression signatures pertinent to<br />
cancer (Fig. 1E). 6 Finally, as with any statistical analyses,<br />
genomic analyses require careful cross-validation to ensure<br />
the relevance <strong>of</strong> the inferred signatures to future datasets<br />
collected from different biologic samples in similar experimental<br />
conditions (Fig. 1F). Further, directed experimentation<br />
is required to prove the functional mechanisms<br />
suggested by the inferred signatures.<br />
In general, there are three main analysis approaches:<br />
unsupervised, supervised, and class-prediction algorithms.<br />
Unsupervised (or class-discovery) algorithms seek gene- or<br />
protein-expression signatures that distinguish samples<br />
without regard for the biologic conditions in the experimental<br />
design. Hierarchical clustering has become perhaps the<br />
most widely adopted algorithm for unsupervised genomic<br />
analysis. 12 These algorithms identify distances between sets<br />
<strong>of</strong> genes or samples. Visualization <strong>of</strong> these clustering across<br />
samples in dendrograms (Fig. 2A) can compare biologic<br />
samples measured under different technical conditions,<br />
making them particularly useful for identifying batch effects<br />
and success <strong>of</strong> algorithms removing those effects (Fig. 1D).<br />
Once removed, the clustering will classify samples to ideally<br />
distinguish patients with different clinical outcomes. Likewise,<br />
genes that clustered together can be hypothesized to<br />
function similarly. Similar utility can be gained from another<br />
form <strong>of</strong> clustering, k-means clustering. This algorithm<br />
divides genes or samples from the data into a predetermined<br />
379
number <strong>of</strong> sets separated by a specified distance measure<br />
(Fig. 2B). However, k-means clustering requires knowledge<br />
<strong>of</strong> the number <strong>of</strong> sets a priori. Furthermore, all clustering<br />
algorithms assign genes only to a single set, making it<br />
impossible to account for the gene reuse that occurs in all<br />
biochemical systems and notably in the processes enabled in<br />
cancer.<br />
In addition, matrix factorization algorithms can simultaneously<br />
infer a fixed number <strong>of</strong> relationships between genes<br />
that are active to varying degrees in subsets <strong>of</strong> the samples.<br />
For example, principal component analysis (PCA) has been<br />
widely applied to the data to infer genomic signatures that<br />
explain the majority <strong>of</strong> the variance in the data. 13 However,<br />
because <strong>of</strong> the underlying statistical assumptions, each<br />
signature inferred from PCA will mix genomic responses<br />
from all active processes, unable to disentangle the genomic<br />
response <strong>of</strong> each active biochemical process. Non-negative<br />
matrix factorization (NMF) 14 and Bayesian Decomposition<br />
(BD) 15 were developed simultaneously to learn the parts <strong>of</strong><br />
systems, equivalent to signatures for biochemical processes.<br />
These algorithms yield continuous, non-negative relationships<br />
between genes and samples that are explicitly allowed<br />
to contain overlap across patterns (Fig. 2C). They have<br />
subsequently been adapted in open-source s<strong>of</strong>tware: the<br />
CRAN package NMF for NMF 16 and the Bioconductor package,<br />
CoGAPS, as an alternative to BD. 17 Both algorithms<br />
are computationally intensive, and their accuracy depends<br />
on the number <strong>of</strong> patterns, analogous to selecting the number<br />
<strong>of</strong> clusters in k-means clustering.<br />
Supervised genomic algorithms seek relationships in the<br />
genomic data that distinguish the biologic conditions in the<br />
experimental design. In their simplest form, the set <strong>of</strong> class<br />
380<br />
FERTIG, SLEBOS, AND CHUNG<br />
Fig 2. Visualization <strong>of</strong> unsupervised analysis<br />
algorithms. (a) A dendrogram provides a visualization<br />
<strong>of</strong> hierarchical clustering algorithms. The<br />
length <strong>of</strong> vertical lines on the y-axis indicates<br />
distances between samples or groups <strong>of</strong> samples<br />
separated by a horizontal line. (b) K-means clustering<br />
assigns samples to clusters shown in the<br />
three ovals colored in red (triangle samples), blue<br />
(plus samples), or purple (circle samples). The xand<br />
y-axes are two dimensions representing the<br />
majority <strong>of</strong> the variability in the samples, typically<br />
determined with PCA. (c) Matrix factorization algorithms<br />
divide the measurement data into two<br />
matrices. Columns in the “Amplitude” matrix represent<br />
genes with similar genomic properties<br />
across the samples in the corresponding rows <strong>of</strong><br />
the “Pattern” matrix.<br />
comparison algorithms typically compare measurements <strong>of</strong><br />
each gene measured in the data. Frequently, these analyses<br />
are applied to find genes with significant differential expression<br />
in cancer samples, as compared to normal samples. For<br />
continuous genomic measurements, such as gene expression<br />
and methylation, class-comparison algorithms typically use<br />
a t test to quantify gene-by-gene differences between the<br />
experimental groups. Because the sample sizes are <strong>of</strong>ten<br />
small, the p values obtained from these test statistics are<br />
most accurately inferred from statistical models such as<br />
permutation tests (e.g., SAM 18 ) or through empirical Bayes<br />
moderated t-statistics (e.g., LIMMA 19 ). Because the empirical<br />
Bayes approach can also infer genomic differences<br />
between multiple classes (e.g., tumor stage) or continuous<br />
variables (e.g., drug-resistance measures), it has widely<br />
replaced permutation tests in analyses. Discrete measurements<br />
<strong>of</strong> sequence alterations require more complex methods,<br />
such as hidden Markov models used to detect copy<br />
number variation. 20 Coordinated changes in these statistics<br />
among groups <strong>of</strong> genes can further implicate activity in<br />
specific biologic processes, such as signaling activity.<br />
Regardless <strong>of</strong> the class discovery algorithm, raw, gene-bygene<br />
p values will identify huge numbers <strong>of</strong> false discoveries.<br />
By definition, a p value <strong>of</strong> 5% means that the statistical test<br />
will fail to reject the null hypothesis 5% <strong>of</strong> the time by<br />
chance alone. Therefore, if a genomics data set measured<br />
1000 genes, 50 genes will falsely be inferred as statistically<br />
different between the groups. Simply dividing the desired<br />
p-value threshold by the number <strong>of</strong> statistical tests can<br />
reduce the number <strong>of</strong> false positives (the per-comparison<br />
error rate). However, this correction is <strong>of</strong>ten too stringent<br />
for genomics and proteomics data, resulting in few—if any—
GENOMIC AND PROTEOMIC TECHNOLOGIES AND BIOMARKERS<br />
genes being detected. As a result, more complex but permissive<br />
algorithms, such as the Bonferroni correction,<br />
Benjamini-Hotchberg adjustment, or q-value have been developed<br />
to correct for the multiple-testing problem. 21<br />
Whereas unsupervised or class-discovery algorithms usually<br />
aim to uncover the biologic mechanisms underlying<br />
differences in samples, class-prediction algorithms seek<br />
genomic signatures that can accurately classify future samples.<br />
Numerous algorithms have been adapted for class<br />
discovery in genomics from the field <strong>of</strong> machine-learning,<br />
including neural networks and support-vector machines<br />
among others. 22 All <strong>of</strong> these techniques provide genomic<br />
signatures that correlate with the desired phenotypic outcome.<br />
As a result, the genomic signatures are <strong>of</strong>ten complex<br />
and difficult to interpret clinically. Validation <strong>of</strong> class discovery<br />
algorithms rests in the performance <strong>of</strong> these signatures<br />
in new data sets, rather than validation <strong>of</strong> the inferred<br />
biologic mechanisms. The parameters <strong>of</strong> these algorithms<br />
will be tuned to maximize true-positives (sensitivity) and<br />
minimize false-negatives (specificity). For cancer diagnostics,<br />
class-discovery algorithms are <strong>of</strong>ten biased toward false<br />
positives over false-negatives, erring on the side <strong>of</strong> additional<br />
patient care. However, this bias may result in excessive<br />
treatments.<br />
Ideally, one set <strong>of</strong> data (training set) will be used to<br />
develop the genomic signature with class-discovery algorithms.<br />
The resulting genomic signature will undergo final<br />
validation in this independent test set and cannot undergo<br />
any further correction. Often, these data sets have too few<br />
samples to feasibly divide the data into test and training<br />
sets that are large enough for accurate statistics. In these<br />
cases, cross-validation techniques can test the ability <strong>of</strong> the<br />
experiment and class discovery algorithm to predict the<br />
phenotypes. However, there is no substitute for a truly<br />
independent data set for testing the inferred signature<br />
before clinical application. Such validation techniques are<br />
also required for clinical application <strong>of</strong> signatures identified<br />
with unsupervised or class-comparison algorithms.<br />
Examples <strong>of</strong> Biomarker Discovery Using Genomic and<br />
Proteomic Technologies<br />
Recent completion <strong>of</strong> whole-exome sequencing <strong>of</strong> HNSCC<br />
with NGS technologies revealed that it has a complex<br />
genome with numerous genetic aberrations, such as mutations<br />
and gene amplification. 3,4 The frequent mutations in<br />
HNSCC are mostly tumor suppressors rather than oncogenes,<br />
which are difficult to develop as biomarkers and also<br />
impose therapeutic challenges. Comprehensive analyses <strong>of</strong><br />
the tumors using genomic and proteomic technologies beyond<br />
the sequencing data can accelerate the rate and number<br />
<strong>of</strong> biomarker discoveries to improve biology-driven<br />
classification <strong>of</strong> tumors for prognosis and patient selection<br />
for a specific therapy. However, clinical application <strong>of</strong> these<br />
technologies is still in the early stage with HNSCC. A few<br />
examples are presented here.<br />
Molecular Classification for Prognosis Using Genomics<br />
Although HNSCC is classified as a histopathological<br />
group <strong>of</strong> squamous cell carcinoma, it is a highly heterogeneous<br />
disease, including several anatomic subsites such as<br />
oral cavity, oropharynx, hypopharynx, and larynx. Clinicians<br />
routinely observe varying outcomes within patients<br />
with presumably comparable clinical prognostic factors,<br />
such performance status and disease staging. Therefore,<br />
identification <strong>of</strong> patients at a high-risk for recurrence and<br />
poor survival will potentially improve the selective intensification<br />
<strong>of</strong> treatments. In a study by Chung and colleagues,<br />
60 tumors obtained from patients with HNSCC were characterized<br />
using microarrays and found that there were at<br />
least four groups within HNSCC segregated by distinct<br />
expression pr<strong>of</strong>iles with different survival outcomes. 23 By<br />
examining the expression signatures, molecular characteristics<br />
<strong>of</strong> these groups could be inferred. For example, the<br />
group with a high-risk <strong>of</strong> poor survival had a signature<br />
reflective <strong>of</strong> EGFR activation whereas the group with an<br />
intermediate risk had higher expression <strong>of</strong> genes involved<br />
in epithelial-to-mesenchymal transition (EMT). The clinical<br />
significance <strong>of</strong> the EMT process was further validated by the<br />
association with a high risk <strong>of</strong> recurrence in HNSCC. 24,25<br />
Prediction <strong>of</strong> Local Recurrence Based on Proteomic<br />
Pr<strong>of</strong>ile in HNSCC<br />
Even with the advancement <strong>of</strong> the intense multimodality<br />
therapies for the management <strong>of</strong> patients with locally advanced<br />
HNSCC, the majority <strong>of</strong> the treatment failure is with<br />
locoregional recurrence. Identification <strong>of</strong> the patients at risk<br />
<strong>of</strong> local recurrence at the time <strong>of</strong> diagnosis may allow further<br />
intensification <strong>of</strong> the therapy. Schaaij-Visser and colleagues<br />
examined this clinical problem by applying differential proteomics.<br />
26 They compared paired normal, precursor and<br />
tumor samples from eight patients using 2D-DIGE, and<br />
identified differentially expressed proteins. After identifying<br />
four proteins as potential biomarkers (keratin 4, keratin 13,<br />
cornulin, and small proline-rich protein 3), they confirmed<br />
that a lack <strong>of</strong> keratin 4 and cornulin expression as the<br />
most discriminate proteins to predict local recurrence in<br />
an independent data set. Equivalent studies using mass<br />
spectrometry-based approaches are still limited, although<br />
several abundant proteins have been studied as potential<br />
biomarkers for HNSCC. 27 Further research using modern<br />
proteomic analysis tools has great potential to generate<br />
novel biomarkers to predict clinical outcome and can potentially<br />
affect future patient care.<br />
Future Directions<br />
In conclusion, genomic and proteomic technologies have a<br />
great potential to contribute to clinical research by generating<br />
an unbiased snapshot <strong>of</strong> the activity and composition <strong>of</strong><br />
cancer cells. Integration <strong>of</strong> these measurements is essential<br />
to infer the underlying and complex mechanisms in cancer<br />
development and maintenance. For example, simultaneous<br />
analyses <strong>of</strong> gene and protein expression using a same<br />
sample set can provide post-transcriptional regulation that<br />
could not be detected with either genomic or proteomic<br />
measurements alone. 28 Once genes and proteins <strong>of</strong> interest<br />
are identified, targeted assays can be developed for rapid<br />
clinical translation. However, erroneous genomic and proteomic<br />
signatures that early on were claimed to be predictive<br />
<strong>of</strong> patient response to chemotherapy or early cancer diagnosis<br />
cast a shadow on genomic and proteomic analyses.<br />
Because <strong>of</strong> the complexity <strong>of</strong> these types <strong>of</strong> data, minor and<br />
unintentional processing errors can render many findings<br />
nonreproducible. 29 Despite the steep learning curve, members<br />
<strong>of</strong> the genomics community are adopting open source-<br />
381
ased s<strong>of</strong>tware and tools, such as R for all analyses.<br />
Releasing scripts that formally document all stages <strong>of</strong> the<br />
genomic and proteomic analyses with raw data when submitting<br />
for publication can provide an additional quality<br />
assurance <strong>of</strong> the analyses at the manuscript-review stage. 30<br />
Although disclosure <strong>of</strong> the analysis scripts and subsequent<br />
review cannot eliminate all mistakes from complex data<br />
analyses, they allow others to reproduce the results and can<br />
Authors’ Disclosures <strong>of</strong> Potential Conflicts <strong>of</strong> Interest<br />
Author<br />
Employment or<br />
Leadership<br />
Positions<br />
Consultant or<br />
Advisory Role<br />
Elana J. Fertig*<br />
Robbert Slebos*<br />
Christine H. Chung Boehringer<br />
Ingelheim;<br />
Bristol-Myers<br />
Squibb<br />
*No relevant relationships to disclose.<br />
1. McPherson JD, Marra M, Hillier L, et al. A physical map <strong>of</strong> the human<br />
genome. Nature. 2001;409:934-941.<br />
2. Venter JC, Adams MD, Myers EW, et al. The sequence <strong>of</strong> the human<br />
genome. Science. 2001;291:1304-1351.<br />
3. Stransky N, Egl<strong>of</strong>f AM, Tward AD, et al. The mutational landscape <strong>of</strong><br />
head and neck squamous cell carcinoma. Science. 2011;333:1157-1160.<br />
4. Agrawal N, Frederick MJ, Pickering CR, et al. Exome sequencing <strong>of</strong> head<br />
and neck squamous cell carcinoma reveals inactivating mutations in<br />
NOTCH1. Science. 2011;333:1154-1157.<br />
5. Brown PO, Botstein D. Exploring the new world <strong>of</strong> the genome with DNA<br />
microarrays. Nat Genet. 1999;21:33-37.<br />
6. Chung CH, Levy S, Chaurand P, Carbone DP. Genomics and proteomics:<br />
emerging technologies in clinical cancer research. Crit Rev Oncol Hematol.<br />
2007;61:1-25.<br />
7. Ledford H. The death <strong>of</strong> microarrays? Nature. 2008;455:847.<br />
8. Metzker ML. Sequencing technologies—the next generation. Nat Rev<br />
Genet. 2010;11:31-46.<br />
9. Trapnell C, Salzberg SL. How to map billions <strong>of</strong> short reads onto<br />
genomes. Nat Biotechnol. 2009;27:455-457.<br />
10. Yates JR, Ruse CI, Nakorchevsky A. Proteomics by mass spectrometry:<br />
approaches, advances, and applications. Annu Rev Biomed Eng. 2009;11:49-<br />
79.<br />
11. Leek JT, Scharpf RB, Bravo HC, et al. Tackling the widespread and<br />
critical impact <strong>of</strong> batch effects in high-throughput data. Nat Rev Genet.<br />
2011;11:733-739.<br />
12. Eisen MB, Spellman PT, Brown PO, et al. Cluster analysis and display<br />
<strong>of</strong> genome-wide expression patterns. Proc Natl Acad Sci USA.1998;95:<br />
14863-14868.<br />
13. Ringner M. What is principal component analysis? Nat Biotechnol.<br />
2008;26:303-304.<br />
14. Lee DD, Seung HS. Learning the parts <strong>of</strong> objects by non-negative<br />
matrix factorization. Nature. 1999;401:788-791.<br />
15. Ochs MF, Stoyanova RS, Arias-Mendoza F, et al. A new method for<br />
spectral decomposition using a bilinear Bayesian approach. J Magn Reson.<br />
1999;137:161-176.<br />
16. Gaujoux R, Seoighe C. A flexible R package for nonnegative matrix<br />
factorization. BMC Bioinformatics. 2010;11:367.<br />
382<br />
further facilitate validation <strong>of</strong> any genomic signatures before<br />
clinical application.<br />
Acknowledgment<br />
The authors appreciate the assistance <strong>of</strong> Michael F. Ochs,<br />
Sarah Wheelan, and Luigi Marchioni in writing this manuscript.<br />
Stock<br />
Ownership Honoraria<br />
REFERENCES<br />
Research<br />
Funding<br />
AstraZeneca;<br />
Bayer; Lilly<br />
FERTIG, SLEBOS, AND CHUNG<br />
Expert<br />
Testimony<br />
Other<br />
Remuneration<br />
17. Fertig EJ, Ding J, Favorov AV, et al. CoGAPS: an R/C�� package to<br />
identify patterns and biological process activity in transcriptomic data.<br />
Bioinformatics. 2010;26:2792-2793.<br />
18. Tusher VG, Tibshirani R, Chu G. Significance analysis <strong>of</strong> microarrays<br />
applied to the ionizing radiation response. Proc Natl Acad Sci U S A.<br />
2001;98:5116-5121.<br />
19. Smyth GK. Bioinformatics and Computational Biology Solutions Using<br />
R and Bioconductor (ed XIX), Springer, 2005:397-420.<br />
20. Winchester L, Yau C, Ragoussis J. Comparing CNV detection methods<br />
for SNP arrays. Brief Funct Genomic Proteomic. 2009;8:353-366.<br />
21. Noble WS. How does multiple testing correction work? Nat Biotechnol.<br />
2009;27:1135-1137.<br />
22. Ressom HW, Varghese RS, Zhang Z, et al. Classification algorithms for<br />
phenotype prediction in genomics and proteomics. Front Biosci. 2008;13:691-<br />
708.<br />
23. Chung CH, Parker JS, Karaca G, et al. Molecular classification <strong>of</strong> head<br />
and neck squamous cell carcinomas using patterns <strong>of</strong> gene expression. Cancer<br />
Cell. 2004;5:489-500.<br />
24. Chung CH, Parker JS, Ely K, et al. Gene expression pr<strong>of</strong>iles identify<br />
epithelial-to-mesenchymal transition and activation <strong>of</strong> nuclear factor-kappaB<br />
signaling as characteristics <strong>of</strong> a high-risk head and neck squamous cell<br />
carcinoma. Cancer Res. 2006;66:8210-8218.<br />
25. Pramana J, Van den Brekel MW, van Velthuysen ML, et al. Gene<br />
expression pr<strong>of</strong>iling to predict outcome after chemoradiation in head and neck<br />
cancer. Int J Radiat Oncol Biol Phys. 2007;69:1544-1552.<br />
26. Schaaij-Visser TB, Graveland AP, Gauci S, et al. Differential proteomics<br />
identifies protein biomarkers that predict local relapse <strong>of</strong> head and neck<br />
squamous cell carcinomas. Clin Cancer Res. 2009;15:7666-7675.<br />
27. Matta A, Ralhan R, DeSouza LV, et al. Mass spectrometry-based<br />
clinical proteomics: head-and-neck cancer biomarkers and drug-targets discovery.<br />
Mass Spectrom Rev. 2010;29:945-961.<br />
28. Halvey PJ, Zhang B, C<strong>of</strong>fey RJ, et al. Proteomic consequences <strong>of</strong> a<br />
single gene mutation in a colorectal cancer model. J Proteome Res. <strong>2012</strong>;11:<br />
1184-1195.<br />
29. Ioannidis JP, Allison DB, Ball CA, et al. Repeatability <strong>of</strong> published<br />
microarray gene expression analyses. Nat Genet. 2009;41:149-155.<br />
30. Baggerly K. Disclose all data in publications. Nature. 2010; 467:401.
TREATMENT OF THYROID CANCERS: NEW<br />
INFORMATION FOR MEDICAL ONCOLOGISTS<br />
CHAIR<br />
A. Dimitrios Colevas, MD<br />
Stanford University School <strong>of</strong> Medicine<br />
Stanford, CA<br />
SPEAKERS<br />
Michael Xing, MD, PhD<br />
Johns Hopkins University<br />
Baltimore, MD<br />
Manisha H. Shah, MD<br />
The Ohio State University Comprehensive Cancer Center<br />
Columbus, OH<br />
Keith C. Bible, MD, PhD<br />
Mayo Clinic<br />
Rochester, MN
Evaluation <strong>of</strong> Patients with Disseminated or<br />
Locoregionally Advanced Thyroid Cancer:<br />
A Primer for Medical Oncologists<br />
By A. Dimitrios Colevas, MD, and Manisha H. Shah, MD<br />
Overview: Historically, patients with thyroid cancers are<br />
managed by endocrinologists, surgeons and radiation oncologists.<br />
Due to recent progress in this field with advances in<br />
treatment <strong>of</strong> thyroid cancer, medical oncologists are now<br />
commonly involved in care <strong>of</strong> patients with advanced thyroid<br />
TRADITIONALLY, TREATMENT <strong>of</strong> patients with differentiated<br />
thyroid carcinomas (DTC), which includes<br />
papillary, follicular, and Hurthle cell carcinomas, has been<br />
highly constrained by the lack <strong>of</strong> effective treatment <strong>of</strong><br />
patients with metastatic or recurrent locoregional disease<br />
after surgical treatments have been exhausted or are not<br />
feasible. Treatments in this setting have been mostly limited<br />
to external-beam radiation treatment focused on the area <strong>of</strong><br />
concern or systemic radioactive iodine (RAI) for patients<br />
with rising serum thyroglobulin levels or evidence <strong>of</strong> iodineavid<br />
evaluable metastases. Because there are no convincing<br />
data on extending survival for patients with disease that is<br />
evaluable only by thyroglobulin levels, endocrinologists and<br />
nuclear medicine physicians have struggled with the selection<br />
<strong>of</strong> patients for whom RAI is indicated. 1 Although some<br />
advocate treatment with RAI for most patients with thyroglobulin<br />
levels higher than 10 ng/mL, others support a more<br />
focused approach <strong>of</strong> treatment for patients who are thought<br />
to be at high risk for clinically relevant sequellae <strong>of</strong> progressive<br />
disease, as disease volume can be reduced even in<br />
patients with thyroglobulin-positive but imaging-negative<br />
disease. 2,3 Therefore, patients referred to the medical oncologist<br />
for RAI-refractory thyroid cancer have commonly received<br />
RAI in the adjuvant setting and <strong>of</strong>ten with an<br />
additional one or two doses <strong>of</strong> RAI in the metastatic setting,<br />
defined by elevated thyroglobulin levels in the absence <strong>of</strong><br />
radiographically evaluable metastatic disease. Historically,<br />
it was not uncommon to encounter patients with relatively<br />
indolent disease who had received cumulative RAI doses<br />
exceeding 600 mCi. More recently, because <strong>of</strong> an increased<br />
awareness <strong>of</strong> long-term adverse events from RAI, such as<br />
bone marrow suppression, salivary gland compromise, and<br />
gonadal dysfunction, many patients treated with RAI, even<br />
those with clinically indolent disease, are referred to the<br />
medical oncologist with lower prior cumulative RAI doses, in<br />
the range <strong>of</strong> 200–400 mCi.<br />
Traditional cytotoxic chemotherapy has played a minimal<br />
role in the treatment <strong>of</strong> patients with RAI-refractory thyroid<br />
cancer. Despite reports or small retrospective experiential<br />
reviews suggesting promise, especially in the induction<br />
setting for patients with no prior cancer treatment, most<br />
well-documented studies demonstrate little to no response to<br />
conventional chemotherapy in patients with RAI-refractory<br />
disease. 4,5 Doxorubicin is <strong>of</strong>ten cited as a standard <strong>of</strong> care on<br />
the basis <strong>of</strong> a 37% response rate, but that rate was reported<br />
in an oldstudy, and neither the time to progression nor the<br />
methods used to determine responses were reported. 6 Subsequent<br />
studies in patients with RAI-refractory disease have<br />
yielded a 5% response rate. 7 Combination chemotherapy has<br />
384<br />
cancers. In this manuscript, we describe general principles in<br />
management <strong>of</strong> patients with various types <strong>of</strong> thyroid cancers<br />
including differentiated, medullary and anaplastic thyroid cancers.<br />
been equally disappointing. In a small study in which the<br />
combination <strong>of</strong> cisplatin and doxorubicin was compared with<br />
doxorubicin alone, the combination had a numerically lower<br />
response rate than single-agent doxorubicin. 8 There are<br />
hints that newer combinations <strong>of</strong> cytotoxic chemotherapy<br />
may be achieving better results. In a study <strong>of</strong> 14 patients<br />
treated with gemcitabine plus oxaliplatin, a complete response<br />
was achieved in one patient and a partial response<br />
was achieved in seven. 9<br />
The medical oncologist caring for patients with advanced<br />
thyroid cancer should be familiar with alternatives to systemic<br />
cytotoxic chemotherapy for patients who have received<br />
RAI (Sidebar 1). Patients with locoregionally recurrent disease<br />
following thyroidectomy and RAI should be considered<br />
for surgery or external-beam radiation treatment. Occasionally,<br />
patients benefit from other locoregional modalities,<br />
such as the use <strong>of</strong> ethanol injection or radi<strong>of</strong>requency ablation<br />
for lymph nodes involved with clinically apparent recurrent<br />
disease.<br />
In patients with locoregionally recurrent or metastatic<br />
disease, the medical oncologist should also ensure that<br />
aggressive suppression <strong>of</strong> thyroid-stimulating hormone has<br />
been achieved. Some data suggest that aggressive thyroid<br />
stimulating hormone (TSH) suppression leads to improved<br />
outcomes, although one large randomized controlled study <strong>of</strong><br />
TSH suppression compared with replacement in the adjuvant<br />
setting after surgery did not demonstrate superiority <strong>of</strong><br />
suppression. 10,11 There is consensus among experts that, for<br />
patients with known recurrent differentiated thyroid cancer<br />
(as opposed to patients being treated in the adjuvant setting),<br />
the target TSH level should be less than 0.1 mU/L.<br />
For patients who have clinically meaningful progression<br />
<strong>of</strong> disease during submaximal TSH suppression (TSH � 0.1<br />
mU/L), the trade-<strong>of</strong>f <strong>of</strong> side effects from more aggressive<br />
TSH suppression must be weighed against the possibility <strong>of</strong><br />
antitumor benefit. The levothyroxine dose necessary to<br />
achieve suppression is typically higher than 1.6 mcg/kg/day,<br />
<strong>of</strong>ten cited in the literature as a minimal thyroid suppressive<br />
dose for patients with benign thyroid conditions. One<br />
study found that 2.56 mcg/kg/d was the average dose neces-<br />
From the Stanford Cancer Institute, Stanford University, Palo Alto, CA; The Ohio State<br />
University Comprehensive Cancer Center, Columbus, OH.<br />
Authors’ disclosures <strong>of</strong> potential conflicts <strong>of</strong> interest are found at the end <strong>of</strong> this article.<br />
Address reprint requests to Manisha H. Shah, MD, Ohio State University, 320 W. 10 th<br />
Avenue, A438 Starling-Loving Hall, Columbus, OH 43210; email: manisha.shah@osumc.edu.<br />
© <strong>2012</strong> by <strong>American</strong> <strong>Society</strong> <strong>of</strong> <strong>Clinical</strong> <strong>Oncology</strong>.<br />
1092-9118/10/1-10
EVALUATION OF ADVANCED THYROID CANCER<br />
Sidebar 1. Treatment Considerations for Recurrent or Metastatic Differentiated Thyroid Carcinoma<br />
Consideration<br />
Surgery<br />
External-beam radiation<br />
Thyrotropin suppression<br />
Confirmation <strong>of</strong> optimal<br />
preparation for prior<br />
radioactive iodine<br />
treatments<br />
Bone lesions<br />
sary to adequately suppress TSH in patients with thyroid<br />
cancer. 12<br />
It is essential that the medical oncologist seeing patients<br />
with recurrent or metastatic DTC be familiar with issues<br />
usually addressed by endocrinologists or nuclear medicine<br />
physicians. For example, before considering treatment with<br />
either cytotoxic chemotherapy or molecularly targeted systemic<br />
agents, the medical oncologist should determine<br />
whether RAI is an option for the patient. There is no<br />
universally accepted definition <strong>of</strong> RAI-refractory DTC (Sidebar<br />
2). Some clinicians accept a rise in thyroglobulin level<br />
following administration <strong>of</strong> RAI as evidence <strong>of</strong> refractory<br />
KEY POINTS<br />
● Spectrum <strong>of</strong> thyroid follicular-cell derived cancers<br />
ranges from indolent differentiated thyroid cancers<br />
(DTC) to very aggressive undifferentiated thyroid<br />
cancer commonly known as anaplastic thyroid cancer<br />
(ATC).<br />
● Medullary thyroid cancer (MTC) originates from<br />
parafollicular c-cells <strong>of</strong> thyroid and is one <strong>of</strong> the best<br />
characterized solid tumors.<br />
● Surgery plays critical role in management <strong>of</strong> all types<br />
<strong>of</strong> thyroid cancers while targeted systemic therapies<br />
with levo-thyroxine for thyroid stimulating hormonesuppression<br />
and radio-iodine are restricted to management<br />
<strong>of</strong> differentiated thyroid cancer.<br />
● Tyrosine kinase inhibitors are promising novel therapies<br />
for patients with DTC and MTC. Vandetanib<br />
was approved by the U.S. Food and Drug Administration<br />
in 2011 for treatment <strong>of</strong> progressive or symptomatic<br />
advanced MTC.<br />
● Despite multimodality treatment with surgery, radiation<br />
and cytotoxic chemotherapy, prognosis remains<br />
dismal with median survival <strong>of</strong> 6 months in patients<br />
with advanced ATC.<br />
Specific Issues<br />
Selected patients may benefit from repeated neck dissections or metastatectomies<br />
<strong>of</strong> limited disease.<br />
Often useful to control inoperable relapses in the neck and treatment <strong>of</strong> foci <strong>of</strong><br />
metastatic disease, especially central nervous system or bone lesions<br />
Target serum thyroid stimulating hormone level �0.1 mU/L. May require 2.5<br />
mcg/kg/d or more <strong>of</strong> levothyroxine<br />
Was a low iodine diet maintained?<br />
Was the TSH adequately elevated via thyroid hormone withdrawal or recombinant<br />
TSH administration?<br />
Did the patient receive iodinated contrast medium in the 3 mos. preceding<br />
treatment with radioactive iodine?<br />
Was a low-iodine diet maintained prior to treatment with radioactive iodine?<br />
Bisphosphonates or RANK-ligand inhibitors reduce the risk <strong>of</strong> fracture.<br />
disease. Others cite lack <strong>of</strong> detectable radioactivity on a<br />
whole-body scan following diagnostic or therapeutic administration<br />
<strong>of</strong> RAI as definition <strong>of</strong> RAI-refractory disease.<br />
Although progression after RAI treatment as documented by<br />
conventional imaging with computerized tomography (CT),<br />
magnetic resonance imaging, or bone scans is probably the<br />
most conservative method <strong>of</strong> declaring a patient’s tumor to<br />
be RAI refractory, conventional imaging comparisons are<br />
<strong>of</strong>ten not available for patients who have been followed up<br />
exclusively by serum thyroglobulin levels and whole-body<br />
scans. Because <strong>of</strong> the inverse relationship between<br />
fluorodeoxyglucose-positron emission tomography (FDG-<br />
PET) avidity and RAI uptake, some clinicians consider a<br />
positive FDG-PET scan itself as a definition <strong>of</strong> RAIrefractory<br />
disease. 13,14<br />
Under certain circumstances, it is important to review<br />
how therapeutic RAI was administered to ensure that it was<br />
given in a setting in which efficacy was maximized. In order<br />
to facilitate maximal RAI uptake by the disease, there must<br />
not be a sink for the RAI. In general, normal thyroid tissue<br />
is more iodine avid than DTC; because <strong>of</strong> this, the presence<br />
Sidebar 2. Various Definitions <strong>of</strong> Radioactive Iodine<br />
(RAI) Refractory Differentiated Thyroid Carcinoma<br />
1. Lack <strong>of</strong> uptake seen on a whole-body scan following<br />
diagnostic RAI dosing<br />
2. Lack <strong>of</strong> uptake seen on a whole-body scan following<br />
therapeutic RAI dosing.<br />
3. Rising thyroglobulin following therapeutic RAI dosing<br />
4. Progression <strong>of</strong> lesions as documented by conventional<br />
imaging (e.g., computerized tomography,<br />
magnetic resonance imaging or bone scan) following<br />
therapeutic RAI dosing.<br />
5. Cumulative dose <strong>of</strong> � 600 mCi <strong>of</strong> 131-iodine ( 131 I)<br />
6. Fluorodeoxyglucose (FDG)–avid lesions on FDGpositron<br />
emission tomography.<br />
385
Table 1. Prognostic Considerations for Patients with Advanced<br />
Differentiated Thyroid Carcinoma<br />
Variable Issue<br />
BRAF mutation in tumor Mutation found in 29–69% <strong>of</strong> papillary thyroid cancer;<br />
associated with a poor prognosis. 18<br />
RET/PTC oncogenic<br />
rearrangement in<br />
tumor<br />
Abnormality present in 5–30% <strong>of</strong> papillary thyroid<br />
cancer and 60–70% <strong>of</strong> radiation-induced papillary<br />
thyroid cancer. 19<br />
FDG-PET scan The median survival for atients with FDG-avid disease is<br />
approximately 4 years. 20<br />
Serum thyroglobulin Thyroglobulin level �10 ng/mL 6–12 mos. after ablation<br />
is associated with an odds ratio for relapse <strong>of</strong> 16 vs.<br />
those lower levels.<br />
The doubling time <strong>of</strong> thyroglobulin strongly predicts<br />
survival. 21<br />
VEGF High immunostaining for VEGF in tumor is associated<br />
with risk <strong>of</strong> metastases. 22<br />
Histology Follicular variant <strong>of</strong> papillary thyroid carcinoma is<br />
associated with shorter survival in patients with<br />
metastatic disease. 23<br />
Sites <strong>of</strong> metastasis Sites other than lung portend a worse prognosis. 24,25<br />
Age Patients �40 yr old have a better prognosis. 25,26<br />
Abbreviations: FDG-PET, fluorodeoxyglucose-positron emission tomography;<br />
VEGF, vascular endothelial growth factor.<br />
<strong>of</strong> residual normal thyroid tissue can diminish the effective<br />
dose taken up by the cancer. Therefore, when residual<br />
normal thyroid tissues cannot be removed surgically, a<br />
two-stage approach to RAI dosing is <strong>of</strong>ten used. An initial<br />
smaller dose <strong>of</strong> RAI is administered with the intent <strong>of</strong><br />
ablating normal thyroid tissue, followed by the anticancer<br />
therapeutic dose after the normal thyroid tissue iodine sink<br />
has been eliminated. A phenomenon known as thyroid<br />
stunning, a situation in which residual DTC may take up<br />
less iodine after ablative doses <strong>of</strong> RAI, has been described,<br />
but the impact <strong>of</strong> stunning and whether it is <strong>of</strong> meaningful<br />
Category Principles <strong>of</strong> Management<br />
Table 2. Principles in Management <strong>of</strong> Medullary Thyroid Carcinoma<br />
COLEVAS AND SHAH<br />
therapeutic relevance has been contested by nuclear medicine<br />
experts. 15 Therefore, when patients have metastatic<br />
DTC that is not thought to be amenable to surgical resection,<br />
a surgical debulking <strong>of</strong> normal and cancerous thyroid<br />
tissue in the neck followed by low or intermediate doses <strong>of</strong><br />
RAI in preparation for definitive RAI treatment may be an<br />
approach preferable to systemic treatment with cytotoxic or<br />
molecularly targeted agents.<br />
Because uptake <strong>of</strong> iodine by both normal thyroid tissue<br />
and DTC is related to serum TSH levels, it is important to<br />
achieve levels <strong>of</strong> less than 30 mU/L before RAI dosing.<br />
Historically, this level was achieved through withdrawal <strong>of</strong><br />
thyroid hormone replacement in patients without an intact<br />
thyroid gland. With the recent availability <strong>of</strong> recombinant<br />
thyrotropin (rTSH), many clinicians have moved from a<br />
levothyroxine withdrawal strategy to administration <strong>of</strong><br />
rTSH administration. Although some data suggest that<br />
there may be a reduction in sensitivity <strong>of</strong> radioiodine scans<br />
using rTSH, a comparison <strong>of</strong> the therapeutic efficacy <strong>of</strong> these<br />
two approaches in the metastatic setting failed to show a<br />
difference in survival. 16,17 In addition to rendering the DTC<br />
maximally avid for iodine, it is also important that the<br />
patient not be exposed to high levels <strong>of</strong> iodine before RAI<br />
dosing. One common occurrence is the use <strong>of</strong> CT scans with<br />
iodinated contrast medium in patients being evaluated for<br />
cancer before the diagnosis <strong>of</strong> DTC has been made. The load<br />
<strong>of</strong> iodine administered as part <strong>of</strong> a thoracic CT with contrast<br />
medium is on the order <strong>of</strong> 30 g, which is 200 times the<br />
recommended daily intake <strong>of</strong> iodine <strong>of</strong> 150 mcg/day. If there<br />
is a question concerning a patient’s compliance with a low<br />
iodine diet, urinary iodine levels can be used as a marker <strong>of</strong><br />
iodine consumption, but it is not well established to what<br />
extent dietary iodine levels influence the efficacy <strong>of</strong> RAI in<br />
this setting.<br />
Once it has been determined that the patient has RAI-<br />
RET genetic testing Genetic testing for germline mutation in RET is recommended for all patients with disease, regardless <strong>of</strong> positive family history<br />
Serum tumor markers Calcitonin and carcinoembryonic antigen<br />
Staging studies Computerized tomography <strong>of</strong> the neck and chest; triple-phase scan <strong>of</strong> the abdomen-pelvis (or magnetic resonance imaging <strong>of</strong><br />
abdomen).<br />
Magnetic resonance imaging for suspected bone metastasis; x-ray for suspected bone metastasis in long bones.<br />
Bone scan and positron emission tomography are not reliable and should not be routinely used for staging medullary thyroid<br />
carcinoma.<br />
Treatment <strong>of</strong> hormonal symptoms Antidiarrheal medication:<br />
(flushing/diarrhea)<br />
short-acting octreotide given subcutaneously; if treatment for 1–2 wks is effective in ameliorating symptoms, can use octreotide<br />
LAR, given intramuscularly.<br />
Treatment <strong>of</strong> tumor, including debulking surgery, also palliates these symptoms.<br />
Treatment <strong>of</strong> primary tumor Surgery can be curative for early-stage disease.<br />
External-beam radiation therapy is <strong>of</strong> questionable benefit, even in the setting <strong>of</strong> high risk for local recurrence.<br />
Resecting bulky primary tumor, even in setting <strong>of</strong> metastatic disease, is <strong>of</strong> value for palliation <strong>of</strong> refractory symptoms.<br />
Treatment <strong>of</strong> locally recurrent disease Surgery for locally recurrent disease is <strong>of</strong> value in selected patients.<br />
External-beam radiation therapy is <strong>of</strong> questionable benefit.<br />
Treatment <strong>of</strong> bone metastasis Orthopedic or neurosurgery team should evaluate patients with metastatic lytic lesion in the long bones or fpatients with high-risk<br />
spine metastasis.<br />
External-beam radiation therapy is <strong>of</strong> great palliative value for symptomatic bone metastasis.<br />
Treatment <strong>of</strong> metastatic disease Wait-and-watch approach with periodic restaging studies and monitoring is feasible for patients who have low tumor burden,<br />
asymptomatic disease, andno evidence <strong>of</strong> progression.<br />
<strong>Clinical</strong> trial enrollment is encouraged for patients with progressive or symptomatic disease or for patients with high tumor burden.<br />
Cytotoxic chemotherapy (doxorubicin- or dacarbazine-based) is associated with a 0–30% response rate and is <strong>of</strong> limited value.<br />
Vandetanib was approved by the Food & Drug Administration in 2011 for patients with progressive or symptomatic metastatic<br />
disease.<br />
386
EVALUATION OF ADVANCED THYROID CANCER<br />
refractory DTC and the options discussed here have been<br />
exhausted, use <strong>of</strong> targeted agents can be considered. Three<br />
questions to answer are (1.) Does the patient currently have<br />
symptoms from DTC that would be relieved from reduction<br />
<strong>of</strong> the tumor? (2.) Based on the clinical history and imaging<br />
studies, are there lesions, although asymptomatic now, that<br />
pose an imminent threat? (3.) Is there evidence from the<br />
patient history, imaging studies, or thyroglobulin levels that<br />
the velocity <strong>of</strong> tumor progression is such that deferred<br />
treatment might result in a rapid decline in performance<br />
status? Some clinical features can be useful guides for<br />
predicting which patients are likely to have more aggressive<br />
disease (Table 1). As is the case with many common solid<br />
tumors such as lung, colorectal, and breast cancers, <strong>of</strong>ten<br />
the decision to treat in the metastatic or recurrent setting is<br />
made on the basis <strong>of</strong> the extent <strong>of</strong> disease without knowledge<br />
<strong>of</strong> the progression velocity. DTC, on the other hand, even in<br />
patients with widely disseminated disease or a high thyroglobulin<br />
level, can be indolent, and the timing <strong>of</strong> treatment<br />
initiation can be challenging in an asymptomatic patient. As<br />
analogies, one can think <strong>of</strong> patients with prostate cancer and<br />
a high prostate-specific antigen level but no symptoms or<br />
measurable disease or patients with low-grade neuroendocrine<br />
tumors with high volume but minimally symptomatic<br />
disease; the art <strong>of</strong> deciding whom and when to treat remains<br />
a challenge.<br />
Once the decision has been made that a patient with DTC<br />
is appropriate for systemic drug treatment, ample data<br />
support the use <strong>of</strong> several tyrosine kinase inhibitors (TKIs).<br />
Medullary Thyroid Carcinoma<br />
Unlike DTC, medullary thyroid carcinoma (MTC) is a rare<br />
type <strong>of</strong> thyroid cancer and accounts for approximately 2% to<br />
8% <strong>of</strong> all thyroid cancer. However, it is one <strong>of</strong> the best<br />
characterized solid tumors. MTC is derived from parafollicular<br />
cells <strong>of</strong> the thyroid and exhibit well-differentiated neuroendocrine<br />
carcinoma histology. Given that MTC does not<br />
originate from follicular cells <strong>of</strong> the thyroid, it is not iodine<br />
avid and does not secrete thyroglobulin. MTC occurs in the<br />
sporadic (75%) and hereditary (25%) setting. Germline mutation<br />
in the RET proto-oncogene is a critical pathogenetic<br />
defect associated with nearly 95% <strong>of</strong> all cases <strong>of</strong> the hereditary<br />
type <strong>of</strong> MTC (MEN-2A, MEN-2B, or familial MTC).<br />
Somatic mutation in the RET proto-oncogene is observed in<br />
30% to 40% <strong>of</strong> tumors <strong>of</strong> sporadic MTC. Serum calcitonin<br />
and carcinoembryonic antigen (CEA) are reliable tumor<br />
markers with prognostic significance. Patients with MTC may<br />
Authors’ Disclosures <strong>of</strong> Potential Conflicts <strong>of</strong> Interest<br />
Author<br />
Employment or<br />
Leadership<br />
Positions<br />
Consultant or<br />
Advisory Role<br />
have hormone-associated symptoms <strong>of</strong> flushing and diarrhea.<br />
MTC is typically a systemic disease and is known to spread in<br />
the setting <strong>of</strong> small primary tumors. In addition to lymph<br />
nodes in the neck, common sites <strong>of</strong> metastasis include the<br />
lung, mediastinal lymph nodes, liver, and bones. While cytotoxic<br />
chemotherapies are ineffective, systemic therapies with<br />
TKIs are promising. Over last decade, numerous clinical trials<br />
ranging from phase I and phase III trials have been conducted<br />
in patients with advanced MTC. Vandetanib is approved by<br />
FDA in 2011 for patients with progressive advanced MTC. We<br />
summarize the principles <strong>of</strong> diagnosis and management in<br />
Table 2.<br />
Anaplastic Thyroid Carcinoma<br />
Anaplastic thyroid cancer (ATC) accounts for 2% <strong>of</strong> all<br />
thyroid cancer. It is one <strong>of</strong> the most aggressive cancers among<br />
all solid tumors and its progression can be clinically evident in<br />
a matter <strong>of</strong> days to weeks. ATC is classified only as stage IV<br />
and is divided into stage IVA, IVB, and IVC. Despite multimodality<br />
aggressive therapy, the prognosis for patients with<br />
ATC is poor, with a median survival <strong>of</strong> 6 months. Treatment<br />
options usually include combinations <strong>of</strong> surgery, cytotoxic<br />
chemotherapy, and radiation therapy. Cytotoxic chemotherapy<br />
agents such as taxanes, platinums, or doxorubicin can be<br />
used. Concurrent or sequential chemoradiation treatment is<br />
an option. Supportive care and hospice should be mentioned<br />
to the patient when discussing management <strong>of</strong> the disease.<br />
In addition to multimodality therapy, other topics that<br />
should be addressed are airway management and nutritional<br />
status, with consideration <strong>of</strong> a percutaneous endoscopic<br />
gastrostomy tube. Despite recent advances in the understanding<br />
<strong>of</strong> the pathogenetics <strong>of</strong> ATC, novel targeted therapies<br />
tested to date have proved to be ineffective. Patients should be<br />
encouraged to participate in clinical trials.<br />
Stock<br />
Ownership Honoraria<br />
Sidebar 3. Guidelines for Management <strong>of</strong> Thyroid<br />
Cancer<br />
The following organizations provide guidelines for<br />
the management <strong>of</strong> various types <strong>of</strong> thyroid cancer.<br />
● National Comprehensive Cancer Network (NCCN):<br />
www.nccn.org<br />
● <strong>American</strong> Thyroid Association (ATA): http://thyroid<br />
guidelines.com<br />
● British Thyroid Association (BTA): www.britishthyroid-association.org/Guidelines<br />
Research<br />
Funding<br />
A. Dimitrios Colevas ActoGeniX;<br />
Bayer/Onyx;<br />
Boehringer<br />
Ingelheim;<br />
Exelixis;<br />
GlaxoSmithKline;<br />
Roche<br />
Manisha H. Shah Bayer Daiichi Sankyo;<br />
Eisai; Exelixis<br />
Expert<br />
Testimony<br />
Other<br />
Remuneration<br />
387
1. Mazzaferri EL. Treating high thyroglobulin with radioiodine: a magic<br />
bullet or a shot in the dark? J Clin Endocrinol Metab. 1995;80:1485-1487.<br />
2. Schlumberger M, Mancusi F, Baudin E, et al. 131I therapy for elevated<br />
thyroglobulin levels. Thyroid. 1997;7:273-276.<br />
3. Pineda JD, Lee T, Ain K, et al. Iodine-131 therapy for thyroid cancer<br />
patients with elevated thyroglobulin and negative diagnostic scan. J Clin<br />
Endocrinol Metab. 1995;80:1488-1492.<br />
4. Besic N, Auersperg M, Gazic B, et al. Neoadjuvant chemotherapy in 29<br />
patients with locally advanced follicular or Hürthle cell thyroid carcinoma: a<br />
phase 2 study. Thyroid. <strong>2012</strong>;22:131-137.<br />
5. Bernhardt B. Follicular thyroid carcinoma: response to chemotherapy.<br />
Am J Med Sci. 1981;282:45-46.<br />
6. Gottlieb JA, Hill CS, Jr. Chemotherapy <strong>of</strong> thyroid cancer with Adriamycin.<br />
Experience with 30 patients. N Engl J Med. 1974;290:193-197.<br />
7. Matuszczyk A, Petersenn S, Bockisch A, et al. Chemotherapy with<br />
doxorubicin in progressive medullary and thyroid carcinoma <strong>of</strong> the follicular<br />
epithelium. Horm Metab Res. 2008;40:210-213.<br />
8. Shimaoka K, Schoenfeld DA, DeWys WD, et al. A randomized trial <strong>of</strong><br />
doxorubicin versus doxorubicin plus cisplatin in patients with advanced<br />
thyroid carcinoma. Cancer. 1985;56:2155-2160.<br />
9. Spano JP, Vano Y, Vignot S, et al. GEMOX regimen in the treatment <strong>of</strong><br />
metastatic differentiated refractory thyroid carcinoma. Med Oncol. 2011 Sept<br />
25.<br />
10. Jonklaas J, Sarlis NJ, Lit<strong>of</strong>sky D, et al. Outcomes <strong>of</strong> patients with<br />
differentiated thyroid carcinoma following initial therapy. Thyroid. 2006;16:<br />
1229-1242.<br />
11. Sugitani I, Fujimoto Y. Does postoperative thyrotropin suppression<br />
therapy truly decrease recurrence in papillary thyroid carcinoma? A randomized<br />
controlled trial. J Clin Endocrinol Metab. 2010;95:4576-4583.<br />
12. Burmeister LA, Goumaz MO, Mariash CN, et al. Levothyroxine dose<br />
requirements for thyrotropin suppression in the treatment <strong>of</strong> differentiated<br />
thyroid cancer. J Clin Endocrinol Metab. 1992;75:344-350.<br />
13. Nanni C, Rubello D, Fanti S, et al. Role <strong>of</strong> 18F-FDG-PET and PET/CT<br />
imaging in thyroid cancer. Biomed Pharmacother. 2006;60:409-413.<br />
14. Mosci C, Iagaru A. PET/CT imaging <strong>of</strong> thyroid cancer. Clin Nucl Med.<br />
2011;36:e180-e185.<br />
15. McDougall IR, Iagaru A. Thyroid stunning: fact or fiction? Semin Nucl<br />
Med. 2011;41:105-112.<br />
388<br />
REFERENCES<br />
COLEVAS AND SHAH<br />
16. Ladenson PW, Braverman LE, Mazzaferri EL, et al. Comparison <strong>of</strong><br />
administration <strong>of</strong> recombinant human thyrotropin with withdrawal <strong>of</strong> thyroid<br />
hormone for radioactive iodine scanning in patients with thyroid carcinoma.<br />
N Engl J Med. 1997;337:888-896.<br />
17. Tala H, Robbins R, Fagin JA, et al. Five-year survival is similar in<br />
thyroid cancer patients with distant metastases prepared for radioactive<br />
iodine therapy with either thyroid hormone withdrawal or recombinant<br />
human TSH. J Clin Endocrinol Metab. 2011;96:2105-2111.<br />
18. Xing M, Westra WH, Tufano RP, et al. BRAF mutation predicts a<br />
poorer clinical prognosis for papillary thyroid cancer. J Clin Endocrinol<br />
Metab. 2005;90:6373-6379.<br />
19. Lanzi C, Cassinelli G, Nicolini V, et al. Targeting RET for thyroid<br />
cancer therapy. Biochem Pharmacol. 2009;77:297-309.<br />
20. Robbins RJ, Wan Q, Grewal RK, et al. Real-time prognosis for metastatic<br />
thyroid carcinoma based on 2-[18F]fluoro-2-deoxy-D-glucose-positron<br />
emission tomography scanning. J Clin Endocrinol Metab. 2006;91:498-505.<br />
21. Miyauchi A, Kudo T, Miya A, et al. Prognostic impact <strong>of</strong> serum<br />
thyroglobulin doubling-time under thyrotropin suppression in patients with<br />
papillary thyroid carcinoma who underwent total thyroidectomy. Thyroid.<br />
2011;21:707-716.<br />
22. Klein M, Vignaud JM, Hennequin V, et al. Increased expression <strong>of</strong><br />
the vascular endothelial growth factor is a pejorative prognosis marker in<br />
papillary thyroid carcinoma. J Clin Endocrinol Metab. 2001;86:656-658.<br />
23. Chrisoulidou A, Boudina M, Tzemailas A, et al. Histological subtype is<br />
the most important determinant <strong>of</strong> survival in metastatic papillary thyroid<br />
cancer. Thyroid Res. 2011;4:12.<br />
24. Ito Y, Higashiyama T, Takamura Y, et al. <strong>Clinical</strong> outcomes <strong>of</strong> patients<br />
with papillary thyroid carcinoma after the detection <strong>of</strong> distant recurrence.<br />
World J Surg. 2010;34:2333-2337.<br />
25. Shoup M, Stojadinovic A, Nissan A, et al. Prognostic indicators <strong>of</strong><br />
outcomes in patients with distant metastases from differentiated thyroid<br />
carcinoma. J Am Coll Surg. 2003;197:191-197.<br />
26. Showalter TN, Siegel BA, Moley JF, et al. Prognostic factors in patients<br />
with well-differentiated thyroid cancer presenting with pulmonary metastasis.<br />
Cancer Biother Radiopharm. 2008;23:655-659.<br />
27. Wells SA, Jr, Robinson BG, Gagel RF, et al. Vandetanib in patients<br />
with locally advanced or metastatic medullary thyroid cancer: a randomized,<br />
double-blind phase III trial. J Clin Oncol. <strong>2012</strong>;30:134-141.
Systemic Therapeutic Approaches to<br />
Advanced Thyroid Cancers<br />
By Michael E. Menefee, MD, Robert C. Smallridge, MD, and Keith C. Bible, MD, PhD,<br />
and the Mayo Clinic Endocrine Malignancies Disease-Oriented Group<br />
Overview: Until only recently, few effective systemic therapies<br />
were available to treat patients with metastatic thyroid<br />
cancers. Recent advances in better understanding the pathogenesis<br />
and altered signaling pathways—especially in medullary<br />
and differentiated thyroid cancers (MTCs and DTCs)—<br />
have begun to change this situation substantially. Vandetanib,<br />
an orally bioavailable inhibitor <strong>of</strong> the RET kinase that is<br />
constitutively activated in MTC, has now been approved by the<br />
U.S. Food and Drug Administration (FDA) for use in progressive<br />
and symptomatic metastatic MTC; it has been shown to<br />
delay time to progression relative to placebo in a randomized<br />
phase III trial. Further, vascular endothelial growth factor<br />
receptor (VEGF-R) inhibitory agents including sorafenib,<br />
sunitinib, pazopanib, and axitinib that are already approved in<br />
THYROID CANCER has become increasingly more common<br />
worldwide, doubling in incidence during the last<br />
decade in the United States to place it among the 10 most<br />
common cancers 1 and prompting increasing media and public<br />
attention. Although thyroid cancers have historically<br />
been more in the purview <strong>of</strong> endocrinologists than <strong>of</strong> medical<br />
oncologists, recent therapeutic innovations have made it<br />
increasingly important that medical oncologists develop<br />
expertise in the management <strong>of</strong> thyroid cancers, especially<br />
when advanced. 2 This expertise not only must include familiarity<br />
with the array <strong>of</strong> available and emerging therapeutics<br />
for various histologic subtypes <strong>of</strong> thyroid cancer, but also<br />
requires a working knowledge <strong>of</strong> when to apply available<br />
systemic therapies. In this regard, because most patients<br />
even with advanced differentiated and medullary thyroid<br />
cancers (DTCs and MTCs) are subject to slowly progressive<br />
disease compared with most other cancers encountered in<br />
medical oncology practice, optimal management <strong>of</strong>ten involves<br />
the delayed and conservative application <strong>of</strong> systemic<br />
therapies. Anaplastic thyroid cancer (ATC), in contrast, is<br />
perhaps the most aggressive <strong>of</strong> all solid tumors with a<br />
doubling time as brief as 10 days—thereby demanding<br />
immediate action. The following sections review recent progress<br />
in developing more effective systemic therapies for<br />
advanced thyroid cancers (Table 1), discussing each <strong>of</strong> the<br />
three major histotypes in separate sections, and providing<br />
general rationale for the application <strong>of</strong> available systemic<br />
therapies.<br />
Systemic Therapeutic Approaches to Advanced DTC<br />
DTC arises from the thyroxine-producing follicular cells <strong>of</strong><br />
the thyroid gland and occurs in two primary histotypes:<br />
papillary thyroid cancer (PTC), the most incident <strong>of</strong> all<br />
thyroid cancers that accounts for the majority <strong>of</strong> all cases<br />
and is the most indolent histotype overall; and follicular<br />
thyroid cancer (FTC), encompassing also the Hürthle cell<br />
(HCC) histologic variant, which tends to be somewhat more<br />
aggressive in behavior. Because DTC arises from follicular<br />
thyroid cells that accumulate iodine as required for thyroid<br />
hormone production, most DTCs at least initially retain the<br />
ability to concentrate iodine facilitated by a functional<br />
sodium iodine symporter (NIS). As a consequence, therapeu-<br />
the United States for use in advanced renal cell carcinoma<br />
have shown high response rates in treating advanced DTCs in<br />
multiple phase II trials, and have become commonly used in<br />
progressive radioiodine-refractory metastatic DTC. Yet additional<br />
agents are now in development, with several including<br />
XL184 (cabozantinib) also showing promise in DTC and MTC.<br />
In anaplastic thyroid cancer (ATC), progress has been slower,<br />
with the greatest apparent gains resulting more from the<br />
application <strong>of</strong> systemic therapies earlier in the disease<br />
course, especially when used in conjunction with initial surgical<br />
and radiation therapies. Despite recent progress, additional<br />
effective systemic therapeutic approaches remain<br />
sorely needed for treating metastatic MTC, DTC, and ATC.<br />
tic RAI remains the standard <strong>of</strong> care for initial therapy in<br />
metastatic DTC.<br />
It is critical to realize that only a small minority <strong>of</strong> all<br />
DTC patients have 1) RAI-refractory metastatic disease,<br />
2) disease not amenable to focal palliative approaches, and<br />
3) rapidly progressive, symptomatic, and imminently threatening<br />
disease sufficient to prompt consideration <strong>of</strong> additional<br />
systemic therapies beyond RAI. In decisions related to<br />
systemic therapy it is, therefore, important to carefully weigh<br />
the anticipated risks and benefits <strong>of</strong> a candidate treatment<br />
compared with the risks imposed by the disease itself.<br />
Because cytotoxic chemotherapy has historically had minimal<br />
efficacy in DTC, recent efforts have instead focused<br />
primarily on alternative approaches, summarized Harris<br />
and Bible. 2 In part, recent advances in DTC therapy arose<br />
in parallel with the development <strong>of</strong> vandetanib for use<br />
in treating advanced MTC. Vandetanib is a small-molecule<br />
inhibitor <strong>of</strong> tyrosine kinases—including the RET kinase that<br />
is <strong>of</strong> pathogenic importance in MTC because <strong>of</strong> its constitutive<br />
activation. Because RET is also upregulated in some<br />
DTCs, interest consequently also developed in parallel in<br />
applying kinase inhibitors in DTCs, resulting in multiple<br />
therapeutic clinical trials <strong>of</strong> tyrosine kinase inhibitors<br />
(TKIs) also in DTC.<br />
Strikingly, response rates to TKIs in advanced DTC have<br />
ranged from 10% to 50%, with attained responses <strong>of</strong>ten quite<br />
durable. Generally, all tested VEGF-R inhibitory TKIs have<br />
activity in DTC, with pazopanib, 3 sorafenib, 4,5 and axitinib 6<br />
seeming to have the highest response rates and most favorable<br />
toxicity pr<strong>of</strong>iles according to published results. Currently,<br />
a randomized phase III trial <strong>of</strong> sorafenib compared<br />
with placebo is ongoing in RAI-refractory DTC (clinicaltrials.<br />
From the Division <strong>of</strong> Medical <strong>Oncology</strong>; Division <strong>of</strong> Endocrinology, Mayo Clinic Florida,<br />
Jacksonville, FL; Division <strong>of</strong> Medical <strong>Oncology</strong>, Mayo Clinic, Rochester, MN.<br />
Authors’ disclosures <strong>of</strong> potential conflicts <strong>of</strong> interest are found at the end <strong>of</strong> this article.<br />
Address reprint requests to Keith C. Bible, MD, PhD, Mayo Clinic, 200 First Street SW,<br />
Rochester, MN 55901; email: bible.keith@mayo.edu.<br />
© <strong>2012</strong> by <strong>American</strong> <strong>Society</strong> <strong>of</strong> <strong>Clinical</strong> <strong>Oncology</strong>.<br />
1092-9118/10/1-10<br />
389
gov identifier NCT00984282), with other TKI trials also<br />
ongoing and planned in RAI-refractory metastatic DTC.<br />
Another antiangiogenic strategy that has been successful,<br />
but less well studied, has been the use <strong>of</strong> immunomodulatory<br />
agents (IMiDs). The effects <strong>of</strong> the IMiDs in thyroid<br />
cancer may extend beyond their impact on angiogenesis.<br />
Thalidomide, 7 and more recently the thalidomide analog<br />
lenalidomide, 8 have produced clinical benefit in patients<br />
with DTC, with response rates ranging between 18% and<br />
22% and stable disease between 32% and 44%. The IMiDs<br />
are another therapeutic option in the limited, although<br />
rapidly expanding, armamentarium against DTC.<br />
The paradigm indicating that cytotoxic therapies are<br />
ineffective in DTC has also been changing. For example, the<br />
combination <strong>of</strong> gemcitabine and oxaliplatin (GEMOX) was<br />
recently reported to produce a 57% response rate, with 7%<br />
complete responses 9 ; these results rival or are superior to<br />
outcomes reported from the use <strong>of</strong> TKIs in DTC. Hence,<br />
newer cytotoxic agents and combinations such as GEMOX<br />
KEY POINTS<br />
● Vandetanib, an orally bioavailable inhibitor <strong>of</strong> the<br />
RET kinase that is constitutively activated in medullary<br />
thyroid cancer (MTC), has been approved by the<br />
U.S. Food and Drug Administration for use in treating<br />
progressive and symptomatic metastatic MTC; it<br />
has been shown to delay time to progression compared<br />
with placebo in a randomized phase III trial.<br />
● Vascular endothelial growth factor receptor<br />
(VEGF-R) inhibitory and antiangiogenic agents including<br />
sorafenib, sunitinib, axitinib, and pazopanib<br />
have shown high response rates in treating advanced<br />
differentiated thyroid cancers (DTCs) in multiple<br />
phase II trials, and are now commonly used in treating<br />
patients with progressive radioiodine (RAI)refractory<br />
metastatic DTC.<br />
● In the decision to initiate systemic therapies in metastatic<br />
and progressive DTC or MTC, considerable<br />
restraint should be exercised, as many patients will<br />
have slowly progressive disease not requiring active<br />
systemic therapy. Alternatives for focal palliation <strong>of</strong><br />
areas <strong>of</strong> threatening disease (especially in the neck<br />
and in bone) should first be prominently considered<br />
before initiation <strong>of</strong> systemic therapies. Moreover, the<br />
risk imposed by the disease versus by the systemic<br />
approaches under consideration should be carefully<br />
weighed.<br />
● Application <strong>of</strong> systemic therapies earlier in the course<br />
<strong>of</strong> anaplastic thyroid cancer (ATC) has begun to show<br />
promise, especially when used in conjunction with<br />
initial surgical and radiation therapies.<br />
● Although considerable recent progress has been<br />
made in better understanding candidate therapeutic<br />
targets and approaches to treating advanced thyroid<br />
cancers, additional and more effective systemic therapies<br />
for these cancers remain sorely needed.<br />
390<br />
Table 1. Selected Therapeutics for Advanced Thyroid Cancers<br />
Histotype Agent Molecular Target(s)<br />
Differentiated thyroid<br />
cancer<br />
Medullary thyroid<br />
cancer<br />
Anaplastic thyroid<br />
cancer<br />
Axitinib6 Kinases, including VEGF-R<br />
Gemcitabine/oxaliplatin9 DNA<br />
Lenalidomide8 Incompletely defined<br />
Pazopanib3 Kinases, including VEGF-R<br />
Thalidomide7 Incompletely defined<br />
Sorafenib4,5 Kinases, including VEGF-R<br />
Sunitinib23 Kinases, including VEGF-R<br />
Vandetanib10 Kinases, including RET<br />
Cabozantinib11 Kinases, including RET, MET,<br />
VEGF-R<br />
Crolibulin17,18 Microtubules<br />
Docetaxel16 Microtubules<br />
Doxorubicin � cisplatin19 Topoisomerase II/DNA<br />
Paclitaxel15 Microtubules<br />
Abbreviation: VEGF-R, vascular endothelial growth factor receptor.<br />
MENEFEE ET AL<br />
provide an alternative approach to TKIs to consider in<br />
patients with RAI-refractory DTC.<br />
Although there has been considerable progress in developing<br />
therapies for patients with rapidly progressive and<br />
imminently threatening radioactive iodine-insensitive DTC,<br />
most patients with DTC (even those with metastatic disease)<br />
are subject to overall indolent disease courses. As a<br />
result, there should be prominent consideration <strong>of</strong> focal<br />
palliation <strong>of</strong> locally threatening disease (especially in the<br />
neck and in bone) before initiation <strong>of</strong> systemic therapies.<br />
Moreover, the risks imposed by the disease versus those<br />
imposed by contemplated systemic therapies must be carefully<br />
weighed so as to ensure that the initiation <strong>of</strong> systemic<br />
therapy is in a particular patient’s best interests. Almost all<br />
patients receiving treatment with TKIs will experience<br />
adverse effects, with fatigue, diarrhea, induced hypertension,<br />
and cutaneous toxicities overall quite common, and<br />
with serious adverse effects also encountered in a minority<br />
<strong>of</strong> patients. Similarly, the application <strong>of</strong> cytotoxic chemotherapy<br />
is <strong>of</strong>ten associated with nausea, cytopenias, and<br />
other adverse effects that have potential to adversely affect<br />
patient quality <strong>of</strong> life and impose risks to patient welfare.<br />
Overall, judicious application <strong>of</strong> available systemic therapies<br />
is therefore required in advanced DTC.<br />
Systemic Therapeutic Approaches to Advanced MTC<br />
As noted earlier herein, MTC has proven to be an excellent<br />
model system in the development and study <strong>of</strong> novel therapeutics<br />
in thyroid cancer. In particular, MTC is commonly<br />
heritable and characterized by constitutive activation <strong>of</strong> the<br />
RET kinase—triggered by a germ-line mutation in the case<br />
<strong>of</strong> heritable MTC, and alternatively by tumor-specific mutation<br />
in the case <strong>of</strong> sporadic MTC. RET activation drives<br />
proliferation in affected parafollicular “C”-cells within the<br />
thyroid, contributing thereby to MTC pathogenesis and<br />
serving as a specific mutational therapeutic molecular target<br />
in MTC.<br />
The RET inhibitor vandetanib has been evaluated in<br />
phase II trials and in a randomized phase III trial compared<br />
with placebo, with results indicating a high response rate<br />
and improved progression-free survival in patients receiving<br />
vandetanib compared with those receiving placebo (as a<br />
result <strong>of</strong> the cross-over trial design; however, the effects <strong>of</strong><br />
vandetanib on overall survival are not reliably assessable). 10<br />
As a result, vandetanib was approved by the U.S. Food and
APPROACHES TO ADVANCED THYROID CANCERS<br />
Drug Administration (FDA) for use in progressive and<br />
symptomatic metastatic MTC in 2011.<br />
Other RET-inhibitory small molecules have also shown<br />
promise in MTC, including XL184 (cabozantinib), 11 which<br />
has also been subject to a recent randomized phase III trial<br />
in MTC that has yielded apparently promising results,<br />
but that has not yet been published. Furthermore, VEGFinhibitory<br />
TKIs also appear to have activity in MTC and are<br />
the subject <strong>of</strong> further study in advanced MTC. 2 It should<br />
also be noted that cytotoxic chemotherapy has shown some<br />
activity in MTC, albeit perhaps more modest in extent than<br />
seen in response to TKI therapy. 12-14<br />
As noted above in the case <strong>of</strong> DTC, potential risks and<br />
benefits <strong>of</strong> candidate systemic therapeutic approaches<br />
should be carefully considered before their initiation in<br />
patients with MTC because many such patients will have<br />
indolent disease courses most <strong>of</strong>ten not requiring systemic<br />
therapies.<br />
Systemic Therapeutic Approaches to Advanced ATC<br />
Cytotoxic chemotherapy remains the most effective, albeit<br />
marginally so, approach to treating advanced ATC. The two<br />
classes <strong>of</strong> agents with highest activity overall in ATC are<br />
anthracyclines and antimicrotubule inhibitors, with the<br />
greatest amount <strong>of</strong> data available in the case <strong>of</strong> paclitaxel. In<br />
particular, the CATCHIT trial demonstrated single-agent<br />
paclitaxel to produce a transient response rate <strong>of</strong> 53% in<br />
ATC—but the requirement for confirmation <strong>of</strong> response had<br />
to be shortened to 2 weeks to yield this result 15 ; to be sure,<br />
responses to most any systemic therapy in advanced ATC<br />
tend to be very brief. The related taxane docetaxel has<br />
also shown some activity in ATC. 16 EPC2407 (Crolibulin;<br />
EpiCept Corporation, Tarrytown, NY) have also shown<br />
some promise in ATC, 17,18 with a phase II trial <strong>of</strong> EPC2407<br />
combined with cisplatin now ongoing (clinicaltrials.gov identifier<br />
NCT01240590).<br />
Doxorubicin, approved by the U.S FDA for use in advanced<br />
thyroid cancer in the 1970s, has been used as a single<br />
agent in thyroid cancer, as well as combined with cisplatin,<br />
with somewhat better outcomes resulting from the twoagent<br />
combination. 19<br />
Although results from evaluation <strong>of</strong> TKIs in ATC have<br />
been disappointing overall, imatinib monotherapy was recent<br />
reported to induce responses in several patients with<br />
ATC. 20 Further, the combination <strong>of</strong> the TKI pazopanib with<br />
Authors’ Disclosures <strong>of</strong> Potential Conflicts <strong>of</strong> Interest<br />
Author<br />
Michael E. Menefee*<br />
Robert C. Smallridge*<br />
Keith C. Bible*<br />
*No relevant relationships to disclose.<br />
Employment or<br />
Leadership<br />
Positions<br />
Consultant or<br />
Advisory Role<br />
1. Siegel R, Naishadham D, Jemal A. Cancer statistics, <strong>2012</strong>. CA Cancer<br />
J Clin. <strong>2012</strong>;62:10-29.<br />
2. Harris PJ, Bible KC. Emerging therapeutics for advanced thyroid<br />
malignancies: rationale and targeted approaches. Expert Opin Investig Drugs.<br />
2011;20:1357-1275.<br />
3. Bible KC, Suman VJ, Molina JR, et al. Efficacy <strong>of</strong> pazopanib in<br />
paclitaxel has produced sufficiently encouraging preclinical<br />
results to be subject to evaluation in a randomized Radiation<br />
<strong>Oncology</strong> Treatment Group (RTOG) <strong>of</strong> intensity-modulated<br />
radiotherapy plus paclitaxel with or without pazopanib<br />
(clinicaltrials.gov identifier NCT01236547). Yet additional<br />
agents and combinations are actively being evaluated in<br />
advanced ATC. 2<br />
Also <strong>of</strong> potential promise in ATC is the application <strong>of</strong><br />
systemic therapies in conjunction with the initial treatment<br />
<strong>of</strong> neck-confined disease—especially when systemic therapy<br />
is combined with radiation therapy. In applying this approach,<br />
several groups have observed unexpectedly favorable<br />
outcomes and survival, 21,22 suggesting that it may be<br />
fruitful in patients with good performance status seeking an<br />
aggressive approach to their initial therapy.<br />
Conclusion<br />
Historically, progress had been slow in developing more<br />
effective systemic approaches to treating advanced thyroid<br />
cancers. This situation has recently begun to change with<br />
the elaboration <strong>of</strong> an increasing amount <strong>of</strong> information<br />
related to the molecular pathways contributing to thyroid<br />
cancer pathogenesis, thereby resulting in the identification<br />
<strong>of</strong> an ever-increasing array <strong>of</strong> candidate therapeutic molecular<br />
targets that has already begun to lead to therapeutic<br />
advances in thyroid cancers. In particular, the RET kinase<br />
inhibitor vandetanib has been identified as sufficiently<br />
promising to merit its approval by the U.S. FDA for use in<br />
metastatic progressive and symptomatic medullary thyroid<br />
cancer, because it has been shown to delay time to progression<br />
and to induce a high rate <strong>of</strong> clinical response in MTC.<br />
Additionally, VEGF-R–inhibitory multi-targeted kinase inhibitors<br />
including sunitinib, sorafenib, axitinib, and pazopanib<br />
have demonstrated high rates <strong>of</strong> durable clinical<br />
responses in DTC, prompting their application as a new<br />
“standard <strong>of</strong> care” in treating patients with metastatic<br />
RAI-refractory DTC. Although therapeutic progress in ATC<br />
has been slow, emerging evidence suggests promise in the<br />
early application <strong>of</strong> cytotoxic chemotherapy in conjunction<br />
with other up-front therapies (e.g., radiation therapy and<br />
surgery) for locoregionally confined ATC. Nevertheless, despite<br />
considerable recent progress, further basic advances<br />
and therapeutic progress related to thyroid cancers are still<br />
very much needed, especially for patients with rapidly<br />
progressive metastatic disease.<br />
Stock<br />
Ownership Honoraria<br />
REFERENCES<br />
Research<br />
Funding<br />
Expert<br />
Testimony<br />
Other<br />
Remuneration<br />
progressive, radioiodine-refractory, metastatic differentiated thyroid cancers:<br />
results <strong>of</strong> a phase 2 consortium study. Lancet Oncol. 2010;11:962-972.<br />
4. Kloos RT, Ringel MD, Knopp MV, et al. Phase II trial <strong>of</strong> sorafenib in<br />
metastatic thyroid cancer. J Clin Oncol. 2009;27:1675-1684.<br />
5. Gupta-Abramson V, Troxel AB, Nellore A, et al. Phase II trial <strong>of</strong><br />
sorafenib in advanced thyroid cancer. J Clin Oncol. 2008;26:4714-4719.<br />
391
6. Cohen EE, Rosen LS, Vokes EE, et al. Axitinib is an active treatment<br />
for all histologic subtypes <strong>of</strong> advanced thyroid cancer: results from a phase II<br />
study. J Clin Oncol. 2008;26:4708-4713.<br />
7. Ain KB, Lee C, Williams KD. Phase II trial <strong>of</strong> thalidomide for therapy<br />
<strong>of</strong> radioiodine-unresponsive and rapidly progressive thyroid carcinomas.<br />
Thyroid. 2007;17:663-670.<br />
8. Ain KB, Lee C, Holbrook KM, et al. Phase II study <strong>of</strong> lenalidomide in<br />
distantly metastatic, rapidly progressive, and radioiodine-unresponsive thyroid<br />
carcinomas: preliminary results. J Clin Oncol. 2008;26:XXX (suppl; abstr 6027).<br />
9. Spano JP, Vano Y, Vignot S, et al. GEMOX regimen in the treatment<br />
<strong>of</strong> metastatic differentiated refractory thyroid carcinoma. Med Oncol. Epub<br />
2011 Sep 25.<br />
10. Wells SA Jr, Robinson BG, Gagel RF, et al. Vandetanib in Patients<br />
With Locally Advanced or Metastatic Medullary Thyroid Cancer: A Randomized,<br />
Double-Blind Phase III Trial. J Clin Oncol. <strong>2012</strong>;30:134-141.<br />
11. Kurzrock R, Sherman SI, Ball DW, et al. Activity <strong>of</strong> XL184 (Cabozantinib),<br />
an oral tyrosine kinase inhibitor, in patients with medullary thyroid<br />
cancer. J Clin Oncol. 2011;29:2660-2666.<br />
12. Nocera M, Baudin E, Pellegriti G, et al. Treatment <strong>of</strong> advanced<br />
medullary thyroid cancer with an alternating combination <strong>of</strong> doxorubicinstreptozocin<br />
and 5 FU-dacarbazine. Groupe d’Etude desTumeurs à Calcitonine<br />
(GETC). Br J Cancer. 2000;83:715-718.<br />
13. Orlandi F, Caraci P, Berruti A, et al. Chemotherapy with dacarbazine<br />
and 5-fluorouracil in advanced medullary thyroid cancer. Ann Oncol. 1994;5:<br />
763-765.<br />
14. Wu LT, Averbuch SD, Ball DW, et al. Treatment <strong>of</strong> advanced medullary<br />
thyroid carcinoma with a combination <strong>of</strong> cyclophosphamide, vincristine, and<br />
dacarbazine. Cancer. 1994;73:432-436.<br />
15. Ain KB, Egorin MJ, DeSimone PA. Treatment <strong>of</strong> anaplastic thyroid<br />
392<br />
MENEFEE ET AL<br />
carcinoma with paclitaxel: phase 2 trial using ninety-six-hour infusion.<br />
Collaborative Anaplastic Thyroid Cancer Health Intervention Trials (CATCHIT)<br />
Group. Thyroid. 2000;10:587-594.<br />
16. Kawada K, Kitagawa K, Kamei S, et al. The feasibility study <strong>of</strong><br />
docetaxel in patients with anaplastic thyroid cancer. Jpn J Clin Oncol.<br />
2010;40:596-599.<br />
17. Dowlati A, Robertson K, Cooney M, et al. A phase I pharmacokinetic<br />
and translational study <strong>of</strong> the novel vascular targeting agent combretastatin<br />
a-4 phosphate on a single-dose intravenous schedule in patients with advanced<br />
cancer. Cancer Res. 2002;62:3408-3416.<br />
18. Mooney CJ, Nagaiah G, Fu P, et al. A phase II trial <strong>of</strong> fosbretabulin in<br />
advanced anaplastic thyroid carcinoma and correlation <strong>of</strong> baseline serumsoluble<br />
intracellular adhesion molecule-1 with outcome. Thyroid. 2009;19:<br />
233-240.<br />
19. Shimaoka K, Schoenfeld DA, DeWys WD, et al. A randomized trial <strong>of</strong><br />
doxorubicin versus doxorubicin plus cisplatin in patients with advanced<br />
thyroid carcinoma. Cancer. 1985;56:2155-2160.<br />
20. Ha HT, Lee JS, Urba S, et al. A phase II study <strong>of</strong> imatinib in patients<br />
with advanced anaplastic thyroid cancer. Thyroid. 2010;20:975-980.<br />
21. Foote RL, Molina JR, Kasperbauer JL, et al. Enhanced survival in<br />
locoregionally confined anaplastic thyroid carcinoma: a single-institution<br />
experience using aggressive multimodal therapy. Thyroid. 2011;21:25-30.<br />
22. Troch M, Koperek O, Scheuba C, et al. High efficacy <strong>of</strong> concomitant<br />
treatment <strong>of</strong> undifferentiated (anaplastic) thyroid cancer with radiation and<br />
docetaxel. J Clin Endocrinol Metab. 2010;95:E54-57.<br />
23. Carr LL, Mank<strong>of</strong>f DA, Goulart BH, et al. Phase II study <strong>of</strong> daily<br />
sunitinib in FDG-PET-positive, iodinerefractory differentiated thyroid cancer<br />
and metastatic medullary carcinoma <strong>of</strong> the thyroid with functional imaging<br />
correlation. Clin Cancer Res. 2010;16:5260-5268.
AVOIDING OVERDIAGNOSIS AND<br />
OVERTREATMENT OF COMMON CANCERS<br />
CHAIR<br />
Ian M. Thompson Jr, MD<br />
University <strong>of</strong> Texas Health Science Center at San Antonio<br />
San Antonio, TX<br />
SPEAKERS<br />
Ruth D. Etzioni, PhD<br />
University <strong>of</strong> Washington<br />
Seattle, WA<br />
Laura Esserman, MD, MBA<br />
University <strong>of</strong> California, San Francisco<br />
San Francisco, CA
Overdiagnosis and Overtreatment <strong>of</strong><br />
Prostate Cancer<br />
Overview: Prostate cancer is a ubiquitous disease, affecting<br />
as many as two-thirds <strong>of</strong> men in their 60s. Through widespread<br />
prostate-specific antigen (PSA) testing, increasing rates <strong>of</strong><br />
prostate biopsy, and increased sampling <strong>of</strong> the prostate, a<br />
larger fraction <strong>of</strong> low-grade, low-volume tumors have been<br />
detected, consistent with tumors <strong>of</strong>ten found at autopsy.<br />
These tumors have historically been treated in a manner<br />
similar to that used for higher-grade tumors but, more recently,<br />
it has become evident that with a plan <strong>of</strong> active<br />
surveillance that reserves treatment for only those patients<br />
whose tumors show evidence <strong>of</strong> progression, very high<br />
disease-specific survival can be achieved. Unfortunately, the<br />
frequency <strong>of</strong> recommendation <strong>of</strong> an active surveillance strat-<br />
WITH THE diffusion <strong>of</strong> PSA screening that first began<br />
in the mid-1980s and with subsequent changes in<br />
the way we diagnose prostate cancer, the tumors <strong>of</strong> today<br />
are remarkably different from those <strong>of</strong> the 1980s or even<br />
1990s. At its most basic level, the opportunity for overdiagnosis<br />
<strong>of</strong> prostate cancer (overdiagnosis can be defined as<br />
detection <strong>of</strong> cancers that will ultimately not harm the host)<br />
must be understood to be related to the background rate <strong>of</strong><br />
the disease. The best insight into this potential opportunity<br />
for diagnosis <strong>of</strong> prostate cancer comes from autopsy studies<br />
<strong>of</strong> men who died as a result <strong>of</strong> other causes. Careful sectioning<br />
<strong>of</strong> prostates in these men gives us an idea as to the lower<br />
bound estimate for the risk <strong>of</strong> the disease. (It is the lower<br />
bound estimate because men who have been treated with<br />
surgery or radiation are effectively censored from these<br />
analyses.) Nonetheless, these rates are substantial, as summarized<br />
in Table 1. 1 Given that a man’s life expectancy<br />
approaches 80 years in the United States, it would seem the<br />
infrequent aging man who does not harbor at least a small<br />
tumor in his prostate that, if struck with a prostate biopsy<br />
needle, would result in a prostate cancer diagnosis. Indeed,<br />
to make a diagnosis <strong>of</strong> prostate cancer, which is generally<br />
asymptomatic until metastases develop, there are three<br />
prerequisites: 1) the physician must suspect cancer, 2) the<br />
patient must accept a prostate biopsy, and 3) the biopsy<br />
needle must strike the tumor. As will be seen, these rates<br />
have changed variably over the years.<br />
How Is the Prostate Cancer <strong>of</strong> <strong>2012</strong> Different from<br />
the Prostate Cancer <strong>of</strong> Prior Decades?<br />
Physician Suspicion <strong>of</strong> Prostate Cancer<br />
Before PSA, the only method for a physician to suspect the<br />
presence <strong>of</strong> prostate cancer (at least, early prostate cancer)<br />
was through an abnormal digital rectal examination (DRE).<br />
This was generally uncommon. In one series we conducted<br />
prospectively in the early 1980s, in 2,005 men undergoing<br />
biopsy, only 65 abnormal examinations were identified. 2<br />
Of these, only 17 proved to be cancer. With the advent <strong>of</strong><br />
PSA testing, it was not necessarily a substantially greater<br />
percentage <strong>of</strong> men in whom prostate cancer was suspected<br />
(approximately 8% <strong>of</strong> the population has a PSA greater than<br />
4.0 ng/mL) but a far greater fraction <strong>of</strong> the population was<br />
interested in PSA testing, likely as a result <strong>of</strong> the poor<br />
By Ian M. Thompson Jr., MD<br />
egy in the United States is low. An alternative strategy to<br />
improve prostate cancer detection is through selected biopsy<br />
<strong>of</strong> those men who are at greater risk <strong>of</strong> harboring high-grade,<br />
potentially lethal cancer. This strategy is currently possible<br />
through the use <strong>of</strong> risk assessment tools such as the Prostate<br />
Cancer Prevention Trial Risk Calculator (www.prostate.cancer.<br />
risk.calculator.com) as well as others. These tools can predict<br />
with considerable accuracy a man’s risk <strong>of</strong> low-grade and<br />
high-grade cancer, allowing informed decision making for the<br />
patient with a goal <strong>of</strong> detection <strong>of</strong> high-risk disease. Ultimately,<br />
other biomarkers including PCA3, TMPRSS2:ERG, and<br />
[-2]proPSA will likely aid in discriminating these two types <strong>of</strong><br />
cancer before biopsy.<br />
acceptance <strong>of</strong> DRE by U.S. males. It was then with this<br />
testing that we witnessed a spike in prostate cancer diagnosis<br />
in the early 1990s, which subsequently fell back (after a<br />
harvest <strong>of</strong> prevalent tumors) to a rate approximately double<br />
that <strong>of</strong> the pre-PSA 1980s. Presently, approximately 75<br />
percent <strong>of</strong> at-risk men have undergone a PSA test and<br />
approximately 50% are tested regularly.<br />
Further changes in suspicion <strong>of</strong> prostate cancer then<br />
developed. These included 1) the recognition that prostate<br />
cancer was quite prevalent at lower levels <strong>of</strong> PSA, 2) the<br />
recognition that other factors such as family history, race,<br />
and age could be incorporated into a risk assessment to lead<br />
to a biopsy in a man with a PSA less than 4.0 ng/mL, and<br />
3) the use <strong>of</strong> changes in PSA (PSA velocity) to prompt biopsy<br />
at even lower levels <strong>of</strong> PSA. 3-5 All <strong>of</strong> these biopsy prompts<br />
have led to dramatic increases in the use <strong>of</strong> prostate biopsy.<br />
The sum total <strong>of</strong> these events has dramatically increased the<br />
likelihood that a man, if a PSA is obtained, will receive a<br />
recommendation to consider prostate biopsy.<br />
Patient Acceptance <strong>of</strong> Prostate Biopsy<br />
This variable can dampen the rate <strong>of</strong> prostate cancer<br />
diagnosis because many men who receive PSA results that<br />
are greater than 4.0 ng/mL will opt to not have a prostate<br />
biopsy. In the Prostate, Lung, Colorectal and Ovarian<br />
(PLCO) Cancer Screening Trial in the United States, for<br />
example, at the baseline PSA test, if a man’s PSA exceeded<br />
4.0 ng/mL, within 1 and 3 years, the likelihood that he<br />
underwent prostate biopsy was 41% and 64%, respectively. 6<br />
Almost certainly, the way the PSA data are presented by the<br />
physician will play a part in whether the patient undergoes<br />
prostate biopsy.<br />
From the Cancer Therapy and Research Center, University <strong>of</strong> Texas Health Science<br />
Center at San Antonio, San Antonio, TX.<br />
Author’s disclosure <strong>of</strong> potential conflicts <strong>of</strong> interest are found at the end <strong>of</strong> this article.<br />
Address reprint requests to Ian M. Thompson Jr., MD, Cancer Therapy and Research<br />
Center, University <strong>of</strong> Texas Health Science Center at San Antonio, 7979 Wurzbach Road,<br />
San Antonio, TX; email: thompsoni@uthscsa.edu.<br />
© <strong>2012</strong> by <strong>American</strong> <strong>Society</strong> <strong>of</strong> <strong>Clinical</strong> <strong>Oncology</strong>.<br />
1092-9118/10/1-10<br />
e35
Table 1. Prevalence <strong>of</strong> Prostate Cancer at Autopsy 1<br />
Age (Years) U.S. Prevalence–Whites (%) U.S. Prevalence–Blacks (%)<br />
41–50 37 43<br />
51–60 44 46<br />
61–70 65 70<br />
71–80 83 81<br />
The Prostate Needle Must Strike the Tumor<br />
Urologists who practiced in the pre-PSA era had the<br />
challenging experience <strong>of</strong> performing prostate biopsies with<br />
instruments that <strong>of</strong>tentimes poorly sampled the prostate.<br />
Additionally, it was not uncommon for a physician to obtain<br />
as few as two biopsy cores from the prostate in determining<br />
whether cancer was present. Certainly, because prostate<br />
cancer is multifocal and <strong>of</strong>tentimes occupies a small fraction<br />
<strong>of</strong> the gland, this poor sampling <strong>of</strong> the prostate missed many<br />
small tumors. On the other hand, if cancer was detected, it<br />
<strong>of</strong>ten was large. Before the late 1980s, the most common<br />
biopsy scheme was four biopsy cores: two from each side <strong>of</strong><br />
the gland (left/right) and, on each side, one core from the<br />
base and one core from the apex (cranial and caudal,<br />
respectively). In the late 1980s, Hodge and colleagues suggested<br />
the use <strong>of</strong> a “sextant” biopsy, arguing that six cores<br />
were more likely to identify cancer than two or four. 7 Over<br />
the ensuing years, authors found that increasing the biopsies<br />
to 10 or 12 cores further increased the likelihood <strong>of</strong><br />
cancer diagnosis. 8 More recently, so-called saturation biopsies<br />
have been advocated in which, in some schemes, cores<br />
are obtained from every cubic centimeter <strong>of</strong> the prostate,<br />
totaling 20–40 biopsy samples. 9<br />
To even the most naïve observer, this ever-increasing<br />
number <strong>of</strong> biopsy cores would clearly increase the risk <strong>of</strong><br />
KEY POINTS<br />
● Prostate cancer is very common in the aging U.S.<br />
male population and is most commonly a low-grade,<br />
low-volume tumor that poses a very low risk <strong>of</strong><br />
progression and death.<br />
● PSA testing and prostate biopsy have increasingly<br />
resulted in detection <strong>of</strong> low-risk tumors; nonetheless,<br />
these tumors are very <strong>of</strong>ten treated radically with<br />
surgery or radiation.<br />
● It is possible, through the use <strong>of</strong> risk assessment<br />
tools, to identify men who are most likely to have<br />
high-risk cancer (for biopsy) and those who are most<br />
likely to have low-risk, potentially indolent cancer (to<br />
not have a biopsy).<br />
● A range <strong>of</strong> novel biomarkers are in the process <strong>of</strong><br />
validation and are predictive <strong>of</strong> high-risk prostate<br />
cancer.<br />
● The future <strong>of</strong> prostate cancer detection will likely<br />
incorporate prostate-specific antigen and other biomeasures<br />
(e.g., digital rectal examination, age, body<br />
mass index), as well as rationally incorporate novel<br />
biomarkers associated with risk <strong>of</strong> high-grade cancer,<br />
to identify the man who is most likely to benefit from<br />
diagnosis and, ultimately, treatment <strong>of</strong> cancer.<br />
e36<br />
IAN M. THOMPSON JR.<br />
detection <strong>of</strong> very small, potentially-inconsequential prostate<br />
cancer. It must be understood, however, that in the early<br />
days <strong>of</strong> PSA testing, physicians were frustrated by patients<br />
who underwent a prostate biopsy in which a small amount <strong>of</strong><br />
cancer was found, and then, on prostatectomy, it was found<br />
that the tumor was large and extraprostatic. It was thus<br />
with good intentions that this increase in the number <strong>of</strong><br />
biopsy cores occurred; the natural result, however, was that<br />
smaller and smaller tumors were detected.<br />
The cumulative effect <strong>of</strong> these changes during the last<br />
30 years has been a growing fraction <strong>of</strong> patients with<br />
small, low-grade cancers detected. Increasing this rate<br />
have been several other events. First, many patients are<br />
now having multiple sets <strong>of</strong> biopsies as they see physicians<br />
who continue to react to elevated PSA levels. Effectively<br />
then, a man who has had one negative 12-core biopsy in<br />
2007, followed by another negative 12-core biopsy in 2009,<br />
and then has another 12-core biopsy in 2011 that shows<br />
cancer in one core, has effectively had a 36-core biopsy with<br />
one <strong>of</strong> 36 cores showing a low-grade cancer. For many <strong>of</strong><br />
these men, it is highly likely that the cancer was indeed<br />
present in 2007 and 2009 and just not struck by the needle.<br />
The second event is that high-volume prostate cancers that<br />
were present in the 1990s and early 2000s and detected on a<br />
first biopsy after screening, have generally been treated and<br />
effectively removed from the screened population. Thus,<br />
patients who are currently being screened <strong>of</strong>ten have undergone<br />
multiple PSA tests and may have had multiple negative<br />
biopsies, and thus are at intrinsically lower risk <strong>of</strong> a<br />
high-volume tumor.<br />
What Is the Evidence that Prostate Cancers Are<br />
Overdetected or Overtreated?<br />
Regarding overdetection, a fairly clear piece <strong>of</strong> evidence in<br />
support <strong>of</strong> this is the ratio <strong>of</strong> annual incidence to mortality.<br />
In 2011, it was expected that 241,740 prostate cancers would<br />
be detected and that 28,170 men would die as a result <strong>of</strong><br />
their cancer. 10 If no overdetection occurred, the ratio would<br />
be closer to 1:1 (by contrast, the rates for lung cancer are<br />
226,160 and 160,340, respectively). An additional piece <strong>of</strong><br />
evidence can be seen through the lifetime risk <strong>of</strong> prostate<br />
cancer (now 16.48%) as opposed to the lifetime risk <strong>of</strong><br />
prostate cancer death (now 2.77%). 11 It is important to<br />
recognize that this lifetime risk <strong>of</strong> prostate cancer is this<br />
high with only approximately 50% <strong>of</strong> the population being<br />
screened for prostate cancer on a regular basis; one might<br />
expect a rate <strong>of</strong> 20% to 30% if all men underwent screening<br />
annually.<br />
There is also clear evidence <strong>of</strong> overtreatment. For the<br />
purposes <strong>of</strong> this article, we will define overtreatment as an<br />
active treatment (e.g., surgery, radiation, hormones, or a<br />
combination <strong>of</strong> these) in a patient who would never have<br />
had symptoms nor have died as a result <strong>of</strong> his prostate<br />
cancer. The primary evidence <strong>of</strong> this comes from the many<br />
series <strong>of</strong> patients with prostate cancer who have been<br />
managed with either watchful waiting (WW) or active surveillance<br />
(AS). It is important to distinguish these two<br />
management plans. In the case <strong>of</strong> WW, the patient undergoes<br />
no monitoring and is not <strong>of</strong>fered an intervention for<br />
cure; he receives treatment only should symptoms or evidence<br />
<strong>of</strong> metastatic disease develop. In the case <strong>of</strong> AS, a<br />
low-risk patient is monitored with serial PSA and DRE<br />
testing and on a periodic basis (e.g., every 1–2 years) and he
OVERDIAGNOSIS AND OVERTREATMENT OF PROSTATE CANCER<br />
undergoes prostate biopsy; if there is evidence <strong>of</strong> development<br />
<strong>of</strong> a more aggressive or larger tumor, treatment can<br />
then be instituted. Thus, AS seeks to reserve treatment for<br />
patients in whom there is evidence <strong>of</strong> a developing risk <strong>of</strong> an<br />
aggressive tumor.<br />
Many early series <strong>of</strong> WW attested to the fact that even<br />
tumors detected with DRE (<strong>of</strong>ten large and extraprostatic<br />
tumors), could be observed and, with extended follow-up,<br />
most patients did not develop metastases nor did they die<br />
as a result <strong>of</strong> prostate cancer. 12 Since the advent <strong>of</strong> PSA<br />
testing, there has been growing evidence that these lowergrade,<br />
lower-volume tumors can be watched carefully,<br />
treated only in the event <strong>of</strong> evidence <strong>of</strong> a more aggressive or<br />
larger tumor, thus avoiding treatment in many patients and<br />
still showing low rates <strong>of</strong> metastasis and cancer death.<br />
There are many examples <strong>of</strong> these series, including that <strong>of</strong><br />
Klotz and colleagues. 13 In their series, the risk <strong>of</strong> prostate<br />
cancer death at 5 years was 0.3% and at 10 years was 2.8%.<br />
It was important to recognize in this series as well that 17%<br />
<strong>of</strong> the patients undergoing AS actually had cancer with a<br />
Gleason score <strong>of</strong> 7.<br />
Sealing the evidence <strong>of</strong> overtreatment are the data from<br />
the CAPSURE prostate cancer registry, a collection <strong>of</strong> a<br />
range <strong>of</strong> urologic practices designed to evaluate practice<br />
patterns related to prostate cancer. In one <strong>of</strong> their studies,<br />
Cooperberg and colleagues found that 92% to 98% <strong>of</strong> patients<br />
who had the lowest tumor risk scores (and presumably<br />
were eligible for AS) were treated with radical surgery,<br />
radiation, or hormone therapy. 14 This stunning observation<br />
that at least 92% <strong>of</strong> men who had a 10-year risk <strong>of</strong> prostate<br />
cancer death <strong>of</strong> 2.8% received aggressive management,<br />
strongly suggests that overtreatment is indeed occurring in<br />
this disease. On the other hand, it is certain that some <strong>of</strong><br />
these men are making an individualized decision that this<br />
risk is high enough to justify treatment.<br />
What Is the Solution to Overdiagnosis and<br />
Overtreatment in Prostate Cancer?<br />
Prostate Cancer Prevention<br />
Many institutions are increasingly focusing on methods to<br />
reduce these problems in the U.S. population that cause<br />
substantial resources to be expended unnecessarily and that<br />
lead to substantial morbidity including sexual dysfunction,<br />
urinary obstruction or incontinence, as well as GI complications.<br />
Certainly, one <strong>of</strong> the methods to reduce this problem is<br />
through prostate cancer prevention. Currently available are<br />
two agents that have been clearly demonstrated to reduce a<br />
man’s risk <strong>of</strong> detection <strong>of</strong> a low-grade prostate cancer:<br />
dutasteride and finasteride, members <strong>of</strong> the class <strong>of</strong> five<br />
alpha reductase inhibitors. These two agents have been<br />
tested in phase III trials and have been found to reduce the<br />
risk <strong>of</strong> prostate cancer by between 22% and 25%. 15,16 Patients<br />
in both studies who received active treatment were,<br />
however, more commonly found to have high-grade cancer.<br />
Although the interpretation <strong>of</strong> some parties has been that<br />
the agents induce high-grade cancer, substantial data show<br />
that both agents facilitate the detection <strong>of</strong> cancer and<br />
high-grade cancer. This occurs in, for example, the case <strong>of</strong><br />
finasteride, through an improved sensitivity <strong>of</strong> PSA for<br />
overall cancer detection, improved sensitivity <strong>of</strong> DRE for<br />
overall cancer detection, and likely through a reduction in<br />
prostate volume leading to a greater likelihood <strong>of</strong> detection<br />
Table 2. Potential Benefits and Harms <strong>of</strong> Prostate Cancer<br />
Chemoprevention<br />
Benefit Harm<br />
Reduction in morbidity <strong>of</strong><br />
treatment<br />
Reduction in cost <strong>of</strong><br />
treatment<br />
Reduction in psychologic<br />
burden <strong>of</strong> diagnosis<br />
Unnecessary exposure to preventive drug <strong>of</strong> healthy<br />
subjects who will never develop prostate cancer.<br />
Side effects <strong>of</strong> chemoprevention agent. In the<br />
example <strong>of</strong> finasteride and dutasteride, these<br />
include gynecomastia, decreased libido, loss <strong>of</strong><br />
ejaculate, erectile dysfunction, and potential<br />
increase in risk <strong>of</strong> high-grade cancer.<br />
Increase in cost. (It is not known whether<br />
chemoprevention would increase or decrease total<br />
cost <strong>of</strong> the disease to society.)<br />
<strong>of</strong> high-grade cancer at the time <strong>of</strong> prostate biopsy by more<br />
comprehensive sampling <strong>of</strong> a smaller prostate. 17,18,19 Multiple<br />
studies have demonstrated that, when these biases<br />
increasing cancer detection with finasteride are taken into<br />
account, the overall impact on the entire range <strong>of</strong> high-grade<br />
tumors (Gleason scores <strong>of</strong> 7–10) is a reduction in risk with<br />
finasteride. Clearly, then, one option for reducing the risk <strong>of</strong><br />
overtreatment through a reduction in overdetection is<br />
through chemoprevention. Table 2 lists the potential benefits<br />
and harms <strong>of</strong> chemoprevention <strong>of</strong> prostate cancer.<br />
Screening and Biopsy <strong>of</strong> Men Who Are Most Likely to Harbor<br />
Consequential Prostate Cancer<br />
Through the use <strong>of</strong> risk assessment tools that provide<br />
physicians with not only the risk <strong>of</strong> cancer detection but the<br />
risk <strong>of</strong> detection <strong>of</strong> a high-grade cancer, it is possible to do a<br />
better job <strong>of</strong> identifying the man who more likely has a<br />
high-grade cancer. It is these men who have the greatest<br />
potential for disease progression and death, best illustrated<br />
by the work <strong>of</strong> Albertsen and colleagues, who examined the<br />
outcomes <strong>of</strong> prostate cancer by age and grade. 20 Let’s examine<br />
these risks in two men using the Prostate Cancer<br />
Prevention Trial Risk Calculator (www.prostate.cancer.<br />
risk.calculator.com), which was developed based on PCPT<br />
data and has been validated in a number <strong>of</strong> external<br />
populations. 21-23 Mr. Smith went to see his primary care<br />
physician who felt a prostate nodule and sent him to his<br />
urologist. Every single guideline in urology at this time<br />
recommends biopsy for such a man, regardless <strong>of</strong> any other<br />
risk factors. If this man is 55 and white, has a PSA <strong>of</strong> 0.5<br />
ng/mL, has no family history <strong>of</strong> prostate cancer, and had a<br />
biopsy <strong>of</strong> the nodule last year, his risk <strong>of</strong> low-grade cancer is<br />
11.9% and his risk <strong>of</strong> high-grade cancer is 0.8%. Thus, there<br />
is a 15-fold greater likelihood <strong>of</strong> finding an inconsequential<br />
cancer in this man and his risk <strong>of</strong> having an aggressive<br />
cancer is about 1 in 125. Because the detection <strong>of</strong> a lowgrade<br />
cancer probably has a net negative impact (the bulk <strong>of</strong><br />
these tumors are <strong>of</strong> low malignant potential and, even with<br />
surveillance, there is a substantial degree <strong>of</strong> morbidity,<br />
anxiety, and cost <strong>of</strong> this strategy), in this particular man,<br />
the net potential benefit (probably 1 in 125) is likely to be far<br />
outweighed by the net potential harm (about 1 in 8). Conversely,<br />
Mr. Jones, a healthy 73-year-old black man with no<br />
other comorbidities who had a father with prostate cancer, is<br />
found to have a new prostate nodule and whose PSA is 5.8<br />
ng/mL, has a 56% risk <strong>of</strong> high-grade disease and a 15% risk<br />
<strong>of</strong> low-grade disease. In this particular man, his risk is 1 in<br />
2 that he may benefit from detection <strong>of</strong> an aggressive cancer,<br />
whereas his risk <strong>of</strong> potential overdetection <strong>of</strong> a low grade<br />
e37
Table 3. Potential Risks and Benefits <strong>of</strong> Early Detection <strong>of</strong><br />
Prostate Cancer<br />
Risks Benefits<br />
Detection and ultimate treatment <strong>of</strong><br />
cancer never destined to cause<br />
harm to the patient<br />
False-positive test: e.g., elevated PSA<br />
in patient without cancer. This then<br />
leads to unnecessary anxiety,<br />
biopsy, and side effects <strong>of</strong> biopsy.<br />
Risk <strong>of</strong> biopsy: bleeding, 2%-4% risk<br />
<strong>of</strong> sepsis<br />
Risk <strong>of</strong> missing prostate cancer due<br />
to sampling error<br />
Risk <strong>of</strong> missing high-grade prostate<br />
cancer (diagnosing only low-grade<br />
cancer) and placing patient on<br />
surveillance when more aggressive<br />
treatment would be appropriate<br />
Abbreviation: PSA, prostate-specific antigen.<br />
cancer is about 1 in 7. It should be clear that the risk/benefit<br />
ratios are utterly different in these two men. The potential<br />
risks and benefits <strong>of</strong> early detection are displayed in Table 3,<br />
and the potential side effects <strong>of</strong> the treatments themselves<br />
are shown in Table 4.<br />
The Future <strong>of</strong> Early Detection<br />
Ideally, we are most likely to benefit the general population<br />
if we can 1) tell men at very low risk that they can skip<br />
several years <strong>of</strong> screening and 2) consider biopsy only if a<br />
high-grade cancer is present and, ideally, when the tumor<br />
is still confined to the prostate and potentially curable.<br />
Although PSA is directly related to risk <strong>of</strong> high-grade<br />
disease, as can be seen in the examples above, when we use<br />
this marker, we simply cannot tell in advance which man<br />
will have which disease; as a result, we serendipitously—<br />
and probably unfortunately—find low-grade disease in the<br />
process.<br />
It will be the inclusion <strong>of</strong> diagnostic markers that are<br />
highly related to high-grade cancer that will allow us to not<br />
perform a biopsy on the man with low-grade disease while<br />
Author’s Disclosure <strong>of</strong> Potential Conflicts <strong>of</strong> Interest<br />
Author<br />
Ian M. Thompson Jr.*<br />
*No relevant relationships to disclose.<br />
Employment or<br />
Leadership<br />
Positions<br />
Consultant or<br />
Advisory Role<br />
1. Delongchamps NB, Singh A, Haas GP. The role <strong>of</strong> prevalence in the<br />
diagnosis <strong>of</strong> prostate cancer. Cancer Control. 2006;13:158-168.<br />
2. Thompson, Ernst JJ, Spence CR, Gangai MP. Adenocarcinoma <strong>of</strong> the<br />
prostate: Results <strong>of</strong> routine urological screening. J Urology. 1984;132:690-<br />
692.<br />
3. Thompson IM, Pauler DK, Goodman PJ, et al. Prevalence <strong>of</strong> prostate<br />
cancer among men with a prostate-specific antigen level � or � 4.0 ng per<br />
milliliter. N Engl J Med. 2004;350:2239-2246.<br />
4. Roobol MJ, Schröder FH, Hugosson J, et al. Importance <strong>of</strong> prostate<br />
volume in the European Randomised Study <strong>of</strong> Screening for Prostate Cancer<br />
(ERSPC) risk calculators: results from the prostate biopsy collaborative<br />
group. World J Urol. Epub 2011 Dec 28.<br />
e38<br />
Detection and cure <strong>of</strong> cancer ultimately<br />
destined to cause morbidity and/or<br />
mortality<br />
Identification <strong>of</strong> very low risk population<br />
who may not require frequent testing.<br />
An example could be man with very<br />
low PSA whose risk <strong>of</strong> prostate cancer<br />
is so low that PSA testing could be<br />
every 2 to 5 years or longer.<br />
recommending biopsy in the man with high-grade disease.<br />
Several biomarkers appear to have this relationship including<br />
PCA3 and TMPRSS2:ERG (urine markers) as well as<br />
[-2]proPSA (a serum marker). The Early Detection Research<br />
Network (EDRN)—Genitourinary Working Group has conducted<br />
several validation trials and will shortly complete the<br />
evaluation <strong>of</strong> all three biomarkers. It is anticipated that the<br />
EDRN will be able to incorporate these markers, either at<br />
the time <strong>of</strong> initial screening or as a reflex test after an<br />
intermediate risk is established by traditional PSA testing.<br />
For example, if the man has a 20% risk <strong>of</strong> a low-grade cancer<br />
and an 8% risk <strong>of</strong> high-grade cancer, a reflex test may be<br />
used to exclude biopsy in the 20% with the low-grade<br />
tumors. Ultimately, it will likely be a panel <strong>of</strong> markers and<br />
what we have deemed biomeasures (e.g., age, race, ethnicity,<br />
body mass index), all collected in a rational series <strong>of</strong> steps so<br />
as to achieve the greatest detection <strong>of</strong> lethal cancers while<br />
avoiding detection <strong>of</strong> inconsequential cancers, that will be<br />
the future <strong>of</strong> prostate cancer detection. This detection process<br />
will then minimize both overdiagnosis and overtreatment.<br />
It is through the efforts <strong>of</strong> the biomarker science<br />
community, as seen through the current risk assessment<br />
tools and through upcoming incorporation <strong>of</strong> new markers,<br />
that this vision will be achieved.<br />
Stock<br />
Ownership Honoraria<br />
REFERENCES<br />
Table 4. Potential Harms <strong>of</strong> Treatment<br />
Active surveillance Quarterly PSA testing: false positives, repeated visits<br />
to lab.<br />
Semi-annual rectal exams: cost, inconvenience,<br />
embarrassment, discomfort.<br />
Every 1- to 2-year prostate biopsy: pain,<br />
bleeding, blood in semen, 2%-4% risk <strong>of</strong> sepsis.<br />
Radical prostatectomy Intraoperative bleeding or anesthetic complications,<br />
infection, impotence, incontinence, anastomotic<br />
stricture, rectal injury, infertility, loss <strong>of</strong> ejaculate,<br />
disease recurrence, need for adjuvant or salvage<br />
therapy.<br />
Radiotherapy (seeds, beam,<br />
combination)<br />
Research<br />
Funding<br />
Expert<br />
Testimony<br />
IAN M. THOMPSON JR.<br />
Urinary dysfunction (retention, difficulty emptying,<br />
urgency, and urge incontinence), impotence,<br />
bowel dysfunction, radiation cystitis and proctitis,<br />
blood in stool or urine, urethral stricture,<br />
secondary malignancies <strong>of</strong> colon or bladder,<br />
disease recurrence, salvage treatment.<br />
Other<br />
Remuneration<br />
5. Fang J, Metter EJ, Landis P, et al. PSA velocity for assessing prostate<br />
cancer risk in men with PSA levels between 2.0 and 4.0 ng/ml. Urology.<br />
2002;59:889-893.<br />
6. Pinsky PF, Andriole GL, Kramer BS, et al. Prostate biopsy following a<br />
positive screen in the prostate, lung, colorectal and ovarian cancer screening<br />
trial. J Urol. 2005;173:746-750.<br />
7. Hodge KK, McNeal JE, Terris MK, Stamey TA. Random systematic<br />
versus directed ultrasound guided transrectal core biopsies <strong>of</strong> the prostate.<br />
J Urol. 1989;142:71-74.<br />
8. Levine MA, Ittman M, Melamed J, Lepor H. Two consecutive sets <strong>of</strong><br />
transrectal ultrasound guided sextant biopsies <strong>of</strong> the prostate for the detection<br />
<strong>of</strong> prostate cancer. J Urol. 1998;159:471-475.
OVERDIAGNOSIS AND OVERTREATMENT OF PROSTATE CANCER<br />
9. Abdollah F, Scattoni V, Raber M, et al. The role <strong>of</strong> transrectal saturation<br />
biopsy in tumour localization: pathological correlation after retropubic radical<br />
prostatectomy and implication for focal ablative therapy. BJU Int. 2011;108:<br />
366-371.<br />
10. Siegel R, Naishadham D, Jemal A. Cancer Statistics, <strong>2012</strong>. CA: Cancer<br />
J Clin. <strong>2012</strong>;62:10-29.<br />
11. National Cancer Institute. SEER Cancer Statistics Review 1975-2008.<br />
Lifetime Risk (Percent) <strong>of</strong> Dying from Cancer by Site and Race/Ethnicity:<br />
Males, Total US, 2006-2008 (Table 1.18) and Females, Total US, 2006-2008<br />
(Table 1.19). http://seer.cancer.gov/csr/1975_2008/results_merged/topic_life<br />
time_risk_death.pdf. Accessed on December 8, 2011.<br />
12. Whitmore WF Jr, Warner JA, Thompson IM Jr. Expectant management<br />
<strong>of</strong> localized prostatic cancer. Cancer. 1991;67:1091-1096.<br />
13. Klotz L, Zhang L, Lam A, et al. <strong>Clinical</strong> results <strong>of</strong> long-term follow-up<br />
<strong>of</strong> a large, active surveillance cohort with localized prostate cancer. J Clin<br />
Oncol. 2010;28:126-131.<br />
14. Cooperberg MR, Broering JM, Carroll PR. Time trends and local<br />
variation in primary treatment <strong>of</strong> localized prostate cancer. J Clin Oncol.<br />
2010;28:1117-1123.<br />
15. Andriole GL, Bostwick DG, Brawley OW, et al. Effect <strong>of</strong> dutasteride on<br />
the risk <strong>of</strong> prostate cancer. N Engl J Med. 2010;362:1192-1202.<br />
16. Thompson IM, Goodman PJ, Tangen CM, et al. The influence <strong>of</strong><br />
finasteride on the development <strong>of</strong> prostate cancer. N Engl J Med. 2003;349:<br />
215-224.<br />
17. Thompson IM, Chi C, Ankerst DP, et al. Effect <strong>of</strong> finasteride on the<br />
sensitivity <strong>of</strong> PSA for detecting prostate cancer. J Natl Cancer Inst. 2006;98:<br />
1128-1133.<br />
18. Thompson IM, Tangen CM, Goodman PJ, et al. Finasteride improves<br />
the sensitivity <strong>of</strong> digital rectal examination for prostate cancer detection.<br />
J Urol. 2007;177:1749-1752.<br />
19. Lucia MS, Epstein JI, Goodman PJ, et al. Finasteride and high-grade<br />
prostate cancer in the Prostate Cancer Prevention Trial. J Natl Cancer Inst.<br />
2007;99:1375-83.<br />
20. Lu-Yao GL, Albertsen PC, Moore DF et al. Outcomes <strong>of</strong> localized<br />
prostate cancer following conservative management. JAMA. 2009;302:1202-<br />
1209.<br />
21. Thompson IM, Ankerst DP, Chi C, et al. Assessing prostate cancer risk:<br />
results from the Prostate Cancer Prevention Trial. J Natl Cancer Inst.<br />
2006;98:529-534.<br />
22. Parekh DJ, Ankerst DP, Higgins BA, et al. External validation <strong>of</strong> the<br />
Prostate Cancer Prevention Trial risk calculator in a screened population.<br />
Urology. 2006;68:1152-1155.<br />
23. Eyre SJ, Ankerst DP, Wei JT, et al. Validation in a multiple urology<br />
practice cohort <strong>of</strong> the Prostate Cancer Prevention Trial calculator for predicting<br />
prostate cancer detection. J Urol. 2009;182:2653-2658.<br />
e39
Overdiagnosis and Overtreatment <strong>of</strong><br />
Breast Cancer<br />
By Michael Alvarado, MD, Elissa Ozanne, PhD, and Laura Esserman, MD, MBA<br />
Overview: Breast cancer is the most common cancer in<br />
women. Through greater awareness, mammographic screening,<br />
and aggressive biopsy <strong>of</strong> calcifications, the proportion <strong>of</strong><br />
low-grade, early stage cancers and in situ lesions among<br />
all breast cancers has risen substantially. The introduction <strong>of</strong><br />
molecular testing has increased the recognition <strong>of</strong> lower<br />
risk subtypes, and less aggressive treatments are more commonly<br />
recommended for these subtypes. Mammographically<br />
detected breast cancers are much more likely to have lowrisk<br />
biology than symptomatic tumors found between screenings<br />
(interval cancers) or that present as clinical masses.<br />
Recognizing the lower risk associated with these lesions<br />
and the ability to confirm the risk with molecular tests<br />
should safely enable the use <strong>of</strong> less aggressive treatments.<br />
Importantly, ductal carcinoma in situ (DCIS) lesions, or what<br />
have been called stage I cancers, in and <strong>of</strong> themselves are<br />
BREAST CANCER incidence has changed substantially<br />
over the past 25 years. The risk has gone from one<br />
in 12 to one in eight women being diagnosed in their<br />
lifetimes. In contrast, the absolute risk <strong>of</strong> dying <strong>of</strong> breast<br />
cancer has remained somewhat the same. As a proportion <strong>of</strong><br />
the women in whom breast cancer is diagnosed, mortality<br />
has gone down considerably. But in absolute terms, the<br />
number <strong>of</strong> women who die each year has changed less. In<br />
the late 1980s, the number <strong>of</strong> women dying ranged from<br />
45,000–50,000. Today, we expect about 42,000 women to die<br />
annually.<br />
The biology <strong>of</strong> the mammographically detected disease<br />
appears to have changed, particularly among postmenopausal<br />
women. In a retrospective study in which nodenegative<br />
tumors diagnosed in Europe in the era before the<br />
advent <strong>of</strong> mammography (1980–1991) were compared<br />
with those detected during the era when screening was<br />
prevalent (2004–2006), the chance <strong>of</strong> low-risk biology did<br />
not change for women younger than age 40 but changed<br />
substantially for those age 49–60. 1 Using a 70-gene signature,<br />
25% to 30% <strong>of</strong> women younger than 40 were diagnosed<br />
with low-risk disease, regardless <strong>of</strong> the time period in<br />
which they were diagnosed. In contrast, among women<br />
49–60, 40% were diagnosed with low-risk disease before<br />
screening, but 58% were diagnosed as having low-risk disease<br />
during the later time period. Of the women with<br />
mammographically detected tumors, 67% were classified as<br />
having low risk by the 70-gene signature. A number <strong>of</strong><br />
factors are likely to contribute to the differences in the<br />
cancers found today. Screening is clearly one factor that<br />
enables the identification <strong>of</strong> small tumors; greater awareness<br />
is another. Hormone-replacement therapy also contributed<br />
to the greater chance <strong>of</strong> being diagnosed with breast<br />
cancer, although with the publication <strong>of</strong> the Women’s Health<br />
Study in 2002, the widely publicized finding that combined<br />
estrogen and progesterone increased breast cancer incidence<br />
resulted in the plummeting <strong>of</strong> the number <strong>of</strong> prescriptions<br />
for hormone-replacement therapy, accompanied by a rapid<br />
decline in incidence. 2<br />
e40<br />
not life-threatening. In situ lesions have been treated in a<br />
manner similar to that <strong>of</strong> invasive cancer, but there is little<br />
evidence to support that this practice has improved mortality.<br />
It is also being recognized that DCIS lesions are<br />
heterogeneous, and a substantial proportion <strong>of</strong> them may in<br />
fact be precursors <strong>of</strong> more indolent invasive cancers. Increasing<br />
evidence suggests that these lesions are being overtreated.<br />
The introduction <strong>of</strong> molecular tests should be able<br />
to help usher in a change in approach to these lesions.<br />
Reclassifying these lesions as part <strong>of</strong> the spectrum <strong>of</strong> highrisk<br />
lesions enables the use <strong>of</strong> a prevention approach. Learning<br />
from the experience with active surveillance in prostate<br />
cancer should empower the introduction <strong>of</strong> new approaches,<br />
with a focus on preventing invasive cancer, especially given<br />
that there are effective, United States Food and Drug Administration<br />
(FDA)-approved breast cancer preventive interventions.<br />
What Is the Evidence that Breast Cancer May Be<br />
Overdiagnosed or Overtreated?<br />
From 1980 to the present, the incidence <strong>of</strong> breast cancer<br />
has risen substantially. 3 Much like prostate cancer, the<br />
proportion <strong>of</strong> early-stage disease accounts for the bulk <strong>of</strong> the<br />
increase in incidence, whereas the rates <strong>of</strong> locally advanced<br />
cancers have not changed considerably. Screening is likely<br />
to be a contributor to this phenomenon (Fig. 1).<br />
Interestingly, we are also seeing a change in the likelihood<br />
<strong>of</strong> local recurrence after breast cancer treatment. Among<br />
women with node-negative tumors who were randomly assigned<br />
to radiation therapy or no radiation therapy after<br />
quadrantectomy, the local recurrence rate was 10% for<br />
women with mammographically detected tumors in the<br />
group who did not receive radiation. 4 Over the past decade,<br />
the recurrence rate for postmenopausal women who receive<br />
radiation therapy has dropped substantially, to the 1% to 2%<br />
range. 5,6 It is likely that low-risk tumors are contributing to<br />
the drop in local recurrence rates reported in many trials. In<br />
the recent update <strong>of</strong> the Canadian trial <strong>of</strong> women treated<br />
with hormone therapy and randomly assigned to radiation<br />
therapy or no radiation therapy, the recurrence risk was 2%<br />
for women with luminal A tumors, even in the absence <strong>of</strong><br />
radiation. 7<br />
We are clearly detecting more low-risk disease. Some<br />
argue that screening has led to the detection <strong>of</strong> tumors that<br />
would never have come to clinical attention. 8 Observational<br />
evidence has shown that screening mammography is associated<br />
with increases in the incidence <strong>of</strong> breast cancer for<br />
women <strong>of</strong> screening age but does not appear to contribute to<br />
any decrease in the incidence <strong>of</strong> breast cancer among older<br />
women after sustained screening for specific time periods. 9<br />
From the Department <strong>of</strong> Surgery, University <strong>of</strong> California, San Francisco.<br />
Authors’ disclosures <strong>of</strong> potential conflicts <strong>of</strong> interest are found at the end <strong>of</strong> this article.<br />
Address reprint requests to Laura Esserman, MD, University <strong>of</strong> California, San Francisco,<br />
Department <strong>of</strong> Surgery, 1600 Divisadero St., 2nd Floor, Box 1710, San Francisco, CA<br />
94115; email: laura.esserman@ucfsmedctr.org.<br />
© <strong>2012</strong> by <strong>American</strong> <strong>Society</strong> <strong>of</strong> <strong>Clinical</strong> <strong>Oncology</strong>.<br />
1092-9118/10/1-10
OVERDIAGNOSIS AND OVERTREATMENT OF BREAST CANCER<br />
Fig 1. The SEER incidence <strong>of</strong> ductal carcinoma in situ (DCIS), stage I<br />
(localized), stage II (regional), and stage IV (distant) breast cancer<br />
over time. The red line at the bottom represents the incidence <strong>of</strong> DCIS<br />
in 1980. The large red arrow represents the change in incidence in<br />
DCIS by the year 2002. A concomitant drop in the incidence <strong>of</strong><br />
invasive cancer (top red arrow <strong>of</strong> equal magnitude) has not been<br />
observed.<br />
This evidence has even led to the hypothesis that some<br />
screen-detected invasive breast cancers would likely regress<br />
and be clinically unimportant in the future. 8,10 Some note<br />
that screening <strong>of</strong> all kinds has led to an epidemic <strong>of</strong> overdiagnosis.<br />
11 It is difficult to say with absolute certainty that<br />
the increase in cancers detected would never become a<br />
clinical problem, but there is ample evidence that the biology<br />
<strong>of</strong> screen-detected tumors differs substantially from that <strong>of</strong><br />
interval cancers or nonscreening-detected cancers, 12,13 although<br />
some say that the improvement in outcome is explained<br />
by lead-time bias. 14<br />
What is clear is that the lower risk biology associated with<br />
these tumors should safely allow less treatment. In particular,<br />
there are groups <strong>of</strong> women for whom local therapy, in<br />
particular, can be less, and for whom breast conservation<br />
with adjuvant hormone therapy alone or with intraoperative<br />
radiation therapy is associated with a low risk for local<br />
recurrence and low mortality. 5,7,15 Surveillance Epidemiology<br />
and End Results (SEER) data indicate that as many as<br />
50% <strong>of</strong> all cancers occur in postmenopausal women and are<br />
KEY POINTS<br />
● SEER data suggest that a meaningful amount <strong>of</strong> both<br />
invasive and noninvasive breast cancers are likely to<br />
be overdiagnosed and overtreated.<br />
● Further risk stratification is needed in the treatment<br />
<strong>of</strong> breast cancer to avoid overtreatment.<br />
● Focusing on the prevention <strong>of</strong> invasive disease may<br />
be one option to reduce the overtreatment <strong>of</strong> DCIS.<br />
● Learning from experiences in prostate cancer, active<br />
surveillance should be considered for some cases <strong>of</strong><br />
DCIS.<br />
● Additional efforts towards the prevention <strong>of</strong> breast<br />
cancer are greatly needed.<br />
node-negative and low or intermediate grade. 16 Given that<br />
the inclusion <strong>of</strong> external-beam radiation therapy is considered<br />
a marker <strong>of</strong> quality care for breast cancer, there is<br />
clearly evidence <strong>of</strong> overtreatment.<br />
For many years, it has been recognized that there are<br />
various pathologic characteristics and a spectrum <strong>of</strong> changes<br />
in breast cell morphology, which range from hyperplasia to<br />
atypia to in situ carcinoma. 17,18 Many researchers believed<br />
that each <strong>of</strong> these changes carried a distinct increase in risk<br />
for future invasive breast cancer. The classic “precursor<br />
pathway,” was accepted and therefore it was assumed that<br />
all low-risk in situ cancers should be treated aggressively to<br />
decrease the risk <strong>of</strong> a future invasive breast cancer. However,<br />
it has been shown that this is likely not the case, both<br />
in epidemiology studies as well as in cohort studies <strong>of</strong><br />
specific “high-risk” lesions.<br />
Considering that evidence suggests overdiagnosis in the<br />
setting <strong>of</strong> invasive breast cancer, it is reasonable to consider<br />
the possibility <strong>of</strong> overdiagnosis <strong>of</strong> DCIS. 9,19 In the context <strong>of</strong><br />
DCIS, overdiagnosis may occur if DCIS lesions are precursors<br />
<strong>of</strong> invasive breast cancer that turn out to be indolent<br />
disease or if the diagnosed DCIS is indolent itself without<br />
the potential to progress to invasive cancer even in the<br />
absence <strong>of</strong> treatment.<br />
Similar to the situation with invasive breast cancer, SEER<br />
data show a dramatic rise in the incidence <strong>of</strong> DCIS since the<br />
introduction <strong>of</strong> screening mammography. However, there is<br />
no corresponding drop in the rate <strong>of</strong> invasive breast cancer<br />
that could account for successful detection and removal <strong>of</strong><br />
DCIS. In contrast, there is an overall increase in the incidence<br />
<strong>of</strong> invasive breast cancer since the introduction <strong>of</strong><br />
screening mammography. After more than 25 years <strong>of</strong> widespread<br />
efforts to increase rates <strong>of</strong> screening mammography,<br />
the rate <strong>of</strong> invasive breast cancer incidence has continued to<br />
increase despite widespread treatment <strong>of</strong> DCIS, except for a<br />
recently observed leveling in incidence rates. It is possible<br />
that this recent change in breast cancer incidence rates is<br />
due to effective screening programs. However, a number <strong>of</strong><br />
epidemiologic studies have shown that this change is directly<br />
associated with the decreased use <strong>of</strong> hormone replacement<br />
therapy among postmenopausal women. 20-22<br />
These observations raise the question <strong>of</strong> whether DCIS is<br />
an obligate precursor to invasive breast cancer and would<br />
progress to invasive cancer if left untreated. 23 A recent<br />
modeling analysis explored the relationship between DCIS<br />
progression rates and the assumed increase in the background<br />
rate <strong>of</strong> invasive breast cancer. 24 The results <strong>of</strong> this<br />
study demonstrated that different sets <strong>of</strong> assumptions related<br />
to DCIS and its progression rates could produce<br />
similar projected trends. The study found that, if most DCIS<br />
lesions were destined to progress to invasive breast cancer,<br />
it would then be reasonable to assume that there is a<br />
substantial increase in the baseline rate <strong>of</strong> invasive breast<br />
cancer over what we have seen. In contrast, if there is only<br />
a modest increase in the baseline rate <strong>of</strong> invasive breast<br />
cancer after the introduction <strong>of</strong> screening mammography,<br />
more in line with the observed trends before screening, it<br />
then follows that most DCIS lesions would not have progressed<br />
to invasive breast cancer, even in the absence <strong>of</strong><br />
treatment.<br />
Similarly, there is evidence in support <strong>of</strong> regression for<br />
invasive breast cancer, which may be more relevant for<br />
DCIS than for invasive cancer. 8,11,19 This conclusion has<br />
e41
een reached by other modeling efforts designed to examine<br />
trends in the incidence <strong>of</strong> breast cancer. With their model,<br />
Mandelblatt and colleagues 25 consistently underestimated<br />
rates <strong>of</strong> DCIS as compared with rates drawn from SEER<br />
data, suggesting that not all DCIS will become clinically<br />
relevant. In a similar model, Fryback and colleagues 26<br />
identified a substantial proportion <strong>of</strong> tumors having limited<br />
potential to become malignant in order to account for the<br />
observed DCIS rate.<br />
Although there are few data, evidence indicates that<br />
low-grade DCIS treated by biopsy alone does not progress to<br />
invasive breast cancer. 27,28 Additionally, autopsy data suggest<br />
that untreated DCIS may not develop into invasive<br />
breast cancer in a time period that will affect a woman’s<br />
life. 29 It appears feasible that there exists a reservoir <strong>of</strong><br />
DCIS in the population that is never diagnosed and never<br />
attains clinical relevance, compelling us to ask if we have<br />
overestimated the potential <strong>of</strong> DCIS to progress to invasive<br />
breast cancer. If so, are we overdiagnosing and therefore<br />
overtreating DCIS?<br />
Researchers have tried to identify which in situ cancers<br />
are the likely precursors <strong>of</strong> invasive disease and, <strong>of</strong> those,<br />
which are <strong>of</strong> such low-risk that they might require minimal<br />
local treatment (i.e., surgery alone or even active surveillance).<br />
Until recently, researchers have tried to identify<br />
these low-risk groups by standard clinical-pathologic features.<br />
30 For example, high-grade palpable lesions have been<br />
shown to carry a higher risk for recurrence in the first 5<br />
years after diagnosis. 31 A number <strong>of</strong> molecular markers<br />
have been proposed to stratify risk but have not yet been<br />
validated in large studies. 32 The Van Nuys score, based on<br />
combinations <strong>of</strong> patient age, tumor size, tumor grade, and<br />
surgical margins, has been used to determine the value <strong>of</strong><br />
radiation therapy. A particular problem however, is the<br />
current use <strong>of</strong> nuclear grade as overall grade, causing as<br />
much as a 50% increase in the classification <strong>of</strong> high-grade<br />
lesions.<br />
Interestingly, a recent analysis <strong>of</strong> grade and molecular<br />
pr<strong>of</strong>iling <strong>of</strong> in situ lesions suggested that low-grade lesions<br />
were not the precursors <strong>of</strong> low-grade tumors, but high-grade<br />
DCIS lesions were consistently predictive <strong>of</strong> high-grade<br />
invasive tumors. However, when molecular subtyping was<br />
performed, the vast majority (85%) <strong>of</strong> the lobular carcinoma<br />
in situ (LCIS) and DCIS lesions were luminal A subtype.<br />
The molecular subtype <strong>of</strong> the subsequent invasive tumor<br />
was largely maintained: 85% for LCIS, and 69% for DCIS. 33<br />
Although the numbers in the study were small, the findings<br />
suggest the possibility that grade is not a reliable indicator<br />
<strong>of</strong> risk and that many DCIS lesions are in fact precursors <strong>of</strong><br />
more indolent cancers.<br />
The unmet need for better risk stratification is now being<br />
served with molecular pr<strong>of</strong>iling in the same way it has for<br />
invasive cancers. Researchers at Genomic Health have now<br />
applied their expertise in molecular pr<strong>of</strong>iling to DCIS. Recent<br />
data from their studies showed that they are able to<br />
identify a low-risk category for which surgery alone would be<br />
adequate local therapy. In fact, not only were they able to<br />
identify risk for local recurrence but they were also able to<br />
distinguish between total recurrences (in situ plus invasive)<br />
and invasive cancer recurrence only. These data suggest<br />
that a substantial number <strong>of</strong> low-risk lesions have, on<br />
average, a 5% risk <strong>of</strong> an invasive local recurrence at 10 years<br />
after lumpectomy alone. This risk is the equivalent <strong>of</strong> a Gail<br />
e42<br />
Risk Score <strong>of</strong> 2.5, indicating a 2.5% risk <strong>of</strong> invasive cancer at<br />
5 years. This risk is also similar to the average risk <strong>of</strong><br />
invasive cancer for a woman in her mid-60s. Again, these<br />
findings mean that DCIS, with excision alone, is appropriately<br />
categorized as a high-risk lesion such as atypia. These<br />
new data are extremely important because they may finally<br />
provide a more precise way <strong>of</strong> identifying women who are<br />
being overtreated for a diagnosis that may have little bearing<br />
on their life for the next 20–30 years.<br />
It is important to recognize that when screening was first<br />
introduced, the goal was to identify invasive cancers. Over<br />
the years, the focus has broadened to include calcifications<br />
and the identification <strong>of</strong> DCIS. In the United States, the<br />
biopsy rates are higher than they are in Europe, 34 and the<br />
cancer-to-biopsy ratios are in the range <strong>of</strong> 25% and 50%,<br />
respectively. The proportion <strong>of</strong> DCIS detected is higher, and<br />
more cases <strong>of</strong> detected DCIS are lower grade. These findings<br />
are important to keep in mind when evaluating results <strong>of</strong><br />
different trials. For example, in the [United Kingdom New<br />
Zealand trial], the majority <strong>of</strong> DCIS detected was high<br />
grade, which may explain why tamoxifen was not found to be<br />
effective in reducing the risk <strong>of</strong> ipsilateral invasive cancer.<br />
The introduction <strong>of</strong> more standard molecular tools may<br />
prove important in comparing DCIS types and in allowing a<br />
different approach to DCIS in the future.<br />
Despite the recognition that some breast cancers we<br />
detect are indolent in nature, there have been almost no<br />
trials <strong>of</strong> active surveillance, as there have been in prostate<br />
cancer. Even DCIS, which is thought to be a precursor <strong>of</strong><br />
invasive cancer, has rarely been approached with active<br />
surveillance. It is hoped that the availability <strong>of</strong> tools to<br />
pr<strong>of</strong>ile DCIS lesions in a standard manner will usher in a<br />
more progressive approach to DCIS, not unlike what has<br />
been adopted in the treatment <strong>of</strong> prostate cancer, especially<br />
given the availability <strong>of</strong> preventive interventions.<br />
What Is the Solution?<br />
ALVARADO, OZANNE, AND ESSERMAN<br />
Change the Approach to DCIS from Cancer Treatment to<br />
Prevention <strong>of</strong> Invasive Cancer<br />
The evidence for overtreatment <strong>of</strong> DCIS is extremely<br />
compelling, given the associated risk. Classic treatment<br />
strategies have almost always included either breast conservation<br />
with whole breast radiation or mastectomy. It is now<br />
becoming apparent that this type <strong>of</strong> “precursor lesion” is<br />
being overtreated. With continued evidence that we are not<br />
only overdetecting but also overtreating breast cancer in<br />
this country and abroad, researchers need to think seriously<br />
about developing protocols for women at the lowest risk.<br />
With regard specifically to DCIS, it would not be unreasonable<br />
to develop algorithms to identify women who could have<br />
active surveillance after carcinoma in situ is found on<br />
evaluation <strong>of</strong> a biopsy specimen. 35 Many groups have discussed<br />
ways in which to stratify risk, and with the availability<br />
<strong>of</strong> molecular markers, it should be an immediate<br />
action item to finally address the overtreatment <strong>of</strong> earlystage<br />
breast cancer. Previous research attempted to identify<br />
risk groups that were typically based on classic, clinicalpathologic<br />
features such as patient age, tumor size, and<br />
tumor grade. Numerous data with molecular pr<strong>of</strong>iling have<br />
indicated that standard clinical-pathologic features are a<br />
poor way <strong>of</strong> truly risk stratifying newly diagnosed breast<br />
cancer. In fact, grade appears to be nonreproducible across
OVERDIAGNOSIS AND OVERTREATMENT OF BREAST CANCER<br />
Table 1. Short- and Long-term Risk <strong>of</strong> Breast Cancer by Lesion/Risk Factor. 36<br />
Abbreviations: LCIS, lobular carcinoma in situ; DCIS, ductal carcinoma in situ.<br />
multiple laboratories and different pathologists. Therefore,<br />
one would expect that future treatment decisions are based<br />
on molecular markers and not subjective data such as grade,<br />
age, and tumor size.<br />
Focus on Prevention<br />
The Athena Breast Health Network is an innovative<br />
collaboration across the five University <strong>of</strong> California medical<br />
centers and includes the UCSF Institute for Health Policy<br />
Studies and the Graduate School <strong>of</strong> Public Health at Berkeley.<br />
Athena is creating a 21st century “knowledge economy”<br />
that will integrate clinical care and research to drive innovation<br />
in prevention, screening, treatment, and management<br />
<strong>of</strong> breast cancer and, at the same time, revolutionize<br />
the delivery <strong>of</strong> care. By working together as a community,<br />
the University <strong>of</strong> California medical centers, their affiliates,<br />
and primary care and specialty physicians will work to<br />
change the options for patients today and create a better<br />
future for all women at increased risk for breast cancer.<br />
Athena will implement a comprehensive Web-based informatics<br />
strategy to modernize the way in which clinical<br />
information is collected, tracked, and integrated with research.<br />
Athena will also develop and implement standard<br />
Web-based tools that will be integrated with the existing<br />
clinical information infrastructure to optimize the capture <strong>of</strong><br />
structured clinical information at the point <strong>of</strong> care from<br />
clinicians and patients. The development <strong>of</strong> an automated<br />
risk assessment tool integrated into Web-based decision<br />
support (BreastHealthDecisions.org, or BHD) is the culmination<br />
<strong>of</strong> 10 years <strong>of</strong> collaborative work by faculty involved<br />
in Athena. BHD was designed to widely disseminate<br />
evidence-based personalized options for breast cancer risk<br />
reduction and to increase awareness and use among the<br />
primary care community and hard-to-reach populations in a<br />
cost-efficient manner. This tool, which includes all <strong>of</strong> the<br />
validated breast cancer risk models, was built on an awardwinning<br />
decision-making framework 15,16 and is a decision<br />
aid that our team has developed and tested with patients<br />
and experts in breast cancer risk and prevention over the<br />
past 5 years. 17 The tool is designed for use by breast<br />
specialists, gynecologists, and primary care physicians, and<br />
incorporates the standard and improved models for breast<br />
cancer risk assessment that include breast density, hereditary<br />
risk, and atypia. 18<br />
Opportunities to Change Now<br />
Given that many cases <strong>of</strong> DCIS are associated with a 5%<br />
risk <strong>of</strong> invasive cancer after 10 years (the equivalent <strong>of</strong> a<br />
Gail Risk score <strong>of</strong> 2.5 or the average risk <strong>of</strong> a woman in her<br />
mid-60s) as shown in Table 1, these lesions too should likely<br />
be considered to be markers <strong>of</strong> elevated risk and should<br />
prompt thoughts <strong>of</strong> chemoprevention strategies. Coopey and<br />
colleagues 37 recently reported the outcome for nearly 3000<br />
patients with atypical breast lesions, ranging from atypical<br />
lobular hyperplasia to borderline DCIS lesions. The data<br />
demonstrated that these lesions conferred a risk <strong>of</strong> cancer in<br />
both breasts and that for this group <strong>of</strong> patients, chemoprevention<br />
with tamoxifen or raloxifene reduced the risk by<br />
about two-thirds, on average. This finding is consistent with<br />
data from the National Surgical Adjuvant Breast and Bowel<br />
Project (NSABP) P-01 trial showing that women with atypia<br />
had an 85% reduction in risk with chemoprevention. 38<br />
Ironically, the only lesions we approach with surveillance<br />
are those with greater risk than that posed by many DCIS<br />
cases. Unlike the case in the prostate community, we in the<br />
e43
east cancer community have FDA-approved risk-reducing<br />
agents that can be <strong>of</strong>fered.<br />
It is possible that the lesions associated with the equivalent<br />
<strong>of</strong> a low DCIS score according to molecular pr<strong>of</strong>iling<br />
with Oncotype DX (Genomic Health, Redwood City, CA)<br />
could be treated with a chemoprevention approach rather<br />
than surgery. The ALLIANCE is opening a trial for nonhighgrade<br />
DCIS in postmenopausal women, with initial treatment<br />
<strong>of</strong> an aromatase inhibitor for 6 months. Magnetic<br />
resonance imaging (MRI) will be done before therapy is<br />
initiated, to rule out the presence <strong>of</strong> invasive cancer and to<br />
document the extent <strong>of</strong> disease by MRI-measured volume.<br />
MRI will be repeated after 3 months <strong>of</strong> therapy, and if there<br />
is no evidence <strong>of</strong> progression, therapy will continue for an<br />
additional 3 months. The endpoint is the change in MRImeasured<br />
volume. For high-grade DCIS, we should be thinking<br />
about screening compounds for their biologic effect, with<br />
the goal <strong>of</strong> generating neoadjuvant trials for the compounds<br />
that have the greatest effect on mitotic activity or other<br />
surrogate markers. Targets should not only be the tumor<br />
itself but changes in the stroma (e.g., density), as these may<br />
be measures <strong>of</strong> impact.<br />
Conclusion: Future <strong>of</strong> Early Detection<br />
Our goal should be to learn who is at risk for what type <strong>of</strong><br />
cancer and to use the available predictors <strong>of</strong> risk to determine<br />
the type and frequency <strong>of</strong> screening. In addition, risk<br />
assessment, using existing and emerging models, should be<br />
part <strong>of</strong> screening. Our efforts should be targeted to preventing<br />
breast cancer in the highest risk groups. Not only should<br />
we work on making interventions available where appropriate<br />
and informing women <strong>of</strong> their options (eg, with tools<br />
such as BHD) but we should also be using emerging markers<br />
<strong>of</strong> risk, such as breast density, as a potential surrogate for<br />
effectiveness <strong>of</strong> prevention agents, as has been shown for<br />
tamoxifen. 39 Other markers should be developed to help<br />
guide the testing <strong>of</strong> new interventions, especially for women<br />
at risk for aggressive tumor biology. In situ lesions should be<br />
used as markers <strong>of</strong> risk and tools to evaluate the effect <strong>of</strong><br />
Authors’ Disclosures <strong>of</strong> Potential Conflicts <strong>of</strong> Interest<br />
Author<br />
Employment or<br />
Leadership<br />
Positions<br />
Consultant or<br />
Advisory Role<br />
therapies. Breast cancer is clearly a heterogeneous disease.<br />
Our treatment approaches need to reflect the differences in<br />
biology even more than they do today, and in the future, we<br />
will need to tailor our screening and prevention recommendations<br />
as well. We can improve the use <strong>of</strong> early detection<br />
efforts in the United States by following the U.S. Preventive<br />
Services Task Force guidelines. Screening every other year<br />
is not associated with a substantial increase in the number<br />
<strong>of</strong> locally advanced cancers, reduces the false-positive results<br />
<strong>of</strong> biopsy evaluations by about a third, 40 and could save<br />
billions <strong>of</strong> dollars annually. 41 Efforts are already underway<br />
to develop risk-based screening guidelines. 42 Not only would<br />
risk-based screening be more effective and a much better<br />
way to apply resources it may also be an important way to<br />
move past the controversy about annual or biennial screening<br />
and encourage women to accept that there may be a<br />
more appropriate way to apply screening.<br />
We can learn from the prostate cancer experience with<br />
active surveillance to recognize the safety <strong>of</strong> following more<br />
indolent disease and apply that to the setting <strong>of</strong> DCIS, which<br />
is not invasive cancer. The first step is to recognize that most<br />
<strong>of</strong> the DCIS lesions detected are not likely to progress. Even<br />
if they do, the period <strong>of</strong> risk spans several years, and many<br />
lesions are likely to be precursors <strong>of</strong> relatively indolent<br />
disease. More importantly, the subsequent invasive events<br />
appear to be largely preventable. We should be thinking<br />
more like how the cardiology community in how they view<br />
the opportunity to reduce the incidence <strong>of</strong> cardiac and stroke<br />
events. We have the markers <strong>of</strong> risk that can be used; these<br />
markers should include not just atypia but lower risk DCIS.<br />
We now have preventive agents that we can use. We need to<br />
refine our ability to determine whether specific patients are<br />
benefiting from chemoprevention interventions, but we have<br />
much that we can do today that is likely better than what we<br />
have been doing. The emergence <strong>of</strong> molecular tools to more<br />
characterize DCIS in a more standard manner should be<br />
used as an opportunity to reframe options and work together<br />
to reduce the morbidity associated with early detection.<br />
Stock<br />
Ownership Honoraria<br />
Michael Alvarado*<br />
Elissa Ozanne*<br />
Laura Esserman Agendia<br />
*No relevant relationships to disclose.<br />
1. Esserman LJ, Shieh Y, Rutgers EJ, et al. Impact <strong>of</strong> mammographic<br />
screening on the detection <strong>of</strong> good and poor prognosis breast cancers. Breast<br />
Cancer Res Treat. 2011;130:725-734.<br />
2. Chlebowski RT, Kuller LH, Prentice RL, et al. Breast cancer after use <strong>of</strong><br />
estrogen plus progestin in postmenopausal women. N Engl J Med. 2009;360:<br />
573-587.<br />
3. Lin C, Moore D, DeMichele A, et al. Detection <strong>of</strong> locally advanced breast<br />
cancers in the I-SPY TRIAL (CALGB 150007/150012, ACRIN 6657) in the<br />
interval between routine screening. J Clin Oncol. 2009;27:15s (suppl; abstr<br />
1503).<br />
4. Malmström P, Holmberg L, Anderson H, et al. Breast conservation<br />
surgery, with and without radiotherapy, in women with lymph node-negative<br />
breast cancer: a randomised clinical trial in a population with access to public<br />
mammography screening. Eur J Cancer. 2003;39:1690-1697.<br />
5. Vaidya JS, Joseph DJ, Tobias JS, et al. Targeted intraoperative radio-<br />
e44<br />
REFERENCES<br />
Research<br />
Funding<br />
ALVARADO, OZANNE, AND ESSERMAN<br />
Expert<br />
Testimony<br />
Other<br />
Remuneration<br />
therapy versus whole breast radiotherapy for breast cancer (TARGIT-A trial):<br />
an international, prospective, randomised, non-inferiority phase 3 trial.<br />
Lancet. 2010;376:91-102.<br />
6. START Trialists’ Group, Bentzen SM, Agrawal RK, et al. The UK<br />
Standardisation <strong>of</strong> Breast Radiotherapy (START) Trial A <strong>of</strong> radiotherapy<br />
hyp<strong>of</strong>ractionation for treatment <strong>of</strong> early breast cancer: a randomised trial.<br />
Lancet. 2008;9:331-341.<br />
7. Fyles A, McCready D, Pintilie M, et al. Luminal A subtype predicts<br />
radiation response in patients with T1N0 breast cancer enrolled in a randomized<br />
trial <strong>of</strong> tamoxifen with or without breast radiation. Paper presented at:<br />
San Antonio Breast Cancer Symposium; December 2011; San Antonio, TX.<br />
8. Zahl PH, Maehlen J, Welch HG. The natural history <strong>of</strong> invasive breast<br />
cancers detected by screening mammography. Arch Intern Med. 2008;168:<br />
2311-2316.<br />
9. Jørgensen, KJ, Gøtzsche PC. Overdiagnosis in publicly organised mam-
OVERDIAGNOSIS AND OVERTREATMENT OF BREAST CANCER<br />
mography screening programmes: systematic review <strong>of</strong> incidence trends.<br />
BMJ. 2009;339:b2587.<br />
10. Welch HG, Frankel BA. Likelihood that a woman with screen-detected<br />
breast cancer has had her “life saved” by that screening. Arch Intern Med.<br />
2011;171:2043-2046.<br />
11. Welch HG, Black WC. Overdiagnosis in cancer. J Natl Cancer Inst.<br />
2010;102:605-613.<br />
12. Mook S, Van ‘t Veer LJ, Rutgers EJ, et al. Independent prognostic value<br />
<strong>of</strong> screen detection in invasive breast cancer. J Natl Cancer Inst. 2011;103:<br />
585-597.<br />
13. Lin C, Buxton MB, Moore D, et al. Locally advanced breast cancers are<br />
more likely to present as interval cancers: results from the I-SPY 1 TRIAL<br />
(CALGB 150007/150012, ACRIN 6657, InterSPORE Trial). Breast Cancer Res<br />
Treat. Epub 2011 Jul 28.<br />
14. Allgood PC, Duffy SW, Kearins O, et al. Explaining the difference in<br />
prognosis between screen-detected and symptomatic breast cancers. Br J<br />
Cancer. 2011;104:1680-1685.<br />
15. Hughes KS, Schnaper LA, Berry D, et al. Lumpectomy plus tamoxifen<br />
with or without irradiation in women 70 years <strong>of</strong> age or older with early<br />
breast cancer. N Engl J Med. 2004;351:971-977.<br />
16. Esserman LJ, Mohan AJ, Park C, et al. Projecting the impact <strong>of</strong><br />
adopting trial results. Poster presented at: San Antonio Breast Care Symposium;<br />
December 2011; San Antonio, TX.<br />
17. Allred DC, Mohsin SK, Fuqua SA. Histological and biological evolution<br />
<strong>of</strong> human premalignant breast disease. Endocr Relat Cancer. 2001;8:47-61.<br />
18. Dupont WD, Page DL. Risk factors for breast cancer in women with<br />
proliferative breast disease. N Engl J Med. 1985;312:146-151.<br />
19. Morrell S, Barratt A, Irwig L, et al. Estimates <strong>of</strong> overdiagnosis <strong>of</strong><br />
invasive breast cancer associated with screening mammography. Cancer<br />
Causes Control. 2010;21:275-282.<br />
20. Ravdin PM, Cronin KA, Howlader N, et al. The decrease in breastcancer<br />
incidence in 2003 in the United States. N Engl J Med. 2007;356:1670-<br />
1674.<br />
21. Kerlikowske K, Miglioretti DL, Buist DS, et al. Declines in invasive<br />
breast cancer and use <strong>of</strong> postmenopausal hormone therapy in a screening<br />
mammography population. J Natl Cancer Inst. 2007;99:1335-1339.<br />
22. Glass AG, Lacey JV Jr, Carreon JD, et al. Breast cancer incidence,<br />
1980-2006: combined roles <strong>of</strong> menopausal hormone therapy, screening mammography,<br />
and estrogen receptor status. J Natl Cancer Inst. 2007;99:1152-<br />
1161.<br />
23. Welch HG. Should I Be Tested for Cancer? Maybe Not and Here’s Why.<br />
Berkeley: University <strong>of</strong> California Press; 2004.<br />
24. Ozanne EM, Shieh Y, Barnes J, et al. Characterizing the impact <strong>of</strong> 25<br />
years <strong>of</strong> DCIS treatment. Breast Cancer Res Treat. 2011;129:165-173.<br />
25. Mandelblatt J, Schechter CB, Lawrence W, et al. The SPECTRUM<br />
population model <strong>of</strong> the impact <strong>of</strong> screening and treatment on U.S. breast<br />
cancer trends from 1975 to 2000: principles and practice <strong>of</strong> the model<br />
methods. J Natl Cancer Inst Monogr. 2006;(36):47-55.<br />
26. Fryback DG, Stout NK, Rosenberg MA, et al. The Wisconsin Breast<br />
Cancer Epidemiology Simulation Model. J Natl Cancer Inst Monogr. 2006;<br />
(36):37-47.<br />
27. Sanders ME, Schuyler PA, Dupont WD, et al. The natural history <strong>of</strong><br />
low-grade ductal carcinoma in situ <strong>of</strong> the breast in women treated by biopsy<br />
only revealed over 30 years <strong>of</strong> long-term follow-up. Cancer. 2005;103:2481-<br />
2484.<br />
28. Lagios MD, Margolin FR, Westdahl PR, et al. Mammographically<br />
detected duct carcinoma in situ. Frequency <strong>of</strong> local recurrence following<br />
tylectomy and prognostic effect <strong>of</strong> nuclear grade on local recurrence. Cancer.<br />
1989;63:618-624.<br />
29. Welch HG, Black WC. Using autopsy series to estimate the disease<br />
“reservoir” for ductal carcinoma in situ <strong>of</strong> the breast: how much more breast<br />
cancer can we find? Ann Intern Med. 1997;127:1023-1028.<br />
30. Silverstein MJ, Lagios MD. Choosing treatment for patients with<br />
ductal carcinoma in situ: fine tuning the University <strong>of</strong> Southern California/<br />
Van Nuys Prognostic Index. J Natl Cancer Inst Monogr. 2010; 2010(41):193-<br />
196.<br />
31. Kerlikowske K, Molinaro A, Cha I, et al. Characteristics associated<br />
with recurrence among women with ductal carcinoma in situ treated by<br />
lumpectomy. J Natl Cancer Inst. 2003;95:1692-1702.<br />
32. Kerlikowske K, Molinaro AM, Gauthier ML, et al. Biomarker expression<br />
and risk <strong>of</strong> subsequent tumors after initial ductal carcinoma in situ<br />
diagnosis. J Natl Cancer Inst. 2010;102:627-637.<br />
33. King TA, Sakr RA, Muhsen S, et al. Is there a low-grade precursor<br />
pathway in breast cancer? Ann Surg Oncol. Epub 2011 Sep 21.<br />
34. Esserman L, Cowley H, Eberle C, et al. Improving the accuracy <strong>of</strong><br />
mammography: volume and outcome relationships. J Natl Cancer Inst.<br />
2002;94:369-375.<br />
35. Meyerson AF, Lessing JN, Itakura K, et al. Outcome <strong>of</strong> long term active<br />
surveillance for estrogen receptor-positive ductal carcinoma in situ. Breast.<br />
2011;20:529-533.<br />
36. Esserman L, Sepucha K, Ozanne EM, Hwang ES. Applying the neoadjuvant<br />
paradigm to ductal carcinoma in situ. Annals <strong>of</strong> Surgical <strong>Oncology</strong>.<br />
2004;11(1 Suppl):28S-36S.<br />
37. Coopey SB, Mazzola E, Buckley JM, et al. Clarifying the risk <strong>of</strong> breast<br />
cancer in women with atypical breast lesions. Paper presented at: San<br />
Antonio Breast Cancer Symposium; December 2011; San Antonio, TX.<br />
38. Fisher B, Costantino JP, Wickerham DL, et al. Tamoxifen for prevention<br />
<strong>of</strong> breast cancer: report <strong>of</strong> the National Surgical Adjuvant Breast and<br />
Bowel Project P-1 Study. J Natl Cancer Inst. 1998;90:1371-1388.<br />
39. Cuzick J, Warwick J, Pinney E, et al. Tamoxifen-induced reduction in<br />
mammographic density and breast cancer risk reduction: a nested casecontrol<br />
study. J Natl Cancer Inst. 2011;103:744-752.<br />
40. Hubbard RA, Kerilkowske K, Flowers CI, et al. Cumulative probability<br />
<strong>of</strong> false-positive recall or biopsy recommendation after 10 years <strong>of</strong> screening<br />
mammography: a cohort study. Ann Intern Med. 2011;155:481-492.<br />
41. Burnside E, Belkora J, Esserman L. The impact <strong>of</strong> alternative practices<br />
on the cost and quality <strong>of</strong> mammographic screening in the United States. Clin<br />
Breast Cancer. 2001;2:145-152.<br />
42. Schousboe JT, Kerilkowske K, Loh A, et al. Personalizing mammography<br />
by breast density and other risk factors for breast cancer: analysis <strong>of</strong><br />
health benefits and cost-effectiveness. Ann Intern Med. 2011;155:10-20.<br />
e45
COSTS OF CANCER CARE: AFFORDABILITY,<br />
ACCESS, AND POLICY<br />
CHAIR<br />
Thomas J. Smith, MD<br />
Sidney Kimmel Comprehensive Cancer Center at Johns Hopkins University<br />
Baltimore, MD<br />
SPEAKERS<br />
Richard Sullivan, PhD, MBBS<br />
Kings Health Partners Integrated Cancer Centre, Guys Hospital<br />
London, United Kingdom<br />
Sean R. Tunis, MD, MSc<br />
Center for Medical Technology Policy<br />
Baltimore, MD
Reducing the Cost <strong>of</strong> Cancer Care: How to<br />
Bend the Curve Downward<br />
By Thomas J. Smith, MD, Bruce E. Hillner, MD, and Ronan J. Kelly, MD, MBA<br />
Overview: Health care and cancer care costs are rising<br />
unsustainably such that insurance costs have doubled in 10<br />
years. Oncologists find themselves both victims <strong>of</strong> high costs<br />
and the cause <strong>of</strong> high-cost care by what we do and what we<br />
do not do. We previously outlined five ways that oncologists<br />
could personally bend the cost curve downward and five<br />
societal attitudes that would require change to lower costs.<br />
Here, we present some practical ways to reduce costs while<br />
THE RISING cost <strong>of</strong> cancer care is unsustainable. Medical<br />
care costs more in the United States than anywhere<br />
else in the world, as shown in Fig. 1. The impact <strong>of</strong> this<br />
rising cost is felt throughout the health care system, from<br />
patients and families to insurers and the government. The<br />
high cost <strong>of</strong> medical care is a concern for manufacturers who<br />
must build that cost into their goods. We illustrate some <strong>of</strong><br />
the impact in human terms in Sidebar 1.<br />
Cancer doctors <strong>of</strong>ten think we are the victims <strong>of</strong> the rising<br />
cost <strong>of</strong> cancer care, when we are both victims and causal<br />
agents. In fact, we are responsible for what we do and what<br />
we do not do, and the consequences <strong>of</strong> that action or inaction.<br />
We order the tests and prescribe the chemotherapy and<br />
supportive care drugs, and make the decisions to continue<br />
chemotherapy, involve palliative care or hospice early, have<br />
discussions about goals <strong>of</strong> care and advance directives—or<br />
not. In our review 4 we explored five changes in oncologist<br />
behavior that would bend the cost curve (Sidebar 2), and<br />
five attitudes that needed to change for this to happen<br />
(Sidebar 3).<br />
Here, we want to think about recognizing the coming<br />
problems, and propose some concrete solutions. First, the<br />
world is changing from fee-for-service to bundled payments<br />
to take away the incentive to administer more pr<strong>of</strong>itable<br />
chemotherapy. We have never maintained that oncologists<br />
administer chemotherapy just to make money. Interestingly,<br />
patients in countries like Sweden and Portugal, where<br />
oncologists do not make money giving chemotherapy, still<br />
receive chemotherapy near the end <strong>of</strong> life. 7 However, there is<br />
no way to generate an income <strong>of</strong> nearly $400,000/year from<br />
cognitive services alone, and there is an inherent conflict <strong>of</strong><br />
interest when we must choose between chemotherapy that<br />
gives us little pr<strong>of</strong>it compared with major pr<strong>of</strong>it. Second,<br />
more people will be uninsured or have less coverage with<br />
higher copays and deductibles, and shift between plans as<br />
insurance becomes more portable. Third, value is missing<br />
in some <strong>of</strong> our spending, if we define value as quality/cost.<br />
Finally, all <strong>of</strong> us have to realize that change is risky and<br />
disruptive, with major implications for our salaries and<br />
ability to support an enterprise.<br />
We propose practical ways to improve health, quality <strong>of</strong><br />
care, and value. First, we propose redesign <strong>of</strong> the National<br />
Comprehensive Cancer Network (NCCN) and other pathways<br />
to incorporate cost and value. Second, we propose an<br />
audit <strong>of</strong> current patterns <strong>of</strong> care for under- and overuse. We<br />
can help improve electronic medical record (EMR) prompts<br />
that promote best practices. Finally, we want to redesign<br />
NCCN and other pathways to incorporate palliative care<br />
e46<br />
maintaining or improving quality, including: 1) evidence-based<br />
surveillance after curative therapy; 2) reduced use <strong>of</strong> white<br />
cell stimulating factors (filgrastim and pegfilgrastim); 3) better<br />
integration <strong>of</strong> palliative care into usual oncology care; and<br />
4) use <strong>of</strong> evidence-based, cost-conscious clinical pathways<br />
that allow appropriate care and lead to equal or better<br />
outcomes at one-third lower cost.<br />
early and concurrently to provide the best care at a cost we<br />
can afford.<br />
This is part <strong>of</strong> a major national effort to restrain costs<br />
while maintaining quality. The <strong>American</strong> Board <strong>of</strong> Internal<br />
Medicine is promoting “Choose Wisely” to have each specialty<br />
find at least five ways to reduce costs 8 has been endorsed<br />
by at least eight major organizations (http://choosing<br />
wisely.org/wp-content/uploads/2011/12/about_choosingwisely.<br />
pdf). Accountable care organizations and medical homes<br />
are part <strong>of</strong> all new health care legislation and attempts to<br />
restrain costs while maintaining quality. 9 The <strong>American</strong><br />
College <strong>of</strong> Physicians has recognized the inherent tension<br />
between doing all that one can for an individual patient<br />
compared with careful use <strong>of</strong> societal resources: “Physicians<br />
have a responsibility to practice effective and efficient health<br />
care and to use health care resources responsibly. Parsimonious<br />
care that utilizes the most efficient means to<br />
effectively diagnose a condition and treat a patient respects<br />
the need to use resources wisely and to help ensure that<br />
resources are equitably available…. Physicians’ considered<br />
judgments should reflect the best available evidence including<br />
data on the cost-effectiveness <strong>of</strong> different clinical approaches.”<br />
10<br />
Some Specific Examples <strong>of</strong> Actions Under<br />
Oncologist Control<br />
Target Surveillance Tests or Imaging to Those<br />
Situations Where a Benefit Has Been Shown<br />
This one should be easy. There are no data suggesting<br />
better medical outcomes for patients with breast or ovary<br />
cancer treated with curative intent who are followed by<br />
other than routine exams. For breast cancer, two large<br />
European studies showed that routine follow-up compared<br />
with more intense schedules <strong>of</strong> scans gave equal outcomes<br />
and equal quality <strong>of</strong> life. There are no data that suggest<br />
early identification <strong>of</strong> metastatic breast cancer leads to<br />
better outcomes. In one small study, carcinoembryonic antigen<br />
(CEA) could detect breast cancer recurrence, but early<br />
From the Palliative Medicine Program, Sidney Kimmel Comprehensive Cancer Center,<br />
The Johns Hopkins University, Baltimore, MD; Massey Cancer Center, Virginia Commonwealth<br />
University, Richmond, VA.<br />
Authors’ disclosures <strong>of</strong> potential conflicts <strong>of</strong> interest are found at the end <strong>of</strong> this article.<br />
Address reprint requests to Thomas J. Smith, MD, Director <strong>of</strong> Palliative Medicine for<br />
Johns Hopkins Medicine, Sidney Kimmel Comprehensive Cancer Center at Johns Hopkins,<br />
600 N. Wolfe Street, Blaylock 369, Baltimore, MD 21287-0005; email: tsmit136@jhmi.edu.<br />
© <strong>2012</strong> by <strong>American</strong> <strong>Society</strong> <strong>of</strong> <strong>Clinical</strong> <strong>Oncology</strong>.<br />
1092-9118/10/1–10
REDUCING THE COST OF CANCER CARE<br />
Fig. 1. Expenditure per person on health care, in 2009 U.S. dollars.<br />
From OECD Health Data 2011, http://stats.oecd.org/Index.<br />
aspx?DataSetCode�SHA. 1<br />
treatment with tamoxifen in estrogen receptor–positive<br />
patients made absolutely no difference compared with treatment<br />
when patients were symptomatic. 11 Early identification<br />
<strong>of</strong> ovarian cancer recurrence with cancer antigen (CA)-<br />
125 testing did not lead to better medical outcomes in a<br />
large, well-designed randomized clinical trial. 12 There are<br />
no data suggesting benefit in lung or prostate cancer for<br />
early detection <strong>of</strong> recurrent disease, and screening is not<br />
recommended. In fact, the only common disease in which one<br />
can make a compelling argument is in colorectal cancer CEA<br />
testing. 13<br />
KEY POINTS<br />
● Cancer costs are rising at an unsustainable rate—by<br />
any measure.<br />
● Oncologists are directly responsible for some <strong>of</strong> the<br />
increase in costs, by what we do and what we do not<br />
do, and are one key to bending the cost curve downward.<br />
● We can reduce the cost <strong>of</strong> care by restricting surveillance<br />
<strong>of</strong> patients who have received treatment to<br />
those tests that have been shown to improve outcomes;<br />
this will save several billions <strong>of</strong> dollars in<br />
breast cancer alone.<br />
● We can save several billions <strong>of</strong> dollars a year in<br />
patients with metastatic solid tumors by reducing<br />
the use <strong>of</strong> white cell growth factors to the indications<br />
approved by ASCO, European Organisation for Research<br />
and Treatment <strong>of</strong> Cancer, and National Comprehensive<br />
Cancer Network, and with dose reduction<br />
instead <strong>of</strong> colony-stimulating factor use where appropriate.<br />
● We can improve care near the end <strong>of</strong> life by involving<br />
palliative care and hospice earlier—3 to 6 months<br />
before death—and, for most diseases by not administering<br />
chemotherapy to patients with poor performance<br />
status, or after progression despite three lines<br />
<strong>of</strong> chemotherapy.<br />
Sidebar 1. Some Facts about the Cost <strong>of</strong> Cancer<br />
Care in the United States<br />
● Medical care costs more in the United States than<br />
any other country—<strong>of</strong>ten twice as much—with no<br />
better survival. We spend $8,000/person/year versus<br />
Canada’s $4500. 1<br />
● Nearly one million 14 families suffered medical<br />
bankruptcy in 2010. (Commonwealth Fund) Over<br />
half <strong>of</strong> those bankruptcies happened to people<br />
with insurance, and most families are middle<br />
class. 2 Approximately 15% to 20% <strong>of</strong> insurance<br />
payments are for cancer, so we caused several<br />
them. Eight percent <strong>of</strong> families with a member<br />
with lung cancer are bankrupt because <strong>of</strong> the<br />
disease. 15<br />
● The cost <strong>of</strong> insurance for a family <strong>of</strong> four has risen<br />
from $6,000 to more than $15,000/year in the last<br />
10 years. 16 No wonder more people are underinsured—up<br />
to nearly 60 million—and others are<br />
underinsured.<br />
● Drug costs are rising at a rate <strong>of</strong> 3%, but represent<br />
only 13% <strong>of</strong> the total cost <strong>of</strong> care. 3<br />
● Much <strong>of</strong> the rest is under our control including<br />
imaging, chemotherapy choices, integration <strong>of</strong> palliative<br />
care, use <strong>of</strong> hospice, and avoiding chemotherapy<br />
and hospitalization near the end <strong>of</strong> life. 4<br />
● Oncologists’ salaries (median $381,992) are<br />
among the highest for medical specialists compared<br />
with primary care doctors ($202,392) and<br />
rose 4% in 2010. This disparity is unique in<br />
developed countries. 5<br />
● At least half <strong>of</strong> oncologists’ income, 50% to 80%,<br />
comes from drug sales which represent an inherent<br />
potential conflict <strong>of</strong> interest, and is unique in<br />
U.S. medicine. 5<br />
● Hospitals have an inherent interest in keeping<br />
beds full and pr<strong>of</strong>it margins high on drugs and<br />
services.<br />
● Approximately 25% <strong>of</strong> all Medicare funds are<br />
spent in the last year <strong>of</strong> life, and more than 9%<br />
(upwards <strong>of</strong> $50 billion) in the last month <strong>of</strong> life, 6<br />
with similar patterns in commercial insurance.<br />
The reason for the lack <strong>of</strong> effect on early recurrence has<br />
been stated eloquently by Dr. Tito Fojo: “Until we have<br />
treatments that kill essentially all the recurrent cancer,<br />
rather than just some <strong>of</strong> it, we will not have a chance <strong>of</strong> cure”<br />
(written communication, December 2011).<br />
The reasons patients and maybe doctors want these tests<br />
are complicated. Some patients want to be “doing something”<br />
even if something is not <strong>of</strong> proven benefit. Explaining<br />
the reasons behind not testing is more difficult and time<br />
consuming than ordering the test. Using breast cancer as an<br />
example, the ASCO printed guidelines that state “no testing”<br />
can be reviewed in less than 10 minutes 17 and patients<br />
can understand the rationale for not testing, 18 but it takes a<br />
conversation that may be difficult for some oncologists.<br />
ASCO is taking the lead in advocating for best practices<br />
in cancer surveillance. Adherence to the ASCO and NCCN<br />
e47
Sidebar 2. Five Changes in Oncologist Behavior<br />
that Will Bend the Cancer Cost Curve<br />
1. Target surveillance tests or imaging to those<br />
situations in which a benefit has been shown.<br />
2. For most solid tumors limit second- and thirdline<br />
treatment for metastatic cancer to sequential<br />
monotherapies. Based on the evidence,<br />
single drugs are indicated in breast, lung and<br />
prostate but not colorectal cancer.<br />
3. For patients with cancer that has progressed<br />
despite treatment, limit future chemotherapy to<br />
patients with good performance status.<br />
4. Replace the routine use <strong>of</strong> white-cell–stimulating<br />
factors in the treatment <strong>of</strong> metastatic solid<br />
cancers with chemotherapy dose reduction.<br />
5. For patients not experiencing response to three<br />
consecutive regimens, limit further chemotherapy<br />
to patients entering clinical trials.<br />
breast cancer guidelines with minimal testing will likely be<br />
one <strong>of</strong> ASCO’s five topics in the <strong>American</strong> Board <strong>of</strong> Internal<br />
Medicine “Choose Wisely” campaign.<br />
Replace the Routine Use <strong>of</strong> White-Cell Stimulating Factors in the<br />
Treatment <strong>of</strong> Metastatic Solid Cancers with Chemotherapy<br />
Dose Reduction<br />
The use <strong>of</strong> colony stimulating factors (CSFs) is one area<br />
in which the United States and other countries have very<br />
disparate patterns <strong>of</strong> care but no difference in mortality. 19<br />
The United States represents 3% <strong>of</strong> the world population but<br />
purchases 75% <strong>of</strong> the granulocyte (G-) CSF and pegylated<br />
G-CSF produced by Amgen (Thousand Oaks, CA). We have<br />
recently explored some <strong>of</strong> the reasons why this disparity<br />
occurs, including dislike <strong>of</strong> febrile neutropenia, marketing,<br />
fear <strong>of</strong> malpractice, and the pr<strong>of</strong>its made—<strong>of</strong>ten several<br />
hundred dollars for each injection. The only proven recommended<br />
uses are in dose-dense treatment for estrogenpositive<br />
breast cancer and for lymphoma treatment when<br />
the risk <strong>of</strong> febrile neutropenia is high. 20 A recent study<br />
showed that for patients older than age 65, there was no<br />
proven clinical benefit to the use <strong>of</strong> primary prophylactic<br />
CSFs and that the cost-effectiveness ratio was more than<br />
$900,000 per quality-adjusted life year saved, 21 which may<br />
make us question current practice or demand lower CSF<br />
prices.<br />
We continue to recommend that the United States follow<br />
the ASCO, NCCN, and European Organisation for Research<br />
and Treatment <strong>of</strong> Cancer (EORTC) guidelines and use CSFs<br />
for curative care. However, for the treatment <strong>of</strong> metastatic<br />
solid tumors in which no clinical benefit has been shown, we<br />
should follow ASCO and NCCN guidelines and use less toxic<br />
regimens, not use primary prophylaxis even in diseases such<br />
as small cell lung cancer (in which it is not recommended by<br />
NCCN guidelines), and if needed reduce the doses per the<br />
original protocol. We simply cannot afford $2200–4800 each<br />
cycle for supportive care that does not improve survival<br />
without the consequences in Sidebar 1.<br />
e48<br />
Sidebar 3. Five Attitudes that Must Change for<br />
Better Cost Consciousness<br />
1. Recognition that oncologists drive the costs <strong>of</strong><br />
care by what we do and do not do.<br />
2. Both doctors and patients need more realistic<br />
expectations.<br />
3. Realignment <strong>of</strong> compensation to rebalance cognitive<br />
services with chemotherapy use.<br />
4. Better integration <strong>of</strong> end-<strong>of</strong>-life non–chemotherapy-oriented<br />
palliative care.<br />
5. Acceptance <strong>of</strong> the necessity for cost-effectiveness<br />
analysis and the need for some limits on care.<br />
Changing Attitudes<br />
SMITH, HILLNER, AND KELLY<br />
There are oncologist attitudes that also drive the cost <strong>of</strong><br />
care. We highlight two: recognition that oncologists drive<br />
the costs <strong>of</strong> care by what we do and not do, and better<br />
integration <strong>of</strong> end-<strong>of</strong>-life, non–chemotherapy-oriented palliative<br />
care.<br />
For instance, having a discussion about impending death<br />
<strong>of</strong> a patient improves the pattern <strong>of</strong> care. Data show clearly<br />
that although we discuss curability or not, 22 only 37% <strong>of</strong> the<br />
time do we discuss the fact <strong>of</strong> an impending death. 23 Data<br />
also show clearly that having this discussion was associated<br />
with no more depression or anxiety in the patient; less<br />
depression in the caregiver; far less end-<strong>of</strong>-life intubation,<br />
resuscitation, and intensive care unit use; and longer hospice<br />
use. Just having that discussion lead to better medical<br />
care, better outcomes for the family, and $1,000 less spent in<br />
the last week <strong>of</strong> life. 24<br />
Sidebar 4. Integrating Best Practices: Use the<br />
Medical Record to Help Your Practice<br />
1. List the treatments used, so that we can know<br />
when it is time to switch to non-chemotherapy<br />
based care. Build in some prompts to trigger<br />
consultation.<br />
2. Put a prompt to remind you to discuss<br />
Goals <strong>of</strong> care<br />
Prognosis<br />
Advance medical directives<br />
“Code status”<br />
Hospice referral<br />
3. Have all patients with incurable cancer receive<br />
a hospice information visit when they have 3 to 6<br />
months to live, to make the transition smoother.<br />
4. If you are not comfortable discussing these difficult<br />
issues, appoint someone in your practice<br />
(an advance practice nurse or social worker) or<br />
get training.<br />
5. After a patient is referred to hospice, pencil in<br />
appointments to call them every week, just to<br />
check on them and make sure they do not feel<br />
abandoned.
REDUCING THE COST OF CANCER CARE<br />
This recognition <strong>of</strong> when it is time to switch treatments<br />
away from chemotherapy is essential to best quality <strong>of</strong> care,<br />
but we are not good at it. At the University <strong>of</strong> Iowa, 60% <strong>of</strong><br />
the patients who died in the hospital were eligible for<br />
hospice on their penultimate admission, yet this was recognized<br />
only 14% <strong>of</strong> the cases and few were enrolled. 25 Sixty<br />
percent <strong>of</strong> oncologists prefer to wait until there are no more<br />
chemotherapy options left to discuss hospice, advance medical<br />
directives, and “code status.” 22 This may explain why<br />
no doctor has mentioned hospice to half <strong>of</strong> all patients with<br />
lung cancer when they have 8 weeks left to live 26 and why<br />
one-third <strong>of</strong> all patients with cancer enter hospice with less<br />
than a week to live, according to the National Hospice and<br />
Palliative Care Organization. 27 Data are also clear that<br />
hospice does not worsen survival <strong>of</strong> patients with non–small<br />
cell lung cancer, 28,29 chemotherapy in the last 2 weeks <strong>of</strong> life<br />
does not improve survival, 28 and fourth-line chemotherapy<br />
has not worked with a response rate <strong>of</strong> 0%. 26 Despite this,<br />
a large percentage <strong>of</strong> people with incurable solid tumors<br />
receive chemotherapy within 2 weeks <strong>of</strong> their death, including<br />
at academic centers. 31<br />
We can improve care by incorporating palliative care<br />
early, as recommended by ASCO in the Provisional <strong>Clinical</strong><br />
Opinion. 32 Studies clearly show no harm, substantial benefit<br />
in most patient reported outcomes, and even better survival<br />
in lung cancer. The strongest correlations with increased<br />
survival in that study were not more chemotherapy but less<br />
intravenous chemotherapy in the last 60 days <strong>of</strong> life, and<br />
better understanding <strong>of</strong> prognosis and goals <strong>of</strong> treatment. 33<br />
Of note, the usual oncology arm used hospice for only 4 days<br />
if they used it at all.<br />
How can we improve care? We have put some simple<br />
recommendations in Sidebar 4.<br />
Bringing it All Together to Improve Care and Reduce<br />
Cost at the Same Time<br />
Non–small cell lung cancer is a good example <strong>of</strong> how we<br />
can modify existing guidelines to improve care. There are<br />
Fig. 2. Changing our practice to incorporate evidence-based<br />
guidelines and best practices about palliative care, using lung cancer<br />
as an example. These suggestions are superimposed on the survival<br />
curve <strong>of</strong> lung cancer patients treated on-pathway or <strong>of</strong>f-pathway;<br />
note that the survival curves are identical but the pathway patients’<br />
care cost 33% less. See Neubauer M et al. 35<br />
Abbreviations: ADs, advance directives; DPMA, durable power <strong>of</strong><br />
medical attorney; PC, palliative care.<br />
several important points that would improve care; we have<br />
illustrated them on the survival curve from U.S. <strong>Oncology</strong><br />
patients on and <strong>of</strong>f their clinical pathways in Figure 2. Note<br />
that the survival curves are superimposable, but the cost<br />
was 35% less with the use <strong>of</strong> the pathways. 34,35 The figure<br />
shows how this might work.<br />
In the first phase, the pathway restricts choice for patients<br />
with nonadenocarcinoma lung cancer to carboplatin plus<br />
paclitaxel with bevacizumab if indicated. This is the standard<br />
treatment arm <strong>of</strong> most cooperative groups. It does not<br />
allow the use <strong>of</strong> high-cost regimens such as carboplatin plus<br />
pemetrexed plus bevacizumab before pro<strong>of</strong> <strong>of</strong> effectiveness<br />
in a randomized clinical trial, which has been the standard<br />
for guidelines. 36<br />
In the second phase, in the first several visits, palliative<br />
care and hospice concepts are introduced by someone on the<br />
team. This can be the oncologist, social worker, or nurse, and<br />
is in line with best current practices. This ensures that the<br />
patient will have time to plan for when the disease grows,<br />
whether in weeks or years. We can discuss the regimens,<br />
outcomes, and costs with patients. Many patients will have<br />
satisfied their $5,000 deductible and not care about additional<br />
costs after that, but some may be paying high dollar<br />
amounts and we could manage their care with generic drugs<br />
and a chest film or standard computed tomography (CT)<br />
scan rather than a positron emission tomography CT scan at<br />
three times the cost. 37<br />
In the third phase we can follow ASCO and NCCN<br />
guidelines and not provide treatment to patients whose<br />
non–small cell lung cancer has grown despite three treatments,<br />
or whose Eastern Cooperative <strong>Oncology</strong> Group performance<br />
status greater than 2 (remembering that 3 is “in<br />
bed or chair over half the time”). If we have started palliative<br />
care (PC) early, it will be easier to transition to full hospice<br />
care. PC consultation improves appropriate hospice enrollment.<br />
Morrison and colleagues compared 1,427 patients who<br />
did not have a PC consult with 296 patients who did,<br />
matched in every other respect, and found that if the PC<br />
team consulted, 30% <strong>of</strong> hospice-eligible patients were discharged<br />
to hospice compared with just 1% <strong>of</strong> patients not<br />
seen by PC. 38 Patients enrolled in PC programs at Sutter<br />
Health enrolled in hospice 47% <strong>of</strong> the time, compared with<br />
20% <strong>of</strong> similar patients. 39,40 Appropriate transition to hospice<br />
not only improves care, 41 but is associated with longer<br />
survival 29 and lower costs. 42 PC “<strong>of</strong>floads” PC service provision<br />
from too-busy oncologists (e.g., surgeons, radiation<br />
oncologists, leukemia specialists, etc.) to PC specialist providers,<br />
43 and the model can save money for oncology practices<br />
and be financially self-sustaining. 44 As noted, PC<br />
consultation leads to increased hospice referrals, and exposure<br />
to palliative care before hospice is strongly associated<br />
with better caregiver quality <strong>of</strong> life, likely resulting from<br />
symptom management and better understanding <strong>of</strong> prognosis<br />
and goals. 45 The use <strong>of</strong> PC, with its attendant planning<br />
about goals <strong>of</strong> care and follow-up, reduced readmissions to<br />
the hospital per patient from 1.15 to 0.7 (a decrease <strong>of</strong> 36%)<br />
in the last 6 months <strong>of</strong> life. 46<br />
To make this reality will require an audit <strong>of</strong> charts and<br />
provision <strong>of</strong> feedback to the doctor and practice. We propose<br />
that we add these “overuse” criteria to ASCO’s Quality<br />
<strong>Oncology</strong> Practice Initiative (QOPI), which clearly works to<br />
reduce end-<strong>of</strong>-life chemotherapy. 47<br />
e49
Conclusion<br />
Costs are rising at an unsustainable rate and, if continued,<br />
will lead to more uninsured, less access, and more<br />
disparity. We can bend the cost curve downward by changing<br />
our practice patterns to provide evidence-based care,<br />
use less-expensive drugs and tests, and increase the use <strong>of</strong><br />
palliative care and hospice. This will maintain quality and<br />
decrease costs to pay for new and expensive advances.<br />
Authors’ Disclosures <strong>of</strong> Potential Conflicts <strong>of</strong> Interest<br />
Author<br />
Employment or<br />
Leadership<br />
Positions<br />
Consultant or<br />
Advisory Role<br />
ACKNOWLEDGMENT<br />
The authors gratefully acknowledge the National Cancer<br />
Institute, who supported this research through ACS Grant<br />
#PEP-10-174-01 (T.S.), R01CA116227-01 (T.J.S.), 2R01CA106370-<br />
05A1 (T.J.S.), and RC2CA148259 (B.E.H.).<br />
Stock<br />
Ownership Honoraria<br />
Research<br />
Funding<br />
Thomas J. Smith <strong>American</strong><br />
Cancer <strong>Society</strong><br />
Bruce E. Hillner*<br />
Ronan J. Kelly*<br />
*No relevant relationships to disclose.<br />
1. Organization for Economic Co-Operation and Development. OECD<br />
Health at a Glance 2011: Health expenditure and financing. http://stats.oecd.<br />
org/Index.aspx?DataSetCode�SHA. Acessed February 24, <strong>2012</strong>.<br />
2. Himmelstein DU, Thorne D, Warren E, et al. Medical bankruptcy in the<br />
United States, 2007: results <strong>of</strong> a national study. Am J Med. 2009;122:741-746.<br />
3. Hillner BE, Smith TJ. Efficacy does not necessarily translate to cost<br />
effectiveness: a case study in the challenges associated with 21st-century<br />
cancer drug pricing. J Clin Oncol. 2009;27:2111-2113.<br />
4. Smith TJ, Hillner BE. Bending the cost curve in cancer care. N Engl<br />
J Med. 2011;364:2060-2065.<br />
5. Gatesman ML, Smith TJ. The shortage <strong>of</strong> essential chemotherapy drugs<br />
in the United States. N Engl J Med. 2011;365:1653-1655.<br />
6. Riley GF, Lubitz JD. Long-term trends in Medicare payments in the last<br />
year <strong>of</strong> life. Health Serv Res. 2010;45:565-576.<br />
7. Braga S. Why do our patients get chemotherapy until the end <strong>of</strong> life? Ann<br />
Oncol. 2011;22:2345-2348.<br />
8. Brody H. Medicine’s ethical responsibility for health care reform: the top<br />
five list. N Engl J Med. 2010;362:283-285.<br />
9. Emanuel EJ, Pearson SD. Physician autonomy and health care reform.<br />
JAMA. <strong>2012</strong>;307:367-368.<br />
10. Snyder L. <strong>American</strong> College <strong>of</strong> Physicians Ethics Manual. Ann Intern<br />
Med. <strong>2012</strong>;156:73-104.<br />
11. Merimsky O, Kovner F, Inbar M, et al. Tamoxifen for disease-negative<br />
but MCA-positive breast cancer patients. Oncol Rep. 1997;4:843-847.<br />
12. Rustin GJ, van der Burg ME, Griffin CL, et al. Early versus delayed<br />
treatment <strong>of</strong> relapsed ovarian cancer (MRC OV05/EORTC 55955): a randomised<br />
trial. Lancet. 2010;376:1155-1163.<br />
13. Desch CE, Benson AB, Somerfield MR, et al. Colorectal cancer surveillance:<br />
2005 update <strong>of</strong> an <strong>American</strong> <strong>Society</strong> <strong>of</strong> <strong>Clinical</strong> <strong>Oncology</strong> practice<br />
guideline. J Clin Oncol. 2005;23:8512-8519.<br />
14. Himmelstein DU, Thorne D, Warren E, et al. Medical bankruptcy in the<br />
United States, 2007: results <strong>of</strong> a national study. Am J Med. 2009;122:741-746.<br />
15. Ramsey SD, McCune JS, Blough DK, et al. Colony-stimulating factor<br />
prescribing patterns in patients receiving chemotherapy for cancer. Am J<br />
Manag Care. 2010;16:678-686.<br />
16. Claxton G, DiJulio B, Whitmore H, et al. Health benefits in 2010:<br />
premiums rise modestly, workers pay more toward coverage. Health Aff<br />
(Millwood). 2010; Epub 2010 Sept. 2.<br />
17. Smith TJ. The <strong>American</strong> <strong>Society</strong> <strong>of</strong> <strong>Clinical</strong> <strong>Oncology</strong> recommended<br />
breast cancer surveillance guidelines can be done in a routine <strong>of</strong>fice visit.<br />
J Clin Oncol. 2005;23:6807.<br />
18. Loprinzi CL, Hayes D, Smith T. Doc, shouldn’t we be getting some<br />
tests? J Clin Oncol. 2003;21:108-111.<br />
19. Meropol NJ, Schulman KA. Cost <strong>of</strong> cancer care: issues and implications.<br />
J Clin Oncol. 2007;25:180-186.<br />
20. Smith TJ, Hillner BE. A way forward on the medically appropriate use<br />
<strong>of</strong> white cell growth factors (CSFs). J Clin Oncol. In press.<br />
21. Chan KKW, Siu E, Krahn MD, et al. A cost-utility analysis <strong>of</strong> primary<br />
prophylaxis versus secondary prophylaxis with granulocute colony-<br />
e50<br />
REFERENCES<br />
SMITH, HILLNER, AND KELLY<br />
Expert<br />
Testimony<br />
Other<br />
Remuneration<br />
stimulating factor in elderly patients wiht diffuse aggressive lymphoma<br />
receiving curative-intent chemotherapy. J Clin Oncol. Epub <strong>2012</strong> Mar 5.<br />
22. Keating NL, Beth Landrum M, Arora NK, et al. Cancer patients’ roles<br />
in treatment decisions: do characteristics <strong>of</strong> the decision influence roles?<br />
J Clin Oncol. 2010;28:4364-4370.<br />
23. Wright AA, Zhang B, Ray A, et al. Associations between end-<strong>of</strong>-life<br />
discussions, patient mental health, medical care near death, and caregiver<br />
bereavement adjustment. JAMA. 2008;300:1665-1673.<br />
24. Zhang B, Wright AA, Huskamp HA, et al. Health care costs in the last<br />
week <strong>of</strong> life: associations with end-<strong>of</strong>-life conversations. Arch Intern Med.<br />
2009;169:480-488.<br />
25. Freund K, Weckmann MT, Casarett DJ, et al. Hospice eligibility in<br />
patients who died in a tertiary care center. J Hosp Med. Epub 2011 Nov 15.<br />
26. Huskamp HA, Keating NL, Malin JL, et al. Discussions with physicians<br />
about hospice among patients with metastatic lung cancer. Arch Intern Med.<br />
2009;169:954-962.<br />
27. NHPCO Facts and Figures: Hospice Care in America. Alexandria, VA:<br />
National Hospice and Palliative Care Organization. http://www.nhpco.org/<br />
files/public/Statistics_Research/2011_Facts_Figures.pdf. Accessed January<br />
<strong>2012</strong>.<br />
28. Saito AM, Landrum MB, Neville BA, et al. The effect on survival <strong>of</strong><br />
continuing chemotherapy to near death. BMC Palliative Care. 2011;10:14.<br />
29. Connor SR, Pyenson B, Fitch K, et al. Comparing hospice and nonhospice<br />
patient survival among patients who die within a three-year window.<br />
J Pain Symptom Manage. 2007;33:238-246.<br />
30. Massarelli E, Andre F, Liu DD, et al. A retrospective analysis <strong>of</strong> the<br />
outcome <strong>of</strong> patients who have received two prior chemotherapy regimens<br />
including platinum and docetaxel for recurrent non-small-cell lung cancer.<br />
Lung Cancer. 2003;39:55-61.<br />
31. Dy SM, Asch SM, Lorenz KA, et al. Quality <strong>of</strong> end-<strong>of</strong>-life care for<br />
patients with advanced cancer in an academic medical center. J Palliat Med.<br />
2011;14:451-457.<br />
32. Smith TJ, Temin S, Alesi E, et al. <strong>American</strong> <strong>Society</strong> <strong>of</strong> <strong>Clinical</strong> <strong>Oncology</strong><br />
provisional clinical opinion: the integration <strong>of</strong> palliative care into standard<br />
oncology care. J Clin Oncol. Epub <strong>2012</strong> Feb 6.<br />
33. Temel JS, Greer JA, Admane S, et al. Longitudinal perceptions <strong>of</strong><br />
prognosis and goals <strong>of</strong> therapy in patients with metastatic non-small-cell lung<br />
cancer: results <strong>of</strong> a randomized study <strong>of</strong> early palliative care. J Clin Oncol.<br />
2011;29:2319-2326.<br />
34. Hoverman JR, Cartwright TH, Patt DA, et al. Pathways, outcomes, and<br />
costs in colon cancer: retrospective evaluations in two distinct databases.<br />
J Oncol Pract. 2011;7:52s-59s.<br />
35. Neubauer MA, Hoverman JR, Kolodziej M, et al. Cost effectiveness <strong>of</strong><br />
evidence-based treatment guidelines for the treatment <strong>of</strong> non-small-cell lung<br />
cancer in the community setting. J Oncol Pract. 2010;6:12-18.<br />
36. Azzoli CG, Baker S, Temin S, et al. <strong>American</strong> <strong>Society</strong> <strong>of</strong> <strong>Clinical</strong><br />
<strong>Oncology</strong> clinical practice guideline update on chemotherapy for stage IV non<br />
small-cell lung cancer. J Clin Oncol. 2009;27:6251-6266.<br />
37. McFarlane J, Riggins J, Smith TJ. SPIKE$: A six-step protocol for
REDUCING THE COST OF CANCER CARE<br />
delivering bad news about the cost <strong>of</strong> medical care. J Clin Oncol. 2008;26:<br />
4200-4204.<br />
38. Morrison RS, Dietrich J, Ladwig S, et al. Palliative care consultation<br />
teams cut hospital costs for medicaid beneficiaries. Health Aff. 2011;30:454-463.<br />
39. Meyers FJ, Linder J, Beckett L, et al. Simultaneous care: a model<br />
approach to the perceived conflict between investigational therapy and<br />
palliative care. J Pain Symptom Manage. 2004;28:548-556.<br />
40. Meyers FJ, Carducci M, Loscalzo MJ, et al. Effects <strong>of</strong> a problem-solving<br />
intervention (COPE) on quality <strong>of</strong> life for patients with advanced cancer on<br />
clinical trials and their caregivers: Simultaneous Care Educational Intervention<br />
(SCEI): linking palliation and clinical trials. J Palliat Med. 2011;14:465-<br />
473.<br />
41. Peppercorn JM, Smith TJ, Helft PR, et al. <strong>American</strong> <strong>Society</strong> <strong>of</strong> <strong>Clinical</strong><br />
<strong>Oncology</strong> statement: toward individualized care for patients with advanced<br />
cancer. J Clin Oncol. 2011;29:755-760.<br />
42. Pyenson B, Connor S, Fitch K, et al. Medicare cost in matched hospice<br />
and non-hospice cohorts. J Pain Symptom Manage. 2004;28:200-210.<br />
43. Alesi E, Fletcher D, Muir C, et al. Palliative care and oncology<br />
partnerships in real practice. <strong>Oncology</strong>. 2011;25:1287-1193.<br />
44. Muir JC, Daly F, Davis MS, et al. Integrating palliative care into the<br />
outpatient, Private Practice <strong>Oncology</strong> Setting. J Pain Symptom Manage.<br />
2010;40:126-135.<br />
45. Wittenburg-Lyles EM. What patients and families don’t hear: Backstage<br />
communication in hospice interdisciplinary team meetings. J Hous<br />
Elderly. 2009;23:92-105.<br />
46. Nelson C, Chand P, Soratais J, et al. Inpatient palliative care consults<br />
and the probability <strong>of</strong> hospital readmission. Perm J. 2011;15:48-51.<br />
47. Blayney DW, McNiff K, Hanauer D, et al. Implementation <strong>of</strong> the<br />
Quality <strong>Oncology</strong> Practice Initiative at a university comprehensive cancer<br />
center. J Clin Oncol. 2009;27:3802-3807.<br />
e51
NEW DIAGNOSTICS AND DEVICES IN THE ERA OF<br />
COMPARATIVE EFFECTIVENESS<br />
CHAIR<br />
Daniel F. Hayes, MD<br />
University <strong>of</strong> Michigan Medical Center<br />
Ann Arbor, MI<br />
SPEAKERS<br />
Muin J. Khoury, MD, PhD<br />
Office <strong>of</strong> Public Health Genomics, Centers for Disease Control and Prevention<br />
Atlanta, GA<br />
David Ransoh<strong>of</strong>f, MD<br />
University <strong>of</strong> North Carolina at Chapel Hill<br />
Chapel Hill, ND
Why Hasn’t Genomic Testing Changed the<br />
Landscape in <strong>Clinical</strong> <strong>Oncology</strong>?<br />
By Daniel F. Hayes, MD, Muin J. Khoury, MD, PhD, and David Ransoh<strong>of</strong>f, MD<br />
Overview: The “omics” revolution produced great optimism<br />
that tumor biomarker tests based on high-order analysis <strong>of</strong><br />
multiple (sometimes thousands) <strong>of</strong> factors would result in truly<br />
personalized oncologic care. Unfortunately, 10 years into the<br />
revolution, the promise <strong>of</strong> omics-based research has not yet<br />
been realized. The factors behind the slow progress in omicsbased<br />
clinical care are many. First, over the last 15 years,<br />
there has been a gradual recognition <strong>of</strong> the importance <strong>of</strong><br />
conducting tumor biomarker science with the kind <strong>of</strong> rigor that<br />
has traditionally been used for therapeutic research. However,<br />
this recognition has only recently been applied widely, and<br />
therefore most tumor biomarkers have insufficiently high<br />
levels <strong>of</strong> evidence to determine clinical utility. Second, omicsbased<br />
research <strong>of</strong>fers its own particular set <strong>of</strong> concerns,<br />
TUMOR BIOMARKERS are used to determine a patient’s<br />
current status and more importantly to predate<br />
future events that might be modified by intervention. 1<br />
Tumor biomarkers can be analyzed in cancer or healthy<br />
tissue, in secretions, and circulating in blood. Tumor biomarkers<br />
usually represent somatic changes that have<br />
emerged during the process <strong>of</strong> carcinogenesis. However, it is<br />
also reasonable to consider inherited germ-line differences<br />
between individuals that predict higher risk <strong>of</strong> developing a<br />
new malignancy or for estimating differential distribution,<br />
metabolism, or response to a drug (“pharmacogenomics”). 2,3<br />
Assays for tumor biomarkers can identify changes, or<br />
individual differences, in nucleic acids (DNA, RNA), proteins,<br />
lipids, whole cells, or tissue processes. Until recently,<br />
an assay for a biomarker usually analyzed a single analyte,<br />
or substance. Perhaps one <strong>of</strong> the best examples is the<br />
development <strong>of</strong> assays for the estrogen receptor (ER) to<br />
predict both prognosis and likelihood <strong>of</strong> responding to antiestrogen,<br />
or “endocrine” therapies. 4,5 The biology <strong>of</strong> ER was<br />
determined in the 1960s, and the first assay was a cumbersome<br />
test to biochemically measure binding <strong>of</strong> radioactively<br />
labeled estrogen to the receptor. Subsequent tests using<br />
specific antibodies to perform either enzyme-linked immunosorbent<br />
assays (ELISAs) and more recently immunohistochemistry<br />
(IHC) replaced the original ligand-binding<br />
assays, and are now almost uniformly used in clinical<br />
medicine. However, recently, newer assays that measure<br />
RNA expression have been introduced. Regardless, these<br />
are all tests that assay for a single analyte, ER, and not for<br />
several analytes that might be combined into a unified index<br />
designed to make a clinical decision.<br />
High-Dimensional Biomarkers<br />
In addition to measuring a single analyte, assessment <strong>of</strong><br />
complex processes, such as counting vessels to determine<br />
levels <strong>of</strong> angiogenesis, or development <strong>of</strong> a multifactoral<br />
index, such as tumor grade (which combines estimates <strong>of</strong><br />
relative gland formation, nuclear appearance, and mitotic<br />
rate), represent higher-order forms <strong>of</strong> a single test. In this<br />
regard, during the last decade, advances in molecular biology,<br />
technology, and bioinformatics have led to a new field<br />
loosely described as “omics.” This field encompasses several<br />
disciplines, generating high-dimensional data from global<br />
e52<br />
especially in regard to overfitting computational models and<br />
false discovery rates. Researchers and clinicians need to<br />
understand the importance <strong>of</strong> analytic validity, and the difference<br />
between clinical/biologic validity and clinical utility. The<br />
latter is required to introduce a tumor biomarker test <strong>of</strong> any<br />
kind (single analyte or omics-based), and are ideally generated<br />
by carefully planned and properly conducted “prospective<br />
retrospective” or truly prospective clinical trials. Only<br />
carefully planned studies, which take all three <strong>of</strong> these into<br />
account and in which the investigators are aware and recognize<br />
the enormous risk <strong>of</strong> unintended bias and overfitting<br />
inherent in omics-based test development, will ultimately<br />
result in translation <strong>of</strong> the exciting new technologies into<br />
better care for patients with cancer.<br />
sets <strong>of</strong> biologic molecules such as DNAs (“genomics”), RNAs<br />
(“transcriptomics”), proteins (“proteomics”), and metabolites<br />
(“metabolomics”). 6 Massive amounts <strong>of</strong> data are used to<br />
produce a pr<strong>of</strong>ile, or “signature,” generated by a computational<br />
mathematical function model. This model may be<br />
unsupervised, meaning that specimens are grouped computationally<br />
by apparent similarities in the omics patterns,<br />
without regard to preconceived biologic or clinical associations.<br />
Alternatively, generation <strong>of</strong> the pr<strong>of</strong>iles can be “supervised”;<br />
in this case, the signature is “pegged” to some sort <strong>of</strong><br />
prospectively defined biologic or clinical characteristic <strong>of</strong><br />
interest. Ultimately, at least in regard to clinical care, one or<br />
more omics-based test that reputedly has clinical utility in<br />
guiding patient care is generated.<br />
In an online continuous horizon scanning review <strong>of</strong> the<br />
literature from 2009 to the present, researchers from the<br />
Centers for Disease Control and Prevention found more than<br />
400 new genomic and other omics-based tests in transition<br />
from bench to bedside, <strong>of</strong> which the vast majority are related<br />
to cancer. 7 However, in <strong>2012</strong>, few if any omics-based tests<br />
have actually been widely adopted or embraced in the clinic.<br />
Why not? There are several obstacles that block introduction<br />
and use <strong>of</strong> an omics-based test. These relate to generation<br />
and validation <strong>of</strong> tumor biomarker tests in general, but<br />
in addition omics-based tests have special considerations<br />
that have impeded progress in the field. 8 It is essential that<br />
basic, translational, clinical, and computational scientists,<br />
and importantly clinicians caring for patients with cancer,<br />
understand these obstacles and work to overcome them so<br />
that patients receive better, more personalized oncologic<br />
care than they do now.<br />
From the University <strong>of</strong> Michigan Comprehensive Cancer Center, Ann Arbor, MI; Epidemiology<br />
and Genomics Research Program, Division <strong>of</strong> Cancer Control and Population<br />
Sciences, National Cancer Institute, Bethesda, MD; Office <strong>of</strong> Public Health Genomics,<br />
Centers for Disease Control and Prevention, Atlanta, GA; Departments <strong>of</strong> Medicine and<br />
Epidemiology, University <strong>of</strong> North Carolina at Chapel Hill, Chapel Hill, NC.<br />
Authors’ disclosures <strong>of</strong> potential conflicts <strong>of</strong> interest are found at the end <strong>of</strong> this article.<br />
Address reprint requests to Daniel F. Hayes, MD, Breast <strong>Oncology</strong> Program, University<br />
<strong>of</strong> Michigan Comprehensive Cancer Center, 6312 Cancer Center, 1500 E. Medical Center<br />
Drive, Ann Arbor, MI 48109-0942; email: hayesdf@umich.edu.<br />
© <strong>2012</strong> by <strong>American</strong> <strong>Society</strong> <strong>of</strong> <strong>Clinical</strong> <strong>Oncology</strong>.<br />
1092-9118/10/1-10
GENOMIC TESTING IN CLINICAL ONCOLOGY<br />
What Are the Problems?<br />
A Matter <strong>of</strong> Semantics<br />
The word “validation” means many things to many people.<br />
8,9 As in all scientific endeavors, carefully used semantics<br />
and precisely defined meanings are critical to<br />
development <strong>of</strong> tumor markers and incorporation into clinical<br />
care. In this regard, the independent multidisciplinary<br />
Evaluation <strong>of</strong> Genomic Applications in Practice and Prevention<br />
(EGAPP) Working Group, convened by the Centers for<br />
Disease Control and Prevention, has defined three important<br />
semantics that have greatly helped to organize the<br />
field. 10 They have defined three types <strong>of</strong> “validation.”<br />
Analytic validity. Analytic validity refers to the preanalytic<br />
and analytic factors that contribute to accuracy, reliability,<br />
and reproducibility <strong>of</strong> the specific assay test for the<br />
biomarker<br />
<strong>Clinical</strong> validity. <strong>Clinical</strong> validity implies that the biomarker<br />
assay distinguishes two or more subgroups within a<br />
population that have different biologic or clinical outcomes.<br />
<strong>Clinical</strong> validity does not imply that a tumor biomarker<br />
assay should be used to care for patients, but it is unlikely<br />
the assay will be useful if it does not at least have clinical<br />
validity.<br />
<strong>Clinical</strong> utility. <strong>Clinical</strong> utility requires that an assay<br />
with analytic and clinical validity be shown to be useful in<br />
improving patient outcomes to the extent that the patient’s<br />
care should differ than if the marker results were not<br />
available.<br />
What is “<strong>Clinical</strong> Utility?” Every clinician makes clinical<br />
decisions on a daily basis. He/she carefully weighs the<br />
relative benefits <strong>of</strong> a planned intervention against the relative<br />
risks and costs, in terms <strong>of</strong> both dollars and inconvenience<br />
to the patient. Diagnostics tests are used to narrow<br />
the funnel <strong>of</strong> differential diagnosis so that any therapeutic<br />
plan is focused as carefully as possible on the individual<br />
characteristics <strong>of</strong> the patient at hand. For example, administering<br />
chemotherapy to every person in our society would<br />
probably result in benefiting those who happened to have<br />
cancer, but would induce prohibitive toxicity and costs for<br />
the vast majority <strong>of</strong> individuals who do not. So, <strong>of</strong> course,<br />
clinicians use history, physical examination, blood tests,<br />
radiologic evaluation, and pathology to focus our therapy on<br />
those who have cancer, and even more so, on those who have<br />
a life-threatening cancer and for whom the benefits <strong>of</strong> a<br />
specific therapy are likely to outweigh the risks. For an<br />
individual who has a cancer and experiences disease response,<br />
the benefit is the same (100%), regardless <strong>of</strong> what we<br />
KEY POINTS<br />
● Tumor biomarker research requires a careful understanding<br />
<strong>of</strong> analytical and clinical validity and clinical<br />
utility.<br />
● Omics-based research is fraught with special problems<br />
related to over-fitting and lack <strong>of</strong> well-designed<br />
studies to generate validity and utility.<br />
● Specific pathways have been proposed to generate<br />
clinical utility by either conducting prospective, retrospective,<br />
or truly prospective clinical trials.<br />
know about that patient or whether everyone else is treated.<br />
However, the magnitude <strong>of</strong> the overall benefit for society is<br />
enhanced when the therapy is administered only to these<br />
patients and not to everyone else. Thus, biomarkers help<br />
identify those patients who either do not need or will not<br />
benefit from therapy. <strong>Clinical</strong> utility <strong>of</strong> a tumor marker can<br />
be established if the marker identifies those patients who<br />
are so unlikely to experience an event (a new cancer, or a<br />
recurrence or death as a result <strong>of</strong> an established cancer), or<br />
those so unlikely to respond or benefit, that the risks are the<br />
same or outweigh the benefits <strong>of</strong> treatment and they should<br />
forgo therapy. Determination <strong>of</strong> clinical utility requires<br />
judgment, on the part <strong>of</strong> the caregiver, the patient, and<br />
society, in regard to the relative benefits comapred with the<br />
risks and costs <strong>of</strong> the therapeutic plan under consideration.<br />
For example, adjuvant endocrine therapy for patients<br />
with breast cancer clearly reduces the odds <strong>of</strong> recurrence<br />
and death. 11 In general, adjuvant endocrine therapy is<br />
reasonably well tolerated with few life-threatening toxicities,<br />
but it is relatively expensive, and it does produce<br />
annoying adverse effects, which occasionally can be intolerable.<br />
ER is a powerful predictive factor for endocrine therapy.<br />
5 However, ER does not predict who will benefit. Indeed,<br />
adjuvant endocrine therapy decreases relapse in only approximately<br />
one-half <strong>of</strong> ER-positive patients who are destined<br />
to experience recurrence. ER does, though, strongly<br />
identify patients who will not benefit. The overall survival<br />
curves for adjuvant tamoxifen compared with nil in ER-poor<br />
patients are almost completely overlapping. 11 Are there<br />
some patients with ER-negative cancers who benefit? Almost<br />
certainly, but clinicians, patients, and guideline panels<br />
feel that the marker is sufficiently strong that the potential<br />
benefits <strong>of</strong> adjuvant endocrine therapy for these few patients<br />
with ER-negative breast cancer do not outweigh the risks<br />
that would be accrued for all <strong>of</strong> the others. 12<br />
Demonstrate <strong>Clinical</strong> Utility <strong>of</strong> a Tumor Biomarker<br />
In medicine, adoption <strong>of</strong> a new therapeutic strategy,<br />
especially related to a new drug, requires high levels <strong>of</strong><br />
evidence <strong>of</strong> a sufficiently large benefit that the intervention<br />
outweighs the risk. In most cases, to recommend adoption<br />
into standard <strong>of</strong> care, guideline bodies require one or more<br />
prospective randomized trials in which use <strong>of</strong> the new<br />
strategy is compared with standard <strong>of</strong> care without it.<br />
Unfortunately, although it should be, the same has not been<br />
true for clinical research <strong>of</strong> diagnostic devices, especially<br />
related to tumor biomarkers. Therefore, several markers<br />
that might be very helpful to personalize oncologic care have<br />
languished in uncertainty regarding their clinical utility.<br />
Perhaps just as unfortunately, markers with unproven clinical<br />
utility have been used to manage care, <strong>of</strong>ten with<br />
adverse outcomes. Both <strong>of</strong> these circumstances lead to poor<br />
disease management, and highlight the phrase that “a bad<br />
tumor marker is as bad as a bad drug.”<br />
In 1996, members <strong>of</strong> the ASCO Tumor Marker Guideline<br />
Committee proposed a tumor marker utility grading system,<br />
which included a level <strong>of</strong> evidence (LOE) scale that could be<br />
used to determine the relative quality <strong>of</strong> data supporting<br />
claims for clinical utility. 13 At that time, almost all tumor<br />
biomarker research fell into the LOE III category or worse,<br />
principally because most studies were performed out <strong>of</strong><br />
convenience <strong>of</strong> having available specimens and a novel<br />
assay. Ideally, to generate higher levels <strong>of</strong> evidence, inves-<br />
e53
tigators should develop prospectively designed and conducted<br />
trials to directly “test the test.” Indeed, a few such<br />
trials have been or are being conducted in North America,<br />
principally in breast and colorectal cancers. Several different<br />
trial designs to test tumor biomarkers have been proposed,<br />
but the details <strong>of</strong> these are outside the scope <strong>of</strong> this<br />
review and the reader is referred to two very well-written<br />
reviews <strong>of</strong> this subject. 14,15<br />
Prospective trials are expensive and time consuming. One<br />
advantage <strong>of</strong> tumor biomarker compared with therapeutic<br />
agent research is the opportunity to use archived specimens<br />
to generate high levels <strong>of</strong> evidence that might support, or<br />
refute, clinical utility <strong>of</strong> a tumor biomarker. Indeed, the<br />
appeal <strong>of</strong> this apparent advantage has worked to the disfavor<br />
<strong>of</strong> many tumor biomarkers, leading to lower-level evidence<br />
studies using conveniently available specimens<br />
without regard to trial design or proper statistical analysis.<br />
13 However, the ASCO Tumor Marker Guidelines Committee<br />
and others, such as the National Cancer Center<br />
(NCCN) committees, have relied as much as possible on<br />
having LOE I data to recommend use <strong>of</strong> a tumor biomarker<br />
to direct care. This publically stated strategy has led to<br />
better and more rigorous studies addressing the true clinical<br />
utility (as opposed to clinical validity) <strong>of</strong> specific tumor<br />
biomarker tests, including those that have been generated<br />
from omics research. Recently, Simon, Paik, and Hayes have<br />
proposed a refinement <strong>of</strong> the original ASCO LOE scale that<br />
establishes a hierarchy <strong>of</strong> tumor biomarker studies, ranging<br />
from very poor (retrospective studies <strong>of</strong> convenience using<br />
archived specimens that were not collected, processed, or<br />
stored with clinical trial–grade annotated data) to highlevel–evidence<br />
“prospective retrospective” studies. The latter<br />
type <strong>of</strong> studies use archived specimens from previously<br />
conducted clinical trials, and although not as well-regarded<br />
as true prospective studies, can, if performed carefully and<br />
validated properly, generate LOE I data to determine clinical<br />
utility. 16<br />
Embedded in such studies are the fundamental concepts<br />
<strong>of</strong> the Evaluation <strong>of</strong> Genomic Applications in Practice and<br />
Prevention (EGAPP) semantics. 10 An assay for a biomarker<br />
must be analytically validated in regard to technical accuracy,<br />
reproducibility, and reliability, especially in regard to<br />
the types <strong>of</strong> specimens to be used in the clinic. Furthermore,<br />
if the assay does not at least have clinical/biologic validity,<br />
it is unlikely to have clinical utility. However, proper study<br />
design must be used to generate the LOE needed to show<br />
clinical utility. To develop clinically useful tumor biomarkers,<br />
it is essential to carefully plan the experiment with<br />
well-developed, prospectively stated hypotheses regarding<br />
the intended use: risk assessment, detection <strong>of</strong> occult tumor,<br />
prognosis, prediction <strong>of</strong> response or resistance to specific<br />
therapy, or monitoring. 17 Moreover, analytic strategies<br />
should be determined in writing before the investigation<br />
is begun. These mandates do not preclude exploratory analyses<br />
and hypothesis-generating discovery studies, but the<br />
latter do not constitute high LOEs that demonstrate clinical<br />
utility.<br />
How Do Omics-Based Tests Differ from Other Tumor Biomarkers?<br />
In addition to the generic considerations related to tumor<br />
biomarker research in general, research <strong>of</strong> omics-based<br />
tests is saddled with several specific concerns. First, highdimensional<br />
data are particularly prone to overfitting. 9,18-20<br />
e54<br />
HAYES, KHOURY, AND RANSOHOFF<br />
Computational models <strong>of</strong>ten appear to perform beautifully<br />
during the discovery phase <strong>of</strong> omics-based research, but<br />
frequently fail when applied to an independent specimen/<br />
data set. 9 There are several factors that lead to these types<br />
<strong>of</strong> false discovery. Not uncommonly the pr<strong>of</strong>ile, or signature,<br />
has been generated without supervision <strong>of</strong> a clinical outcome<br />
or biologic hypothesis. Even if the discovery phase is performed<br />
with supervision to a biologic or clinical factor, the<br />
enormous number <strong>of</strong> factors that have been put into the<br />
original model (for example, expression <strong>of</strong> � 10,000 genes)<br />
and the relatively small data sets that are <strong>of</strong>ten used, make<br />
the lack <strong>of</strong> subsequent validation the exception rather than<br />
the rule.<br />
Has There Been Any Success in Translating Omics-Based Tests to<br />
<strong>Clinical</strong> <strong>Oncology</strong>?<br />
Although disappointing, the field has not been bereft <strong>of</strong><br />
success. Perhaps the best examples are illustrated in two<br />
separate, although linked, investigations that span the<br />
Atlantic Ocean. In the late 1990s, by using gene expression<br />
microarray technology, investigators from Amsterdam reported<br />
generation <strong>of</strong> a 70-gene signature that divided patients<br />
with early-stage breast cancer into very good versus<br />
poor prognosis. 21 Subsequent single and multi-institutional<br />
studies demonstrated clinical validity <strong>of</strong> this assay. 22,23<br />
Unfortunately, the original and follow-up validation studies<br />
were conducted using specimens that had been banked<br />
and stored as part <strong>of</strong> routine care (as opposed to part <strong>of</strong> a<br />
prospective clinical trial), without regard to a specifically<br />
identifiable question that would demonstrate clinical utility,<br />
and use <strong>of</strong> this assay has not been recommended by clinical<br />
guidelines committees (ASCO or NCCN). 12,24 A prospective<br />
randomized trial (the MINDACT trial) is now underway in<br />
Europe to determine whether this omics-based test should<br />
or should not be used to guide administration <strong>of</strong> adjuvant<br />
chemotherapy in node-negative patients.<br />
A second omics-based test was developed, in part, by<br />
incorporating some <strong>of</strong> the genes identified in the 70-gene<br />
signature, and in part by choosing logical and biologically<br />
based candidate genes. In this manner, investigators from<br />
the National Surgical Adjuvant Bowel and Breast Project<br />
(NSABP) generated a 21-gene assay specifically to identify a<br />
group <strong>of</strong> women with node-negative, ER-positive breast<br />
cancer treated with adjuvant tamoxifen whose prognosis<br />
was so good that, even if chemotherapy were effective, so few<br />
patients would benefit that its risks would outweigh its<br />
utility. 25 This test was shown to have superb analytic and<br />
clinical validity in a blinded analysis <strong>of</strong> a prestated hypothesis<br />
applying the assay to formalin-fixed, paraffin embedded<br />
tissue. 26,27 Moreover, the investigations used to determine<br />
clinical utility were “prospective retrospective” studies using<br />
material collected from women who participated in two<br />
prospective clinical trials addressing the benefits <strong>of</strong> tamoxifen<br />
in this patient group (NSABP B20 and B14). Thus,<br />
although the Simon-Paik-Hayes criteria had not yet been<br />
proposed at the time <strong>of</strong> the original publications <strong>of</strong> this<br />
assay, the ASCO Tumor Marker Guidelines Panel as well as<br />
the NCCN Breast Cancer Guidelines Panel have both recommended<br />
the use <strong>of</strong> the 21-gene assay for this clinical<br />
use. 12,24<br />
The predictive role <strong>of</strong> either <strong>of</strong> these assays for benefit or<br />
resistance to chemotherapy has been proposed on the basis<br />
<strong>of</strong> data from other prospective retrospective studies. 28,29
GENOMIC TESTING IN CLINICAL ONCOLOGY<br />
However, the specimen/data sets to validate the clinical<br />
utility <strong>of</strong> these exploratory observations are not available.<br />
Thus, the Eastern Cooperative <strong>Oncology</strong> Group and Southwest<br />
<strong>Oncology</strong> Group have completed or are conducting<br />
prospective clinical trials (TailoRX and RxPonder, respectively)<br />
within the North <strong>American</strong> Breast Cancer Group to<br />
address the predictive role <strong>of</strong> the 21-gene recurrence score in<br />
ER-positive patients with breast cancer.<br />
Conclusion<br />
Unfortunately, the field <strong>of</strong> tumor biomarker research has<br />
been chaotic and haphazard, leading to many published<br />
papers in the peer-reviewed literature, but very few markers<br />
Authors’ Disclosures <strong>of</strong> Potential Conflicts <strong>of</strong> Interest<br />
Author<br />
Employment or<br />
Leadership<br />
Positions<br />
Consultant or<br />
Advisory Role<br />
Daniel F. Hayes Biomarker<br />
Strategies;<br />
Chugai Pharma<br />
Muin J. Khoury*<br />
David Ransoh<strong>of</strong>f*<br />
*No relevant relationships to disclose.<br />
1. Khleif SN, Doroshow JH, Hait WN. AACR-FDA-NCI Cancer Biomarkers<br />
Collaborative consensus report: advancing the use <strong>of</strong> biomarkers in cancer<br />
drug development. Clin Cancer Res. 2010;16:3299-3318.<br />
2. Weinshilboum R. Inheritance and drug response. N Engl J Med. 2003;<br />
348:529-537.<br />
3. Wang L, McLeod HL, Weinshilboum RM. Genomics and drug response.<br />
N Engl J Med. 2011;364:1144-1153.<br />
4. Hammond ME, Hayes DF, Dowsett M, et al. <strong>American</strong> <strong>Society</strong> <strong>of</strong> <strong>Clinical</strong><br />
<strong>Oncology</strong>/College <strong>of</strong> <strong>American</strong> Pathologists guideline recommendations for<br />
immunohistochemical testing <strong>of</strong> estrogen and progesterone receptors in<br />
breast cancer (unabridged version). Arch Pathol Lab Med. 2010;134:e48-e72.<br />
5. Hammond ME, Hayes DF, Dowsett M, et al. <strong>American</strong> <strong>Society</strong> <strong>of</strong> <strong>Clinical</strong><br />
<strong>Oncology</strong>/College Of <strong>American</strong> Pathologists guideline recommendations for<br />
immunohistochemical testing <strong>of</strong> estrogen and progesterone receptors in<br />
breast cancer. J Clin Oncol. 2010;28:2784-2795.<br />
6. Field D, Glockner FO, Garrity GM, et al. Meeting report: the fourth<br />
Genomic Standards Consortium (GSC) workshop. OMICS. 2008;12:101-108.<br />
7. Gwinn M, Grossniklaus DA, Yu W, et al. Horizon scanning for new<br />
genomic tests. Genet Med. 2011;13:161-165.<br />
8. Ioannidis JP, Khoury MJ. Improving validation practices in “omics”<br />
research. Science. 2011;334:1230-1232.<br />
9. Ransoh<strong>of</strong>f DF. Rules <strong>of</strong> evidence for cancer molecular-marker discovery<br />
and validation. Nat Rev Cancer. 2004;4:309-314.<br />
10. Teutsch SM, Bradley LA, Palomaki GE, et al. The Evaluation <strong>of</strong><br />
Genomic Applications in Practice and Prevention (EGAPP) Initiative: methods<br />
<strong>of</strong> the EGAPP Working Group. Genet Med. 2009;11:3-14.<br />
11. Early Breast Cancer Trialists’ Collaborative Group (EBCTCG). Effects<br />
<strong>of</strong> chemotherapy and hormonal therapy for early breast cancer on recurrence<br />
and 15-year survival: an overview <strong>of</strong> the randomised trials. Lancet. 2005;365:<br />
1687-1717.<br />
12. Harris L, Fritsche H, Mennel R, et al. <strong>American</strong> <strong>Society</strong> <strong>of</strong> <strong>Clinical</strong><br />
<strong>Oncology</strong> 2007 update <strong>of</strong> recommendations for the use <strong>of</strong> tumor markers in<br />
breast cancer. J Clin Oncol. 2007;25:5287-5312.<br />
13. Hayes DF, Bast RC, Desch CE, et al. Tumor marker utility grading<br />
system: a framework to evaluate clinical utility <strong>of</strong> tumor markers. J Natl<br />
Cancer Inst. 1996;88:1456-1466.<br />
14. Sargent DJ, Conley BA, Allegra C, et al. <strong>Clinical</strong> trial designs for<br />
predictive marker validation in cancer treatment trials. J Clin Oncol. 2005;<br />
23:2020-2027.<br />
15. Freidlin B, McShane LM, Korn EL. Randomized clinical trials with<br />
biomarkers: design issues. J Natl Cancer Inst. 2010;102:152-160.<br />
16. Simon RM, Paik S, Hayes DF. Use <strong>of</strong> archived specimens in evaluation<br />
that truly have clinical utility or that can be recommended<br />
for routine patient care. This situation has led to two<br />
problematic circumstances: 1) use <strong>of</strong> biomarkers in the<br />
absence <strong>of</strong> high levels <strong>of</strong> evidence supporting their clinical<br />
utility; and 2) lack <strong>of</strong> biomarkers that can lead to truly<br />
personalized cancer care with high confidence. It is imperative<br />
that the field take major actions to break the vicious<br />
cycle that has led to these circumstances and to do the<br />
rigorous research needed to provide proper assessments,<br />
because “a bad tumor marker is as bad as a bad drug.” When<br />
these factors are recognized and incorporated into tumor<br />
biomarkers studies, the dream <strong>of</strong> personalized oncologic care<br />
will be more likely to become a reality.<br />
Stock<br />
Ownership Honoraria<br />
Research<br />
Funding<br />
OncImmune Novartis; Pfizer;<br />
Veridex<br />
REFERENCES<br />
Expert<br />
Testimony<br />
Other<br />
Remuneration<br />
<strong>of</strong> prognostic and predictive biomarkers. J Natl Cancer Inst. 2009;101:1446-<br />
1452.<br />
17. Henry NL, Hayes DF. Uses and abuses <strong>of</strong> tumor markers in the<br />
diagnosis, monitoring, and treatment <strong>of</strong> primary and metastatic breast<br />
cancer. Oncologist. 2006;11:541-552.<br />
18. Simon R. Roadmap for developing and validating therapeutically relevant<br />
genomic classifiers. J Clin Oncol. 2005;23:7332-7341.<br />
19. Simon R. Development and validation <strong>of</strong> therapeutically relevant<br />
multi-gene biomarker classifiers. J Natl Cancer Inst. 2005;97:866-867.<br />
20. Simon R. Development and evaluation <strong>of</strong> therapeutically relevant<br />
predictive classifiers using gene expression pr<strong>of</strong>iling. J Natl Cancer Inst.<br />
2006;98:1169-1171.<br />
21. van’t Veer LJ, Dai H, van de Vijver MJ, et al. Gene expression pr<strong>of</strong>iling<br />
predicts clinical outcome <strong>of</strong> breast cancer. Nature. 2002;415:530-536.<br />
22. van de Vijver MJ, He YD, van’t Veer LJ, et al. A gene-expression<br />
signature as a predictor <strong>of</strong> survival in breast cancer. N Engl J Med.<br />
2002;347:1999-2009.<br />
23. Buyse M, Loi S, van’t Veer L, et al. Validation and clinical utility <strong>of</strong> a<br />
70-gene prognostic signature for women with node-negative breast cancer.<br />
J Natl Cancer Inst. 2006;98:1183-1192.<br />
24. Carlson RW, Allred DC, Anderson BO, et al. Invasive breast cancer.<br />
J Natl Compr Canc Netw. 2011;9:136-222.<br />
25. Paik S, Shak S, Tang G, et al. A multi-gene RT-PCR assay using fixed,<br />
paraffin-embedded tumor tissue to predict the likelihood <strong>of</strong> breast cancer<br />
recurrence in node negative, estrogen receptor positive, tamoxifen-treated<br />
patients. N Engl J Med. 2004;351:2817-2826.<br />
26. Cronin M, Pho M, Dutta D, et al. Measurement <strong>of</strong> gene expression in<br />
archival paraffin-embedded tissues: development and performance <strong>of</strong> a 92gene<br />
reverse transcriptase-polymerase chain reaction assay. Am J Pathol.<br />
2004;164:35-42.<br />
27. Cronin M, Sangli C, Liu ML, et al. Analytical validation <strong>of</strong> the Oncotype<br />
DX genomic diagnostic test for recurrence prognosis and therapeutic response<br />
prediction in node-negative, estrogen receptor-positive breast cancer. Clin<br />
Chem. 2007;53:1084-1091.<br />
28. Albain KS, Barlow WE, Shak S, et al. Prognostic and predictive value<br />
<strong>of</strong> the 21-gene recurrence score assay in postmenopausal women with nodepositive,<br />
oestrogen-receptor-positive breast cancer on chemotherapy: a retrospective<br />
analysis <strong>of</strong> a randomised trial. Lancet Oncol. 2010;11:55-65.<br />
29. Straver ME, Glas AM, Hannemann J, et al. The 70-gene signature as a<br />
response predictor for neoadjuvant chemotherapy in breast cancer. Breast<br />
Cancer Res Treat. 2010;119:551-558.<br />
e55
MANAGEMENT OF CHRONIC LYMPHOCYTIC<br />
LEUKEMIA<br />
CHAIR<br />
Neil E. Kay, MD<br />
Mayo Clinic<br />
Rochester, MN<br />
SPEAKERS<br />
Peter Hillmen, PhD<br />
Leeds General Infirmary<br />
Leeds, United Kingdom<br />
John G. Gribben, DSc, MD<br />
St. Bartholomew’s Hospital<br />
London, United Kingdom
Predicting <strong>Clinical</strong> Outcome in B-Chronic<br />
Lymphocytic Leukemia<br />
Overview: B-Chronic lymphocytic leukemia (CLL) is a relatively<br />
common B-cell malignancy that has a very heterogeneous<br />
clinical course, despite carrying the designation <strong>of</strong><br />
“chronic,” which is a gross oversimplification. Being able to<br />
give some estimate <strong>of</strong> the rates <strong>of</strong> disease progression and<br />
overall survival (OS) at first diagnosis is, therefore, important<br />
in CLL. The ability to accurately predict response to therapy,<br />
as well as subsequent duration <strong>of</strong> response to therapy, is<br />
required given the variability <strong>of</strong> current therapies to induce<br />
and sustain treatment responses. The holy grail <strong>of</strong> prognostics<br />
THE MAJOR features <strong>of</strong> CLL that dictate need for<br />
prognostics include the following points: it is a relatively<br />
common leukemia (approximately one in 100,000<br />
patients in North America), 1 the disease is incurable with<br />
the exception <strong>of</strong> transplant, clinical courses are notoriously<br />
variable, and the patients have considerable anxiety with<br />
this diagnosis. The latter is now well documented and it is<br />
attributable to the fact that we <strong>of</strong>ten do not treat these<br />
patients with Rai stage 0–1 on diagnosis. This practice is<br />
based on clinical trial data suggesting treatment is not a<br />
benefit in the early stages <strong>of</strong> disease. 2 It is also partially<br />
because we do not have a ready portfolio <strong>of</strong> relatively<br />
nontoxic agents for treating our early-stage patients with<br />
CLL. The anxiety found in CLL is pervasive and has been<br />
quantified, with clear evidence that the anxiety may exist for<br />
years postdiagnosis. 3,4 The cause <strong>of</strong> this psychologic distress<br />
is multifactorial but one causal aspect is the current dogma<br />
to watch and evaluate early-stage CLL as best as is possible<br />
for a given patient. The case for prognostics, therefore, is<br />
that it provides useful guides for initial patient assurance<br />
(or not) and subsequent treatment. It is this need for<br />
prognostic assistance in early-stage CLL that primarily<br />
drove the research to develop more effective and powerful<br />
methods that inform us. The result <strong>of</strong> this research is now<br />
evident in that there are prognostic parameters used alone<br />
or in model systems that assist us in counseling and/or<br />
management <strong>of</strong> the majority <strong>of</strong> patients with CLL. These<br />
models can incorporate either clinical-based factors or molecular<br />
parameters that reflect the biology <strong>of</strong> CLL B cell and<br />
its microenvironment. These prognostic features or models<br />
are helpful for predicting time to first therapy, extent <strong>of</strong><br />
response to treatments, and duration <strong>of</strong> response. In some<br />
cases, these features or models guide us in limited ways to<br />
treatment choices. Although not ideal, the wise use <strong>of</strong> these<br />
prognostics is and should be a major assist to us in CLL<br />
practice.<br />
In this article, I will survey the available maneuvers and<br />
tests that can be used for most patients and that will provide<br />
for the practitioner and the patient a guidepost to risk<br />
stratification. These tests can tell us whether a given patient<br />
has a lower or higher risk <strong>of</strong> progressing over time. In<br />
addition, it is possible to use these tests to guide patient<br />
counseling, including determining how <strong>of</strong>ten a patient<br />
should be seen by the practitioner and potentially some<br />
treatment selection.<br />
394<br />
By Neil E. Kay, MD<br />
would be to state with accuracy which therapy or types <strong>of</strong><br />
therapy are best for a given patient. Although there is no<br />
complete answer to prognostic counseling, there is a continued<br />
development <strong>of</strong> markers specific to the CLL B cell and/or<br />
to its environment, as well as <strong>of</strong> testing <strong>of</strong> prognostic models.<br />
These models use both traditional and novel prognostic markers<br />
that can aid in the dissection <strong>of</strong> outcome for early-stage<br />
CLL in terms <strong>of</strong> progression risk and time to therapy. This has<br />
resulted in significant enhancement <strong>of</strong> our ability to guide and<br />
predict outcome for our patients with CLL.<br />
<strong>Clinical</strong> Course and Prognostic Parameters<br />
The current dogma is that for every 100 patients with<br />
CLL, approximately one-third will not progress to treatment<br />
even over decades, one-third will eventually progress, and<br />
one-third will need more urgent treatment than the rest. 5 In<br />
addition, because <strong>of</strong> intensive research on early-stage CLL it<br />
is known that approximately 50% <strong>of</strong> those patients will have<br />
high risk based on the presence <strong>of</strong> adverse prognostic features.<br />
6,7 Indeed, we can now identify a cohort <strong>of</strong> patients<br />
with high-risk CLL at diagnosis who will have rapid disease<br />
progression, poor response to treatment, and poor survival<br />
based on prognostic methods developed from an improved<br />
understanding <strong>of</strong> the biology <strong>of</strong> CLL. The prognostic parameters<br />
that are used to define this risk can be subdivided into<br />
both traditional/clinical and novel prognostic factors. The<br />
traditional and clinical factors are usually based on quantifiable<br />
plasma factors (beta-2 microglobulin, lactic dehydrogenase<br />
[LDH]), Rai stage, or hematologic features such as<br />
bone marrow features (diffuse infiltration) or levels <strong>of</strong> blood<br />
lymphocyte counts over time. The novel prognostic parameters<br />
that are in routine practice are leukemic cell based.<br />
Examples <strong>of</strong> the latter include the presence <strong>of</strong> membrane<br />
proteins such as CD38 or CD49 days, cytoplasmic presence<br />
<strong>of</strong> ZAP-70, and prognostic nuclear features including immunoglobulin<br />
variable heavy chain (IgVH) gene mutation status<br />
and cytogenetic abnormalities on fluorescent in situ<br />
hybridization (iFISH). In the assessment <strong>of</strong> these prognostic<br />
factors for CLL it is critical to consider and clarify the<br />
clinical features that are studied for association with the<br />
particular prognostic factor(s). Although the focus has been<br />
on using prognostics for previously untreated CLL, there is<br />
a special need for more information on subsequent course for<br />
relapsed patients. The most helpful information in these<br />
cases is the time <strong>of</strong> the patient’s relapse from initial therapy.<br />
Here we also need better prognostic variable to predict their<br />
subsequent clinical outcomes. However, for now, the treatment<br />
responses following initial therapy can be best predicted<br />
by clinical features that include extent <strong>of</strong> response<br />
From the College <strong>of</strong> Medicine, Mayo Clinic, Rochester, MN.<br />
Author’s disclosure <strong>of</strong> potential conflicts <strong>of</strong> interest are found at the end <strong>of</strong> this article.<br />
Address reprint requests to Neil E. Kay, MD, College <strong>of</strong> Medicine, Mayo Clinic, Stabile<br />
6-28, 200 First Street SW, Rochester, MN 55905; email: kay.neil@mayo.edu.<br />
© <strong>2012</strong> by <strong>American</strong> <strong>Society</strong> <strong>of</strong> <strong>Clinical</strong> <strong>Oncology</strong>.<br />
1092-9118/10/1-10
PREDICTIVE PARAMETERS IN CLL<br />
Table 1. Prognostic Factors That Are <strong>of</strong> Help in Defining<br />
Risk <strong>of</strong> Progression<br />
Factor Biologic Role Test Assay 1<br />
CD38 Cell activation Flow cytometry<br />
ZAP-70 Cell signaling Flow cytometry<br />
CD49d Cell adhesion Flow cytometry<br />
iFISH panel Common recurring<br />
genetic defects<br />
to first-line therapy and the first remission duration. This<br />
article primarily focuses on the most recent work on prognostic<br />
parameters that best predict the clinical course for<br />
newly diagnosed patients, their subsequent response to<br />
upfront therapy, and its duration.<br />
Useful Prognostic Variables in CLL<br />
Traditional Prognostic Features<br />
The most useful prognostic features incorporate clinical<br />
features <strong>of</strong> the patient at diagnosis and have been used for<br />
decades as the Rai and Binet staging systems. Although<br />
these are still helpful, they do not identify good from bad<br />
prognostic cohorts in the patient with early-stage CLL.<br />
Because we now identify at least 70% <strong>of</strong> patients in earlystage<br />
disease, these staging systems are no longer satisfactory<br />
as sole predictors (Table 1). To improve on earlier<br />
predictors for patients with CLL, a number <strong>of</strong> routine blood<br />
and/or serum factors have been assessed for their utility.<br />
The lymphocyte doubling time (LDT), calculated by determining<br />
the number <strong>of</strong> months it takes the absolute lymphocyte<br />
count (ALC) to double in number, is a marker <strong>of</strong> disease<br />
kinetics that has been found to correlate with both<br />
progression-free (PFS) and OS rates. 8 Beyond this, the ALC<br />
KEY POINTS<br />
Probes detected by microscopic<br />
fluorescence<br />
1. None <strong>of</strong> the prognostic factors are approved by the U.S. Food and Drug<br />
Administration for use in CLL with the exception <strong>of</strong> some iFISH probes.<br />
High Risk factors for adverse clinical outcomes<br />
● Presence <strong>of</strong> del(17p13) or del(11q22) or complex iFISH (more than one iFISH<br />
detectable defect)<br />
o predict for shorter time to therapy (del(17p13)) except if seen close to<br />
diagnosis<br />
o del(11q22) is associated with bulky disease and less response to therapy<br />
● Unmutated status with or without ZAP-70<br />
o predict for shorter time to therapy<br />
o predict for less durable response to therapy<br />
● del(17p13) with or without inactivating p53 mutations<br />
o predict for poor response to chemoimmunotherapy<br />
o <strong>of</strong>ten need stem cell transplant ultimately or experimental drug approaches<br />
● Prognostic information is very helpful in initial treatment<br />
<strong>of</strong> patients with B-chronic lymphocytic leukemia.<br />
● Biomarkers that reflect key biologic aspects <strong>of</strong> the<br />
CLL B cell and/or its environment are highly informative<br />
<strong>of</strong> disease course.<br />
● Prognosis for time to treatment from first diagnosis is<br />
now feasible for a given patient.<br />
● Models for prediction <strong>of</strong> durability <strong>of</strong> response to<br />
standard therapies are available.<br />
● Targeted therapy based on prognostics are emerging<br />
but still immature.<br />
has also been shown to predict both PFS and OS. 9,10 LDT<br />
values have been shown to be predictive <strong>of</strong> outcome for<br />
patients with early-stage CLL. A median PFS <strong>of</strong> 20 months<br />
for Binet stage A patients with an LDT <strong>of</strong> 12 months or less<br />
was found, compared with 75 months for those with an LDT<br />
<strong>of</strong> more than 12 months. The same study found a median<br />
PFS <strong>of</strong> 17 months for patients with an ALC <strong>of</strong> greater than<br />
30 � 10 9 /L compared with 88 months for those with an<br />
ALC <strong>of</strong> less than 30 � 10 9 /L 9 . The serum level <strong>of</strong> beta-2<br />
microglobulin, when the level is greater than 4.0 mg/dL at<br />
presentation, can be an adverse prognostic feature, 9,10 but<br />
this measure is not valuable as a sequential tool and may be<br />
confounded by other clinical variables such as renal disease<br />
and infectious or inflammatory conditions. One study found<br />
that the median PFS <strong>of</strong> patients with beta-2 microglobulin<br />
levels greater than 3.5 mg/L was 13 months compared with<br />
75 months for those with a beta-2 microglobulin levels less<br />
than 3.5 mg/L 9 . Similarly the use <strong>of</strong> a LDT can be difficult<br />
because <strong>of</strong> associated disease conditions that will also raise<br />
lymphocyte counts.<br />
Novel Prognostic Features<br />
Here we refer to prognostic features that are clearly<br />
related to leukemic B-cell biology. The novel prognostic<br />
markers that are in widespread use in clinical practice<br />
include: immunoglobulin heavy-chain variable region<br />
(IgVH) mutational status, interphase fluorescence in situ<br />
hybridization (iFISH) abnormalities, CD38, and zetaassociated<br />
protein (ZAP)-70. These markers on an individual<br />
basis can be used to predict clinical outcome where patients<br />
with mutated IgVH genes, low CD38 expression, low ZAP-70<br />
expression, and the absence <strong>of</strong> del(17p13) del(11q22) are all<br />
associated with a good prognosis. 11-20 Each <strong>of</strong> these markers<br />
has been shown to be a predictor <strong>of</strong> time to treatment and<br />
<strong>of</strong> OS on univariate analysis. Of special interest, it is only<br />
iFISH defects <strong>of</strong> novel prognosis that are known to change<br />
over time in CLL. This latter aspect is <strong>of</strong> importance as<br />
clonal evolution (acquisition <strong>of</strong> new genetic changes or <strong>of</strong><br />
increasing percentages <strong>of</strong> abnormal leukemic cells by iFISH)<br />
is associated with more aggressive disease. Clonal evolution<br />
is associated with short survival, but the difference in<br />
survival was limited to patients who acquired a del(17p13)<br />
or del(11q22) where the median survival postclonal evolution<br />
was only 1.3 years. 21 In addition newer information on<br />
iFISH defects initially associated with good clinical outcome,<br />
such as 13q-, may not always hold true. Thus for patients<br />
with CLL with 13q- those with higher percentages or those<br />
with structural defects larger than the typical defect (shown<br />
by single nucleotide polymorphisms [SNP] chips) have a<br />
greater tendency to have worse clinical outcome. 22<br />
Important issues regarding these prognostic parameters<br />
are that the pivotal studies identifying these prognostic<br />
markers used cohorts <strong>of</strong> patients with all stages <strong>of</strong> CLL. Also<br />
there is no currently accepted or agreed on mandate for use<br />
<strong>of</strong> these markers for prediction <strong>of</strong> disease outcome for CLL<br />
despite their usefulness. The need for prognostic evaluation<br />
approaches, however, is now buttressed by many studies<br />
conducted in patients with recently diagnosed CLL. The use<br />
<strong>of</strong> these novel prognostics can help better define the heterogeneity<br />
in outcome and can at least promote enhanced<br />
counseling information and, thus, potentially alleviate significant<br />
anxiety for many patients.<br />
The following can be most useful in predicting more<br />
395
aggressive disease for early-stage disease: presence <strong>of</strong> highrisk<br />
iFISH (del(17p13) del(11q22) or complex iFISH (multiple<br />
iFISH defects), p53 mutations, unmutated IgVH status,<br />
or combinations <strong>of</strong> these defects. The studies <strong>of</strong> IgVH status<br />
are most useful in showing the relative power <strong>of</strong> a single<br />
prognostic factor related to the biology <strong>of</strong> the CLL B cell.<br />
Thus the pivotal study <strong>of</strong> IgVH demonstrated that patients<br />
with CLL who had mutated immunoglobulin IgVH had a<br />
survival <strong>of</strong> more than 20 years, whereas those with unmutated<br />
IgVH had median survival <strong>of</strong> around 8-years. 11,12 The<br />
ability <strong>of</strong> IgVH mutation status to stratify OS remains when<br />
applied exclusively to patients with early-stage disease. 12<br />
For prediction <strong>of</strong> response to therapy, patients with<br />
del(17p13) or del(11q22) defects are less likely to have a<br />
vigorous response to even the most aggressive upfront treatments.<br />
For patients with unmutated and/or del(17p13), the<br />
durability <strong>of</strong> response to therapy is markedly reduced.<br />
Recent work has shown that clinical course in patients with<br />
iFISH-detectable del(17p13) is pr<strong>of</strong>oundly affected by the<br />
presence or absence <strong>of</strong> TP53 mutations. 23 If TP53 mutations<br />
(detected by SNP microarray or by sequencing <strong>of</strong> the p53<br />
gene) are present along with the del(17p13), OS is adversely<br />
affected. 24 Most recently there have been advances in putting<br />
together prognostic models for prediction <strong>of</strong> clinical<br />
courses and response to therapy.<br />
Overview <strong>of</strong> Prognostic Models in CLL<br />
Most patients in the 21 st century are diagnosed with<br />
early-stage disease and within this early-stage group, the<br />
vast majority do not require immediate treatment. There is,<br />
however, a small subgroup <strong>of</strong> patients who do present with<br />
early-stage disease and with symptoms such as fatigue or<br />
night sweats or with advanced-stage disease, requiring<br />
treatment. Although the latter cohort is not a clinical dilemma<br />
in terms <strong>of</strong> treatment decisions, the majority <strong>of</strong><br />
patients with early-stage CLL do need better prognostication<br />
and subsequent enhanced counseling with regard to<br />
follow-up times and to their individual risk <strong>of</strong> progression.<br />
There have been a limited but important set <strong>of</strong> published<br />
work on the development <strong>of</strong> prognostic models for important<br />
clinical end points such as time to first treatment, clinical<br />
levels <strong>of</strong> responses, and duration <strong>of</strong> responses. These models<br />
are important to develop and validate because they may be<br />
very useful for patient clinical trial stratification and for<br />
providing patient-comparison platforms across clinical trials<br />
(Table 2).<br />
Table 2. Prognostic Models<br />
Factors in Model Risk Groups Defined <strong>Clinical</strong> Application Focus Reference<br />
* Age Low a) To predict survival (5 and 10 yr) Ref 10<br />
* Beta-2 microglobulin Intermediate b) Useful for all Rai stages<br />
* Absolute lymphocyte count High<br />
* Sex<br />
* Rai stage<br />
* Number <strong>of</strong> involved lymph node groups<br />
* Lymphocyte doubling time Multivariable models developed for early stage Useful for predicting TTFT and OS Ref 30<br />
* IGHV Mutation Status<br />
*CD38<br />
* Age at Diagnosis<br />
* Treatment Regimen Nomograms Developed a) Predict response to initial therapy Ref 29<br />
* Beta-2 Microglobulin b) Useful for prediction <strong>of</strong> CR, TTFT, and OS<br />
* Age<br />
Abbreviations: CR, complete response; OS, overall survival, TTFT, time to first therapy.<br />
396<br />
NEIL E. KAY<br />
A pivotal and pioneering study <strong>of</strong> clinical outcome was<br />
conducted in an initial large series <strong>of</strong> patients with CLL<br />
patients. This analysis identified six factors including age,<br />
stage, sex, ALC, beta-2 microglobulin, and number <strong>of</strong> lymph<br />
node regions involved that were independently associated<br />
with patient survival. 10 From this set <strong>of</strong> data these various<br />
factors were then combined in a prognostic index that was<br />
able to predict OS more accurately than clinical stage alone.<br />
Further work validating the prognostic index was subsequently<br />
published by two groups indicating that this is a<br />
valuable and practical model. 25,26 In one <strong>of</strong> these studies it<br />
was extended to show that the index remained useful even<br />
when applied just to Rai stage 0 for the prediction <strong>of</strong> time to<br />
treatment and <strong>of</strong> OS. 25<br />
Subsequent prognostic models have been developed that<br />
also incorporated the biologic factors shown to reflect the<br />
pathobiology <strong>of</strong> the CLL B cell. The initial work on this<br />
included an analysis <strong>of</strong> four novel markers: ZAP-70, CD38,<br />
iFISH, and IgVH mutation status for predicting time to<br />
treatment in a study <strong>of</strong> more than 1,000 patients with<br />
CLL. 19 The analysis found three groups: low-, intermediate-,<br />
and high-risk groups based on ZAP-70 and IgVH mutation<br />
status. In summary ZAP-70–positive patients were high<br />
risk irrespective <strong>of</strong> IgVH mutation status; patients with CLL<br />
who are ZAP-70 negative can be classified as low (mutated<br />
IgVH) or intermediate (unmutated IgVH) risk based on<br />
IgVH status. However widespread application <strong>of</strong> this model<br />
is difficult because <strong>of</strong> the lack <strong>of</strong> standardization <strong>of</strong> ZAP-70<br />
measurements.<br />
Further work on the ability <strong>of</strong> iFISH detectable del(17p13)<br />
and IgVH mutation status in a cohort <strong>of</strong> 99 nontreated<br />
early-stage patients found surprisingly better time to treatment<br />
and survival in some. 27 However if Rai stage was 1 or<br />
greater and/or the patient also had an unmutated IgVH<br />
status, the favorable clinical course dropped <strong>of</strong>f sharply.<br />
A European study found that Binet A patients could be<br />
categorized into low-, intermediate-, and high-risk groups<br />
based on the combination <strong>of</strong> iFISH and IgVH mutation<br />
status. 15 Here patients with del(17p13) were high risk<br />
regardless <strong>of</strong> mutation status, whereas patients with<br />
del(11q22) and/or unmutated IgVH genes were intermediate<br />
risk. Finally patients with mutated IgVH without del(17p13)<br />
or del(11q22) were low risk.<br />
Another recent study analyzed 930 patients with CLL for<br />
time to first treatment where both traditional and new<br />
prognostic factors were measured. 28 For this study the
PREDICTIVE PARAMETERS IN CLL<br />
patients did not have active CLL requiring initiation <strong>of</strong><br />
treatment within 3 months <strong>of</strong> first visit and were observed<br />
for time to first treatment. The results from this study<br />
showed a mix <strong>of</strong> both types <strong>of</strong> prognostic parameters were<br />
found to be independently associated with a shorter time to<br />
first therapy. These parameters included: three involved<br />
lymph node sites, increased size <strong>of</strong> cervical lymph nodes,<br />
presence <strong>of</strong> del(17p13) or del(11q22) increased serum lactate<br />
dehydrogenase, and unmutated IgVH mutation status.<br />
From a subset <strong>of</strong> patients a multivariable model was constructed<br />
and a nomogram was developed using the latter<br />
prognostics to predict the risk <strong>of</strong> time to first treatment. The<br />
authors make the point that this nomogram model system<br />
was constructed from only patients with early-stage CLL<br />
and where they had very prolonged clinical follow-up using<br />
strict criteria to decide on treatment initiation.<br />
The same group <strong>of</strong> investigators has also attempted to<br />
develop prognostic features to better predict response to first<br />
therapy and subsequent clinical course. Rationale for this<br />
is that responses to upfront therapy can be very heterogeneous.<br />
Knowledge <strong>of</strong> critical patient features that are<br />
strongly associated with clinical courses post–therapy will<br />
again aid in counseling and even in performing relevant<br />
clinical trials. For 595 patients who underwent upfront<br />
therapy, researchers looked for predictors <strong>of</strong> three aspects:<br />
complete response, time to treatment failure, and OS. 29 In<br />
patients who achieved a complete response there was a more<br />
favorable durability and survival, but having received combination<br />
chemotherapy with antibody regimen was very<br />
significant for all three clinical outcomes. The use <strong>of</strong> various<br />
clinical parameters (i.e., age, serum beta-2 microglobulin)<br />
generated two nomograms that could be used to predict 5and<br />
10-year OS.<br />
Author’s Disclosure <strong>of</strong> Potential Conflicts <strong>of</strong> Interest<br />
Author<br />
Employment or<br />
Leadership<br />
Positions<br />
Consultant or<br />
Advisory Role<br />
Finally there was a very recent and quite large study <strong>of</strong><br />
1,154 patients with Binet stage A CLL recently reported<br />
that studied both traditional and novel prognostic factors in<br />
association with clinical outcome. 30 This study found that<br />
LDT was most significantly associated with time to first<br />
therapy but that IgVH was strongest in association with OS.<br />
In addition only LDT, mutation status, CD38, and age at<br />
diagnosis were independent prognostic variables for time to<br />
first therapy and OS. Their recommendation was to assess<br />
IgVH mutation status and CD38 expression at initial evaluation<br />
as they have independent prognostic value in earlystage<br />
CLL.<br />
Conclusion<br />
The use <strong>of</strong> both traditional and novel prognostic parameters<br />
can be a very valuable ally in determining the relative<br />
risk <strong>of</strong> the individual patient’s clinical course <strong>of</strong> CLL. There<br />
are multiple parameters other than individual prognostic<br />
factors that can be used at initial prognostic evaluation and<br />
can inform the practitioner about disease progression risk<br />
and time to therapy. However, the routine use <strong>of</strong> these<br />
parameters, either in single use or in models, is not absolutely<br />
mandated in the care <strong>of</strong> the patients with CLL. In<br />
addition, prognostic parameters should include the recognition<br />
<strong>of</strong> other influences on the course <strong>of</strong> all patients with<br />
CLL, such as advanced age and poor biologic fitness (defined<br />
as impaired physical fitness and organ function), which<br />
considerably increase the risk for adverse consequences <strong>of</strong><br />
progressive disease and contribute to the decreased survival<br />
for patients with CLL compared with the age-matched<br />
population.<br />
Stock<br />
Ownership Honoraria<br />
Research<br />
Funding<br />
Neil E. Kay Genentech;<br />
Hospira<br />
1. Jemal A, Siegel R, Ward E, et al. Cancer statistics, 2008. CA Cancer<br />
J Clin. 2008;58(2):71-96. Epub 2008 Feb 20.<br />
2. Dighiero G, Maloum K, Desablens B, et al. Chlorambucil in indolent<br />
chronic lymphocytic leukemia. French Cooperative Group on Chronic Lymphocytic<br />
Leukemia. N Engl J Med. 1998;338(21):1506-1514.<br />
3. Shanafelt TD, Bowen D, Venkat C, et al. Quality <strong>of</strong> life in chronic<br />
lymphocytic leukemia: An international survey <strong>of</strong> 1482 patients. Br J Haematol.<br />
2007;139(2):255-264.<br />
4. Shanafelt TD, Bowen DA, Venkat C, et al. The physician-patient<br />
relationship and quality <strong>of</strong> life: Lessons from chronic lymphocytic leukemia.<br />
Leuk Res. 2009;33(2):263-270. Epub 2008 Jul 25.<br />
5. Rozman C, Bosch F, Montserrat E. Chronic lymphocytic leukemia: A<br />
changing natural history? Leukemia. 1997;11(6):775-778.<br />
6. Zent CS, Kay NE. Management <strong>of</strong> patients with chronic lymphocytic<br />
leukemia with a high risk <strong>of</strong> adverse outcome: The Mayo Clinic approach.<br />
Leuk Lymphoma. 2011;52(8):1425-1434. Epub 2011 Jun 8.<br />
7. Shanafelt TD, Rabe KG, Kay NE, et al. Age at diagnosis and the utility<br />
<strong>of</strong> prognostic testing in patients with chronic lymphocytic leukemia. Cancer.<br />
2010;116(20):4777-4787.<br />
8. Molica S, Alberti A. Prognostic value <strong>of</strong> the lymphocyte doubling time in<br />
chronic lymphocytic leukemia. Cancer. 1987;60(11):2712-2716.<br />
9. Bergmann MA, Eichhorst BF, Busch R, et al. Prospective Evaluation <strong>of</strong><br />
Prognostic Parameters in Early Stage Chronic Lymphocytic Leukemia (CLL):<br />
REFERENCES<br />
Expert<br />
Testimony<br />
Other<br />
Remuneration<br />
Results <strong>of</strong> the CLL1-Protocol <strong>of</strong> the German CLL Study Group (GCLLSG).<br />
ASH Annual Meeting Abstracts. 2007;110(11):625.<br />
10. Wierda WG, O’Brien S, Wang X, et al. Prognostic nomogram and index<br />
for overall survival in previously untreated patients with chronic lymphocytic<br />
leukemia. Blood. 2007;109(11):4679-4685. Epub 2007 Feb 13.<br />
11. Damle RN, Wasil T, Fais F, et al. Ig V gene mutation status and CD38<br />
expression as novel prognostic indicators in chronic lymphocytic leukemia.<br />
Blood. 1999;94(6):1840-1847.<br />
12. Hamblin TJ, Davis Z, Gardiner A, et al. Unmutated Ig V(H) genes are<br />
associated with a more aggressive form <strong>of</strong> chronic lymphocytic leukemia.<br />
Blood. 1999;94(6):1848-1854.<br />
13. Hamblin TJ, Orchard JA, Ibbotson RE, et al. CD38 expression and<br />
immunoglobulin variable region mutations are independent prognostic variables<br />
in chronic lymphocytic leukemia, but CD38 expression may vary during<br />
the course <strong>of</strong> the disease. Blood. 2002;99(3):1023-1029.<br />
14. Dohner H, Stilgenbauer S, Benner A, et al. Genomic aberrations and<br />
survival in chronic lymphocytic leukemia. N Engl J Med. 2000;343(26):1910-<br />
1916. Epub 2008 Mar 26.<br />
15. Krober A, Seiler T, Benner A, et al. V(H) mutation status, CD38<br />
expression level, genomic aberrations, and survival in chronic lymphocytic<br />
leukemia. Blood. 2002;100(4):1410-1416.<br />
16. Crespo M, Bosch F, Villamor N, et al. ZAP-70 expression as a surrogate<br />
for immunoglobulin-variable-region mutations in chronic lymphocytic leukemia.<br />
N Engl J Med. 2003;348(18):1764-1775.<br />
397
17. Orchard JA, Ibbotson RE, Davis Z, et al. ZAP-70 expression and<br />
prognosis in chronic lymphocytic leukaemia. Lancet. 2004;363(9403):105-111.<br />
18. Rassenti LZ, Huynh L, Toy TL, et al. ZAP-70 compared with immunoglobulin<br />
heavy-chain gene mutation status as a predictor <strong>of</strong> disease progression<br />
in chronic lymphocytic leukemia. N Engl J Med. 2004;351(9):893-901.<br />
19. Rassenti LZ, Jain S, Keating MJ, et al. Relative value <strong>of</strong> ZAP-70, CD38,<br />
and immunoglobulin mutation status in predicting aggressive disease in<br />
chronic lymphocytic leukemia. Blood. 2008;112(5):1923-1930. Epub 2008 Jun<br />
24.<br />
20. Del Principe MI, Del Poeta G, Buccisano F, et al. <strong>Clinical</strong> significance <strong>of</strong><br />
ZAP-70 protein expression in B-cell chronic lymphocytic leukemia. Blood.<br />
2006;108(3):853-861. Epub 2006 Apr 6.<br />
21. Shanafelt TD, Witzig TE, Fink SR, et al. Prospective evaluation <strong>of</strong><br />
clonal evolution during long-term follow-up <strong>of</strong> patients with untreated earlystage<br />
chronic lymphocytic leukemia. J Clin Oncol. 2006;24(28):4634-4641.<br />
22. Dal Bo M, Rossi FM, Rossi D, et al. 13q14 deletion size and number <strong>of</strong><br />
deleted cells both influence prognosis in chronic lymphocytic leukemia. Genes<br />
Chromosomes Cancer. 2011;50(8):633-643. Epub 2011 May 11.<br />
23. Zenz T, Krober A, Scherer K, et al. Monoallelic TP53 inactivation is<br />
associated with poor prognosis in chronic lymphocytic leukemia: Results from<br />
a detailed genetic characterization with long-term follow-up. Blood. 2008;<br />
112(8):3322-3329. Epub 2008 Aug 8.<br />
24. Gonzalez D, Martinez P, Wade R, et al. Mutational Status <strong>of</strong> the TP53<br />
398<br />
NEIL E. KAY<br />
Gene as a Predictor <strong>of</strong> Response and Survival in CLL Patients with and<br />
without 17p Deletion. ASH Annual Meeting Abstracts. 2008;112(11):784.<br />
25. Shanafelt TD, Jenkins G, Call TG, et al. Validation <strong>of</strong> a new prognostic<br />
index for patients with chronic lymphocytic leukemia. Cancer. 2009;115(2):<br />
363-372.<br />
26. Gonzalez Rodriguez AP, Gonzalez Garcia E, Fernandez Alvarez C, et al.<br />
B-chronic lymphocytic leukemia: Epidemiological study and comparison <strong>of</strong><br />
MDACC and GIMENA pronostic indexes. Med Clin (Barcelona). 2009;133(5):<br />
161-166. Epub 2009 Jun 18.<br />
27. Tam CS, Shanafelt TD, Wierda WG, et al. De novo deletion 17p13.1<br />
chronic lymphocytic leukemia shows significant clinical heterogeneity:<br />
The M. D. Anderson and Mayo Clinic experience. Blood. 2009;114(5):957-964.<br />
Epub 2009 May 4.<br />
28. Wierda WG, O’Brien S, Wang X, et al. Multivariable model for time to<br />
first treatment in patients with chronic lymphocytic leukemia. J Clin Oncol.<br />
2011;29(31):4088-4095. Epub 2011 Oct 3.<br />
29. Wierda WG, O’Brien S, Wang X, et al. Characteristics associated with<br />
important clinical end points in patients with chronic lymphocytic leukemia<br />
at initial treatment. J Clin Oncol. 2009;27(10):1637-1643. Epub 2009 Feb 17.<br />
30. Pepper C, Majid A, Lin TT, et al. Defining the prognosis <strong>of</strong> early stage<br />
chronic lymphocytic leukaemia patients. Br J Haematol. <strong>2012</strong>;156(4):499-507.<br />
Epub 2011 Dec 15.
Transplant in Chronic Lymphocytic<br />
Leukemia: To Do It or Not and If So, When<br />
and How?<br />
Overview: Most patients with chronic lymphocytic leukemia<br />
(CLL) have an indolent clinical course, but the disease remains<br />
incurable with standard therapy and the prognosis is dismal<br />
for those patients with disease refractory to available treatment<br />
options. The only potentially curative treatment is allogeneic<br />
hematopoietic stem cell transplantation (SCT), but<br />
since CLL is a disease <strong>of</strong> elderly patients, few patients are<br />
candidates for myeloablative allogeneic SCT. Although autologous<br />
SCT is feasible and has low treatment-related mortality,<br />
it is not curative. The widespread adoption <strong>of</strong> reducedintensity<br />
conditioning (RIC) allogeneic SCT has made this<br />
approach applicable to the elderly patient population with<br />
CLL. This approach relies on the documented graft-versus-<br />
SCT IS NOT a suitable treatment option for the majority<br />
<strong>of</strong> patients with CLL in whom the disease follows an<br />
indolent course, and many patients never require therapy.<br />
The outcome has improved dramatically over the past decade<br />
for patients whose disease progresses to require treatment.<br />
1 Most patients with CLL are elderly and not<br />
sufficiently fit for SCT. However, it is possible to identify<br />
suitable candidates for SCT using a number <strong>of</strong> clinical and<br />
biologic features. 2 The role <strong>of</strong> SCT in a number <strong>of</strong> other<br />
hematologic malignancies has been established in prospective<br />
studies, but no studies in CLL have compared the<br />
outcome after standard chemotherapy with allogeneic SCT.<br />
The biggest challenges remain the decision <strong>of</strong> which patients<br />
are eligible for SCT and when in their disease course SCT<br />
should occur.<br />
Patient Selection for SCT<br />
CLL is an extremely heterogeneous disease, with the<br />
clinical course varying from patients who never require<br />
therapy to patients with rapidly progressive and fatal malignancy.<br />
Treatment guidelines state that therapy should be<br />
reserved for those with advanced, symptomatic, or progressive<br />
disease. 3 For those patients whose disease requires<br />
therapy, the results <strong>of</strong> randomized clinical trials have demonstrated<br />
significant improvement over the past decade<br />
with the use <strong>of</strong> combination chemotherapy and now chemoimmunotherapy.<br />
4-9 When assessing the potential role <strong>of</strong><br />
transplantation, these approaches must be considered in<br />
addition to the many exciting novel agents currently in<br />
clinical trials, which may alter our approach as to when and<br />
to whom transplant should be <strong>of</strong>fered.<br />
Major advances have been made in our understanding<br />
<strong>of</strong> CLL pathophysiology, which has led to the emergence <strong>of</strong><br />
a large number <strong>of</strong> prognostic biomarkers cytogenetics, immunoglobulin<br />
heavy chain (IgV H) gene mutational status,<br />
zeta-associated protein 70 (ZAP70) expression, and CD38<br />
expression. 10-13 The emergence <strong>of</strong> prognostic biomarkers<br />
has implications for the selection for which patients might<br />
merit SCT, but it is not yet fully clear how we should use<br />
these factors in CLL management. 14<br />
By John G. Gribben, MD, DSc<br />
leukemia (GVL) effect and is strong in CLL. Steps to further<br />
decrease the morbidity and mortality <strong>of</strong> the RIC SCT and in<br />
particular to reduce the incidence <strong>of</strong> chronic extensive graftversus-host<br />
disease (GVHD) remain a major focus. Many<br />
potential treatments are available for CLL, and appropriate<br />
patient selection and SCT timing remain controversial and the<br />
focus <strong>of</strong> ongoing clinical trials. The use <strong>of</strong> SCT must always<br />
be weighed against the risk <strong>of</strong> the underlying disease, particularly<br />
in a setting where improvements in treatment are leading<br />
to improved outcome. The major challenge remains how<br />
to identify which patients with CLL merit this approach and<br />
where in the treatment course this treatment can be applied<br />
optimally.<br />
Autologous SCT<br />
The role <strong>of</strong> autologous SCT in CLL remains highly controversial<br />
and there is currently no role for autologous SCT for<br />
CLL except in the setting <strong>of</strong> a clinical trial. A number <strong>of</strong><br />
phase II studies have reported outcome following autologous<br />
SCT for CLL, demonstrating that this approach is feasible<br />
with a transplant-related mortality (TRM) <strong>of</strong> 1% to 10%,<br />
with most toxicity occurring late. 15-17 Such studies need to<br />
be considered in terms <strong>of</strong> intention to treat, and this can be<br />
difficult to determine in transplant centers where only<br />
responding might be referred. In a pilot study to assess the<br />
feasibility <strong>of</strong> performing autologous SCT, 115 previously<br />
untreated patients with CLL prospectively enrolled, and<br />
only 65 (56%) proceeded to transplant. 16 TRM was low,<br />
complete remission (CR) rate was 74%, the 5-year estimated<br />
overall survival (OS) was 77.5%, and progression-free survival<br />
(PFS) was 51.5%. Of concern, 8% <strong>of</strong> patients developed<br />
post-transplant acute myeloid leukemia/myelodysplastic<br />
syndrome, a complication also seen in other series. 15 In a<br />
single-center study, among 137 patients who underwent<br />
autologous transplantation, the one-year TRM was 4% but<br />
rose to 10% when late events were taken into account. At the<br />
median follow-up time <strong>of</strong> 6.5 years, OS was 58% after<br />
autologous SCT. There was no TRM among 72 patients who<br />
underwent autologous SCT in five Finnish centers. 17 Initial<br />
enthusiasm for autologous SCT has been tempered since the<br />
observed results demonstrated no plateau in either eventfree<br />
survival (EFS) or OS and because <strong>of</strong> concerns regarding<br />
the risk <strong>of</strong> secondary malignancies. 18<br />
A retrospective matched-pair analysis was performed including<br />
66 patients who had undergone a uniform high-dose<br />
therapy and autologous SCT with a database <strong>of</strong> 291 patients<br />
treated conventionally and suggested a survival advantage<br />
From the Barts Cancer Institute, Queen Mary University <strong>of</strong> London, Charterhouse<br />
Square, London.<br />
Author’s disclosure <strong>of</strong> potential conflicts <strong>of</strong> interest are found at the end <strong>of</strong> this article.<br />
Address reprint requests John G. Gribben, MD, DSc, Barts Cancer Institute, Queen Mary<br />
University <strong>of</strong> London, Charterhouse Square, London EC1M 6BQ, United Kingdom; email:<br />
j.gribben@qmul.ac.uk.<br />
© <strong>2012</strong> by <strong>American</strong> <strong>Society</strong> <strong>of</strong> <strong>Clinical</strong> <strong>Oncology</strong>.<br />
1092-9118/10/1-10<br />
399
for autologous SCT over conventional therapy. 19 Survival<br />
was significantly longer for patients who had undergone<br />
autologous SCT compared with patients conventionally<br />
treated.<br />
Results have now been reported for three phase III randomized<br />
trials examining the role <strong>of</strong> autologous transplant<br />
in CLL. 20-22 The European Intergroup Study randomly<br />
selected 223 patients, 83% after first-line therapy and 17%<br />
after second-line therapy. Only 59% <strong>of</strong> patients were in CR.<br />
Patients were randomly selected to receive autologous SCT<br />
(112 patients) and observation (111 patients). Autologous<br />
SCT significantly improved EFS from 24.4 months in the<br />
observation group to 51.2 months in the patients receiving<br />
autologous SCT, but there was no difference in OS at 5<br />
years. 20 In a second study, 241 patients received three<br />
courses <strong>of</strong> cyclophosphamide, hydroxydaunorubicin, vincristine,<br />
and prednisone/prednisolone (mini-CHOP), and then<br />
three courses <strong>of</strong> fludarabine. Patients in CR were then<br />
randomly selected to receive autologous SCT or observation,<br />
whereas patients not in CR were randomly selected to<br />
receive salvage therapy followed by either autologous SCT<br />
or three courses <strong>of</strong> fludarabine plus cyclophosphamide (FC).<br />
The patients achieving CR were also included for analysis in<br />
the European Intergroup Study. Autologous SCT improved<br />
EFS in patients achieving CR to initial therapy, but no<br />
differences were observed among patients who required<br />
salvage therapy. No difference in OS was found in either<br />
group. 21 The GOELAMS LLC 98 trial compared six monthly<br />
courses <strong>of</strong> cyclophosphamide, doxorubicin hydrochloride,<br />
vincristine, and prednisone (CHOP) followed by six CHOP<br />
courses every 3 months in responding patients with highdose<br />
therapy with autologous SCT used as consolidation in<br />
responding patients after three CHOP courses. A total <strong>of</strong> 86<br />
patients were enrolled, with 43 patients evaluable after<br />
autologous SCT. On an intent-to-treat basis and with a<br />
median follow-up time <strong>of</strong> 77.1 months, median PFS was 22<br />
months after conventional therapy and 53 months after<br />
autologous SCT (p � 0.0001). There was no difference in OS<br />
between the two arms. 25 Concerns regarding this trial<br />
include the relatively small number <strong>of</strong> patients enrolled; the<br />
trial had to be closed early because <strong>of</strong> concerns <strong>of</strong> emerging<br />
better induction therapies, leading to a low statistical power.<br />
The results obtained from these randomized studies all<br />
demonstrate that patients with chemotherapy-sensitive disease<br />
had prolongation in EFS but not OS. The question<br />
KEY POINTS<br />
● Autologous stem cell transplantation (SCT) is not<br />
curative for chronic lymphocytic leukemia (CLL) and<br />
should be performed only in the setting <strong>of</strong> clinical<br />
trials.<br />
● Myeloablative allogeneic SCT is applicable only to a<br />
small population and is associated with unacceptably<br />
high treatment-related morbidity and mortality.<br />
● Reduced-intensity conditioning allogeneic SCT is potentially<br />
curative but complicated by chronic GVHD.<br />
● Identifying patients with CLL who should receive<br />
transplantation and when in their clinical course this<br />
should occur remains a major challenge.<br />
400<br />
Study<br />
Table 1. Outcome after FCR or ASCT<br />
Number <strong>of</strong><br />
Patients<br />
remains if these patients would have had similar results if<br />
they had been <strong>of</strong>fered the type <strong>of</strong> chemoimmunotherapy that<br />
is the standard <strong>of</strong> care today. Whereas great caution must be<br />
taken comparing the outcome <strong>of</strong> separate studies, the outcomes<br />
in terms <strong>of</strong> EFS and OS might be similar for patients<br />
treated with fludarabine, cyclophosphamide, and rituximab<br />
(FCR) and those receiving autologous SCT (Table 1). This<br />
does not address whether improved outcome would be seen<br />
with autologous SCT after FCR. In a phase II study, those<br />
patients who received autologous SCT after front-line FCR<br />
had an inferior outcome to patients who underwent autologous<br />
SCT before chemoimmunotherapy was available. 15<br />
Monoclonal antibodies have been used to increase the<br />
likelihood <strong>of</strong> elimination <strong>of</strong> minimal residual disease (MRD)<br />
after autologous SCT, ex vivo or by in vivo treatment with<br />
alemtuzumab or rituximab. 15 Alemtuzumab was used in the<br />
conditioning regimen for autologous SCT in one arm <strong>of</strong> the<br />
German CLL Study Group CLL3 trial, and 12 <strong>of</strong> 16 patients<br />
(87%) developed a skin rash between 43 and 601 days<br />
post-SCT. In seven patients, biopsy confirmed GVHD that<br />
persisted for a median duration <strong>of</strong> 517 (range, 60 to 867)<br />
days. 23 The trial was discontinued because <strong>of</strong> the TRM, but<br />
addition <strong>of</strong> alemtuzumab led to improved disease control.<br />
When alemtuzumab was used at modified dose (10 mg<br />
subcutaneously three times per week for 6 weeks) in 34<br />
patients who had a clinical response to a fludarabine-based<br />
regimen, the CR rate improved from 35% to 79.5% with 56%<br />
achieving eradication <strong>of</strong> MRD. 24 Peripheral blood stem cell<br />
collection was subsequently successfully performed in 92%.<br />
Most studies reported relatively short follow-up and therefore<br />
focus mostly on TRM early post-SCT, but late consequences—particularly<br />
development <strong>of</strong> secondary<br />
myelodysplasia and acute myeloid leukemia (MDS/AML)—<br />
are <strong>of</strong> concern. Among 65 previously untreated patients who<br />
were treated with fludarabine followed by autologous SCT,<br />
eight developed MDS/AML, with a 5-year actuarial risk <strong>of</strong><br />
12% developing MDS/AML after autologous SCT. 16 Longterm<br />
follow-up reports a high incidence <strong>of</strong> other solid tumors<br />
in 31 (19%) patients. 15<br />
Myeloablative Allogeneic SCT<br />
Median<br />
Age (y)<br />
FCR<br />
M. D. Anderson 8 300 57 51 at 5 y 77 at 6 y<br />
CLL8 9 408 61 65 at 3 y 87 at 3 y<br />
Autologous SCT<br />
European Intergroup 20 112 54 42 at 5 y 86 at 5 y<br />
SFGM-TC/GFLLC 21 52 in CR 56 79.8 at 3 y 96 at 3 y<br />
46 not CR 48.9 at 3 y 82 at 3 y<br />
GOELAMS 22 43 Upper age<br />
limit 60<br />
The major advantage <strong>of</strong> allogeneic SCT is the potential<br />
for a GVL effect. There is strong evidence for a GVL effect in<br />
CLL as demonstrated by a decreased risk <strong>of</strong> relapse in<br />
patients with chronic GVHD, increased risk <strong>of</strong> relapse in<br />
EFS<br />
(%)<br />
PFS 53 at<br />
6.5 y<br />
JOHN G. GRIBBEN<br />
OS<br />
(%)<br />
107.4 at 6.5 y<br />
Abbreviations: FCR, fludarabine, cyclophosphamide, and rituximab; ASCT,<br />
autologous stem cell transplantation; y, years; EFS, event-free survival; OS,<br />
overall survival; SFGM-TC, Societe’ Francaise de Greffe de Moelle et de Therapie<br />
Cellulaire; GFLLC, Groupe Français d’étude de la Leucémie Lymphoïde Chronique;<br />
CR, complete remission; GOELAMS, Groupe Ouest Est d’Etude des<br />
Leucémies et Autres Maladies du Sang.
WHEN TO OFFER TRANSPLANTATION IN CLL<br />
Number <strong>of</strong><br />
Patients<br />
Age<br />
Years<br />
(range)<br />
patients who have undergone T-cell depletion, and clinical<br />
responses to removal <strong>of</strong> immune suppression or to donor<br />
lymphocyte infusion (DLI). 15 Allogeneic SCT has significant<br />
morbidity and TRM, from regimen-related toxicity, GVHD,<br />
and infection, but surviving patients have long-term disease<br />
control. 15,25-27 In registry data, TRM following allogeneic<br />
SCT in patients with CLL was unacceptably high at 46%,<br />
with mortality from GVHD <strong>of</strong> 20%. 28 Currently there is only<br />
a very limited role for myeloablative allogeneic SCT in the<br />
setting <strong>of</strong> very young patients with particularly aggressive<br />
disease. Among 25 patients with CLL who underwent allogeneic<br />
SCT at the Fred Hutchinson Cancer Research Center<br />
(FHCRC), grades 2 through 4 acute GVHD was seen in 14<br />
patients, 10 developed clinically extensive chronic GVHD,<br />
and estimated OS at 5 years was 32%. 29<br />
No randomized studies have compared the outcome <strong>of</strong><br />
autologous SCT with allogeneic SCT. Studies from University<br />
<strong>of</strong> Texas M. D. Anderson Cancer Center (MDACC)<br />
demonstrate improved outcome after allogeneic SCT compared<br />
with autologous SCT, suggesting that allogeneic SCT<br />
can induce durable remission even in patients with refractory<br />
disease. 30 At Dana-Farber Cancer Institute, 162 patients<br />
with high-risk CLL were enrolled in a “biologic<br />
randomization” in which 25 patients with a human leukocyte<br />
antigen (HLA)-matched sibling donor underwent T<br />
cell–depleted myeloablative allogeneic SCT, while 137 with<br />
no HLA-matched sibling donor underwent B cell–purged<br />
autologous SCT, with both groups receiving identical conditioning<br />
regimen using high-dose cyclophosphamide and total<br />
body irradiation. 15 The 100-day TRM was 4% after autologous<br />
or allogeneic SCT, but later TRM had a major effect<br />
on outcome. At the median follow-up <strong>of</strong> 6.5 years, PFS<br />
was significantly longer following autologous than T cell–<br />
depleted allogeneic SCT, but no significant differences were<br />
observed in disease recurrence or deaths without recurrence<br />
by type <strong>of</strong> transplant. There was no difference in OS between<br />
the two groups, and at the median follow-up time <strong>of</strong> 6.5<br />
years, OS was 58% after autologous and 55% after allogeneic<br />
SCT.<br />
RIC SCT for CLL<br />
Prior<br />
Regimens<br />
(range)<br />
Chemorefractory<br />
(%)<br />
Prior<br />
Auto-SCT<br />
A major advance in reducing the short-term morbidity and<br />
mortality <strong>of</strong> allogeneic SCT has been the introduction <strong>of</strong><br />
Table 2. RIC Allogeneic SCT for CLL<br />
Donor<br />
(includes<br />
mismatch) TRM<br />
Acute<br />
Grade 2–4<br />
GVHD<br />
Chronic<br />
Extensive<br />
Survival Reference<br />
82 82 4 87% 4 63% related 25% overall 55% 49% related OS 50% 5 yr Sorror et al 2008 28<br />
(42–72) 37% unrelated 53% unrelated PFS 45%<br />
77 54 3 33% 10 81% related 18% 12 m 34% 58% OS 72% 2 yr Dreger et al 2003 29<br />
(30–66) (0–8) PFS 56%<br />
46 53 5 57% 10 33% related 17% overall 34% 43% OS 54% 2 yr Brown et al 2006 30<br />
(35–67) (1–10) 67% unrelated PFS 34%<br />
41 54 3 27% 11 58% related 5% at 100 d 10% 33%* OS 51 2 yr Delgado et al 2006 31<br />
(37–67) (1–8) 42% unrelated 26% overall (grade 3–4) *after DLI PFS 45%<br />
39 57 3 Not stated 90% related 2% at 100 d 45% 58% OS 48% 4 yr Khouri et al 2006 32<br />
(34–70) (2–8) 10% unrelated PFS 44%<br />
30 50 3 47% 50% related 13% overall 56% 21% OS 72% 2 yr Schetelig et al 2003 33<br />
(12–63) (0–8) 50% unrelated PFS 67%<br />
Abbreviations: RIC, reduced-intensity conditioning; SCT, stem cell transplantation; CLL, chronic lymphocytic leukemia; TRM, transplant-related mortality; GVHD,<br />
graft-versus-host disease; OS, overall survival; PFS, progression-free survival; m, months; d, days.<br />
nonmyeloablative or RIC regimens to allow engraftment <strong>of</strong><br />
allogeneic stem cells. This approach is much more applicable<br />
to the age group with CLL who are potential candidates for<br />
SCT. Most patients reported have been treated on experimental<br />
treatment protocols with enrollment <strong>of</strong> many patients<br />
with chemo-refractory end-stage disease.<br />
RIC regimens allow transplantation in older patients,<br />
making this approach more applicable to increased numbers<br />
<strong>of</strong> patients with CLL and results from the larger reported<br />
studies are shown in Table 2. 29-34 Most patients were<br />
heavily pretreated and refractory to therapy, but despite<br />
these issues, the majority demonstrated donor engraftment<br />
and had a high CR rate. The ability <strong>of</strong> such approaches to<br />
eradicate MRD in patients with advanced CLL and the<br />
observation <strong>of</strong> late remissions in patients treated with low<br />
doses <strong>of</strong> chemotherapy provide direct evidence for a powerful<br />
GVL in CLL. 35 The outcome from the FHCRC multiinstitutional<br />
protocol after RIC allogeneic SCT was reported<br />
for 82 patients with advanced fludarabine-refractory CLL<br />
using related (52 patients) or unrelated donors (30 patients)<br />
median age 56 (range, 42 to 72). 29 TRM was 23% at 5 years,<br />
with significant GVHD a remaining problem. Five-year OS<br />
was 50% and EFS was 39%. Among 46 patients who underwent<br />
RIC transplantation at Dana-Farber Cancer Institute—67%<br />
using unrelated donors—factors associated with<br />
increased risk <strong>of</strong> relapse include low levels <strong>of</strong> donor chimerism<br />
at day 30, chemotherapy-refractory disease, increased<br />
number <strong>of</strong> previous therapies, and adverse cytogenetics. 31<br />
No formal assessment <strong>of</strong> RIC compared with myeloablative<br />
allogeneic SCT has been undertaken, but the outcome<br />
after RIC allogeneic SCT <strong>of</strong> 73 patients who had undergone<br />
RIC was compared with that <strong>of</strong> 82 matched patients who<br />
had undergone standard myeloablative conditioning for CLL<br />
from the European Blood and Marrow Transplantation<br />
(EBMT) registry database during the same time period.<br />
Patients undergoing RIC transplants had significantly reduced<br />
TRM but higher relapse incidence, and there was no<br />
significant difference in OS or PFS between these two<br />
groups. 36 Of particular interest is the report <strong>of</strong> 44 patients<br />
with CLL with deletion <strong>of</strong> 17p and loss <strong>of</strong> p53 in whom<br />
allogeneic SCT has the potential to induce long-term remission<br />
in these very high-risk patients. 37<br />
401
Addition <strong>of</strong> Monoclonal Antibodies to RIC SCT<br />
GVHD remains the major concern after RIC SCT and<br />
attempts have been made to utilize monoclonal antibodies to<br />
reduce the incidence <strong>of</strong> GVHD without increasing the subsequent<br />
risk <strong>of</strong> relapse. Excellent results have been obtained<br />
at MDACC using RIC based on a combination <strong>of</strong> fludarabine<br />
and cyclophosphamide with the addition <strong>of</strong> rituximab, an<br />
approach designed to maximize GVL by early tapering <strong>of</strong><br />
immune suppression with use <strong>of</strong> rituximab and DLI. Among<br />
39 patients treated, median age 57 (range, 34 to 70), median<br />
time from diagnosis to transplantation was 4.5 years. 33 All<br />
patients had recurrent advanced disease, were heavily pretreated<br />
with a median <strong>of</strong> three (range, 2 to 8) chemotherapy<br />
regimens, and had been previously treated with fludarabine/<br />
rituximab-based regimens. At transplant, 34 patients (87%)<br />
had active disease, including nine (23%) with evidence <strong>of</strong><br />
Richter’s transformation. In this series, only four <strong>of</strong> the<br />
donors were unrelated. Fourteen patients required immunomodulation<br />
with rituximab and DLI for persistent disease<br />
after SCT. Only one patient died early, and among the 38<br />
evaluable patients, 27 (71%) achieved CR with a 48% estimated<br />
OS at 4 years and current PFS <strong>of</strong> 44%. Acute grade 2<br />
through 4 GVHD was observed in 45%, but chronic extensive<br />
GVHD was reduced without concomitant increased risk <strong>of</strong><br />
relapse.<br />
GVHD can be decreased using alemtuzumab in the conditioning<br />
regimen to reduce donor lymphocytes, but this is<br />
associated with delayed immune reconstitution, increases<br />
the risk <strong>of</strong> infective complications, and appears to impair<br />
GVL. In 41 consecutive patients with CLL (24 HLA-matched<br />
sibling donors and 17 unrelated volunteer donors, including<br />
4 mismatched) treated with the conditioning regimen alemtuzumab<br />
with fludarabine and melphalan had impressive<br />
antitumor effects with 100% <strong>of</strong> patients with chemotherapysensitive<br />
disease and 86% with chemotherapy-refractory<br />
disease responding. 32 The TRM rate was 26%, OS was 51%,<br />
and relapse risk was 29% at 2 years. GVHD rates were<br />
relatively low with acute GVHD occurring in 17 (41%) and<br />
chronic GVHD in 13 (33%). The unexpectedly high TRM<br />
rate was because <strong>of</strong> a high incidence <strong>of</strong> fungal and viral<br />
infections.<br />
How to Select Who and When to Offer<br />
Allogeneic SCT<br />
All studies <strong>of</strong> SCT have enrolled younger patients with<br />
“high-risk” disease. This term is very loosely defined and it is<br />
difficult to determine precisely the risk factors used in each<br />
<strong>of</strong> the reported studies. EBMT guidelines have been established<br />
outlining indications for SCT in CLL, which conclude<br />
that there is a evidence base for the efficacy <strong>of</strong> allogeneic<br />
SCT in CLL and that this procedure is indicated in patients<br />
with high-risk CLL. 38 Patients at high risk are defined in<br />
Table 3 and include those requiring treatment who have<br />
TP53 abnormalities (who merit allogeneic SCT in first<br />
response), patients who fail to achieve CR, who progress<br />
within 12 months after fludarabine, who relapse within 24<br />
months after having achieved a response with combination<br />
therapy, those who have relapsed after prior autologous<br />
SCT, or those patients who are fludarabine refractory. It<br />
should be noted that the only category that requires assessment<br />
<strong>of</strong> biologic risk is detection <strong>of</strong> TP53 abnormalities.<br />
Ongoing prospective clinical studies will determine the ef-<br />
402<br />
JOHN G. GRIBBEN<br />
Table 3. EBMT Guidelines for Transplantation in CLL 38<br />
Allo-SCT is a reasonable treatment option in poor-risk CLL including:<br />
● Fludarabine resistance—nonresponse or early relapse (� 12 months) after<br />
purine analogue-based therapy<br />
● Relapse � 24 months after purine analogue combinations or auto-SCT (plus<br />
high-risk genetics)<br />
● p53 mutation with treatment indication<br />
Auto-SCT indicated in clinical trial only<br />
Abbreviation: EBMT, European Blood and Marrow Transplantation; Allo-SCT,<br />
allogeneic stem cell transplantation; CLL, chronic lymphocytic leukemia; auto-<br />
SCT, autologous SCT.<br />
fect <strong>of</strong> biomarkers including IgV H mutational status and<br />
other cytogenetic abnormalities in identification <strong>of</strong> patients<br />
at sufficiently high risk to merit use <strong>of</strong> allogeneic SCT in first<br />
CR.<br />
The definition <strong>of</strong> “refractory” CLL is <strong>of</strong> particular importance<br />
and the terms “refractory” CLL and “fludarabinerefractory”<br />
CLL are <strong>of</strong>ten used interchangeably. But with<br />
the large number <strong>of</strong> treatment options, the consideration <strong>of</strong><br />
treatment context is important. The type <strong>of</strong> therapy to which<br />
patients fail to respond and previous therapies received are<br />
<strong>of</strong> major importance. The clinical importance <strong>of</strong> refractory<br />
CLL is based on the fact that these patients have very poor<br />
prognosis (median 1–2 years OS in most studies) despite<br />
various and intense salvage therapy strategies. 2,39-40 Most<br />
trials <strong>of</strong> investigational agents use this definition as an<br />
entry point for early drug development. The most recent<br />
CLL guidelines define refractory CLL as treatment failure<br />
(less than partial remission [PR]) or disease progression<br />
within 6 months <strong>of</strong> the last antileukemic therapy. 3 These<br />
still correspond to the definition coined when CLL treatment<br />
was based on chlorambucil or fludarabine monotherapy. In<br />
current standard practice, patients eligible to consider allogeneic<br />
SCT are most likely to have received combination<br />
chemoimmunotherapy. There is accumulating evidence that<br />
patients who are formally not refractory based on the current<br />
definition but relapse within 3 years after chemoimmunotherapy<br />
also have very poor outcome and are unlikely to<br />
have durable responses to subsequent chemotherapy. 2<br />
The guidelines suggest three groups <strong>of</strong> patients who may<br />
be <strong>of</strong>fered therapy. The first group is those with TP53 loss,<br />
and these are the patients who have the poorest response<br />
and shortest duration <strong>of</strong> remission. Allogeneic SCT is considered<br />
appropriate in first response after initial therapy for<br />
these patients. Those patients who failed to achieve CR to<br />
FCR and/or have very short (� 24 months) response to prior<br />
FCR can be added to this very high-risk group. These<br />
patients are prime candidates for drugs with proven activity<br />
in TP53 deleted/mutant cells, investigational agents in clinical<br />
trials, and then for allogeneic SCT if they respond.<br />
A second scenario <strong>of</strong> “high-risk” CLL is identification <strong>of</strong><br />
subgroups destined to relapse relatively early after standard<br />
treatment and it is in this group that biomarker analysis<br />
will prove useful. Candidates within this group include<br />
patients with high beta-2-microglobulin or thymidine kinase,<br />
unmutated immunoglobulin heavy chain variable region<br />
(IgVH) or 11q deletion. 9,41 These patients are the ones<br />
who gained most by the addition <strong>of</strong> rituximab to FC. 9<br />
Biomarker analysis will also be useful to identify those<br />
patients in whom allogeneic SCT should not be <strong>of</strong>fered (no<br />
11q deletion, no TP53 deletion/mutation, mutated IGHV,<br />
low beta-2-MG, no prior therapy) who have very favorable<br />
outcome despite progressing to indication for treatment.
WHEN TO OFFER TRANSPLANTATION IN CLL<br />
The results <strong>of</strong> phase II studies <strong>of</strong> RIC allogeneic SCT<br />
suggest that whereas patients with refractory disease may<br />
respond to the GVL effect, optimal outcome will be achieved<br />
by consideration <strong>of</strong> transplant for those patients with highrisk<br />
disease before they become truly refractory and those<br />
patients in whom minimal disease state cannot be achieved.<br />
We have demonstrated that CLL cells are inherently immunosuppressive<br />
even to allogeneic donor T cells, and this may<br />
explain why better results are obtained when patients<br />
receive RIC allogeneic SCT with low tumor bulk, which can<br />
only be achieved if patients are <strong>of</strong>fered this approach before<br />
truly refractory disease occurs. 42<br />
Conclusion<br />
SCT has a role to play in selected patients with CLL, with<br />
major focus now on the use <strong>of</strong> RIC allogeneic SCT. Future<br />
Author’s Disclosure <strong>of</strong> Potential Conflicts <strong>of</strong> Interest<br />
Author<br />
Employment or<br />
Leadership<br />
Positions<br />
Consultant or<br />
Advisory Role<br />
approaches to the management <strong>of</strong> this disease must take<br />
into account the balance between the increased morbidity<br />
and mortality <strong>of</strong> SCT in CLL with the curative potential that<br />
these approaches potentially <strong>of</strong>fer, in the setting <strong>of</strong> the<br />
outcome improvements that can now be seen using chemoimmunotherapy.<br />
Although RIC allogeneic SCT results in<br />
high response rates and eradication <strong>of</strong> polymerase chain<br />
reaction–detectable MRD and is potentially curative, the<br />
follow-up <strong>of</strong> most clinical trials remains too short to assess<br />
whether SCT can really cure CLL. In the absence <strong>of</strong> any<br />
other treatment modalities currently capable <strong>of</strong> improving<br />
outcome in this disease, SCT should be considered as a<br />
treatment approach for younger patients with high-risk CLL<br />
early in the course <strong>of</strong> the disease, ideally in the setting <strong>of</strong><br />
well-designed clinical trials assessing the treatment’s effect<br />
on outcome in these patients. Several such trials are underway.<br />
Stock<br />
Ownership Honoraria<br />
John G. Gribben Celgene; Merck Mundipharma;<br />
Roche<br />
1. Gribben JG, O’Brien S. Update on therapy <strong>of</strong> chronic lymphocytic<br />
leukemia. J Clin Oncol. 2011;29:544-550.<br />
2. Stilgenbauer S, Zenz T. Understanding and managing ultra high-risk<br />
chronic lymphocytic leukemia. Hematology Am Soc Hematol Educ Program.<br />
2010;2010:481-488.<br />
3. Hallek M, Cheson BD, Catovsky D, et al. Guidelines for the diagnosis<br />
and treatment <strong>of</strong> chronic lymphocytic leukemia: a report from the International<br />
Workshop on Chronic Lymphocytic Leukemia updating the National<br />
Cancer Institute-Working Group 1996 guidelines. Blood. 2008;111:5446-5456.<br />
4. Rai KR, Peterson BL, Appelbaum FR, et al. Fludarabine compared with<br />
chlorambucil as primary therapy for chronic lymphocytic leukemia. N Engl<br />
J Med. 2000;343:1750-1757.<br />
5. Eichhorst BF, Busch R, Hopfinger G, et al. Fludarabine plus cyclophosphamide<br />
versus fludarabine alone in first-line therapy <strong>of</strong> younger patients<br />
with chronic lymphocytic leukemia. Blood. 2006;107:885-891.<br />
6. Flinn IW, Neuberg DS, Grever MR, et al. Phase III trial <strong>of</strong> fludarabine<br />
plus cyclophosphamide compared with fludarabine for patients with previously<br />
untreated chronic lymphocytic leukemia: US Intergroup Trial E2997.<br />
J Clin Oncol. 2007;25:793-798.<br />
7. Catovsky D, Richards S, Matutes E, et al. Assessment <strong>of</strong> fludarabine<br />
plus cyclophosphamide for patients with chronic lymphocytic leukaemia (the<br />
LRF CLL4 Trial): a randomised controlled trial. Lancet. 2007;370:230-239.<br />
8. Tam CS, O’Brien S, Wierda W, et al. Long-term results <strong>of</strong> the fludarabine,<br />
cyclophosphamide, and rituximab regimen as initial therapy <strong>of</strong> chronic<br />
lymphocytic leukemia. Blood. 2008;112:975-980.<br />
9. Hallek M, Fischer K, Fingerle-Rowson G, et al. Addition <strong>of</strong> rituximab<br />
to fludarabine and cyclophosphamide in patients with chronic lymphocytic<br />
leukaemia: a randomised, open-label, phase 3 trial. Lancet. 2010;376:1164-<br />
1174.<br />
10. Dohner H, Stilgenbauer S, Benner A, et al. Genomic aberrations and<br />
survival in chronic lymphocytic leukemia. N Engl J Med. 2000;343:1910-1916.<br />
11. Damle RN, Wasil T, Fais F, et al. Ig V gene mutation status and CD38<br />
expression as novel prognostic indicators in chronic lymphocytic leukemia.<br />
Blood. 1999;94:1840-1847.<br />
12. Rassenti LZ, Huynh L, Toy TL, et al. ZAP-70 compared with immunoglobulin<br />
heavy-chain gene mutation status as a predictor <strong>of</strong> disease progression<br />
in chronic lymphocytic leukemia. N Engl J Med. 2004;351:893-901.<br />
13. Rassenti LZ, Jain S, Keating MJ, et al. Relative value <strong>of</strong> ZAP-70, CD38,<br />
and immunoglobulin mutation status in predicting aggressive disease in<br />
chronic lymphocytic leukemia. Blood. 2008;112:1923-1930.<br />
14. Gribben JG. Are prognostic factors in CLL overrated? <strong>Oncology</strong> (Williston<br />
Park). 2011;25:703, 706.<br />
15. Gribben JG, Zahrieh D, Stephans K, et al. Autologous and allogeneic<br />
REFERENCES<br />
Research<br />
Funding<br />
Expert<br />
Testimony<br />
Other<br />
Remuneration<br />
stem cell transplantation for poor-risk chronic lymphocytic leukemia. Blood.<br />
2005;106:4389-4396.<br />
16. Milligan DW, Fernandes S, Dasgupta R, et al. Results <strong>of</strong> the MRC pilot<br />
study show autografting for younger patients with chronic lymphocytic<br />
leukemia is safe and achieves a high percentage <strong>of</strong> molecular responses.<br />
Blood. 2005;105:397-404.<br />
17. Jantunen E, Itala M, Siitonen T, et al. Autologous stem cell transplantation<br />
in patients with chronic lymphocytic leukaemia: the Finnish experience.<br />
Bone Marrow Transplant. 2006;37:1093-1098.<br />
18. Montserrat E, Gribben JG. Autografting CLL: the game is over! Blood<br />
2011;117:6057-6058.<br />
19. Dreger P, Stilgenbauer S, Benner A, et al. The prognostic impact <strong>of</strong><br />
autologous stem cell transplantation in patients with chronic lymphocytic<br />
leukemia: a risk-matched analysis based on the VH gene mutational status.<br />
Blood. 2004;103:2850-2858.<br />
20. Michallet M, Dreger P, Sutton L, et al. Autologous hematopoietic stem<br />
cell transplantation in chronic lymphocytic leukemia: Results <strong>of</strong> European<br />
intergroup randomized trial comparing autografting versus observation.<br />
Blood. 2011;117:1516-1521.<br />
21. Sutton L, Chevret S, Tournilhac O, et al. Autologous stem cell transplantation<br />
as a first-line treatment strategy for chronic lymphocytic leukemia:<br />
a multicenter, randomized, controlled trial from the SFGM-TC and<br />
GFLLC. Blood. 2011;117:6109-6119.<br />
22. Brion A, Mahe B, Kolb B, et al. Autologous transplantation in CLL<br />
patients with B and C Binet stages: Final results <strong>of</strong> the prospective randomized<br />
GOELAMS LLC 98 trial. Bone Marrow Transplant. Epub 2011 Jul 4.<br />
23. Zenz T, Ritgen M, Dreger P, et al. Autologous graft-versus-host diseaselike<br />
syndrome after an alemtuzumab-containing conditioning regimen and<br />
autologous stem cell transplantation for chronic lymphocytic leukemia. Blood.<br />
2006;108:2127-2130.<br />
24. Montillo M, Tedeschi A, Miqueleiz S, et al. Alemtuzumab as consolidation<br />
after a response to fludarabine is effective in purging residual disease in<br />
patients with chronic lymphocytic leukemia. J Clin Oncol. 2006;24:2337-2342.<br />
25. Michallet M, Archimbaud E, Bandini G, et al. HLA-identical sibling<br />
bone marrow transplantation in younger patients with chronic lymphocytic<br />
leukemia. European Group for Blood and Marrow Transplantation and the<br />
International Bone Marrow Transplant Registry. Ann Internal Med. 1996;<br />
124:311-315.<br />
26. Khouri I, Champlin R. Allogenic bone marrow transplantation in<br />
chronic lymphocytic leukemia. Ann Intern Med. 1996;125:780-787.<br />
27. Doney KC, Chauncey T, Appelbaum FR. Allogeneic related donor<br />
hematopoietic stem cell transplantation for treatment <strong>of</strong> chronic lymphocytic<br />
leukemia. Bone Marrow Transplant. 2002;29:817-823.<br />
403
28. Khouri IF, Przepiorka D, van Besien K, et al. Allogeneic blood or<br />
marrow transplantation for chronic lymphocytic leukaemia: timing <strong>of</strong> transplantation<br />
and potential effect <strong>of</strong> fludarabine on acute graft-versus-host<br />
disease. Br J Haematol. 1997;97:466-473.<br />
29. Sorror ML, Storer BE, Sandmaier BM, et al. Five-year follow-up <strong>of</strong><br />
patients with advanced chronic lymphocytic leukemia treated with allogeneic<br />
hematopoietic cell transplantation after nonmyeloablative conditioning.<br />
J Clin Oncol. 2008;26:4912-4920.<br />
30. Dreger P, Brand R, Hansz J, et al. Treatment-related mortality and<br />
graft-versus-leukemia activity after allogeneic stem cell transplantation for<br />
chronic lymphocytic leukemia using intensity-reduced conditioning. Leukemia.<br />
2003;17:841-848.<br />
31. Brown JR, Kim HT, Li S, et al. Predictors <strong>of</strong> improved progression-free<br />
survival after nonmyeloablative allogeneic stem cell transplantation for<br />
advanced chronic lymphocytic leukemia. Biol Blood Marrow Transplant.<br />
2006;12:1056-1064.<br />
32. Delgado J, Thomson K, Russell N, et al. Results <strong>of</strong> alemtuzumab-based<br />
reduced-intensity allogeneic transplantation for chronic lymphocytic leukemia:<br />
a British <strong>Society</strong> <strong>of</strong> Blood and Marrow Transplantation Study. Blood.<br />
2006;107:1724-1730.<br />
33. Khouri IF. Reduced-intensity regimens in allogeneic stem-cell transplantation<br />
for non-hodgkin lymphoma and chronic lymphocytic leukemia.<br />
Hematology Am Soc Hematol Educ Program. 2006:390-397.<br />
34. Schetelig J, Thiede C, Bornhauser M, et al. Evidence <strong>of</strong> a graft-versusleukemia<br />
effect in chronic lymphocytic leukemia after reduced-intensity<br />
404<br />
JOHN G. GRIBBEN<br />
conditioning and allogeneic stem-cell transplantation: the Cooperative German<br />
Transplant Study Group. J Clin Oncol. 2003;21:2747-2753.<br />
35. Moreno C, Villamor N, Colomer D, et al. <strong>Clinical</strong> significance <strong>of</strong><br />
minimal residual disease, as assessed by different techniques, after stem cell<br />
transplantation for chronic lymphocytic leukemia. Blood. 2006;107:4563-4569.<br />
36. Dreger P, Brand R, Milligan D, et al. Reduced-intensity conditioning<br />
lowers treatment-related mortality <strong>of</strong> allogeneic stem cell transplantation<br />
for chronic lymphocytic leukemia: a population-matched analysis. Leukemia.<br />
2005;19:1029-1033.<br />
37. Schetelig J, van Biezen A, Brand R, et al. Allogeneic hematopoietic<br />
stem-cell transplantation for chronic lymphocytic leukemia with 17p deletion:<br />
a retrospective European Group for Blood and Marrow Transplantation<br />
analysis. J Clin Oncol. 2008;26:5094-5100.<br />
38. Dreger P, Corradini P, Kimby E, et al. Indications for allogeneic stem<br />
cell transplantation in chronic lymphocytic leukemia: the EBMT transplant<br />
consensus. Leukemia. 2007;21:12-17.<br />
39. Montserrat E, Moreno C, Esteve J, et al. How I treat refractory CLL.<br />
Blood. 2006;107:1276-1283.<br />
40. Tsimberidou AM, Keating MJ. Treatment <strong>of</strong> fludarabine-refractory<br />
chronic lymphocytic leukemia. Cancer. 2009;115:2824-2836.<br />
41. Oscier D, Wade R, Davis Z, et al. Prognostic factors identified three risk<br />
groups in the LRF CLL4 trial, independent <strong>of</strong> treatment allocation. Haematologica.<br />
2010;95:1705-1712.<br />
42. Ramsay AG, Johnson AJ, Lee AM, et al. Chronic lymphocytic leukemia<br />
T cells show impaired immunological synapse formation that can be reversed<br />
with an immunomodulating drug. J Clin Invest. 2008;118:2427-2437.
NEW DEVELOPMENTS IN MYELOPROLIFERATIVE<br />
NEOPLASMS<br />
CHAIR<br />
Srdan Verstovsek, MD, PhD<br />
University <strong>of</strong> Texas M. D. Anderson Cancer Center<br />
Houston, TX<br />
SPEAKERS<br />
Jason Gotlib, MD, MS<br />
Stanford University School <strong>of</strong> Medicine<br />
Stanford, CA<br />
Francesco Passamonti, MD<br />
Fondazione IRCCS Policlinico San Matteo<br />
Pavia, Italy
Therapeutic Advances in Myeloproliferative<br />
Neoplasms: The Role <strong>of</strong> New-Small<br />
Molecule Inhibitors<br />
Overview: The discovery that a somatic point mutation<br />
(JAK2V617F) in the Janus kinase 2 (JAK2) is highly prevalent in<br />
patients with myeloproliferative neoplasms (MPNs) has been a<br />
crucial breakthrough in our understanding <strong>of</strong> the underlying<br />
molecular mechanisms <strong>of</strong> these diseases. Therefore, preclinical<br />
and clinical research in recent years has focused intensely<br />
on the development <strong>of</strong> new therapies targeted to JAK2. These<br />
efforts culminated in recent approval <strong>of</strong> ruxolitinib as the first<br />
<strong>of</strong>ficial therapy for patients with intermediate- or high-risk<br />
myel<strong>of</strong>ibrosis (MF). Therapy with JAK2 inhibitors substantially<br />
improves quality <strong>of</strong> life and reduces organomegaly in MF with<br />
or without JAKV617F mutation. Recent results suggest that<br />
patients with advanced MF may live longer when receiving<br />
PRIMARY MYELOFIBROSIS (PMF) is a Philadelphia<br />
chromosome (Ph)-negative myeloproliferative neoplasm<br />
(MPN) characterized by diffuse bone marrow fibrosis and<br />
osteosclerosis leading to bone marrow failure, extramedullary<br />
hematopoiesis, massive splenomegaly, very poor quality<br />
<strong>of</strong> life as a result <strong>of</strong> debilitating MF-related systemic symptoms,<br />
weight loss, decrease in performance status, and an<br />
increased risk <strong>of</strong> transformation to acute myeloid leukemia.<br />
1,2 Among the classic Ph-negative MPNs (the other two<br />
are polycythemia vera [PV] and essential thrombocythemia<br />
[ET]) PMF has the worse outcome, with a median survival <strong>of</strong><br />
approximately 5 to 7 years. 1,2 MF can also develop secondary<br />
to disease transformation from PV and ET (then called<br />
post-PV or -ET MF). The clinical course and outcome <strong>of</strong><br />
patients with primary and secondary MF appears to be<br />
similar. MF presents a great burden to patients and a<br />
challenge for their treating physicians, and is a focus <strong>of</strong><br />
intense clinical research to develop new effective therapies.<br />
1,2 Until recently, no medication has been approved as<br />
therapy for MF, and our efforts in helping patients battling<br />
the disease have been largely palliative. 1,2<br />
Starting with the discovery <strong>of</strong> the JAKV617F mutation in<br />
approximately 50% <strong>of</strong> patients with MF in 2005, 3 research to<br />
find new treatment options for MF has expanded greatly<br />
over recent years to include not just Janus kinase (JAK) 2<br />
tyrosine kinase but also several other promising molecular<br />
targets. 2,4 These efforts led to a recent approval <strong>of</strong> ruxolitinib,<br />
an oral inhibitor <strong>of</strong> JAK1 and JAK2 tyrosine kinases,<br />
as therapy for patients with intermediate- and highrisk<br />
MF. 5 Ruxolitinib therapy leads to a reduction in<br />
splenomegaly, improvement in systemic MF-related symptoms<br />
and signs, and improved quality <strong>of</strong> life in patients with<br />
advanced MF. 6 Recent evidence suggests that ruxolitinib<br />
may also prolong the life <strong>of</strong> patients with advanced MF. 7<br />
<strong>Clinical</strong> benefits are seen in patients with and without<br />
JAK2V617F mutation (most common among many known<br />
mutations in MPN), which reflects emerging notion that<br />
dysregulated JAK/signal transducers and activators <strong>of</strong><br />
transcription (STAT) pathway is a common pathogenetic<br />
abnormality in MPN. 8 Ruxolitinib, being nonspecific for<br />
JAK2V617F mutation, therefore, may benefit all patients<br />
with MF regardless <strong>of</strong> their mutation status. 6 Many other<br />
406<br />
By Srdan Verstovsek, MD, PhD<br />
therapy with ruxolitinib. However, JAK2 inhibitors do not<br />
eliminate the disease and new medications are needed to<br />
expand on the benefits seen with JAK2 inhibitors. Although<br />
many agents are still in the early stages <strong>of</strong> development, the<br />
wealth <strong>of</strong> publications and presentations has continued to<br />
support our growing understanding <strong>of</strong> the pathophysiology <strong>of</strong><br />
MF as well as the potential short- and long-term outcomes <strong>of</strong><br />
these new and diverse approaches to treatment. Focus <strong>of</strong><br />
ongoing efforts is particularly on the improvements in anemia<br />
and fibrosis, as well as on rational combination trials <strong>of</strong> JAK2<br />
inhibitors and other potentially active agents. Therapeutic<br />
potential and limitations <strong>of</strong> JAK2 inhibitors and other novel<br />
medications in clinical studies are reviewed.<br />
JAK2 inhibitors, as well as histone deacetylase inhibitors<br />
(HDACIs), mammalian target <strong>of</strong> rapamycin (mTOR) inhibitors,<br />
transforming growth factor (TGF)-beta inhibitors, and<br />
others are in clinical studies (Table 1). An overview <strong>of</strong><br />
published information for these medications (including published<br />
abstracts) is provided here; we focus on medications<br />
in active clinical development.<br />
One important caveat when analyzing results <strong>of</strong> clinical<br />
trials for patients with MF is that studies use distinct<br />
response criteria, thus hampering comparisons between<br />
different therapeutic approaches. Uniform criteria for response<br />
in MF were developed several years ago (International<br />
Working Group for Myel<strong>of</strong>ibrosis Research and<br />
Therapy [IWG-MRT] response criteria), 9 but have already<br />
been questioned, and modifications proposed (already in use<br />
in some studies; e.g., “anemia response” as it relates to the<br />
definition <strong>of</strong> transfusion dependency and independency in<br />
patients with MF). 10 Indeed, the importance <strong>of</strong> clinically<br />
relevant and objective trial end points that would support<br />
claims <strong>of</strong> efficacy cannot be emphasized enough. 11 To that<br />
end, use <strong>of</strong> magnetic resonance imaging (MRI) instead <strong>of</strong><br />
physical exam to assess a response in splenomegaly and the<br />
development <strong>of</strong> Myel<strong>of</strong>ibrosis Symptom Assessment Form<br />
(MFSAF), which is an inventory to measure the symptom<br />
burden in MPNs, have been largely adopted in ongoing<br />
clinical trials.<br />
JAK2 inhibitors<br />
Ruxolitinib<br />
Efficacy and safety <strong>of</strong> ruxolitinib has been evaluated in<br />
two phase III clinical trials: the Controlled Myel<strong>of</strong>ibrosis<br />
Study with Oral JAK1/JAK2 Inhibitor Treatment I and II<br />
(COMFORT-I and COMFORT-II). 7,12 Results <strong>of</strong> these studies<br />
led to a recent approval <strong>of</strong> ruxolitinib in the United<br />
From the Department <strong>of</strong> Leukemia, M. D. Anderson Cancer Center, Houston, TX.<br />
Author’s disclosures <strong>of</strong> potential conflicts <strong>of</strong> interest are found at the end <strong>of</strong> this article.<br />
Address reprint requests to Srdan Verstovsek, MD, PhD, M. D. Anderson Cancer Center,<br />
Department <strong>of</strong> Leukemia, Unit 428, 1515 Holcombe Blvd., Houston, TX 77030; email:<br />
sverstov@mdanderson.org.<br />
© <strong>2012</strong> by <strong>American</strong> <strong>Society</strong> <strong>of</strong> <strong>Clinical</strong> <strong>Oncology</strong>.<br />
1092-9118/10/1-10
SMALL-MOLECULE INHIBITORS FOR MPN<br />
States as therapy for patients with intermediate- and highrisk<br />
MF.<br />
COMFORT-I is a double-blind, placebo-controlled study<br />
that enrolled 309 adults in the United States, Canada, and<br />
Australia with palpable splenomegaly and an International<br />
Prognostic Scoring System (IPSS) classification <strong>of</strong> intermediate-2<br />
or high risk myel<strong>of</strong>ibrosis (both primary or secondary<br />
MF). 7 Patients were randomly assigned (1:1) to<br />
receive ruxolitinib or placebo. Starting doses <strong>of</strong> ruxolitinib<br />
depended on platelet count at baseline: 15 mg twice daily for<br />
patients with at least 100 to 200 � 10 9 /L and 20 mg twice<br />
daily for patients with more than 200 � 10 9 /L. The primary<br />
end point was the proportion <strong>of</strong> patients achieving a reduction<br />
in spleen volume <strong>of</strong> at least 35% from baseline to week<br />
24 as measured by MRI or computed tomography (CT). Key<br />
KEY POINTS<br />
Table 1. Agents in Recent or Ongoing <strong>Clinical</strong> Studies for Myel<strong>of</strong>ibrosis<br />
Product MOA Company Phase<br />
Givinostat (ITF2357) HDAC inhibitor Italfarmaco (Milan, Italy) IIA<br />
Panobinostat (LBH589) HDAC inhibitor Novartis Pharmaceuticals (Basel, Switzerland) I/II<br />
Pracinostat (SB939) HDAC inhibitor S*Bio (Singapore, Singapore) II<br />
Saridegib (IPI-926) Hedgehog inhibitor (inhibits smoothened) Infinity Pharmaceuticals (Cambridge, MA) II<br />
GS6624 (AB0024) LOXL2 humanized monoclonal antibody (mAb) Gilead Sciences (Foster City, CA) II<br />
Pomalidomide (CC-4047) Immunomodulating agent Celgene (Summit, NJ) III<br />
AZD1480 JAK2 inhibitor AstraZeneca (Wilmington, DE) I/II<br />
BMS911543 JAK2 Inhibitor Bristol-Myers Squibb (Princeton, NJ) I/II<br />
CYT387 JAK1/2, TYK2 inhibitor YM Biosciences (Cytopia; Ontario, Canada) I/II<br />
LY2784544 JAK2 inhibitor Eli Lilly (Indianapolis, IN) I<br />
Ruxolitinib (Jakafi �U.S. trade name�;<br />
INCB018424; INC424)<br />
● Ruxolitinib, an oral Janus kinase (JAK) 1 and JAK2<br />
inhibitor, has recently been approved in the USA as<br />
the first medication for patients with intermediate- or<br />
high-risk myel<strong>of</strong>ibrosis (MF).<br />
● JAK inhibitor therapy in MF results in a rapid and<br />
sustained reduction in organomegaly (spleen and<br />
liver) and improvement in debilitating MF-related<br />
constitutional symptoms. Quality-<strong>of</strong>-life improvement<br />
is <strong>of</strong> a significant clinical benefit for patients.<br />
● With better control <strong>of</strong> the signs and symptoms <strong>of</strong> the<br />
disease with JAK inhibitors, patients with advanced<br />
MF may live longer.<br />
● JAK2 inhibitors are not specific for JAK2V617F mutation,<br />
and therefore all patients with MF benefit<br />
from therapy regardless whether they have JAK2<br />
mutation.<br />
● JAK2 inhibitors do not eliminate the disease, and<br />
new therapies and combination strategies are being<br />
explored, including histone deacetylase inhibitors,<br />
mammalian target <strong>of</strong> rapamycin inhibitors, and<br />
others.<br />
JAK1/2 inhibitor Incyte Corporation (Wilmington, DE) and<br />
Novartis Pharmaceuticals<br />
SAR302503 (TG101348) JAK2 inhibitor San<strong>of</strong>i (TargeGen; Paris, France) III<br />
Pacritinib (SB1518) JAK2 inhibitor S*Bio I/II<br />
NS-018 JAK2 inhibitor Nippon Shinyaku (Kyoto, Japan) I/II<br />
Plitidepsin (Aplidin) Marine cyclic depsipeptide PharmaMar (Madrid, Spain) II<br />
Everolimus (Afinitor; RAD-001) mTOR inhibitor Novartis Pharmaceuticals I/II<br />
Abbreviations: MOA, mode <strong>of</strong> action; HDAC, histone deacetylase; JAK, janus kinase; mTOR, mammalian target <strong>of</strong> rapamycin.<br />
Approved in<br />
United States<br />
secondary end points included the durability <strong>of</strong> spleen volume<br />
reduction (loss <strong>of</strong> response defined as a reduction <strong>of</strong> less<br />
than 35% from baseline and an increase <strong>of</strong> at least 25% from<br />
nadir), the proportion <strong>of</strong> patients with at least 50% improvement<br />
in symptoms (as measured by a Total Symptom Score<br />
[TSS] from the modified MFSAF version 2.0 electronic<br />
diary), and overall survival. At week 24, 41.9% <strong>of</strong> patients<br />
receiving ruxolitinib and 0.7% <strong>of</strong> patients receiving placebo<br />
achieved a spleen volume reduction <strong>of</strong> at least 35% from<br />
baseline (p � 0.0001). In patients treated with ruxolitinib<br />
who achieved at least 35% reduction in spleen volume,<br />
67.0% maintained the reduction for 48 weeks or more. At<br />
week 24, the proportion <strong>of</strong> patients who experienced a<br />
50% or greater improvement in TSS was 45.9% with ruxolitinib<br />
and 5.3% with placebo (p � 0.0001). In addition,<br />
mean TSS improved by 46.1% with ruxolitinib and worsened<br />
by 41.8% with placebo (p � 0.0001). Additional analyses <strong>of</strong><br />
COMFORT-I showed similar trends in spleen volume reductions<br />
and TSS improvements with ruxolitinib treatment<br />
regardless <strong>of</strong> patient subgroup evaluated, including myel<strong>of</strong>ibrosis<br />
subtype (PMF, post-PV MF [PPV-MF], or post-ET<br />
MF [PET-MF]), age (� 65 or � 65 years), IPSS risk category<br />
(high-risk or intermediate-2), baseline hemoglobin level<br />
(� 10 or � 10 g/dL), baseline palpable spleen length (� 10<br />
or � 10 cm), and JAK2V617F mutation status (positive or<br />
negative). At the time <strong>of</strong> the prospectively defined data<br />
cut<strong>of</strong>f (median follow-up <strong>of</strong> 32 weeks), there were 10 deaths<br />
with ruxolitinib and 14 deaths with placebo (6.5% vs. 9.1%;<br />
hazard ratio [HR] � 0.67; 95% CI, 0.30 to 1.50; p � 0.33). In<br />
a subsequent survival analysis on the basis <strong>of</strong> a planned<br />
data cut<strong>of</strong>f with an additional 4 months <strong>of</strong> follow-up (median<br />
follow-up, 51 weeks), there were 13 deaths with ruxolitinib<br />
and 24 deaths with placebo (8.4% vs. 15.6%; HR � 0.50; 95%<br />
CI, 0.25 to 0.98; p � 0.04).<br />
COMFORT-II is an open-label phase III study that enrolled<br />
patients in Europe with palpable splenomegaly with<br />
IPSS intermediate-2 and high-risk PMF, PPV-MF, or<br />
PET-MF (N � 219). 12 Patients were randomly assigned (2:1)<br />
to receive ruxolitinib or investigator-determined best available<br />
therapy (BAT). The dosing regimen for ruxolitinib<br />
was similar to that in COMFORT-I, and BAT included any<br />
commercially available monotherapy or combination ther-<br />
407
apy or no therapy. The primary end point was the proportion<br />
<strong>of</strong> patients achieving a reduction in spleen volume <strong>of</strong> at least<br />
35% from baseline to week 48, as measured by MRI or CT.<br />
Key secondary end points included spleen volume reduction<br />
at 24 weeks, duration <strong>of</strong> spleen volume reduction,<br />
progression-free survival, leukemia-free survival, and overall<br />
survival. Measurement <strong>of</strong> patient-reported quality <strong>of</strong> life<br />
and reduction in MF-related symptoms by using the European<br />
Organisation for Research and Treatment <strong>of</strong> Cancer<br />
(EORTC) QLQ-C30 and the Functional Assessment <strong>of</strong> Cancer<br />
Therapy Lymphoma (FACT-Lym) questionnaires was<br />
an exploratory end point. At week 48, 28.5% <strong>of</strong> patients who<br />
received ruxolitinib achieved at least 35% reduction from<br />
baseline in spleen volume. In contrast, no patients who<br />
received BAT achieved this end point (p � 0.0001). Similarly,<br />
only patients in the ruxolitinib group achieved a<br />
reduction from baseline <strong>of</strong> at least 35% at 24 weeks (31.9%,<br />
ruxolitinib; 0%, BAT; p � 0.0001). The median duration <strong>of</strong><br />
response with ruxolitinib was not reached; 80% <strong>of</strong> patients<br />
continued to have a response after a median follow-up <strong>of</strong><br />
12 months. Subgroup analyses demonstrated that ruxolitinib<br />
was more effective than BAT at reducing spleen<br />
volume regardless <strong>of</strong> gender, age, JAK2V617F mutation<br />
status, IPSS risk category, baseline spleen size, myel<strong>of</strong>ibrosis<br />
subtype, or ruxolitinib starting dose (15 or 20 mg). There<br />
was no significant difference in progression-free survival,<br />
leukemia-free survival, or overall survival between treatment<br />
groups. However, the study was not powered for<br />
statistical analyses <strong>of</strong> these end points. Compared with<br />
patients treated with BAT, ruxolitinib-treated patients experienced<br />
improvements in quality <strong>of</strong> life and marked reductions<br />
in MF-related symptoms, as measured by the EORTC<br />
QLQ-C30 and FACT-Lym subscales.<br />
The most common hematologic adverse events in both<br />
COMFORT-I and COMFORT-II were thrombocytopenia and<br />
anemia. These events were manageable with dose reduction<br />
or temporary interruption <strong>of</strong> therapy, and rarely led to<br />
discontinuation <strong>of</strong> ruxolitinib (one patient for each event<br />
in COMFORT-I; one patient for thrombocytopenia in<br />
COMFORT-II). The most common nonhematologic adverse<br />
events with ruxolitinib in COMFORT-I were ecchymosis,<br />
dizziness, and headache; these events were primarily grade<br />
1 or 2. The percentage <strong>of</strong> patients that stopped therapy<br />
as a result <strong>of</strong> ruxolitinib-related adverse effects in the<br />
COMFORT-I study was similar to the percentage <strong>of</strong> patients<br />
that stopped therapy as a result <strong>of</strong> placebo-related adverse<br />
effects. After temporary interruption or discontinuation <strong>of</strong><br />
ruxolitinib, MF-related symptoms generally returned to<br />
baseline levels within approximately 1 week. Analysis <strong>of</strong><br />
adverse events that occurred after interruption or discontinuation<br />
<strong>of</strong> ruxolitinib therapy showed no clear pattern to<br />
indicate a specific withdrawal effect. However, gradual tapering<br />
<strong>of</strong> the ruxolitinib may be considered when discontinuing<br />
therapy for reasons other than thrombocytopenia.<br />
Importantly, there are no contraindications to prescribing<br />
ruxolitinib to patients with intermediate- or high-risk MF. 5<br />
SAR302503<br />
A phase I/II clinical trial with SAR302503 has been<br />
published, and updated follow-up data were recently presented.<br />
13,14 Fifty-nine patients with MF were enrolled and<br />
received SAR302503 at doses ranging from 30 to 800 mg<br />
once daily. The maximum tolerated dose (MTD) was 680 mg<br />
408<br />
SRDAN VERSTOVSEK<br />
daily, and dose-limiting toxicities (DLTs) were asymptomatic<br />
grade 3 to 4 hyperamylasemia and hyperlipasemia.<br />
After six cycles <strong>of</strong> therapy, the spleen response in the MTD<br />
cohort was 45% by IWG criteria. No MRI testing was<br />
performed to assess volumetric reduction <strong>of</strong> the spleen;<br />
measurements were done by palpation. Most patients who<br />
presented with leukocytosis and thrombocytosis experienced<br />
normalized counts. There was a significant improvement in<br />
systemic symptoms after two to six cycles <strong>of</strong> therapy: early<br />
satiety (56% complete resolution), fatigue (63% improvement),<br />
night sweats (64% complete resolution), cough (67%<br />
complete resolution), and pruritus (50% complete resolution).<br />
A decrease in the JAK2V617F allele burden was<br />
observed in some patients: patients who presented with high<br />
(� 20%) allele burden at diagnosis had a significant decrease<br />
after 24 months. There was considerable hematologic toxicity:<br />
new-onset transfusion dependent anemia (grade 3 to 4,<br />
35.1%) and thrombocytopenia (grade 3 to 4, 23.7%). Other<br />
adverse effects included diarrhea, nausea, and vomiting.<br />
Current new clinical trials with this JAK2 inhibitor include<br />
a phase II randomized trial exploring alternative<br />
dose schedules to improve tolerability and maintain efficacy,<br />
as well as phase III placebo controlled blinded study for<br />
possible approval <strong>of</strong> this medication as therapy for MF<br />
(patients are randomly assigned to placebo or SAR302503 at<br />
initial doses <strong>of</strong> 400 and 500 mg once daily; JAKARTA;<br />
NCT01437787).<br />
CYT387<br />
Results <strong>of</strong> a phase I/II clinical trial with CYT387 have<br />
been presented in abstract form. 15,16 MTD was determined<br />
to be 300 mg/day, and DLTs included grade 3 headache and<br />
grade 3 hyperlipasemia at 400 mg/day. Most patients experienced<br />
improvements in pruritus, night sweats, bone pain,<br />
and fever. In the last update, an experience in 166 patients<br />
was summarized, after a median follow-up <strong>of</strong> 10.4 months;<br />
32 patients (19%) have discontinued therapy. A reduction<br />
in splenomegaly <strong>of</strong> at least 50% to qualify as a response<br />
by IWG-MRT criteria occurred in 31% <strong>of</strong> patients (by palpation).<br />
More strikingly was the response observed in anemia.<br />
Sixty-eight patients (41%) were transfusion dependent at<br />
baseline. The rate <strong>of</strong> transfusion independence for 12 weeks<br />
and hemoglobin <strong>of</strong> at least 8 g/dL was 46%, and was 62%<br />
among patients receiving treatment at MTD. Median time<br />
to transfusion independence was 84 days, and median duration<br />
has not been reached. The mechanism <strong>of</strong> anemia<br />
response remains a focus <strong>of</strong> active investigation. One complicating<br />
factor is that the IWG response criteria definitions<br />
for transfusion dependence and independence have<br />
been shown inadequate for proper assessment <strong>of</strong> clinical<br />
benefit: In COMFORT-1 study (described earlier, evaluating<br />
ruxolitinib therapy vs. placebo in blinded fashion), those<br />
patients receiving placebo experienced transfusion independence<br />
by IWG criteria in 47% <strong>of</strong> cases. 7 Further maturation<br />
<strong>of</strong> data and results <strong>of</strong> this study are awaited in the near<br />
future. The most common hematologic adverse effect was<br />
thrombocytopenia, which occurred in 33% <strong>of</strong> cases and was<br />
grade 3 to 4 in 17%. The most common nonhematologic<br />
toxicity (20%; grade 1 only) was called the “first dose effect”:<br />
transient lightheadedness and/or dizziness which may be<br />
accompanied by hypotension. Other less common and low<br />
grade nonhematologic adverse effects included peripheral<br />
neuropathy, diarrhea, nausea, and headache.
SMALL-MOLECULE INHIBITORS FOR MPN<br />
LY2784544<br />
In an open-label phase I trial, 19 patients received treatment<br />
with LY2784544. 17 The MTD was 120 mg. Similar to<br />
results with other JAK2 inhibitors, a reduction in spleen<br />
size and systemic symptoms was observed within the first<br />
two to three cycles <strong>of</strong> therapy. No reduction <strong>of</strong> JAK2V617F<br />
allele burden had been noted; however, preliminary results<br />
suggest that the compound could improve the fibrosis associated<br />
with this disorder. Tumor lysis syndrome was observed<br />
at the lower end <strong>of</strong> the dose range associated with<br />
efficacy, and therefore dosing has been amended to include a<br />
lower dose lead-in period. This study is ongoing.<br />
HDACIs<br />
HDACIs are compounds that inhibit activity <strong>of</strong> HDAC and<br />
can lead to increased histone acetylation and gene expression.<br />
Some HDACIs in clinical trials for MPNs include<br />
givinostat (formerly known as ITF2357) and panobinostat<br />
(formerly known as LBH589). A pilot study with givinostat<br />
in MF was recently published. 18 Twenty-nine patients with<br />
JAK2V617F-positive MPNs (PV, n � 12; ET, n � 1; MF, n �<br />
16) received treatment with givinostat 50 mg twice daily.<br />
Responses were seen in 54% <strong>of</strong> PV/ET patients (complete,<br />
n � 1; partial, n � 6) according to European LeukemiaNet<br />
criteria. Responses in MF were more modest: Only three <strong>of</strong><br />
16 patients had a major response. Adverse effects included<br />
diarrhea, anemia, thrombocytopenia, fatigue, and QTc elongation.<br />
Panobinostat was evaluated in a phase I trial for<br />
patients with hematologic malignancies, including 13 patients<br />
with MF. 19 Most patients received panobinostat<br />
thrice weekly, and the MTD was 60 mg/dose. Grade 3 to 4<br />
adverse effects included thrombocytopenia (33%), fatigue<br />
(28%, DLT), neutropenia (28%), and anemia (12%). Four<br />
patients with MF had response by IWG-MRT criteria. These<br />
results were followed by two studies evaluating panobinostat<br />
solely in patients with MF. Mascarenhas and colleagues<br />
reported on a phase I study in 15 patients with MF;<br />
thrombocytopenia was the DLT. 20 Five patients received<br />
more than 6 months <strong>of</strong> therapy, and all achieved clinical<br />
improvement in spleen by IWG response criteria. 21 In another<br />
trial, 31 patients with MF received treatment with<br />
panobinostat (initial dose was 60 mg thrice weekly) but<br />
the toxicity precluded delivery <strong>of</strong> the medication beyond 1<br />
month <strong>of</strong> therapy in almost all patients. 22 A combination<br />
study <strong>of</strong> ruxolitinib and panobinostat at low dose is underway<br />
on the basis <strong>of</strong> exciting preclinical results. 23<br />
mTOR Inhibitor<br />
Activated mTOR is a serine/threonine kinase which regulates<br />
cell growth, proliferation and metabolism. Results for<br />
30 patients from a phase I/II study <strong>of</strong> everolimus (RAD-001)<br />
in high- or intermediate-risk primary or secondary MF were<br />
published during 2011. 24 No DLT was observed up to 10<br />
mg/day. At this dose, toxicities were infrequent; the most<br />
common toxicity was grade 1 to 2 stomatitis. A splenomegaly<br />
reduction <strong>of</strong> more than 50% from baseline occurred in 20% <strong>of</strong><br />
patients, whereas a more than 30% reduction occurred in<br />
44% <strong>of</strong> patients. A total <strong>of</strong> 69% and 80% <strong>of</strong> patients experienced<br />
complete resolution <strong>of</strong> systemic symptoms and pruritus,<br />
respectively. Response in leukocytosis, anemia, and<br />
thrombocytosis occurred in 15% to 25% <strong>of</strong> patients.<br />
Immunomodulatory Inhibitory Drugs<br />
The first published study <strong>of</strong> pomalidomide (an analog <strong>of</strong><br />
thalidomide, formerly CC-4047) evaluated 84 patients with<br />
MF who were randomly assigned among four arms: pomalidomide<br />
2 mg daily (n � 22), pomalidomide 2 mg daily plus<br />
prednisone (n � 19), pomalidomide 0.5 mg daily plus prednisone<br />
(n � 22) and prednisone alone (n � 21). 25 Observed<br />
benefit was limited to improvement in anemia, according to<br />
IWG criteria: Response rates were 23% (pomalidomide 2 mg<br />
plus placebo), 16% (pomalidomide 2 mg plus prednisone),<br />
36% (pomalidomide 0.5 mg plus prednisone), and 19% (prednisone<br />
alone). The drug was very well tolerated. 25 In a phase<br />
I/II trial, the MTD <strong>of</strong> pomalidomide was determined to be<br />
3 mg/day, and DLT was myelosuppression. 26 In accordance<br />
with the results <strong>of</strong> the randomized study, better response<br />
rates were observed in patients who received low-dose pomalidomide<br />
(63% response in anemia). 26 Begna and colleagues<br />
recently reported the results <strong>of</strong> a use <strong>of</strong> low-dose<br />
pomalidomide alone (0.5 mg/day) in 58 patients with MF. 27<br />
The anemia response was 17% by IWG-MRT criteria. An<br />
increase in platelet count in patients with thrombocytopenia<br />
was observed in 58% <strong>of</strong> cases. No patient had an improvement<br />
in splenomegaly. A phase III randomized study <strong>of</strong><br />
pomalidomide compared with placebo in patients with MF<br />
who are transfusion dependent is underway. Patients are<br />
randomly assigned to either pomalidomide (0.5 mg/day) or<br />
placebo; primary end point is rate <strong>of</strong> transfusion independence<br />
after 6 months <strong>of</strong> therapy.<br />
Conclusion<br />
In recent years, the outlook for patients with chronic<br />
Ph-negative MPNs has changed with new discoveries on<br />
molecular biology <strong>of</strong> these diseases and the subsequent<br />
development <strong>of</strong> new compounds directed against those molecular<br />
defects. JAK2 inhibitors and other compounds, such<br />
as pomalidomide, are on the verge <strong>of</strong> making the transition<br />
from clinical trials to routine clinical use. Indeed, ruxolitinib,<br />
an oral JAK1 and JAK2 inhibitor, has recently been<br />
approved in the United States as the first medication ever<br />
for treatment <strong>of</strong> patients with intermediate- and high-risk<br />
MF. 13 The most striking benefits observed with JAK2 inhibitors<br />
are a reduction in spleen size and improvement in<br />
constitutional symptoms. As suggested by COMFORT-I trial<br />
results, with the better control <strong>of</strong> the symptoms and signs<br />
<strong>of</strong> the disease, thus potentially delaying a progression <strong>of</strong> the<br />
disease, patients with advanced MF may live longer. However,<br />
with JAK2 inhibitors a reduction in bone marrow<br />
fibrosis is mostly anecdotal. Similarly, with the majority <strong>of</strong><br />
compounds there is no significant reduction in the JAK2<br />
allele burden. JAK2 inhibitors are not able to clonally<br />
eradicate the disease. Therefore, JAK inhibitors are just<br />
the first building block in our efforts to effectively treat MF,<br />
and there is much room for improvement. We need to better<br />
understand how these drugs work, and through which<br />
mechanism(s) they are producing benefits for individual<br />
patients. We need biomarkers to determine which patients<br />
will respond to a specific agent, as well as to rationally<br />
combine active agents for additional benefits. With coordinated<br />
efforts and appropriate design <strong>of</strong> clinical trials, we can<br />
hope to overcome these challenges and improve outcomes for<br />
patients with MF, not just in controlling disease signs and<br />
symptoms, but potentially eliminating the disease.<br />
409
Author’s Disclosure <strong>of</strong> Potential Conflicts <strong>of</strong> Interest<br />
Author<br />
Employment or<br />
Leadership<br />
Positions<br />
Consultant or<br />
Advisory Role<br />
Stock<br />
Ownership Honoraria<br />
Research<br />
Funding<br />
Srdan Verstovsek AstraZeneca;<br />
Bristol-Myers<br />
Squibb; Celgene;<br />
Geron; Gilead<br />
Sciences; Incyte;<br />
Lilly; NS Pharma;<br />
Roche; S*Bio<br />
1. Cervantes F, Pereira A. Advances in the understanding and management<br />
<strong>of</strong> primary myel<strong>of</strong>ibrosis. Curr Opin Oncol. 2011;23:665-671.<br />
2. Mesa RA. Assessing new therapies and their overall impact in myel<strong>of</strong>ibrosis.<br />
Hematology Am Soc Hematol Educ Program. 2010;2010:115-121.<br />
3. James C, Ugo V, Le Couedic JP, et al. A unique clonal JAK2 mutation<br />
leading to constitutive signalling causes polycythaemia vera. Nature. 2005;<br />
434:1144-1148.<br />
4. Quintas-Cardama A, Kantarjian H, Cortes J, et al. Janus kinase inhibitors<br />
for the treatment <strong>of</strong> myeloproliferative neoplasias and beyond. Nat Rev<br />
Drug Discov. 2011;10:127-140.<br />
5. JAKAFI [Prescribing Information]. Wilmington, DE: Incyte Corporation;<br />
2011.<br />
6. Verstovsek S, Kantarjian H, Mesa RA, et al. Safety and efficacy <strong>of</strong><br />
INCB018424, a JAK1 and JAK2 inhibitor, in myel<strong>of</strong>ibrosis. N Engl J Med.<br />
2010;363:1117-1127.<br />
7. Verstovsek S, Mesa RA, Gotlib J, et al. A double-blind, placebocontrolled<br />
trial <strong>of</strong> ruxolitinib for myel<strong>of</strong>ibrosis. N Engl J Med. <strong>2012</strong>;366:799-<br />
807.<br />
8. Tefferi A. Novel mutations and their functional and clinical relevance<br />
in myeloproliferative neoplasms: JAK2, MPL, TET2, ASXL1, CBL, IDH and<br />
IKZF1. Leukemia. 2010;24:1128-1138.<br />
9. Cervantes F, Dupriez B, Pereira A, et al. New prognostic scoring system<br />
for primary myel<strong>of</strong>ibrosis based on a study <strong>of</strong> the International Working<br />
Group for Myel<strong>of</strong>ibrosis Research and Treatment. Blood. 2009;113:2895-2901.<br />
10. Gale RP, Barosi G, Barbui T, et al. What are RBC-transfusiondependence<br />
and -independence? Leuk Res. 2011;35:8-11.<br />
11. Barosi G, Tefferi A, Barbui T. Do current response criteria in classical<br />
Ph-negative myeloproliferative neoplasms capture benefit for patients? Leukemia.<br />
Epub 2011 Nov 22.<br />
12. Harrison C, Kiladjian JJ, Al-Ali HK, et al. JAK inhibition with<br />
ruxolitinib versus best available therapy for myel<strong>of</strong>ibrosis. N Engl J Med.<br />
<strong>2012</strong>; 366:787-798.<br />
13. Pardanani A, Gotlib J, Jamieson C, et al. SAR302503: interim safety,<br />
efficacy and long-term impact on JAK2 V617F allele burden in a phase I/II<br />
study in patients with myel<strong>of</strong>ibrosis. Blood. 2011;118 (abstr 3838).<br />
14. Pardanani A, Gotlib JR, Jamieson C, et al. Safety and efficacy <strong>of</strong><br />
TG101348, a selective JAK2 inhibitor, in myel<strong>of</strong>ibrosis. J Clin Oncol. 2011;<br />
29:789-796.<br />
15. Pardanani AD, Caramazza D, George G, et al. Safety and efficacy <strong>of</strong><br />
410<br />
REFERENCES<br />
Expert<br />
Testimony<br />
SRDAN VERSTOVSEK<br />
Other<br />
Remuneration<br />
CYT387, a JAK-1/2 inhibitor, for the treatment <strong>of</strong> myel<strong>of</strong>ibrosis. J Clin Oncol.<br />
2011;29 (suppl; abstr 6514).<br />
16. Pardanani A, Gotlib J, Gupta V, et al. An expanded multicenter phase<br />
I/II study <strong>of</strong> CYT387, a JAK- 1/2 inhibitor for the treatment <strong>of</strong> myel<strong>of</strong>ibrosis.<br />
Blood. 2011;118 (abstr 3849).<br />
17. Verstovsek S, Mesa RA, Rhoades SK, et al. Phase I study <strong>of</strong> the JAK2<br />
V617F inhibitor, LY2784544, in patients with myel<strong>of</strong>ibrosis (MF), polycythemia<br />
vera (PV), and essential thrombocythemia (ET). Blood. 2011;118 (abstr<br />
2814).<br />
18. Rambaldi A, Dellacasa CM, Finazzi G, et al. A pilot study <strong>of</strong> the<br />
histone-deacetylase inhibitor givinostat in patients with JAK2V617F positive<br />
chronic myeloproliferative neoplasms. Br J Haematol. 2010;150:446-455.<br />
19. DeAngelo DJ, Spencer A, Fischer T, et al. Activity <strong>of</strong> oral panobinostat<br />
(LBH589) in patients with myel<strong>of</strong>ibrosis. Blood, 2009;114 (abstr 2898).<br />
20. Mascarenhas J, Wang X, Rodriguez A, et al. A phase I study <strong>of</strong> LBH589,<br />
a novel histone deacetylase inhibitor in patients with primary myel<strong>of</strong>ibrosis<br />
(PMF) and post-polycythemia/essential thrombocythemia myel<strong>of</strong>ibrosis (Post-<br />
PV/ET MF). Blood. 2009;114 (abstr 308).<br />
21. Mascarenhas J, Mercado A, Rodriguez A, et al. Prolonged low dose<br />
therapy with a pan-deacetylase inhibtor, panobinostat (LBH589), in patients<br />
with myel<strong>of</strong>ibrosis. Blood. 2011;118 (abstr 794).<br />
22. DeAngelo DJ, Tefferi A, Fiskus W, et al. A phase II trial <strong>of</strong> panobinostat,<br />
an orally available deacetylase inhibitor (DACi), in patients with primary<br />
myel<strong>of</strong>ibrosis (PMF), post essential thrombocythemia (ET), and post polycythemia<br />
vera (PV) myel<strong>of</strong>ibrosis. Blood. 2010;116 (abstr 630).<br />
23. Baffert F, Evrot E, Ebel N, et al. Improved efficacy upon combined<br />
JAK1/2 and pan-deacetylase inhibition using ruxolitinib (INC424) and panobinostat<br />
(LBH589) in preclinical mouse models <strong>of</strong> JAK2V617F-driven disease.<br />
Blood. 2011;118 (abstr 798).<br />
24. Guglielmelli P, Barosi G, Rambaldi A, et al. Safety and efficacy <strong>of</strong><br />
everolimus, a mTOR inhibitor, as single agent in a phase 1/2 study in patients<br />
with myel<strong>of</strong>ibrosis. Blood. 2011;118:2069-2076.<br />
25. Tefferi A, Verstovsek S, Barosi G, et al. Pomalidomide is active in the<br />
treatment <strong>of</strong> anemia associated with myel<strong>of</strong>ibrosis. J Clin Oncol. 2009;27:<br />
4563-4569.<br />
26. Mesa RA, Pardanani AD, Hussein K, et al. Phase1/-2 study <strong>of</strong> pomalidomide<br />
in myel<strong>of</strong>ibrosis. Am J Hematol. 2009;85:129-130.<br />
27. Begna KH, Mesa RA, Pardanani A, et al. A phase-2 trial <strong>of</strong> low-dose<br />
pomalidomide in myel<strong>of</strong>ibrosis. Leukemia. 2011;25:301-304.
Insights into the Molecular Genetics <strong>of</strong><br />
Myeloproliferative Neoplasms<br />
By Huong (Marie) Nguyen, MD, and Jason Gotlib, MD, MS<br />
Overview: The molecular biology <strong>of</strong> the BCR-ABL1-negative<br />
chronic myeloproliferative neoplasms (MPNs) has witnessed<br />
unprecedented advances since the discovery <strong>of</strong> the acquired<br />
JAK2 V617F mutation in 2005. Despite the high prevalence <strong>of</strong><br />
JAK2 V617F in polycythemia vera (PV), essential thrombocythemia<br />
(ET), and primary myel<strong>of</strong>ibrosis (PMF), and the common<br />
finding <strong>of</strong> dysregulated JAK-STAT signaling in these disorders,<br />
it is now appreciated that MPN pathogenesis can reflect<br />
the acquisition <strong>of</strong> multiple genetic mutations that alter several<br />
biologic pathways, including epigenetic control <strong>of</strong> gene expression.<br />
Although certain gene mutations are identified at<br />
higher frequencies with disease evolution to the blast phase,<br />
MPNS ARE clonal hematopoietic disorders that result<br />
in overproduction <strong>of</strong> one or more terminally differentiated<br />
blood cell types arising from the myeloid lineage. In<br />
the 2008 World Health Organization (WHO) classification,<br />
MPNs are divided into eight subtypes: chronic myeloid<br />
leukemia (CML); polycythemia vera (PV); essential thrombocythemia<br />
(ET); primary myel<strong>of</strong>ibrosis (PMF); systemic<br />
mastocytosis (SM); chronic neutrophilic leukemia; chronic<br />
eosinophilic leukemia, not otherwise specified (CEL, NOS);<br />
and MPN-unclassifiable (MPN-U). 1 Myeloid (and lymphoid)<br />
neoplasms associated with eosinophilia and rearrangement<br />
<strong>of</strong> platelet-derived growth factor receptor alpha or beta<br />
(PDGFRA or PDGFRB) and fibroblast growth factor receptor1(FGFR1)<br />
are distinguished by their own major WHO<br />
disease category, but share numerous clinicopathologic features<br />
with MPNs. 1 A pathogenetic hallmark <strong>of</strong> MPNs is<br />
dysregulation <strong>of</strong> tyrosine kinases (TKs), which in turn results<br />
in aberrant downstream signaling and increased cellular<br />
proliferation and/or decreased apoptosis. Table 1<br />
summarizes the molecular lesions (e.g., reciprocal chromosomal<br />
translocations, point mutations, interstitial chromosomal<br />
deletions) that generate oncogenic TKs in MPNs and<br />
myeloid neoplasms associated with eosinophilia. The notion<br />
that PV, ET, or MF is driven by a single TK lesion such as<br />
JAK2 V617F has now been abandoned given the genetic and<br />
epigenetic complexity observed in most patients (Fig. 1). The<br />
cytogenetics <strong>of</strong> MPNs is important to understand disease<br />
pathogenesis and prognosis; however, this topic is not addressed<br />
in this monograph, and readers are directed elsewhere<br />
for reviews on the subject.<br />
Before JAK2 V617F: Clonality, Cytokine Independence,<br />
and Aberrant JAK-STAT Signaling<br />
In a 1951 Blood editorial, Dr. William Dameshek first<br />
conceptualized the inter-relatedness <strong>of</strong> PV, ET, and MF and<br />
postulated a “hitherto undiscovered stimulus” as the biologic<br />
basis <strong>of</strong> their shared myeloproliferative features. 2 In the<br />
1970s, Adamson and colleagues used restriction fragment<br />
length polymorphism analysis <strong>of</strong> the X-linked glucose-6phosphate<br />
dehydrogenase (G6PD) gene in a female patient<br />
to confirm the clonal basis <strong>of</strong> PV, with ensuing studies<br />
demonstrating clonality in ET and MF. 3 These investigations<br />
were followed by two seminal observations: 1) hematopoietic<br />
progenitors from patients with PV (and in some<br />
MPN initiation and progression are not explained by a single,<br />
temporal pattern <strong>of</strong> clonal changes. A complex interplay<br />
between acquired molecular abnormalities and host genetic<br />
background, in addition to the type and allelic burden <strong>of</strong><br />
mutations, contributes to the phenotypic heterogeneity <strong>of</strong><br />
MPNs. At the population level, an inherited predisposition to<br />
developing MPNs is linked to a relatively common JAK2associated<br />
haplotype (referred to as ‘46/1’), but it exhibits a<br />
relatively low penetrance. This review details the current state<br />
<strong>of</strong> knowledge <strong>of</strong> the molecular genetics <strong>of</strong> the classic MPNs<br />
PV, ET, and PMF and discusses the clinical implications <strong>of</strong><br />
these findings.<br />
cases ET and MF) proliferate in the absence <strong>of</strong> exogenous<br />
cytokines such as erythropoietin (Epo) (e.g., endogenous<br />
erythroid colony growth [EEC]), and 2) such cells are hypersensitive<br />
to growth factors such as Epo, thrombopoietin<br />
(Tpo), and interleukin-3 (IL-3). 4<br />
The endogenous, self-stimulatory property <strong>of</strong> MPN cells<br />
and cytokine hypersensitivity led to increasing interest in<br />
JAK2 as a contributor to MPN pathogenesis. JAK2 belongs<br />
to a family <strong>of</strong> four janus kinases, which also includes JAK1,<br />
JAK3, and TYK2. Each JAK protein has an active tyrosine<br />
kinase domain (JAK homology 1[JH1]), an inactive pseudokinase<br />
domain (JAK homology 2 [JH2]), an SRC homology 2<br />
domain (SH2), and an amino terminal FERM (4-point-1,<br />
Erzin, Radixin, Moesin) homology domain, which binds to<br />
the cytoplasmic tail <strong>of</strong> cytokine receptors. JAK2 facilitates<br />
normal myelopoiesis by transmitting signals from type I<br />
receptors for Epo (EpoR), thrombopoietin (TpoR or MPL),<br />
granulocyte-colony-stimulating factor (G-CSFR), and IL-3<br />
(IL-3R). In the normal state, binding <strong>of</strong> ligand to receptor<br />
causes JAK2 to change from a receptor-bound, inactive<br />
conformation to an active catalytic enzyme because <strong>of</strong> escape<br />
from the inhibitory effects <strong>of</strong> the pseudokinase domain on<br />
the kinase domain. 5 Auto-phosphorylation <strong>of</strong> JAK2 and<br />
phosphorylation <strong>of</strong> downstream signaling intermediates results<br />
in recruitment <strong>of</strong> SH2-domain containing proteins<br />
such as STAT3 and STAT5. After phosphorylation by JAK2,<br />
the STAT proteins homodimerize and translocate to the<br />
nucleus, where they activate transcription <strong>of</strong> target genes<br />
involved in regulating a variety <strong>of</strong> cellular processes, including<br />
proliferation, differentiation, and apoptosis. Dampening<br />
<strong>of</strong> JAK-STAT activation occurs via different negative feedback<br />
mechanisms, including the suppressor <strong>of</strong> cytokine<br />
signaling (SOCS) family <strong>of</strong> proteins, LNK, CBL, and various<br />
tyrosine phosphatases.<br />
From the Division <strong>of</strong> Hematology, Department <strong>of</strong> Medicine, Stanford University School <strong>of</strong><br />
Medicine/Stanford Cancer Institute, Stanford, CA.<br />
Authors’ disclosures <strong>of</strong> potential conflicts <strong>of</strong> interest are found at the end <strong>of</strong> this article.<br />
Address reprint requests to Jason Gotlib, MD, MS, Associate Pr<strong>of</strong>essor <strong>of</strong> Medicine,<br />
Stanford Cancer Institute, 875 Blake Wilbur Drive, Room 2324, Stanford, CA 94305-5821;<br />
email: jason.gotlib@stanford.edu.<br />
© <strong>2012</strong> by <strong>American</strong> <strong>Society</strong> <strong>of</strong> <strong>Clinical</strong> <strong>Oncology</strong>.<br />
1092-9118/10/1-10<br />
411
The JAK2 V617F Mutation<br />
Aberrant JAK-STAT signaling had been demonstrated in<br />
various MPN patient subgroups, laying the groundwork for<br />
identification <strong>of</strong> the molecular alterations in this pathway.<br />
In 2005, the somatic activating mutation JAK2 V617F<br />
(V617F) was discovered by several groups using either<br />
high-throughput tyrosine kinase genome sequencing, or<br />
functional methods such as microsatellite mapping to further<br />
define a segment <strong>of</strong> chromosome 9p containing the<br />
JAK2 gene in which loss <strong>of</strong> heterozygosity (LOH) in that<br />
region had previously been identified in some PV and ET<br />
patients. 6-9 The French group’s focus on JAK2 derived from<br />
investigations in which a JAK2 inhibitor and siRNA against<br />
KEY POINTS<br />
● JAK2 V617F is a common pathogenetic mutation in<br />
myeloproliferative neoplasms (MPNs) and is sufficient<br />
to produce a myeloproliferative phenotype in<br />
murine models.<br />
● Molecular abnormalities in positive and negative<br />
regulators <strong>of</strong> the JAK-STAT axis (e.g., MPL, CBL,<br />
LNK) can recapitulate activation <strong>of</strong> this signaling<br />
pathway in patients negative for JAK2 V167F.<br />
● The acquisition <strong>of</strong> multiple pre- and post-JAK2 mutational<br />
events is common in MPNs and likely contributes<br />
to phenotypic diversity and progression to<br />
acute myeloid leukemia.<br />
● Epigenetic dysregulation, including changes in pathways<br />
that control DNA methylation and modification<br />
<strong>of</strong> chromatin, has emerged as an important paradigm<br />
in MPN biology.<br />
● Inherited predisposition to MPNs is partly explained<br />
by a JAK2-associated haplotype. Genome-wide association<br />
studies are in progress to identify additional<br />
susceptibility loci.<br />
Table 1. Tyrosine Kinase Mutations in Myeloproliferative Neoplasms<br />
Prototypic Tyrosine<br />
Kinase Mutation Frequency Chromosome Comments<br />
Myeloproliferative Neoplasms<br />
Chronic Myeloid Leukemia BCR-ABL1 100% t(9;22)(q34;q11.2) BCR-ABL1 defines CML<br />
Polycythemia Vera JAK2 V617F �95% 9p24 Post-MPN AML frequency �50%<br />
Essential Thrombocythemia JAK2 V617F �50–60% 9p24<br />
Primary Myel<strong>of</strong>ibrosis JAK2 V617F �50–60% 9p24<br />
Systemic Mastocytosis KIT D816V �80% 4q12 Juxtamembrane KIT mutations �5%<br />
Chronic Eosinophilic Leukemia, Not Otherwise Specified None N/A N/A No characteristic TK abnormality<br />
Chronic Neutrophilic Leukemia None N/A N/A No characteristic TK abnormality<br />
Myeloproliferative Neoplasm, Unclassifiable None N/A N/A No characteristic TK abnormality<br />
Myeloid and Lymphoid Neoplasms with Eosinophilia and<br />
Abnormalities <strong>of</strong> PDGFRA, PDGFRB, orFGFR1<br />
PDGFRA-Rearranged FIP1L1-PDGFRA �10%* 4q12 cryptic interstitial<br />
chromosome deletion<br />
�5 additional variant fusion partners with<br />
PDGFRA identified<br />
PDGFRB-Rearranged ETV6-PDGFRB Rare t(5;12)(q31�33;p13) �20 additional variant fusion partners with<br />
PDGFRB identified<br />
FGFR1-Rearranged ZNF198-FGFR1 Rare t(8;13)(p11�12;q12) �10 additional variant fusion partners with<br />
FGFR1 identified<br />
Abbreviations: TK, tyrosine kinase; MPN, myeloproliferative neoplasm; AML, acute myeloid leukemia; N/A, not applicable.<br />
* Estimated frequency <strong>of</strong> �10% among patients with idiopathic hypereosinophilia in developed countries.<br />
412<br />
NGUYEN AND GOTLIB<br />
JAK2 blocked terminal differentiation <strong>of</strong> PV erythroid progenitors<br />
and/or blocked EEC growth. 9 Sequencing <strong>of</strong> the<br />
JAK2 gene revealed a somatic point mutation at base pair<br />
1849 (G3T), causing substitution <strong>of</strong> the normal valine to<br />
phenylalanine in codon 617 (V617F) <strong>of</strong> exon 14. This V617F<br />
mutation is found in approximately 95% <strong>of</strong> PV patients and<br />
50% to 60% <strong>of</strong> ET and PMF patients. The mutation frequency<br />
is considerably less (e.g., �5%) in patients with other<br />
MPNs, such as systemic mastocytosis and chronic eosinophilic<br />
leukemias or acute myeloid leukemia. However, in<br />
patients with overlap MDS/MPNs, such as proliferative-type<br />
chronic myelomonocytic leukemia (CMML) and refractory<br />
anemia with ring sideroblasts with thrombocytosis (RARS-<br />
T), the frequency increases to approximately 10% and 50%,<br />
respectively.<br />
The V617F mutation resides in the autoinhibitory pseudokinase<br />
domain <strong>of</strong> JAK2. Modeling studies suggest that<br />
V617F causes a structural change in the pseudokinase<br />
domain, relieving inhibition <strong>of</strong> JAK2 kinase activity. 5 Overexpression<br />
<strong>of</strong> the JAK2 V617F allele in cell lines results<br />
in phosphorylation <strong>of</strong> JAK2 and STAT5 in the absence <strong>of</strong><br />
cytokine stimulation. 6,8,9 The human erythroleukemia<br />
(HEL) cell line, which carries a homozygous V617F mutation,<br />
exhibits constitutive phosphorylation <strong>of</strong> JAK2 and<br />
STAT5; treatment <strong>of</strong> HEL cells with a JAK2 inhibitor led to<br />
reduced phosphorylation <strong>of</strong> JAK2 and STAT5 and inhibition<br />
<strong>of</strong> proliferation. 7 In a V617F knock-in murine model <strong>of</strong><br />
PV, STAT5 was an absolute requirement for the pathogenesis<br />
<strong>of</strong> PV. 10 Deletion <strong>of</strong> STAT5 normalized all the clinical,<br />
blood, and histopathologic features <strong>of</strong> PV, including EEC<br />
formation.<br />
One JAK2 V617F Mutation, Multiple MPN Phenotypes<br />
The high prevalence <strong>of</strong> the V617F mutation in three<br />
clinically distinct MPNs begs the question <strong>of</strong> what additional<br />
factors contribute to phenotypic diversity between PV,<br />
ET, and PMF. First, mutant allele burden <strong>of</strong> V617F may<br />
modulate phenotype. Murine retroviral transplant models<br />
resulting in high levels <strong>of</strong> V617F expression produced a<br />
PV-like phenotype with marked erythrocytosis. 11 Con-
GENETICS OF MYELOPROLIFERATIVE NEOPLASMS<br />
Fig. 1. Genetic and Epigenetic Dysregulation in Myeloproliferative Neoplasms. Adapted from Expert Review <strong>of</strong> Hematology, “JAK2 V617F<br />
and beyond: role <strong>of</strong> genetics and aberrant signaling in the pathogenesis <strong>of</strong> myeloproliferative neoplasms,” Vol. 3, No. 3, June 2010, pages<br />
323-327. Stephen Oh and Jason Gotlib, Figure 1, with permission <strong>of</strong> Expert Reviews Ltd.; and adapted with kind permission from Springer<br />
Science and Business Media: Current Hematologic Malignancy Reports, “Disordered Epigenetic Regulation in the Pathophysiology <strong>of</strong> Myeloproliferative<br />
Neoplasms,” vol. 7, no. 1, March <strong>2012</strong>, pages 34-42, Su-Jiang Zhang and Omar Abdel-Wahab, Figure 1.<br />
versely, transgenic models with more physiologic levels <strong>of</strong><br />
V617F expression resulted in phenotypes resembling ET<br />
and PMF. 12 These animal models corroborate the findings<br />
in patients in which V617F allele burden tends to be the<br />
highest in PV and PMF, with lower levels in ET patients.<br />
For ET patients, positivity for V617F tends to confer a<br />
PV-like phenotype, with higher hemoglobin and lower platelet<br />
counts than in V617F-negative ET patients. Differences<br />
in intracellular signaling arising from V617F may also<br />
explain the development <strong>of</strong> PV compared with ET: preferential<br />
activation <strong>of</strong> STAT1 constrains erythroid differentiation<br />
and promotes megakaryocytic development, leading to an<br />
ET phenotype. 13 In contrast, reduced STAT1 phosphorylation<br />
promotes erythroid development as observed in PV.<br />
The type <strong>of</strong> hematopoietic progenitor(s) targeted by the<br />
V617F mutation may contribute to differences in MPN<br />
subtype. The V617F mutation is found in myeloid, or less<br />
commonly, lymphoid lineage cells from MPN patients, suggesting<br />
that it arises in a hematopoietic stem cell (HSC) or<br />
early progenitor cell. 14 Jamieson and colleagues identified<br />
the V617F mutation in cells from PV patients with an HSC<br />
immunophenotype and demonstrated that these cells were<br />
skewed toward the erythroid lineage. 15 These findings suggest<br />
that either the V617F mutation induces erythroid<br />
differentiation or that the mutation preferentially targets an<br />
HSC subset already committed to an erythroid fate.<br />
Host genetic background may also influence disease presentation.<br />
In retroviral transplant models, disparate phenotypes<br />
were observed depending on the mouse strain. In<br />
C57Bl/6 mice, transplantation with JAK2 V617Ftransduced<br />
cells resulted in a PV-like disease predominantly<br />
characterized by erythrocytosis. 16 However, in Balb/C mice,<br />
similar experiments yielded mice with erythrocytosis, but<br />
also leukocytosis and the subsequent development <strong>of</strong> myel<strong>of</strong>ibrosis.<br />
17<br />
Perhaps the most compelling basis for MPN diversity<br />
comes from the additional molecular abnormalities that<br />
either precede or follow the acquisition <strong>of</strong> V617F (Table 2).<br />
The aggregate data suggest that there is no strict temporal<br />
order <strong>of</strong> mutation occurrence that defines the development<br />
or natural history <strong>of</strong> specific MPNs. However, several lines<br />
<strong>of</strong> evidence support that V617F may arise on a pre-existing<br />
abnormal clonal substrate: 1) in some patients, the V617F<br />
burden is relatively small compared with the proportion <strong>of</strong><br />
cells with a coexistent clonal karyotypic abnormality; and<br />
2) in AML arising from a V617F-positive MPN, the mutant<br />
V617F allele can frequently no longer be detected. 18 This<br />
suggests that the MPN and AML share a clonal origin that<br />
likely preceded the acquisition <strong>of</strong> V617F. Furthermore, in<br />
V617F-positive PV and ET patients, both JAK2 wild-type<br />
and V617F-positive EECs have been detected within the<br />
same patient, suggesting that a separate event may confer<br />
clonal, erythropoietin-independent growth, before V617F.<br />
One study in ET patients demonstrated that V617F was<br />
acquired on separate alleles as multiple, independent<br />
events. 19 Therefore, at least for ET, the V617F mutation<br />
does not necessarily confer clonal dominance, permitting<br />
JAK2 V617F-negative cells to continue to proliferate and<br />
subsequently acquire the V617F mutation independently.<br />
Activated JAK-STAT Signaling without JAK2 V617F<br />
In the absence <strong>of</strong> V617F, activation <strong>of</strong> JAK-STAT signaling<br />
can be demonstrated in some MPN patients, suggesting<br />
that molecular alterations that interdigitate with this axis<br />
may contribute to MPN pathogenesis. This paradigm is<br />
413
exemplified by mutations in exon 12 <strong>of</strong> the JAK2 gene, the<br />
receptor for thrombopoietin (MPL), CBL, and LNK.<br />
Exon 12 JAK2<br />
A combination <strong>of</strong> missense, insertion, or deletion mutations<br />
in exon 12 <strong>of</strong> JAK2 was first described in 10 patients<br />
initially diagnosed with idiopathic erythrocytosis. 20 The<br />
identification <strong>of</strong> these mutations operationally redefine<br />
these patients as having PV. These mutations affect a region<br />
5� <strong>of</strong> the pseudokinase domain, spanning residues 536 to<br />
547. In contrast to the singular V617F mutation in exon 14,<br />
a review published in 2011 catalogued 37 different exon 12<br />
mutations, including 11 nonsynonymous substitutions, 20<br />
deletion variants, and six duplications. 21 Similar to V617F,<br />
the residues affected by exon 12 mutations are located at the<br />
interface <strong>of</strong> the pseudokinase and kinase domains and are<br />
postulated to cause a structural change resulting in JAK2<br />
activation. This notion is supported by cell line experiments,<br />
in which each <strong>of</strong> the mutant alleles conferred IL-3 independent<br />
growth and constitutive phosphorylation <strong>of</strong> JAK2. 29<br />
In addition, in a murine retroviral transplant assay, one <strong>of</strong><br />
the mutant alleles (K539L) caused a pronounced erythrocytosis<br />
and growth <strong>of</strong> Epo-independent erythroid colonies. 20<br />
In contrast to the V617F mutation, exon 12 mutations<br />
appear to be restricted to PV and account for an aggregate<br />
3% mutation frequency among 11 published PV cohorts. 21<br />
MPL<br />
Table 2. Estimated Frequencies <strong>of</strong> Non-JAK2 V617F Gene Mutations in PV, ET, MF, and Post-MPN AML<br />
Gene Chromosome Location PV ET PMF Post-MPN AML<br />
Exon 12 JAK2 9p24 �1–3% Rare Rare Not Reported<br />
MPL 1p34 Rare �1–5% �5–10% Not Reported<br />
CBL 11q23 Rare Rare �5–10% Rare/Reported<br />
LNK 12q24 Rare/Reported �5% �5% �10%<br />
TET2 4q24 �7–16% �4–11% �8–17% �20%<br />
DNMT3A 2p23 �7% �3% �7–15% �17%<br />
IDH1/IDH2 2q33.3/15q26.1 �2% �1% �4% �22%<br />
EZH2 7q35 �3–5% Rare �6–13% Not Reported<br />
ASXL1 20q11.21 �2–5% �5% �13–23% �20%<br />
IKZF1 7p12 Rare Rare Rare �21%<br />
TP53 17p13.1 �3% in chronic phase MPN �27%<br />
Abbreviations: PV, polycythemia vera; ET, essential thrombocythemia; MF, myel<strong>of</strong>ibrosis; MPN, myeloproliferative neoplasms; AML, acute myeloid leukemia.<br />
Mutations <strong>of</strong> MPL, the gene encoding the thrombopoetin<br />
receptor, have been identified in approximately 5% to 10% <strong>of</strong><br />
PMF patients, approximately1% to 5% <strong>of</strong> patients with ET,<br />
and rarely in PV. 22,23 The two most common amino acid<br />
substitutions at codon 515 result in a tryptophan to leucine<br />
(W515L) or lysine (W515K) change, with rare asparagine<br />
and alanine variants having been reported. Although originally<br />
described in familial ET cohorts, acquired MPL S505N<br />
mutations have also been described. All MPL gene mutations<br />
reside in exon 10, which comprises the juxtamembrane,<br />
intracytosolic portion <strong>of</strong> the protein important for<br />
preventing spontaneous receptor activation. Overexpression<br />
<strong>of</strong> the mutant MPL W515L allele in cell lines results in<br />
cytokine-independent growth, TPO hypersensitivity, and<br />
activated JAK-STAT signaling. 22 In a murine retroviral<br />
transplant model, the W515L allele produced a phenotype <strong>of</strong><br />
marked thrombocytosis, splenomegaly, and reticulin fibrosis,<br />
but not erythrocytosis.<br />
414<br />
NGUYEN AND GOTLIB<br />
CBL<br />
Casitas B-lineage lymphoma (CBL) proteins have dual<br />
roles as multifunctional adaptor proteins, which recruit<br />
components to downstream signaling pathways, and as E3<br />
ubiquitin ligases, which are involved in the trafficking and<br />
degradation <strong>of</strong> activated tyrosine kinases. Three mammalian<br />
CBL homologs exist: c-CBL, CBL-b, and CBL-c. c-CBL<br />
consists <strong>of</strong> a tyrosine kinase binding (TKB) domain, a linker<br />
domain, a RING finger domain (RFD), a proline-rich region,<br />
and a ubiquitin-associated (UBA) domain overlapping with<br />
a leucine zipper (LZ) motif. Several groups identified recurrent<br />
acquired uniparental disomy (UPD) at 11q (which<br />
encompasses c-CBL) in a wide spectrum <strong>of</strong> myeloid malignancies,<br />
and the majority <strong>of</strong> these cases were found to<br />
harbor mutations in c-CBL. 24,25 The highest frequency <strong>of</strong><br />
c-CBL mutations have been identified in juvenile myelomonocytic<br />
leukemia (JMML) (approximately 20%), 26 but are<br />
also found in a low proportion <strong>of</strong> patients with MPN (e.g.,<br />
10% <strong>of</strong> MF), MDS/MPN overlap conditions besides JMML,<br />
and secondary AML transformed from MDS or MPN. 27 The<br />
majority <strong>of</strong> mutations in c-CBL are missense mutations<br />
localizing to the linker or RFD domains that result in loss <strong>of</strong><br />
ubiquitin ligase activity. The high frequency <strong>of</strong> homozygous<br />
mutations findings are consistent with the notion that<br />
c-CBL is a tumor suppressor and that loss <strong>of</strong> c-CBL function<br />
leads to hypersensitivity to cytokines via dysregulation <strong>of</strong><br />
cytokine receptor-mediated signaling.<br />
LNK<br />
LNK (SH2B3) belongs to a family <strong>of</strong> adaptor proteins (e.g.,<br />
also SH2-B, APS) that share several structural motifs,<br />
including a proline-rich N-terminus, a pleckstrin homology<br />
(PH) domain, an SH2 domain, and a conserved tyrosine<br />
residue near the C-terminus. LNK binds to MPL via its SH2<br />
domain and colocalizes to the plasma membrane via its PH<br />
domain. On cytokine stimulation with Tpo, LNK binds<br />
strongly to JAK2 and inhibits downstream STAT activation,<br />
thereby providing critical negative feedback regulation.<br />
LNK �/� mice exhibit a phenotype similar to human MF,<br />
including leukocytosis and thrombocytosis, as well as<br />
splenomegaly with marked fibrosis and extramedullary hematopoiesis.<br />
28 An initial study <strong>of</strong> 33 V617F-negative ET and<br />
PMF patients identified two individuals (6%) with mutations<br />
in exon 2 <strong>of</strong> LNK. 29 One patient with PMF exhibited a<br />
5 base-pair deletion and missense mutation leading to a<br />
premature stop codon and loss <strong>of</strong> the pleckstrin homology<br />
(PH) and SH2 domains. A second patient with ET had a
GENETICS OF MYELOPROLIFERATIVE NEOPLASMS<br />
missense mutation (E208Q) in the PH domain. BaF3-MPL<br />
cells transduced with these LNK mutants displayed augmented<br />
and sustained thrombopoietin-dependent growth<br />
and signaling and primary samples from these patients<br />
exhibited aberrant JAK-STAT activation. Follow-up studies<br />
indicate a low frequency <strong>of</strong> LNK mutations in the chronic<br />
phase <strong>of</strong> MPN (�5%), increasing to approximately 10% with<br />
transformation to AML. 30 Mutations have also been found<br />
in a few patients with idiopathic erythrocytosis/PV. 31 Most<br />
mutations cluster in an exon 2 ‘hot spot,’ which are expected<br />
to disrupt the PH domain, but frameshift, missense, and<br />
nonsense mutations in exons 5, 7, and 8 have now been<br />
reported, and may coexist with other MPN mutations in the<br />
same patient. Taken together, these findings indicate that<br />
JAK-STAT activation caused by loss <strong>of</strong> LNK negative feedback<br />
regulation can phenocopy MPN disease.<br />
Mutations in Components <strong>of</strong> the Epigenetic Machinery<br />
TET2<br />
SNP and CGH arrays identified acquired LOH on chromosome<br />
4q in patients with myeloid neoplasms, with further<br />
mapping defining the TET oncogene family member 2<br />
(TET2) gene as a mutated locus. Somatic mutations in TET2<br />
are a mixture <strong>of</strong> deletions, frameshifts, stop codons, or<br />
conserved amino acid substitutions and occur at a patient<br />
frequency <strong>of</strong> 7% to 16% in PV, 4% to 11% in ET, and 8% to<br />
17% in PMF. 32,33 The biologic and biochemical roles <strong>of</strong> TET<br />
family proteins and the functional consequences <strong>of</strong> TET2<br />
mutations are becoming better clarified. TET proteins are<br />
�-ketoglutarate-dependent enzymes that catalyze the conversion<br />
<strong>of</strong> 5-methylcytosine (5mC) to 5-hydroxymethylcytosine<br />
(5hmC). 34 As DNA methylation is known to be an<br />
important modulator <strong>of</strong> gene expression, it has been speculated<br />
that the conversion <strong>of</strong> 5mC to 5hmC may alter chromatin<br />
structure and restrict access to DNA methyltransferases,<br />
which mediate repression <strong>of</strong> gene transcription. It follows<br />
that loss <strong>of</strong> TET enzyme activity results in increased methylation,<br />
and in a sample <strong>of</strong> patients with TET2 mutations,<br />
a reduction in levels <strong>of</strong> 5hmc levels is demonstrated. 35 In<br />
various TET2 knockout murine models, loss <strong>of</strong> TET2 results<br />
in a myeloproliferative phenotype (e.g., splenomegaly, extramedullary<br />
hematopoiesis) with expansion <strong>of</strong> the hematopoietic<br />
stem cell compartment. 36 Expression <strong>of</strong> mutant<br />
TET2 in early hematopoietic precursors skews hematopoiesis<br />
in favor <strong>of</strong> myeloid over lymphoid progenitors. 37<br />
DNMT3A<br />
Whole-genome sequencing <strong>of</strong> patients with AML revealed<br />
recurrent mutations in 22% <strong>of</strong> AML patients in the gene for<br />
DNA methyltransferase 3A (DNMT3A), another epigenetic<br />
regulator. 38 In this study, DNMT3A mutations were associated<br />
with poor outcome. In one analysis <strong>of</strong> patients with<br />
MPN, 12 DNMT3A variants were found in 115 patients<br />
(10%). 39 Mutations were most frequently detected in patients<br />
during blastic phase transformation (6/35, 17%) and<br />
MF (3/20, 15%), followed by PV (2/30, 7%) and ET (1/30, 3%).<br />
In another study <strong>of</strong> 94 patients (46 PMF, 22 post-PV/ET MF,<br />
11 blast-phase MPN, and 15 CMML), only three DNMT3A<br />
mutations were found among 46 PMF patients (7%). 40 In<br />
both studies, mutations were always heterozygous, and in<br />
some cases, coexisting JAK2 V617F, TET2, ASXL1, and IDH<br />
1/2 mutations were found. Although DNMT3A mutations<br />
are ubiquitous among myeloid neoplasms, their mechanistic<br />
consequences are less clear. Using a conditional ablation<br />
model in mice, it was found that loss <strong>of</strong> DNMT3A results in<br />
impairment <strong>of</strong> hematopoietic stem cell (HSC) differentiation<br />
and expansion <strong>of</strong> HSC with serial transplantation. 41 However,<br />
analysis <strong>of</strong> HSC from DNMT3A-null mice showed both<br />
increased and decreased methylation at specific loci. In the<br />
study <strong>of</strong> AML patients, global methylation patterns and<br />
5-methylcytosine content in genomes were not significantly<br />
altered in patients with DNMT3A mutations. The potential<br />
cooperative effects <strong>of</strong> DNMT3A with other epigenetic modifiers,<br />
and the gene’s role in disease transformation requires<br />
further study.<br />
IDH 1/2<br />
The genes IDH1 and IDH2 encode enzymes that catalyze<br />
oxidative decarboxylation <strong>of</strong> isocitrate to �-ketoglutarate.<br />
Mutant IDH has decreased affinity to isocitrate, but displays<br />
neomorphic catalytic activity toward �-ketoglutarate,<br />
resulting in accumulation <strong>of</strong> 2-hydroxyglutarate. IDH1 and<br />
IDH2 mutations were first reported in gliomas/glioblastomas<br />
and AML. In a multi-institutional cohort <strong>of</strong> 1,473<br />
MPN patients, 38 IDH mutations were found, and ranged in<br />
frequency from 0.8% to 4% among chronic phase PV, ET, and<br />
PMF patients (highest in the latter), and significantly increased<br />
to 21% in blast phase MPN. 42 This study corroborates<br />
the results <strong>of</strong> smaller studies, which documented a<br />
higher incidence <strong>of</strong> IDH 1/2 mutations in the leukemic phase<br />
<strong>of</strong> MPNs. In a multivariate analysis, the presence <strong>of</strong> IDH<br />
mutations predicted a worse survival in a subgroup <strong>of</strong> 43<br />
blastic phase MPN patients from the Mayo Clinic. 42<br />
It is noteworthy that TET2 activity is impaired in cells<br />
with mutated IDH 1/2 because its activity depends on<br />
a-ketoglutarate. 43 Mutual exclusivity <strong>of</strong> IDH 1/2 and TET2<br />
mutations has been observed in AML and was speculated<br />
to extend to MPNs as well. However, in the aforementioned<br />
study, IDH mutational frequencies were similar among<br />
patients with JAK2 (3.6%), MPL (4.3%) and TET2 (3.2%)<br />
mutations. 42 Given the coexistence <strong>of</strong> these molecular derangements,<br />
the specific contribution <strong>of</strong> IDH 1/2 mutations<br />
to epigenetic dysregulation, and MPN biology in general,<br />
becomes more difficult to dissect.<br />
Mutations <strong>of</strong> the Polycomb Repressive Complex (PRC2)<br />
The polycomb repressive complex (PRC2) is the major<br />
methyltransferase for histone H3K27 methylation, one type<br />
<strong>of</strong> post-translational histone modification involved in regulation<br />
<strong>of</strong> gene transcription. A low frequency <strong>of</strong> mutations<br />
in the PRC2 enzymatic component EZH2, noncatalytic components<br />
EED and SUZ12, and PRC2-associated protein<br />
JARID2 have been described in either MDS, MPNs, or<br />
MDS/MPNs overlap disorders. 44-46 Acquired uniparental<br />
disomy on chromosome 7q led to identification <strong>of</strong> EZH2<br />
mutations, which are either heterozygous or homozygous<br />
and result in loss-<strong>of</strong>-function. 44 To date, a well-defined<br />
picture <strong>of</strong> the effect <strong>of</strong> PRC2 complex-related mutations on<br />
myeloid pathobiology has not yet emerged. The prognostic<br />
relevance <strong>of</strong> EZH2 mutations was assessed in a survey <strong>of</strong><br />
370 patients with PMF and 148 patients with post PV/ET-<br />
MF. Twenty-five different EZH2 mutations were found in<br />
5.9% <strong>of</strong> PMF, 1.2% <strong>of</strong> post PV-MF, and 9.4% <strong>of</strong> post-ET MF<br />
patients. Leukemia-free and overall survival was significantly<br />
reduced in patients with EZH2-mutated PMF. 47<br />
415
ASXL1 and L3MBTL1<br />
ASXL1 belongs to the Enhancer <strong>of</strong> trithorax and Polycomb<br />
gene family. Its biologic roles are not well understood;<br />
however, one epigenetic function ascribed to the Drosophilia<br />
homolog is modification <strong>of</strong> chromatin complexes, including<br />
ubiquitination <strong>of</strong> histone H2A lysine 119. 48 Similar to other<br />
epigenetic modifiers, mutation <strong>of</strong> ASXL1 (primarily exon 12)<br />
is found in a wide spectrum <strong>of</strong> myeloid malignancies, but<br />
among the MPNs, seems to be preferentially associated with<br />
MF at a rate up to 13% to 23%. 45<br />
Another polycomb family member, L3MBTL1, is the human<br />
homolog <strong>of</strong> the Drosophila lethal, 3 malignant brain<br />
tumor (L3MBT) that functions as a tumor suppressor. It is<br />
located on the long arm <strong>of</strong> chromosome 20 (q12), within a<br />
region commonly deleted in several myeloid malignancies<br />
and is a candidate tumor suppressor gene. Depletion <strong>of</strong><br />
L3MBTL1 from human cells causes replicative stress, DNA<br />
breaks, activation <strong>of</strong> the DNA damage response, and<br />
genomic instability. 49 In addition to being involved in histone<br />
H4 methylation, 50 it has been found that haploinsufficiency<br />
for L3MBTL1 promotes erythroid differentiation,<br />
suggesting a role in PV development. 51<br />
Extra-TK Functions <strong>of</strong> JAK2 V617F<br />
Recently, the unexpected finding was made that normal<br />
JAK2 can translocate to the nucleus. Among its nuclear<br />
roles, JAK2 can phosphorylate histone H3Y41, releasing the<br />
transcriptional repressor HP1alfa from chromatin. 52 Exclusion<br />
<strong>of</strong> HP1alfa resulted in increased gene transcription and<br />
elevated expression <strong>of</strong> the oncogene LMO2 in leukemia<br />
cells, which was reversed with inhibition <strong>of</strong> JAK2. It still<br />
remains to be determined how JAK2’s effects on chromatin<br />
structure translate into expression <strong>of</strong> specific genes and<br />
promotion <strong>of</strong> oncogenesis in certain cellular or disease contexts.<br />
A nucleus-associated gain-<strong>of</strong>-function <strong>of</strong> JAK2 V617F,<br />
separate from its role in activated JAK-STAT signaling, is<br />
its capacity to phosphorylate the protein arginine methyltransferase<br />
5 (PRMT5). 53 This modification reduces PRMT5’s<br />
ability to methylate histones H2A and H4. These chromatin<br />
changes were modeled with knockdown <strong>of</strong> PRMT5 by a short<br />
hairpin RNA in CD34-positive cells, resulting in erythroid<br />
differentiation and increased colony formation. These data<br />
suggest that the inhibitory effects <strong>of</strong> JAK2 V617F on PRMT5<br />
activity may be relevant to MPN pathobiology.<br />
Other Genetic Mutations and Pathway<br />
Alterations in MPNs<br />
Hemizygous deletions encompassing all or part <strong>of</strong> the gene<br />
encoding the Ikaros (IKZF1) transcription factor on chromosome<br />
7 are associated with MPN transformation. 54 Modeling<br />
<strong>of</strong> IKZF1 haploinsufficiency in mice progenitors with SiRNA<br />
knockdown <strong>of</strong> IKZF1 led to an increase in cytokinedependent<br />
growth and STAT5 activation. In another study<br />
<strong>of</strong> post-MPN AML, P53 mutations were found at a frequency<br />
<strong>of</strong> 27%, including independent bi-allelic abnormalities and<br />
homozygous mutations caused by acquired uniparental disomy<br />
<strong>of</strong> chromosome 17p. 55 Other genes mutated in the<br />
leukemic phase <strong>of</strong> MPNs include NRAS and RUNX1. The<br />
role <strong>of</strong> all <strong>of</strong> these genes in disease initiation compared to<br />
MPN transformation remains to be clarified.<br />
416<br />
NGUYEN AND GOTLIB<br />
Genetic Instability and Apoptosis Resistance<br />
Increased DNA damage and evasion <strong>of</strong> apoptosis may<br />
also play a role in MPN pathogenesis. Expression <strong>of</strong> the<br />
antiapoptotic protein Bcl-xL is increased in PV erythroid<br />
progenitors, suggesting that this is one possible mechanism<br />
<strong>of</strong> resistance to apoptosis that normally occurs on withdrawal<br />
<strong>of</strong> Epo. 56 In cell line experiments, overexpression <strong>of</strong><br />
V617F induced an increase in homologous recombination<br />
and genetic instability. 57 Furthermore, a marked increase in<br />
homologous recombination was found in CD34 positive progenitors<br />
from patients with PV and PMF when compared<br />
with controls. Nonenzymatic deamidation <strong>of</strong> Bcl-xL is a key<br />
step in the induction <strong>of</strong> apoptosis in response to DNA<br />
damage. This pathway is inhibited in both CML and PV. 58<br />
JAK2 inhibitors restored the Bcl-xL deamidation pathway<br />
in primary samples from PV patients, suggesting that activated<br />
signaling mediated by V617F may lead to resistance to<br />
apoptosis.<br />
Hypermethylation and Phosphorylation <strong>of</strong> SOCS Family Members<br />
The SOCS proteins are SH2-domain containing proteins<br />
that function as key negative regulators <strong>of</strong> JAK-STAT signaling.<br />
SOCS1 and SOCS3 can bind to the catalytic domain<br />
<strong>of</strong> JAK2 and inhibit its kinase activity, as well as target<br />
JAK2 for ubiquitination and subsequent degradation. Recent<br />
studies have suggested that hypermethylation <strong>of</strong> SOCS<br />
genes may also play a role in MPN pathogenesis. Hypermethylation<br />
<strong>of</strong> SOCS3, and to a lesser extent, SOCS1, has<br />
been observed in either V617F or wild type JAK2 chronic<br />
and blast phase MPN patient samples. 59<br />
Phosphorylation <strong>of</strong> SOCS proteins leads to enhanced<br />
SOCS protein degradation, and this process may be altered<br />
in MPNs. In addition to its previously highlighted gain-<strong>of</strong>functions,<br />
activated JAK2 V617F has also been shown to<br />
overcome normal SOCS regulation by hyper-phosphorylating<br />
SOCS3, thus effectively blocking its inhibitory<br />
activity and perhaps even potentiating the proliferative<br />
capacity <strong>of</strong> the mutant V617F allele. 60 To date, somatic<br />
mutations in SOCS genes in MPNs have not been demonstrated.<br />
Inherited Susceptibility to MPNs<br />
A five- to seven-fold increased risk <strong>of</strong> MPN development is<br />
found in first-degree relatives <strong>of</strong> patients with MPN. Three<br />
studies demonstrated that a germline haplotype (GGCC,<br />
referred to as ‘46/1’) encompassing the 3� region <strong>of</strong> JAK2<br />
gene is associated with a three- to four-fold risk <strong>of</strong> developing<br />
a V617F-positive (as well as MPL-mutated) MPN. 61-63<br />
Patients who were heterozygous for this haplotype preferentially<br />
acquired the V617F mutation in cis with the predisposition<br />
allele, suggesting that the haplotype may lead to<br />
hypermutability at the JAK2 locus. However, the haplotype<br />
was also weakly associated with JAK2 V617F-negative<br />
MPNs, suggesting that it may confer a more generalized<br />
propensity for MPN development independent <strong>of</strong> JAK2<br />
V617F. One possibility is that the germline haplotype may<br />
result in a functional difference such that cells with the risk<br />
haplotype gain a selective advantage. However, evidence to<br />
support this notion has been lacking. No differences in JAK2<br />
expression level or nonsynonymous JAK2 coding polymorphisms<br />
associated with this haplotype have been identified.<br />
Given the low penetrance <strong>of</strong> the haplotype, and a lack <strong>of</strong><br />
correlation with specific-disease related features, testing for
GENETICS OF MYELOPROLIFERATIVE NEOPLASMS<br />
the 46/1 haplotype in first-degree relatives is not recommended.<br />
Recently, Mendelian inheritance <strong>of</strong> non-V617F<br />
germline activating JAK2 variants, V617I and R564Q, was<br />
identified in two families with thrombocytosis. 64,65<br />
<strong>Clinical</strong> Implications <strong>of</strong> the Molecular<br />
Genetics <strong>of</strong> MPNs<br />
Despite the burgeoning genetic data in MPNs, their clinical<br />
utility has thus far been limited. The evaluation <strong>of</strong> an<br />
erythrocytosis, leukocytosis, or thrombocytosis begins with<br />
a differential diagnosis between a reactive condition and<br />
primary bone marrow disorder. The identification <strong>of</strong> an<br />
activating mutation such as V617F, exon 12 JAK2 or MPL,<br />
or clonal cytogenetic abnormality, identifies a myeloid neoplasm.<br />
However, the WHO classification requires a combination<br />
<strong>of</strong> clinical, laboratory, and histopathologic data in<br />
order to diagnose a specific MPN. Despite individual reports<br />
finding statistical significance between the presence (or<br />
allelic burden) <strong>of</strong> certain molecular abnormalities in myel<strong>of</strong>ibrosis,<br />
such data have not been prognostically validated<br />
for overall and leukemia-free survival in the current scoring<br />
systems (e.g., DIPSS-Plus). In addition, risk stratification in<br />
PV and ET is still guided primarily by age and prior history<br />
<strong>of</strong> thrombosis. At this time, this new molecular information<br />
also has minimal influence on treatment decisions. In MF,<br />
JAK inhibitors demonstrate activity in patients with either<br />
wild-type or mutant JAK2, and trial participation is not<br />
dependent on JAK2 mutation status.<br />
Authors’ Disclosures <strong>of</strong> Potential Conflicts <strong>of</strong> Interest<br />
Employment or<br />
Leadership<br />
Positions<br />
Consultant or<br />
Advisory Role<br />
Author<br />
Huong (Marie) Nguyen*<br />
Jason Gotlib Gilead Sciences;<br />
Incyte; Novartis;<br />
YM BioSciences<br />
*No relevant relationships to disclose.<br />
1. Bain BJ, Gilliland DG, Horny H-P, et al. ‘Myeloproliferative Neoplasms’<br />
and ‘Myeloid and lymphoid neoplasms with eosinophilia and abnormalities <strong>of</strong><br />
PDGFRA, PDGFRB, or FGFR1’. In: Swerdlow S, Harris NL, Stein H, Jaffe<br />
ES, Theile J, Vardiman JW (eds). World Health Organization Classification <strong>of</strong><br />
Tumours. Pathology and Genetics <strong>of</strong> Tumours <strong>of</strong> Haematopoietic and Lymphoid<br />
Tissues. Lyon, France: IARC Press, 2008:30-73.<br />
2. Dameshek W. Some speculations on the myeloproliferative syndromes.<br />
Blood. 1951;6:372-375.<br />
3. Adamson JW, Fialkow PJ, Murphy S, et al. Polycythemia vera: Stem-cell<br />
and probable clonal origin <strong>of</strong> the disease. N Engl J Med. 1976;295:913-916.<br />
4. Prchal JF, Axelrad AA. Bone-marrow responses in polycythemia vera.<br />
N Engl J Med. 1974;290:1382.<br />
5. Saharinen P, Silvennoinen O. The pseudokinase domain is required for<br />
suppression <strong>of</strong> basal activity <strong>of</strong> Jak2 and Jak3 tyrosine kinases and for<br />
cytokine-inducible activation <strong>of</strong> signal transduction. J Biol Chem. 2002;277:<br />
47954-47963.<br />
6. Baxter EJ, Scott LM, Campbell PJ, et al. Acquired mutation <strong>of</strong> the<br />
tyrosine kinase JAK2 in human myeloproliferative disorders. Lancet. 2005;<br />
365:1054-1061.<br />
7. Levine R, Wadleigh M, Cools J, et al. Activating mutation in the tyrosine<br />
kinase JAK2 in polycythemia vera, essential thrombocythemia, and myeloid<br />
metaplasia with myel<strong>of</strong>ibrosis. Cancer Cell. 2005;7:387-397.<br />
8. Kralovics R, Passamonti F, Buser AS, et al. A gain-<strong>of</strong>-function mutation<br />
<strong>of</strong> JAK2 in myeloproliferative disorders. N Engl J Med. 2005;352:1779-1790.<br />
9. James C, Ugo V, Le Couedic JP, et al. A unique clonal JAK2 mutation<br />
Conclusion and Future Directions<br />
Although mutations such as V617F and MPL are alone<br />
sufficient to induce myeloproliferative disease in mice, MPN<br />
initiation and progression in humans is genetically more<br />
complex. Current whole genome and exome sequencing<br />
approaches will be useful for unearthing novel molecular<br />
determinants <strong>of</strong> MPN and to further evaluate specific genetic<br />
loci derived from genome-wide association studies. In<br />
the current genomic era, evaluation <strong>of</strong> panels <strong>of</strong> acquired<br />
somatic variants, which are linked to patient outcome data,<br />
will help establish whether particular combinations <strong>of</strong> mutations<br />
generate additional prognostic risk information. The<br />
recognition that both genetic and epigenetic dysregulation<br />
underlie MPN disease justifies the use <strong>of</strong> not only JAK<br />
inhibitors, but also DNA/chromatin-modifying agents such<br />
as hypomethylating drugs and histone deacetylase inhibitors.<br />
An enhanced understanding <strong>of</strong> the molecular genetic<br />
pr<strong>of</strong>ile <strong>of</strong> the leukemic stem cells from which MPNs originate<br />
is critical to developing therapies that eradicate minimal<br />
residual disease and effect cure.<br />
Acknowledgments<br />
The authors wish to thank members <strong>of</strong> the Stanford Division<br />
<strong>of</strong> Hematology and the International Working Group for Myeloproliferative<br />
Neoplasms Research and Treatment (IWG-<br />
MRT) for their support. Because <strong>of</strong> space restrictions, we<br />
apologize to investigators whose work was not included in this<br />
review.<br />
Stock<br />
Ownership Honoraria<br />
REFERENCES<br />
Research<br />
Funding<br />
Gilead Sciences;<br />
Incyte; Infinity;<br />
Novartis; San<strong>of</strong>i;<br />
YM BioSciences<br />
Expert<br />
Testimony<br />
Other<br />
Remuneration<br />
leading to constitutive signalling causes polycythaemia vera. Nature. 2005;<br />
434:1144-1148.<br />
10. Yan D, and Mohi G. Critical Role for Stat5 in the Initiation and<br />
Maintenance <strong>of</strong> Polycythemia Vera in a Jak2V617F Knock-In Mouse Model.<br />
Blood (ASH Annual Meeting Abstracts). 2011;118:121.<br />
11. Lacout C, Pisani DF, Tulliez M, et al. JAK2V617F expression in murine<br />
hematopoietic cells leads to MPD mimicking human PV with secondary<br />
myel<strong>of</strong>ibrosis. Blood. 2006;108:1652-1660.<br />
12. Xing S, Wanting T, Zhao W, et al. Transgenic expression <strong>of</strong> JAK2V617F<br />
causes myeloproliferative disorders in mice. Blood. 2008;111:5109-5117.<br />
13. Chen E, Beer PA, Godfrey AL, et al. Distinct clinical phenotypes<br />
associated with JAK2V617F reflect differential STAT1 signaling. Cancer Cell.<br />
2010;18:524-535.<br />
14. Delhommeau F, Dupont S, Tonetti C, et al. Evidence that the JAK2<br />
G1849T (V617F) mutation occurs in a lymphomyeloid progenitor in polycythemia<br />
vera and idiopathic myel<strong>of</strong>ibrosis. Blood. 2007;109:71-77.<br />
15. Jamieson CH, Gotlib J, Durocher JA, et al. The JAK2 V617F mutation<br />
occurs in hematopoietic stem cells in polycythemia vera and predisposes<br />
toward erythroid differentiation. Proc Natl Acad Sci USA.2006;103:6224-<br />
6229.<br />
16. Wernig G, Mercher T, Okabe R, et al. Expression <strong>of</strong> Jak2V617F causes<br />
a polycythemia vera-like disease with associated myel<strong>of</strong>ibrosis in a murine<br />
bone marrow transplant model. Blood. 2006;107:4274-4281.<br />
17. Bumm T, Elsea C, Corbin A, et al. Characterization <strong>of</strong> murine<br />
JAK2V617F-positive myeloproliferative disease. Cancer Res. 2006;66:11156-<br />
11165.<br />
417
18. Theocharides A, Boissinot M, Girodon F, et al. Leukemic blasts in<br />
transformed JAK2-V617F-positive myeloproliferative disorders are frequently<br />
negative for the JAK2-V617F mutation. Blood. 2007;110:375-379.<br />
19. Lambert JR, Everington T, Linch DC, et al. In essential thrombocythemia,<br />
multiple JAK2 V617F clones are present in most mutant-positive<br />
patients: A new disease paradigm. Blood. 2009;114:3018-3023.<br />
20. Scott LM, Tong W, Levine R, et al. JAK2 exon 12 mutations in<br />
polycythemia vera and idiopathic erythrocytosis. N Engl J Med. 2007;356:<br />
459-468.<br />
21. Scott LM. The JAK2 exon 12 mutations: A comprehensive review. Am J<br />
Hematol. 2011;86:668-676.<br />
22. Pikman Y, Lee BH, Mercher T, et al. MPLW515L is a novel somatic<br />
activating mutation in myel<strong>of</strong>ibrosis with myeloid metaplasia. PLoS Med.<br />
2006;3:e270.<br />
23. Pardanani A, Levine R, Lasho T, et al. MPL515 mutations in myeloproliferative<br />
and other myeloid disorders: A study <strong>of</strong> 1182 patients. Blood.<br />
2006;108:3472-3476.<br />
24. Dunbar AJ, Gondek LP, O’Keefe CL, et al. 250K single nucleotide<br />
polymorphism array karyotyping identifies acquired uniparental disomy and<br />
homozygous mutations, including novel missense substitutions <strong>of</strong> c-Cbl, in<br />
myeloid malignancies. Cancer Res. 2008;68:10349-10357.<br />
25. Grand FH, Hidalgo-Curtis CE, Ernst T, et al. Frequent CBL mutations<br />
associated with 11q acquired uniparental disomy in myeloproliferative neoplasms.<br />
Blood. 2009;113:6182-6192.<br />
26. Loh ML, Sakai DS, Flotho C, et al. Mutations in CBL occur frequently<br />
in juvenile myelomonocytic leukemia. Blood. 2009;114:1859-1863.<br />
27. Sanada M, Suzuki T, Shih LY, et al. Gain-<strong>of</strong>-function <strong>of</strong> mutated C-CBL<br />
tumour suppressor in myeloid neoplasms. Nature. 2009;460:904-908.<br />
28. Velazquez L, Cheng AM, Fleming HE, et al. Cytokine signaling and<br />
hematopoietic homeostasis are disrupted in Lnk-deficient mice. J Exp Med.<br />
2002;195:1599-1611.<br />
29. Oh ST, Simonds EF, Jones C, et al. Novel mutations in the inhibitory<br />
adaptor protein LNK drive JAK-STAT signaling in patients with myeloproliferative<br />
neoplasms. Blood. 2010;116:988-992.<br />
30. Pardanani A, Lasho T, Finke C, et al. LNK mutation studies in<br />
blast-phase myeloproliferative neoplasms, and in chronic-phase disease with<br />
TET2, IDH, JAK2 or MPL mutations. Leukemia. 2010;24:1713-1718.<br />
31. Lasho TL, Pardanani A, Tefferi A. LNK mutations in JAK2 mutationnegative<br />
erythrocytosis. N Engl J Med. 2010;363:1189-1190.<br />
32. Delhommeau F, Dupont S, Della Valle V, et al. Mutation in TET2 in<br />
myeloid cancers. N Engl J Med. 2009;360:2289-2301.<br />
33. Abdel-Wahab O, Mullally A, Hedvat C, et al. Genetic characterization<br />
<strong>of</strong> TET1, TET2, and TET3 alterations in myeloid malignancies. Blood.<br />
2009;114:144-147.<br />
34. Tahiliani M, Koh KP, Shen Y, et al. Conversion <strong>of</strong> 5-methylcytosine to<br />
5-hydroxymethylcytosine in mammalian DNA by MLL partner TET1. Science.<br />
2009;324:930-935.<br />
35. Figueroa ME, Abdel-Wahab O, Lu C, et al. Leukemic IDH1 and IDH2<br />
mutations result in a hypermethylation phenotype, disrupt TET2 function,<br />
and impair hematopoietic differentiation. Cancer Cell. 2010;18:553-567.<br />
36. Li Z, Cai X, Cai C, et al. Deletion <strong>of</strong> Tet2 in mice leads to dysregulated<br />
hematopoietic stem cells and subsequent development <strong>of</strong> myeloid malignancies.<br />
Blood. 2011;118:4509-4518.<br />
37. Moran-Crusio K, Reavie L, Shih A, et al. Tet2 loss leads to increased<br />
hematopoietic stem cell self-renewal and myeloid transformation. Cancer<br />
Cell. 2011;20:11-24.<br />
38. Ley TJ, Ding L, Walter MJ, et al. DNMT3A mutations in acute myeloid<br />
leukemia. N Engl J Med. 2010;363:2424-2433.<br />
39. Stegelmann F, Bullinger L, Schlenk RF, et al. DNMT3A mutations in<br />
myeloproliferative neoplasms. Leukemia. 2011;25:1217-1219.<br />
40. Abdel-Wahab O, Pardanani A, Rampal R, et al. DNMT3A mutational<br />
analysis in primary myel<strong>of</strong>ibrosis, chronic myelomonocytic leukemia and<br />
advanced phases <strong>of</strong> myeloproliferative neoplasms. Leukemia. 2011;25:1219-<br />
1220.<br />
41. Challen GA, Sun D, Jeong M, et al. Dnmt3a is essential for hematopoietic<br />
stem cell differentiation. Nat Genet. 2011;44:23-31.<br />
42. Tefferi A, Lasho TL, Abdel-Wahab O, et al. IDH1 and IDH2 mutation<br />
studies in 1473 patients with chronic-, fibrotic- or blast-phase essential<br />
418<br />
NGUYEN AND GOTLIB<br />
thrombocythemia, polycythemia vera or myel<strong>of</strong>ibrosis. Leukemia. 2010;24:<br />
1302-1309.<br />
43. Xu W, Yang H, Liu Y, et al. Oncometabolite 2-hydroxyglutarate is a<br />
competitive inhibitor <strong>of</strong> alpha-ketoglutarate-dependent dioxygenases. Cancer<br />
Cell. 2011;19:17-30.<br />
44. Ernst T, Chase AJ, Score J, et al. Inactivating mutations <strong>of</strong> the histone<br />
methyltransferase gene EZH2 in myeloid disorders. Nat Genet. 2010;42:722-<br />
726.<br />
45. Abdel-Wahab O, Pardanani A, Patel J, et al. Concomitant analysis <strong>of</strong><br />
EZH2 and ASXL1 mutations in myel<strong>of</strong>ibrosis, chronic myelomonocytic leukemia<br />
and blast-phase myeloproliferative neoplasms. Leukemia. 2011;25:1200-<br />
1202.<br />
46. Score J, Hidalgo-Curtis C, Jones AV, et al. Inactivation <strong>of</strong> polycomb<br />
repressive complex 2 components in myeloproliferative and myelodysplastic/<br />
myeloproliferative neoplasms. Blood. <strong>2012</strong>;119:1208-1213.<br />
47. Guglielmelli P, Biamonte F, Score J, et al. EZH2 mutational status<br />
predicts poor survival in myel<strong>of</strong>ibrosis. Blood. 2011;118:5227-5234.<br />
48. Scheuermann JC, de Ayala Alonso AG, Oktaba K, et al. Histone H2A<br />
deubiquitinase activity <strong>of</strong> the Polycomb repressive complex PR-DUB. Nature.<br />
2010;465:243-247.<br />
49. Gurvich N, Perna F, Farina A, et al. L3MBTL1 polycomb protein, a<br />
candidate tumor suppressor in del(20q12) myeloid disorders, is essential for<br />
genome stability. Proc Natl Acad Sci USA.2010;107:22552-22557.<br />
50. Min J, Allali-Hassani A, Nady N, et al. L3MBTL1 recognition <strong>of</strong> monoand<br />
dimethylated histones. Nat Struct Mol Biol. 2007;14:1229-1230.<br />
51. Perna F, Gurvich N, Hoya-Arias R, et al. Depletion <strong>of</strong> L3MBTL1<br />
promotes the erythroid differentiation <strong>of</strong> human hematopoietic progenitor<br />
cells: Possible role in 20q- polycythemia vera. Blood. 2010;116:2812-2821.<br />
52. Dawson MA, Bannister AJ, Göttgens B, et al. JAK2 phosphorylates<br />
histone H3Y41 and excludes HP1alpha from chromatin. Nature. 2009;461:<br />
819-822.<br />
53. Liu F, Zhao X, Perna F, et al. JAK2V617F-mediated phosphorylation<br />
<strong>of</strong> PRMT5 downregulates its methyltransferase activity and promotes myeloproliferation.<br />
Cancer Cell. 2011;19:283-294.<br />
54. Jäger R, Gisslinger H, Passamonti F, et al. Deletions <strong>of</strong> the transcription<br />
factor Ikaros in myeloproliferative neoplasms. Leukemia. 2010;24:1290-<br />
1298.<br />
55. Harutyunyan A, Klampfl T, Cazzola M, et al. p53 lesions in leukemic<br />
transformation. N Engl J Med. 2011;364:488-490.<br />
56. Silva M, Richard C, Benito A, et al. Expression <strong>of</strong> Bcl-x in erythroid<br />
precursors from patients with polycythemia vera. N Engl J Med. 1998;338:<br />
564-571.<br />
57. Plo I, Nakatake M, Malivert L, et al. JAK2 stimulates homologous<br />
recombination and genetic instability: Potential implication in the heterogeneity<br />
<strong>of</strong> myeloproliferative disorders. Blood. 2008;112:1402-1412.<br />
58. Zhao R, Follows GA, Beer PA, et al. Inhibition <strong>of</strong> the Bcl-xL deamidation<br />
pathway in myeloproliferative disorders. N Engl J Med. 2008;359:2778-<br />
2789.<br />
59. Capello D, Deambrogi C, Rossi D, et al. Epigenetic inactivation <strong>of</strong><br />
suppressors <strong>of</strong> cytokine signalling in Philadelphia-negative chronic myeloproliferative<br />
disorders. Br J Haematol. 2008;141:504-511.<br />
60. Hookham MB, Elliott J, Suessmuth Y, et al. The myeloproliferative<br />
disorder-associated JAK2 V617F mutant escapes negative regulation by<br />
suppressor <strong>of</strong> cytokine signaling 3. Blood. 2007;109:4924-4929.<br />
61. Jones AV, Chase A, Silver RT, et al. JAK2 haplotype is a major risk<br />
factor for the development <strong>of</strong> myeloproliferative neoplasms. Nat Genet.<br />
2009;41:446-449.<br />
62. Kilpivaara O, Mukherjee S, Schram AM, et al. A germline JAK2 SNP is<br />
associated with predisposition to the development <strong>of</strong> JAK2(V617F)-positive<br />
myeloproliferative neoplasms. Nat Genet. 2009;41:455-459.<br />
63. Olcaydu D, Harutyunyan A, Jager R, et al. A common JAK2 haplotype<br />
confers susceptibility to myeloproliferative neoplasms. Nat Genet. 2009;41:<br />
450-454.<br />
64. Mead AJ, Rugless MJ, Jacobsen SE, et al. Germline JAK2 mutation in<br />
a family with hereditary thrombocytosis. N Engl J Med. <strong>2012</strong>;366:967-969.<br />
65. Etheridge L, Corbo LM, Kaushansky K, et al. A novel activating JAK2<br />
mutation, JAK2R564Q, causes familial essential thrombocytosis (fET) via<br />
mechanisms distinct from JAK2V617F. Blood (ASH Annual Meeting Abstracts).<br />
2011;118:123.
Classification <strong>of</strong> Myeloproliferative<br />
Neoplasms and Prognostic Factors<br />
Overview: Myeloproliferative neoplasms (MPNs) are currently<br />
diagnosed according to the World Health Organization<br />
(WHO) criteria. Molecular pr<strong>of</strong>iling should include the analysis<br />
<strong>of</strong> JAK2 V617F (first, exon 12 only in V617F-negative polycythemia<br />
vera [PV]) and MPL mutations (in V617F-negative<br />
essential thrombocythemia [ET] and myel<strong>of</strong>ibrosis [MF]). For<br />
patients with PV and ET, the risk stratification <strong>of</strong> low- and<br />
high-risk disease requires only two parameters: older than<br />
age 60 and prior history <strong>of</strong> thrombosis. Additionally, it might<br />
be important to monitor leukocyte count and know the mutational<br />
pr<strong>of</strong>ile.<br />
MPNs INCLUDE different entities summarized in<br />
Table 1. 1 This article will focus on classical BCR-<br />
ABL1-negative MPNs, including ET, PV, and primary myel<strong>of</strong>ibrosis<br />
(PMF).<br />
Classification <strong>of</strong> MPNs<br />
MPNs are clonal stem cell neoplasms found to share some<br />
relevant biologic features. Recent molecular advances have<br />
demonstrated that the abnormal myeloproliferation arises<br />
from constitutively active signal transduction pathways,<br />
caused by specific mutations affecting protein tyrosine kinases<br />
or related molecules. Current classification <strong>of</strong> MPNs is<br />
based on the WHO criteria established in the 2008 revision<br />
(Tables 2-4), which, besides simple clinical parameters,<br />
convey two novelties: molecular genetics and histopathology.<br />
Complete Blood Count (CBC) for MPN Diagnosis<br />
PV is suspected in men with hemoglobin levels greater<br />
than 18.5 g/dL or 16.5 g/dL in women or hemoglobin levels<br />
greater than 17 g/dL in men or 15 g/dL in women if<br />
associated with a documented and sustained increase <strong>of</strong> at<br />
least 2 g/dL from an individual’s baseline value. It is no<br />
longer necessary to use red cell mass measurement to<br />
exclude secondary polycythemia. In approximately 20% to<br />
40% <strong>of</strong> PV cases, leukocytosis and/or thrombocytosis might<br />
be present. Thrombocytosis remains a criterion for the<br />
diagnosis <strong>of</strong> ET. In the revised WHO criteria, the platelet<br />
threshold for the diagnosis <strong>of</strong> ET was lowered to 450 �<br />
10 9 /L. This level <strong>of</strong> thrombocytosis is not specific for ET and<br />
can be secondary to other conditions (e.g., iron deficiency,<br />
trauma, infection, inflammation, bleeding), which must be<br />
excluded first. Thrombocytosis can also be present in chronic<br />
myeloid leukemia (CML; test for BCR-ABL1 fusion gene), in<br />
refractory anemia with ringed sideroblasts associated with<br />
marked thrombocytosis (RARS-T; signs <strong>of</strong> dyserythropoiesis<br />
at morphological examination), and also in PV cases in<br />
which erythrocytosis is not evident because <strong>of</strong> relative iron<br />
deficiency. Concerning MF, anemia might be accompanied<br />
with leukopenia/leukocytosis and/or thrombocytopenia/<br />
thrombocytosis. Peripheral blood smear should be reviewed<br />
in all MPN cases as microcytic red blood cells in PV/ET<br />
might be a sign <strong>of</strong> iron deficiency, leukoerythroblastosis is<br />
present in almost all PMF cases, and myeloid progenitors<br />
are typical <strong>of</strong> CML.<br />
By Francesco Passamonti, MD<br />
Survival <strong>of</strong> patients with MF is defined by the International<br />
Prognostic Scoring System (IPSS) model at diagnosis and the<br />
Dynamic IPSS (DIPSS) anytime during the course <strong>of</strong> the<br />
disease. The IPSS and the DIPSS are based on patient age<br />
older than age 65, presence <strong>of</strong> constitutional symptoms,<br />
hemoglobin level less than 10 g/dL, leukocyte count greater<br />
than 25 � 10 9 /L, and circulating blast cells 1% or greater. The<br />
DIPSS-plus adds critical prognostic information and suggests<br />
also considering cytogenetic categories, platelet count, and<br />
red blood cell transfusion need.<br />
Molecular Genetics for MPN Diagnosis<br />
Screening assays for MPN genetics are not standardized,<br />
and the possibility <strong>of</strong> false-positives or false-negatives can be<br />
an issue when using highly sensitive allele-specific assays<br />
and in cases <strong>of</strong> low mutant allele burden. The JAK2 V617F<br />
mutation is found in more than 95% <strong>of</strong> patients with PV and<br />
in nearly 50% to 60% <strong>of</strong> those with ET and PMF. It has also<br />
been found in other MPNs but not in nonmyeloid malignancies<br />
or in cases <strong>of</strong> secondary polycythemia. Therefore, JAK2<br />
V617F is a sensitive diagnostic marker <strong>of</strong> PV. Although low<br />
JAK2 V617F allele burden is more distinctive <strong>of</strong> ET than <strong>of</strong><br />
PV and MF, mutational load measurement is not useful for<br />
diagnostic purposes. Less than 3% <strong>of</strong> patients with PV carry<br />
exon 12 mutations <strong>of</strong> JAK2, and, as different mutations do<br />
not confer different phenotypes, a screening test highresolution<br />
melting is adequate. 2 Exon 12 mutations <strong>of</strong> JAK2<br />
should be screened in case <strong>of</strong> JAK2 V617F-negative erythrocytosis<br />
with low erythropoietin level. A small portion <strong>of</strong><br />
patients with ET and PMF who lack a mutated JAK2 may<br />
have activating mutations in MPL (mainly W515K/L). For<br />
the time being, the study <strong>of</strong> JAK2 and MPL mutations is to<br />
be included in the diagnostic workup <strong>of</strong> MPN, while all other<br />
less prevalent mutations (LNK, NF1, cCBL, SOCS1, TET2,<br />
EZH2, ASXL1, IDH1/2, DNMT3A) cannot enter into the<br />
diagnostic process. 2<br />
Serum Erythropoietin<br />
A simple and timeless PV test is serum erythropoietin<br />
dosage, which can discriminate PV (low level) from secondary<br />
erythrocytosis (high level). In the absence <strong>of</strong> any JAK2<br />
mutation or in case <strong>of</strong> unavailability <strong>of</strong> the JAK2 test, low<br />
erythropoietin levels—if accompanied by MPN-consistent<br />
bone marrow features—is <strong>of</strong> diagnostic value for PV.<br />
Bone Marrow Histopathology<br />
In the WHO classification, bone marrow histopathology<br />
has assumed a critical diagnostic role, since the distinction<br />
From the Division <strong>of</strong> Hematology, Department <strong>of</strong> Internal Medicine, University Hospital<br />
Ospedale di Circolo e Fondazione Macchi, Varese, Italy.<br />
Author’s disclosures <strong>of</strong> potential conflicts <strong>of</strong> interest are found at the end <strong>of</strong> this article.<br />
Address reprint requests to Francesco Passamonti, MD, Division <strong>of</strong> Hematology, Department<br />
<strong>of</strong> Internal Medicine, University Hospital Ospedale di Circolo e Fondazione Macchi,<br />
Viale L. Borri 57, 21100 Varese, Italy; email: francesco.passamonti@ospedale.varese.it.<br />
© <strong>2012</strong> by <strong>American</strong> <strong>Society</strong> <strong>of</strong> <strong>Clinical</strong> <strong>Oncology</strong>.<br />
1092-9118/10/1-10<br />
419
etween ET, prefibrotic PMF, and PMF requires bone marrow<br />
evaluation. In PV, marrow is usually hypercellular for<br />
age with trilineage growth and prominent erythroid, granulocytic,<br />
and megakaryocytic proliferation, and bone marrow<br />
fibrosis is generally absent, even though fibrosis may be<br />
found in approximately 15% <strong>of</strong> cases. In ET, bone marrow<br />
examination reveals no or only a slight increase in agematched<br />
cellularity, no significant increase in granulo- and<br />
erythropoiesis or prominent large to giant mature megakaryocytes<br />
with hyperlobulated or deeply folded nuclei,<br />
dispersed or loosely clustered in the marrow space. In PMF,<br />
the bone marrow demonstrates marked megakaryocyte<br />
proliferation with atypia described as small to large megakaryocytes<br />
with aberrant nuclear/cytoplasmic ratio and<br />
hyperchromatic and irregularly folded nuclei with dense<br />
clustering, which is accompanied by reticulin and/or collagen<br />
fibrosis. When reticulin fibrosis is absent, a diagnosis <strong>of</strong><br />
prefibrotic PMF needs to be considered. Prefibrotic PMF<br />
presents with marked increase in age-matched cellularity,<br />
pronounced granulopoiesis precursors, reduced erythroid<br />
precursors, and dense or loose clustering <strong>of</strong> medium-sized to<br />
giant megakaryocytes. The histopathologic distinction between<br />
ET and prefibrotic PMF is <strong>of</strong> prognostic value, as the<br />
latter disease is associated with a higher risk <strong>of</strong> evolution<br />
into MF and acute myeloid leukemia (AML) and worse<br />
survival. 3<br />
Prognostication in PV and ET<br />
Life expectancy <strong>of</strong> patients with PV is reduced when<br />
compared with that <strong>of</strong> the general population, while the<br />
survival <strong>of</strong> patients with ET is not significantly shortened<br />
(831 patients with PV and ET followed for a median <strong>of</strong> 10<br />
years). 4<br />
Disease-related complications affecting survival are<br />
mainly vascular events (thrombosis and hemorrhage) and<br />
transformation to MF or AML. In PV, the incidence <strong>of</strong><br />
thrombosis was estimated at 18 � 1,000 person years and<br />
for MF and AML at 5 � 1,000 person years. 4 In ET, the<br />
incidence <strong>of</strong> thrombosis, MF, and AML was 12, 1.6, and<br />
1.2 � 1,000 person years. 4 The ECLAP study also indicated<br />
that cardiovascular mortality in PV accounted for 45% <strong>of</strong> all<br />
deaths and hematologic transformation accounted for 13%. 5<br />
KEY POINTS<br />
● Myeloproliferative neoplasms (MPNs) are diagnosed<br />
according to the 2008 World Health Organization<br />
criteria.<br />
● Molecular analysis in MPNs should include JAK2<br />
V617F first, exon 12 only in V617F-negative polycythemia<br />
vera (PV) and MPL mutations in V617Fnegative<br />
essential thrombocythemia (ET) and<br />
myel<strong>of</strong>ibrosis (MF).<br />
● Bone marrow biopsy is mandatory to distinguish ET<br />
from primary myel<strong>of</strong>ibrosis.<br />
● Risk stratification <strong>of</strong> PV and ET is based on patient<br />
age older than age 60 and prior history <strong>of</strong> thrombosis.<br />
● Risk stratification <strong>of</strong> MF is based on the IPSS prognostic<br />
model at diagnosis and the DIPSS model<br />
during follow-up.<br />
420<br />
FRANCESCO PASSAMONTI<br />
Thus, as thrombosis is the most frequent complication in PV<br />
and ET, there is a comprehensible rationale for stratifying<br />
these patients according to the risk <strong>of</strong> thrombosis.<br />
Risk Factors for Thrombosis<br />
There is a general agreement among investigators to<br />
consider patient age older than 60 at diagnosis and presence<br />
<strong>of</strong> vascular events in the patients’ history as the two prognostic<br />
factors for thrombosis in PV and ET (Table 5).<br />
Therefore, patients with one or two <strong>of</strong> these risk factors at<br />
diagnosis are considered at high risk, while those with none<br />
<strong>of</strong> them are considered at low risk. This risk classification<br />
has an effect on the therapeutic approach for patients with<br />
PV and ET.<br />
Cardiovascular risk factors (arterial hypertension, smoking,<br />
hypercholesterolemia, diabetes) do not enter in the risk<br />
stratification <strong>of</strong> ET and PV, but an appropriate strategy <strong>of</strong><br />
prevention and management is recommended. 5<br />
Concerning the CBC, hematocrit levels up to 50% and<br />
high platelet count are not associated with thrombosis. 5 On<br />
the other hand, extreme thrombocytosis might provoke<br />
excess bleeding from acquired von Willebrand disease, and<br />
platelet counts over 1,500 � 10 9 /L should be considered as a<br />
criterion to start cytoreduction in ET and PV. The relationship<br />
between leukocytosis and thrombosis was largely studied<br />
in ET. A retrospective analysis seems to indicate that<br />
patients with low-risk ET—most frequently not treated—<br />
have a higher risk <strong>of</strong> thrombosis in the presence <strong>of</strong> leukocytosis<br />
at diagnosis or developed during the disease course. 6<br />
Concerning patients with PV, a subsequent analysis <strong>of</strong> the<br />
ECLAP trial showed that patients with a leukocyte count<br />
exceeding 15 � 10 9 /L had a significant increase <strong>of</strong> myocardial<br />
infarction compared with those patients with leukocyte<br />
counts below 10 � 10 9 /L (p � 0.017). 5<br />
Many studies recently evaluated the relationship between<br />
the presence <strong>of</strong> the JAK2 V617F mutation and thrombosis. 7<br />
Meta-analyses suggest a correlation between the JAK2<br />
V617F mutation and thrombosis in ET. Among patients with<br />
JAK2 V617F-positive ET, those with higher allele burden<br />
seem to have a higher risk <strong>of</strong> thrombosis, while a prospective<br />
observation <strong>of</strong> PV did not disclose any relationship. 8,9<br />
Regarding patients with exon 12 mutations <strong>of</strong> JAK2, the<br />
proposed risk stratification <strong>of</strong> PV may be applied. 10 Rare<br />
mutations involving MPL and TET2 genes seem not to effect<br />
events in patients with ET. 2<br />
Prediction <strong>of</strong> PV and ET Disease Evolution<br />
Approximately 10% <strong>of</strong> patients with PV evolve in post-PV<br />
MF with progressive splenomegaly, MF-related symptoms,<br />
anemia, and leukocytosis. 11 MF evolution is difficult to<br />
predict, but leukocyte count greater than 15 � 10 9 /L has<br />
been documented as a risk factor for disease evolution. 12 In<br />
a prospective study that included 338 patients with PV, an<br />
allele burden greater than 50% implied a higher risk <strong>of</strong> MF<br />
evolution. 9 When PV progresses to post-PV MF, survival<br />
worsens and might be predicted using a dynamic prognostic<br />
score based on three risk factors: hemoglobin level less than<br />
10 g/dL, leukocyte count greater than 30 � 10 9 /L, and<br />
platelet count less than 100 � 10 9 /L. 12 The low-risk group<br />
includes patients without risk factors, while higher-risk<br />
categories include patients with one, two, or three risk<br />
factors. When a single patient acquires one risk factor, his or<br />
her survival worsens 4.2-fold. Despite the availability <strong>of</strong> this
CLASSIFICATION AND PROGNOSIS OF MPNs<br />
Table 1. Classification <strong>of</strong> Chronic Myeloid Neoplasms According to the World Health Organization in 2008<br />
Diseases Main criteria<br />
Chronic myelogenous leukemia, BCR-<br />
ABL1-positive<br />
Positivity for the BCR-ABL1 fusion gene<br />
Myeloproliferative Neoplasms<br />
Chronic neutrophilic leukemia PB leukocytosis, WBC �25 � 10 9 /L<br />
- Segmented PMNs and band forms are � 80% <strong>of</strong> WBCs<br />
- Immature granulocytes � 10% <strong>of</strong> WBCs<br />
- Myeloblasts � 1% <strong>of</strong> WBCs<br />
Hypercellular BM biopsy (myeloblasts � 5% <strong>of</strong> NMCs)<br />
Hepatosplenomegaly<br />
No evidence <strong>of</strong> reactive neutrophila<br />
No Philadelphia chromosome or BCR-ABL1 fusion gene; no rearrangement <strong>of</strong> PDGFRA, PDGFRB, or FGFR1; no evidence <strong>of</strong> PV, ET, or PMF;<br />
no evidence <strong>of</strong> MDS or MDS/MPN<br />
Polycythemia vera See Table 2<br />
Primary myel<strong>of</strong>ibrosis See Table 4<br />
Essential thrombocythemia See Table 5<br />
Chronic eosinophilic leukemia, NOS Eosinophils �1.5 � 10 9 /L<br />
No Philadelphia chromosome or BCR-ABL1 fusion gene or other MPN (PV, ET, or PMF) or MDS/MPN (CMML or aCML)<br />
No t(5;12)(q31-35;p13) or other rearrangement <strong>of</strong> PDGFRB<br />
No FIP1L1-PDGFRA fusion gene or other rearrangement <strong>of</strong> PDGFRA<br />
No rearrangement <strong>of</strong> FGFR1<br />
PB and BM blasts � 20%, no inv(16)(p13.1;q22) or t(16;16)(p13.1;q22) or other feature diagnostic <strong>of</strong> AML<br />
Presence <strong>of</strong> a clonal cytogenetic or molecular genetic abnormality, or blast cells � 2% in the PB or � 5% in the BM<br />
Mastocytosis Cutaneous mastocytosis: mast cell infiltrate in the skin (criteria for SM not met)<br />
SM: mast cell infiltrate in BM and/or other extracutaneous organs (minor diagnostic criteria include mast cell atypia in � 25% <strong>of</strong> cells,<br />
detection <strong>of</strong> a mutation at codon 816 <strong>of</strong> KIT, an aberrant mast cell immunophenotype and elevated serum triptase levels)<br />
MPN, unclassifiable <strong>Clinical</strong>, laboratory, and morphologic features <strong>of</strong> an MPN without meeting the diagnostic criteria for one specific MPN<br />
Myeloid and lymphoid neoplasms<br />
associated with eosinophilia and<br />
abnormalities <strong>of</strong> PDGFRA, PDGFRB,<br />
or FGFR1<br />
Myeloid and lymphoid neoplasms associated with eosinophilia and abnormalities <strong>of</strong> PDGFRA, PDGFRB, or FGFR1<br />
An MPN with prominent eosinophilia (cases presenting with AML or ALL with eosinophilia are also included) AND presence <strong>of</strong> a FIP1L1-<br />
PDGFRA fusion gene or presence <strong>of</strong> t(5;12)(q31�q33;p12) or a variant translocation or demonstration <strong>of</strong> PDGFRB fusion gene or <strong>of</strong><br />
rearrangement <strong>of</strong> PDGFRB or presence <strong>of</strong> t(8;13)(p11;q12) or a variant translocation leading to FGFR1 rearrangement<br />
MDS/MPN<br />
Chronic myelomonocytic leukemia Persistent PB monocytosis � 1 � 10 9 /L<br />
No Philadelphia chromosome or BCR-ABL1 fusion gene<br />
No rearrangement <strong>of</strong> PDGFRA or PDGFRB<br />
� 20% blasts (including myeloblasts, monoblasts, and promonocytes) in PB and BM<br />
Dysplasia in �1 myeloid lineage–if no/minimal myelodysplasia:<br />
- Presence <strong>of</strong> an acquired clonal cytogenetic or molecular genetic abnormality OR<br />
- Persistent monocytosis (�3 mo) and exclusion <strong>of</strong> all other causes <strong>of</strong> monocytosis<br />
Juvenile myelomonocytic leukemia Persistent PB monocytosis � 1 � 10 9 /L<br />
No Philadelphia chromosome or BCR-ABL1 fusion gene<br />
� 20% blasts (including promonocytes) in PB and BM<br />
Two <strong>of</strong> the following: hemoglobin F increased for age; immature granulocytes in PB; WBC count � 10 � 10 9 /L, clonal chromosomal<br />
abnormality (may be monosomy 7); GM-CSF hypersensitivity <strong>of</strong> myeloid progenitors in vitro<br />
Atypical chronic myeloid leukemia, BCR-<br />
ABL1-negative<br />
PB leukocytosis (WBCs �13 � 10 9 /L) because <strong>of</strong> increased numbers <strong>of</strong> neutrophils and their precursors (which constitute �10% <strong>of</strong> leukocytes)<br />
with prominent dysgranulopoiesis<br />
No Philadelphia chromosome or BCR-ABL1 fusion gene<br />
No rearrangement <strong>of</strong> PDGFRA or PDGFRB<br />
Minimal absolute basophilia; basophils usually � 2% <strong>of</strong> leukocytes<br />
No or minimal absolute monocytosis; monocytes � 10% <strong>of</strong> WBCs<br />
Hypercellular BM with granulocytic proliferation and dysplasia, � dysplasia in the other lineages<br />
� 20% blasts in PB and BM<br />
MDS/MPN, unclassifiable <strong>Clinical</strong>, laboratory and morphologic features <strong>of</strong> an MDS AND<br />
- Provisional entity: refractory anemia<br />
with ring sideroblasts associated with<br />
marked thrombocytosis<br />
Prominent myeloproliferative features, e.g., platelet count �450 � 10 9 /L associated with megakaryocytic proliferation, or WBC count �13 �<br />
10 9 /L, � splenomegaly AND<br />
No preceding MDS or MPN, no history <strong>of</strong> cytotoxic or growth factor therapy, no Philadelphia chromosome or BCR-ABL1 fusion gene, no<br />
rearrangement <strong>of</strong> PDGFRA, PDGFRB or FGFR1, and no isolated del(5q), t(3;3)(q21;q26) or inv(3)(q21;q26) OR<br />
De novo disease with mixed myeloproliferative and myelodysplastic features that cannot be assigned to any <strong>of</strong> the MDS, MPN, or MDS/MPN<br />
categories<br />
RARS-T: anemia, ring sideroblasts �15% <strong>of</strong> erythroid precursors, platelet count �450 � 10 9 /L, large and atypical BM megakaryocytes, no PB<br />
blasts, �5% BM blasts AND NO BCR-ABL1 fusion gene, isolated del(5q), t(3;3)(q21;q26), inv(3)(q21;q26)<br />
Myelodysplastic syndromes<br />
Myelodysplastic syndrome PB cytopenia(s), dysplasia in �1 myeloid cell line, PB and BM blasts � 20%, � BM ring sideroblasts—several MDS categories exist: refractory<br />
cytopenias with unilineage dysplasia (refractory anemia, refractory neutropenia, refractory thrombocytopenia), refractory anemia with ring<br />
sideroblasts, refractory cytopenia with multilineage dysplasia, refractory anemia with excess blasts type 1 or 2, myelodysplastic syndrome<br />
unclassified, MDS associated with isolated del(5q)<br />
Abbreviations: PB, peripheral blood; WBCs, white blood cells; PMNs, polymorphonucleated cells; BM, bone marrow; NMCs, nucleated marrow cells; PV: polycythemia<br />
vera; ET, essential thrombocythemia; PMF, primary myel<strong>of</strong>ibrosis; MDS, myelodysplastic syndrome; MDS/MPN: myelodysplastic/myeloproliferative neoplasm; NOS, not<br />
otherwise specified; MPN, myeloproliferative neoplasms; CMML, chronic myelomonocytic leukemia; aCML, atypical chronic myeloid leukemia; AML, acute myeloid<br />
leukemia; SM, systematic mastocytosis; ALL, acute lymphoblastic leukemia; GM-CSF, granulocyte macrophage colony-stimulating factor; RARS-T, refractory anemia<br />
with ring sideroblasts associated with marked thrombocytosis.<br />
421
Table 2. WHO Criteria for Diagnosis <strong>of</strong> Polycythemia Vera a<br />
Major Criteria<br />
Hemoglobin � 18.5 g/dL in men, � 16.5 g/dL in women, or other evidence <strong>of</strong><br />
increased red cell volume<br />
Presence <strong>of</strong> JAK2 V617F or other functionally similar mutation (JAK2 exon 12<br />
mutation)<br />
Minor Criteria<br />
Bone marrow biopsy showing hypercellularity for age with trilineage<br />
myeloproliferation<br />
Serum erythropoietin level below the normal reference range<br />
Endogenous erythroid colony formation in vitro<br />
Abbreviations: WHO, World Health Organization; PV, polycythemia vera.<br />
a Diagnosis <strong>of</strong> PV requires meeting either both major criteria and one minor<br />
criterion or the first major criterion and two minor criteria.<br />
disease-specific model, many investigators are more confident<br />
applying prognostic systems developed in PMF in this<br />
patient subset. 13,14<br />
Evolution to AML is a rare event, and the predictors<br />
include advanced age (sign <strong>of</strong> genomic instability) and a high<br />
leukocyte count (sign <strong>of</strong> myeloproliferation). Concerning the<br />
role <strong>of</strong> chemotherapy, a recent population-based study on<br />
11,039 MPNs proved that 25% <strong>of</strong> patients with post-MPN<br />
AML were never exposed to cytotoxic drugs and that hydroxyurea<br />
at any dose is not associated with an increased<br />
risk <strong>of</strong> AML, whereas an increasing cumulative dose <strong>of</strong><br />
alkylators is. 15<br />
Specific <strong>Clinical</strong> Situations in PV and ET<br />
A retrospective study on the outcome <strong>of</strong> 311 surgical<br />
interventions for patients with PV and ET showed that 7.7%<br />
<strong>of</strong> them were complicated by fatal arterial or venous thromboses<br />
and 7.3% by fatal major hemorrhage within 3 months<br />
from the procedure. 16 Although no prognostic factors could<br />
be identified to predict postsurgery outcome, these data<br />
should mandate watchfulness in the surgical management<br />
<strong>of</strong> these patients.<br />
When treating patients with ET (generally younger than<br />
those with PV and PMF), a potential issue is the approach to<br />
pregnancy. A study on 103 pregnancies in 62 women with<br />
ET reported a 64% live birth rate, with 51% <strong>of</strong> pregnancies<br />
being uneventful. 17 Maternal complications occurred in 9%<br />
<strong>of</strong> cases, while fetal complications occurred in 40% <strong>of</strong> them.<br />
Fetal loss was 3.4-fold higher than that <strong>of</strong> the general<br />
population. The JAK2 V617F mutation was an independent<br />
predictor <strong>of</strong> pregnancy complications.<br />
Prognostication in MF<br />
Among MPNs, PMF has the most heterogeneous clinical<br />
presentation, which may encompass anemia, splenomegaly,<br />
Table 3. WHO Criteria for Diagnosis <strong>of</strong> Essential<br />
Thrombocythemia a<br />
Sustained platelet count over 450 � 10 9 /L<br />
Bone marrow biopsy specimen showing proliferation mainly <strong>of</strong> the megakaryocytic<br />
lineage with increased numbers <strong>of</strong> enlarged, mature megakaryocytes; no<br />
significant increase or left-shift <strong>of</strong> neutrophil granulopoiesis or erythropoiesis<br />
Not meeting WHO criteria for PV or PMF, BCR–ABL1-positive CML, MDS, or other<br />
myeloid neoplasm<br />
Demonstration <strong>of</strong> JAK2 V617F or other clonal marker or, in the absence <strong>of</strong> JAK2<br />
V617F, no evidence <strong>of</strong> reactive thrombocytosis<br />
Abbreviations: WHO, World Health Organization; PV, polycythemia vera; PMF,<br />
primary myel<strong>of</strong>ibrosis; CML, chronic myelogenous leukemia; MDS, myelodysplastic<br />
syndrome.<br />
a The diagnosis <strong>of</strong> ET requires meeting all four criteria.<br />
422<br />
FRANCESCO PASSAMONTI<br />
Table 4. WHO criteria for Diagnosis <strong>of</strong> Primary Myel<strong>of</strong>ibrosis a<br />
Major Criteria<br />
Presence <strong>of</strong> megakaryocyte proliferation and atypia, accompanied by either<br />
reticulin or collagen fibrosis or—in the absence <strong>of</strong> significant reticulin<br />
fibrosis—the megakaryocyte changes must be accompanied by an increased<br />
marrow cellularity characterized by granulocytic proliferation and <strong>of</strong>ten<br />
decreased erythropoiesis (i.e., prefibrotic cellular-phase disease)<br />
Not meeting WHO criteria for PV, CML, MDS, or other myeloid disorders<br />
Demonstration <strong>of</strong> JAK2 V617F or other clonal marker (e.g., MPLW515K/L) or—<br />
in the absence <strong>of</strong> the above clonal markers—no evidence <strong>of</strong> secondary bone<br />
marrow fibrosis<br />
Minor Criteria<br />
Leukoerythroblastosis<br />
Increased serum lactate dehydrogenase level<br />
Anemia<br />
Splenomegaly<br />
Abbreviations: WHO, World Health Organization; PV, polycythemia vera; CML,<br />
chronic myelogenous leukemia; MDS, myelodysplastic syndrome.<br />
a Diagnosis <strong>of</strong> PMF requires all three major criteria and two minor criteria.<br />
leukocytosis or leukopenia, thrombocytosis or thrombocytopenia,<br />
and constitutional symptoms. Median survival in<br />
PMF is estimated at 6 years, but it can range from a few<br />
months to many years. 13,14,18,19 Causes <strong>of</strong> death in MF<br />
include bone marrow failure (severe anemia, bleeding from<br />
thrombocytopenia, and infections from leukopenia) in 25%<br />
to 30% <strong>of</strong> patients, AML transformation in 10% to 20% <strong>of</strong><br />
patients, cardiovascular complications in 15% to 20%, and<br />
portal hypertension in 10%.<br />
Risk Factors for Survival<br />
The following were shown to be associated with poor<br />
outcome in patients with PMF: advanced age, anemia, red<br />
blood cell transfusion need, leukopenia, leukocytosis, thrombocytopenia,<br />
peripheral blast count, constitutional symptoms,<br />
hepatic myeloid metaplasia, decreased marrow<br />
cellularity with higher degree <strong>of</strong> fibrosis, higher degree <strong>of</strong><br />
microvessel density, high number <strong>of</strong> circulating CD34positive<br />
cells, cytogenetic abnormalities, mutational pr<strong>of</strong>ile,<br />
and high level <strong>of</strong> proinflammatory cytokines (IL-8 and IL-<br />
2R). Among these, some parameters require a more detailed<br />
discussion.<br />
Molecular Abnormalities. Within a large international<br />
database, 345 patients had an available JAK2 mutational<br />
status and no association was observed between the JAK2<br />
status and survival. 14 This result parallels other studies<br />
including 199, 186, and 174 patients each that showed<br />
no significant correlation between the presence <strong>of</strong> the<br />
JAK2 V617F mutation and survival or leukemic transformation.<br />
20-22 However, there is no a general agreement on<br />
this matter. Data seem to indicate that having a lower JAK2<br />
V617F allele burden implies a worse survival. 21,22 The<br />
explanation for this association, however, differs as patients<br />
mostly died from infections secondary to myelodepletion in<br />
one study and from AML evolution in the other. 21,22 A study<br />
on 139 patients with MF receiving allogenic stem cell transplantation<br />
(ASCT) reported very intriguing results on the<br />
association between allele burden reduction and post-ASCT<br />
Table 5. Risk Categories in Essential Thrombocythemia<br />
and Polycythemia Vera<br />
Low risk Age � 60 and no history <strong>of</strong> thrombosis<br />
High risk Age � 60 or history <strong>of</strong> thrombosis
CLASSIFICATION AND PROGNOSIS OF MPNs<br />
Table 6. Score Values for the International Prognostic Scoring<br />
System (IPSS) 14 and Dynamic International Prognostic<br />
Scoring System (DIPSS) 13<br />
outcome. 23 Patients who abrogated the JAK2 mutation 6<br />
months after ASCT had a significantly lower relapse rate,<br />
shedding light on the likely clinical benefit <strong>of</strong> reduced V617F<br />
allele burden (p � 0.04).<br />
Concerning less frequent mutations described so far in<br />
MF, studies reported that mutations involving TET2 (approximately<br />
15% to 30% <strong>of</strong> patients with MF) and MPL<br />
(approximately 7% to 9% <strong>of</strong> patients with MF) genes seem<br />
not to affect survival, while IDH1 (approximately 4% <strong>of</strong><br />
patients with MF) implies a worse survival and higher risk<br />
<strong>of</strong> AML, as does EZH2 mutations (approximately 6% <strong>of</strong><br />
patients with MF). 24-27 Nullizygosity for the JAK2 46/1<br />
haplotype was associated with shortened survival, raising<br />
the possibility that non-46/1 haplotypes are associated with<br />
a biologically more aggressive phenotype. 28<br />
Chromosomal Abnormalities. An abnormal karyotype is<br />
present in approximately 30% to 40% <strong>of</strong> MF and implies a<br />
shorter survival. 14,29 A study <strong>of</strong> 433 patients with MF<br />
provided a two-tiered cytogenetic-risk stratification with a<br />
respective 5-year survival <strong>of</strong> 8% (high risk) and 51% (low<br />
risk). 20 Cytogenetic pr<strong>of</strong>iles are discussed in Table 7.<br />
Red Blood Cell Transfusion Dependency. Patients with<br />
the JAK2 V617F mutation seem not as prone to anemia as<br />
those carrying the MPL or TET2 mutations. 24,25 The prognostic<br />
effect <strong>of</strong> red blood cell transfusion need was examined<br />
in 254 consecutive patients. 30 Median survival was 35<br />
months for patients receiving red blood cell transfusions and<br />
117 months for patients who were nontransfused.<br />
Prognostic Models at Diagnosis<br />
The first prognostic model used to stratify PMF was the<br />
Lille score, which included hemoglobin level less than10<br />
g/dL and leukocyte count less than 4 � 10 9 /L or greater than<br />
30 � 10 9 /L. 29 Patients were grouped into three categories<br />
with median survivals <strong>of</strong> 93, 26, and 13 months.<br />
More recently, the International Working Group on MPN<br />
Research and Treatment (IWG-MRT) developed the IPSS to<br />
predict survival in patients with PMF at diagnosis. 14 The<br />
IPSS was developed using a dataset <strong>of</strong> 1,054 patients with<br />
PMF from seven international centers. Currently, the IPSS<br />
Author’s Disclosures <strong>of</strong> Potential Conflicts <strong>of</strong> Interest<br />
Author<br />
Francesco Passamonti*<br />
*No relevant relationships to disclose.<br />
Employment or<br />
Leadership<br />
Positions<br />
Scores<br />
Parameter<br />
IPSS DIPSS<br />
Age � 65 1 1<br />
Hemoglobin � 10 g/dL 1 2<br />
Leukocyte count � 25 � 109 /L 1 1<br />
Blast cells � 1% 1 1<br />
Constitutional symptoms 1 1<br />
IPSS: score 0 for low risk, score 1 for intermediate-1 risk, score 2 for<br />
intermediate-2 risk, score � 3 for high risk; DIPSS: score 0 for low risk, score 1-2<br />
for intermediate-1 risk, score 3-4 for intermediate-2 risk, score 5-6 for high risk.<br />
Consultant or<br />
Advisory Role<br />
Table 7. Score Values for the Dynamic International Prognostic<br />
Scoring System-Plus (DIPSS-Plus)<br />
is the prognostic scoring system used for diagnosis in clinical<br />
practice. Table 6 illustrates risk factors and the score system.<br />
According to the four risk categories, median survivals<br />
were 135, 95, 48, and 27 months.<br />
A composite prognostic model was developed by the University<br />
<strong>of</strong> Texas M. D. Anderson Cancer Center involving<br />
256 patients with PMF at diagnosis. 19 The risk factors were<br />
unfavorable karyotype, hemoglobin level less than 10 g/dL,<br />
platelet count less than 100 � 10 9 /L, and performance<br />
status greater than 1. Overall survival ranged from 7 to 69<br />
months.<br />
Dynamic Prognostic Models<br />
Dynamic prognostic models are based on the knowledge<br />
that the acquisition <strong>of</strong> additional risk factors during the<br />
disease course may substantially modify patients’ outcome.<br />
The DIPSS Model. Among 1,054 patients evaluated to<br />
develop the IPSS model, 525 had adequate follow-up data for<br />
a time-dependent analysis aimed at the definition <strong>of</strong> the<br />
DIPSS. 13 The DIPSS is based on the same factors as the<br />
IPSS and the score system is illustrated in Table 6. Median<br />
survival was not reached in patients at low risk; it was 14.2<br />
years in intermediate-1, 4 years in intermediate-2, and 1.5<br />
years in patients at high risk.<br />
The DIPPS-Plus Model. The DIPSS-plus is a refinement<br />
<strong>of</strong> the DIPSS, which includes additional variables as illustrated<br />
in Table 7. DIPSS-plus includes four categories with<br />
median survivals <strong>of</strong> 185, 78, 35, and 16 months, respectively.<br />
18<br />
Conclusion<br />
MPN diagnosis must be done according to the 2008 WHO<br />
criteria. Molecular analysis should include JAK2 and MPL<br />
mutations. The risk stratification for patients with PV and<br />
ET requires only two parameters: age older than 60 and<br />
prior history <strong>of</strong> thrombosis. As for PMF, IPSS at diagnosis<br />
and DIPSS anytime during disease course easily define<br />
survival <strong>of</strong> these patients very adequately.<br />
Stock<br />
Ownership Honoraria<br />
Parameter Score value<br />
DIPSS intermediate-1 1<br />
DIPSS intermediate-2 2<br />
DIPSS high risk 3<br />
Unfavorable karyotype* 1<br />
Red blood cell need 1<br />
Platelet � 100 � 10 9 /L 1<br />
* Unfavorable karyotype: complex karyotype or sole or two abnormalities that<br />
include �8, �7/7q-, i(17q), �5/5q-, 12p-, inv(3), or 11q23 rearrangement;<br />
favorable karyotype: normal and diploid karyotype, sole or two abnormalities not<br />
included in the unfavorable karyotype category.<br />
DIPSS-plus: score 0 for low risk, score 1 for intermediate-1 risk, score 2-3 for<br />
intermediate-2 risk, score 4-6 for high risk.<br />
Research<br />
Funding<br />
Expert<br />
Testimony<br />
Other<br />
Remuneration<br />
423
1. Tefferi A, Thiele J, Orazi A, et al. Proposals and rationale for revision <strong>of</strong><br />
the World Health Organization diagnostic criteria for polycythemia vera,<br />
essential thrombocythemia, and primary myel<strong>of</strong>ibrosis: Recommendations<br />
from an ad hoc international expert panel. Blood. 2007;110:1092-1097.<br />
2. Tefferi A. How I treat myel<strong>of</strong>ibrosis. Blood. 2011;117:3494-3504.<br />
3. Barbui T, Thiele J, Passamonti F, et al. Survival and disease progression<br />
in essential thrombocythemia are significantly influenced by accurate morphologic<br />
diagnosis: An international study. J Clin Oncol. 2011;29:3179-3184.<br />
4. Passamonti F, Rumi E, Pungolino E, et al. Life expectancy and prognostic<br />
factors for survival in patients with polycythemia vera and essential<br />
thrombocythemia. Am J Med. 2004;117:755-761.<br />
5. Landolfi R, Marchioli R, Kutti J, et al. Efficacy and safety <strong>of</strong> low-dose<br />
aspirin in polycythemia vera. N Engl J Med. 2004;350:114-124.<br />
6. Carobbio A, Antonioli E, Guglielmelli P, et al. Leukocytosis and risk<br />
stratification assessment in essential thrombocythemia. J Clin Oncol. 2008;<br />
26:2732-2736.<br />
7. Passamonti F. Prognostic factors and models in polycythemia vera,<br />
essential thrombocythemia, and primary myel<strong>of</strong>ibrosis. Clin Lymphoma Myeloma<br />
Leuk. 2011;suppl 1:S25-S27.<br />
8. Vannucchi AM, Antonioli E, Guglielmelli P, et al. <strong>Clinical</strong> pr<strong>of</strong>ile <strong>of</strong><br />
homozygous JAK2 617V�F mutation in patients with polycythemia vera or<br />
essential thrombocythemia. Blood. 2007;110:840-846.<br />
9. Passamonti F, Rumi E, Pietra D, et al. A prospective study <strong>of</strong> 338<br />
patients with polycythemia vera: The impact <strong>of</strong> JAK2 (V617F) allele burden<br />
and leukocytosis on fibrotic or leukemic disease transformation and vascular<br />
complications. Leukemia. 2010;24:1574-1579.<br />
10. Passamonti F, Elena C, Schnittger S, et al. Molecular and clinical<br />
features <strong>of</strong> the myeloproliferative neoplasm associated with JAK2 exon 12<br />
mutations. Blood. 2011;117:2813-2816.<br />
11. Barosi G, Mesa RA, Thiele J, et al. Proposed criteria for the diagnosis<br />
<strong>of</strong> post-polycythemia vera and post-essential thrombocythemia myel<strong>of</strong>ibrosis:<br />
A consensus statement from the International Working Group for Myel<strong>of</strong>ibrosis<br />
Research and Treatment. Leukemia. 2008;22:437-438.<br />
12. Passamonti F, Rumi E, Caramella M, et al. A dynamic prognostic model<br />
to predict survival in post-polycythemia vera myel<strong>of</strong>ibrosis. Blood. 2008;111:<br />
3383-3387.<br />
13. Passamonti F, Cervantes F, Vannucchi AM, et al. A dynamic prognostic<br />
model to predict survival in primary myel<strong>of</strong>ibrosis: A study by the IWG-MRT<br />
(International Working Group for Myeloproliferative Neoplasms Research<br />
and Treatment). Blood. 2010;115:1703-1708.<br />
14. Cervantes F, Dupriez B, Pereira A, et al. New prognostic scoring system<br />
for primary myel<strong>of</strong>ibrosis based on a study <strong>of</strong> the International Working<br />
Group for Myel<strong>of</strong>ibrosis Research and Treatment. Blood. 2009;113:2895-2901.<br />
15. Bjorkholm M, Derolf AR, Hultcrantz M, et al. Treatment-related risk<br />
factors for transformation to acute myeloid leukemia and myelodysplastic<br />
syndromes in myeloproliferative neoplasms. J Clin Oncol. 2011;29:2410-2415.<br />
16. Ruggeri M, Rodeghiero F, Tosetto A, et al. Postsurgery outcomes in<br />
424<br />
REFERENCES<br />
FRANCESCO PASSAMONTI<br />
patients with polycythemia vera and essential thrombocythemia: A retrospective<br />
survey. Blood. 2008;111:666-671.<br />
17. Passamonti F, Randi ML, Rumi E, et al. Increased risk <strong>of</strong> pregnancy<br />
complications in patients with essential thrombocythemia carrying the JAK2<br />
(617V�F) mutation. Blood. 2007;110:485-489.<br />
18. Gangat N, Caramazza D, Vaidya R, et al. DIPSS plus: A refined<br />
Dynamic International Prognostic Scoring System for primary myel<strong>of</strong>ibrosis<br />
that incorporates prognostic information from karyotype, platelet count, and<br />
transfusion status. J Clin Oncol. 2011;29:392-397.<br />
19. Tam CS, Abruzzo LV, Lin KI, et al. The role <strong>of</strong> cytogenetic abnormalities<br />
as a prognostic marker in primary myel<strong>of</strong>ibrosis: Applicability at the<br />
time <strong>of</strong> diagnosis and later during disease course. Blood. 2009;113:4171-4178.<br />
20. Caramazza D, Begna KH, Gangat N, et al. Refined cytogenetic-risk<br />
categorization for overall and leukemia-free survival in primary myel<strong>of</strong>ibrosis:<br />
A single center study <strong>of</strong> 433 patients. Leukemia. 2011;25:82-88.<br />
21. Guglielmelli P, Barosi G, Specchia G, et al. Identification <strong>of</strong> patients<br />
with poorer survival in primary myel<strong>of</strong>ibrosis based on the burden <strong>of</strong><br />
JAK2V617F mutated allele. Blood. 2009;114:1477-1483.<br />
22. Tefferi A, Lasho TL, Huang J, et al. Low JAK2V617F allele burden in<br />
primary myel<strong>of</strong>ibrosis, compared to either a higher allele burden or unmutated<br />
status, is associated with inferior overall and leukemia-free survival.<br />
Leukemia. 2008;22:756-761.<br />
23. Alchalby H, Badbaran A, Zabelina T, et al. Impact <strong>of</strong> JAK2V617Fmutation<br />
status, allele burden and clearance after allogeneic stem cell<br />
transplantation for myel<strong>of</strong>ibrosis. Blood. 2010;116:3572-3581.<br />
24. Tefferi A, Pardanani A, Lim KH, et al. TET2 mutations and their<br />
clinical correlates in polycythemia vera, essential thrombocythemia and<br />
myel<strong>of</strong>ibrosis. Leukemia. 2009; 23:905-911.<br />
25. Pardanani A, Guglielmelli P, Lasho TL, et al. Primary myel<strong>of</strong>ibrosis<br />
with or without mutant MPL: Comparison <strong>of</strong> survival and clinical features<br />
involving 603 patients. Leukemia. 2011;25:1834-1839.<br />
26. Tefferi A, Jimma T, Sulai NH, et al. IDH mutations in primary<br />
myel<strong>of</strong>ibrosis predict leukemic transformation and shortened survival: <strong>Clinical</strong><br />
evidence for leukemogenic collaboration with JAK2V617F. Leukemia.<br />
Epub 2011 Sept 13.<br />
27. Guglielmelli P, Biamonte F, Score J, et al. EZH2 mutational status<br />
predicts poor survival in myel<strong>of</strong>ibrosis. Blood. 2011;118:5227-5234.<br />
28. Tefferi A, Lasho TL, Patnaik MM, et al. JAK2 germline genetic<br />
variation affects disease susceptibility in primary myel<strong>of</strong>ibrosis regardless <strong>of</strong><br />
V617F mutational status: Nullizygosity for the JAK2 46/1 haplotype is<br />
associated with inferior survival. Leukemia. 2010;24:105-109.<br />
29. Dupriez B, Morel P, Demory JL, et al. Prognostic factors in agnogenic<br />
myeloid metaplasia: A report on 195 cases with a new scoring system. Blood.<br />
1996;88:1013-1018.<br />
30. Tefferi A, Siragusa S, Hussein K, et al. Transfusion-dependency at<br />
presentation and its acquisition in the first year <strong>of</strong> diagnosis are both equally<br />
detrimental for survival in primary myel<strong>of</strong>ibrosis-prognostic relevance is<br />
independent <strong>of</strong> IPSS or karyotype. Am J Hematol. 2010;85:14-17.
AROUND THE WORLD IN ALMOST 80 MINUTES:<br />
LUNG CANCER CARE AND RESEARCH<br />
CHAIR<br />
Gilberto Schwartsmann, MD, PhD<br />
Federal University <strong>of</strong> Rio Grande do Sul<br />
Porto Alegre, Brazil<br />
SPEAKERS<br />
Tony Mok, MD<br />
Prince <strong>of</strong> Wales Hospital<br />
Shatin, Hong Kong<br />
Tudor E. Ciuleanu, MD<br />
Institute Ion Chiricuta<br />
Cluj-Napoca, Romania
Lung Cancer in Brazil<br />
Overview: Cancer is now the second leading cause <strong>of</strong> death<br />
in Brazil (after cardiovascular diseases) and a public health<br />
problem, with around 500,000 new cases in <strong>2012</strong>. Excluding<br />
nonmelanoma skin cancer, lung cancer is the second most<br />
incident cancer type in men, with 17,210 expected new cases.<br />
In women, it is the fifth most incident cancer, with 10,110<br />
expected new cases. The estimated age-adjusted lung cancer<br />
mortality rate is about 13/100,000 for men and 5.4/100,000 for<br />
women. Lung cancer rates in men increased until the early<br />
1990s and decreased thereafter, especially in the younger<br />
population. In contrast, a steady upward trend was observed<br />
for women. The positive effects in men were probably due to<br />
the successful anti-tobacco campaign conducted in Brazil<br />
over the last decades, which led to a decrease in the adult<br />
WORLDWIDE, LUNG cancer is the number one cause<br />
<strong>of</strong> cancer death in men and the second in women,<br />
with more than 1.6 million new cases and 1.4 million deaths<br />
every year. Notably, the majority <strong>of</strong> lung cancer cases now<br />
occur in developing countries (55%), a substantial increase<br />
since the 1980s, when only one-third <strong>of</strong> cases occurred in<br />
these regions. 1<br />
Considering that cigarette smoking is the main causative<br />
factor for about four <strong>of</strong> five lung cancer deaths in men and for<br />
one-half <strong>of</strong> deaths in women, the variation in incidence<br />
largely reflects the patterns <strong>of</strong> tobacco consumption. Environmental<br />
exposure to other causative factors—such as<br />
asbestos, radon, arsenic, radiation, air pollution, coal smoke,<br />
and indoor emission from unventilated coal-fueled stoves<br />
and cooking—is also relevant in specific regions. 2<br />
Epidemiology<br />
Brazil is the largest country in South America, with an<br />
estimated population <strong>of</strong> more than 190 million. Cancer is<br />
now the second most common cause <strong>of</strong> death in most<br />
geographic regions in the country (Fig. 1). In <strong>2012</strong>, more<br />
than 500,000 new cancer cases are estimated, equally distributed<br />
between genders. Lung cancer will be responsible<br />
for around 17,210 new cancer cases in men and 10,110 new<br />
cases in women (Fig. 2). Male lung cancer increased until the<br />
early 1990s and decreased in the 2000s, especially in the<br />
younger population. In women, there was a steady upward<br />
trend for cancers <strong>of</strong> the lung during the same period. 3<br />
Recent data from the Brazilian National Cancer Institute<br />
estimate that lung cancer is the second and fifth most<br />
common type <strong>of</strong> cancer among men and women, respectively<br />
(not including nonmelanoma skin cancers) (Fig. 2). Trends<br />
in lung cancer mortality showed that the age-adjusted<br />
mortality rate among men increased from 10.6 to 13.1/<br />
100,000 between the years 1979 and 2004, with an increase<br />
from 3.0 to 5.4/100,000 among women during the same<br />
period. As a rule, lung cancer mortality rates in Brazil are<br />
significantly higher among men. Specific rates for men over<br />
the age <strong>of</strong> 64 and for women <strong>of</strong> all ages are increasing.<br />
Notably, there is a lesser risk <strong>of</strong> mortality among men born<br />
after 1950 and an increasing risk across all cohorts among<br />
women. The results regarding younger generations indicate<br />
that present trends are likely to continue. Interestingly, the<br />
cohort effect in women points to an increasing trend in<br />
426<br />
By Gilberto Schwartsmann, MD, PhD<br />
smoking population, from 32% in the early 1980s to 17% in the<br />
2000s. Although the Brazilian National Cancer Institute is<br />
strongly committed to providing excellence in multimodality<br />
care to cancer patients, limitations in availability and adequate<br />
geographic distribution <strong>of</strong> specialists and wellequipped<br />
cancer centers are evident. Major disparities in<br />
patient access to proper staging and state-<strong>of</strong>-the-art treatment<br />
still exist. Considering that World Health Organization<br />
(WHO) <strong>of</strong>ficials estimate that cancer will become the number<br />
one cause <strong>of</strong> death in most developing countries, including<br />
Brazil, in the next decades, it is highly recommended for<br />
government authorities to implement firm actions to face this<br />
tremendous challenge.<br />
mortality rates, whereas the reduction in rates in men<br />
younger than age 65 suggests that the above-mentioned<br />
trend should continue, probably as a consequence <strong>of</strong> the<br />
tobacco control measures adopted after the 1980s. 4<br />
Lung Cancer in Never Smokers<br />
As previously reported, patients who have never smoked<br />
tend to have a better survival rate than patients who did<br />
smoke, regardless <strong>of</strong> gender and histologic type. This was<br />
also illustrated by a report by Brazilian investigators from a<br />
large academic hospital in Sao Paulo, Brazil, where 56 <strong>of</strong><br />
285 (19%) patients with non-small cell lung cancer (NSCLC)<br />
were never smokers. They were more likely to be female<br />
(68% vs. 32%) and have adenocarcinoma (70% vs. 51%). 5<br />
Prevention and Early Diagnosis<br />
Brazil was one <strong>of</strong> the first countries to support the WHO<br />
Framework Convention on Tobacco Control and, over the<br />
last few decades, implemented several tobacco control measures,<br />
including banning smoking in public places, restricting<br />
tobacco advertising and promotion, counter-advertising,<br />
raising the price <strong>of</strong> cigarettes, and providing counseling for<br />
tobacco dependence. As a result, the rate <strong>of</strong> active smokers<br />
dropped from approximately 32% in the late 1980s to 17% in<br />
the 2000s. Brazilian authorities are projecting a further<br />
0.3% annual reduction in smoking, with an expected smoking<br />
population <strong>of</strong> approximately 11% by 2020. 6<br />
Worldwide, late diagnosis is a critical component <strong>of</strong> the<br />
dismal survival rates <strong>of</strong> patients with lung cancer. As<br />
localized disease can be treated with curative intent, new<br />
hope is derived from the identification <strong>of</strong> effective screening<br />
efforts. Recent trials have demonstrated that lung cancer<br />
From the Department <strong>of</strong> Medical <strong>Oncology</strong>, Hospital de Clínicas de Porto Alegre, Federal<br />
University <strong>of</strong> Rio Grande do Sul, Porto Alegre, Brazil; and South-<strong>American</strong> Office for<br />
Anticancer Drug Development, Porto Alegre, Brazil.<br />
Author’s disclosure <strong>of</strong> potential conflicts <strong>of</strong> interest are found at the end <strong>of</strong> this article.<br />
Address reprint requests to Gilberto Schwartsmann, MD, PhD, Department <strong>of</strong> Medical<br />
<strong>Oncology</strong>, Hospital de Clinicas de Porto Alegre, Federal University <strong>of</strong> Rio Grande do Sul,<br />
Rua Ramiro Barcelos 2350/s399, CP 90 035-903, Porto Alegre, RS, Brazil; email:<br />
gilberto.ez@terra.com.br.<br />
© <strong>2012</strong> by <strong>American</strong> <strong>Society</strong> <strong>of</strong> <strong>Clinical</strong> <strong>Oncology</strong>.<br />
1092-9118/10/1-10
LUNG CANCER IN BRAZIL<br />
Fig. 1. Relative distribution <strong>of</strong> causes <strong>of</strong> death in Brazil, 2010. 31<br />
Abbreviation: CNS, central nervous system.<br />
mortality can be reduced by annual screening with low-dose<br />
CT, especially in high-risk populations. However, this strategy<br />
can be associated with possible harm. 7 Currently, there<br />
is still no formal recommendation for the routine use <strong>of</strong><br />
screening CT as standard clinical practice in Brazil.<br />
KEY POINTS<br />
● Cancer is now a public health problem in Brazil, with<br />
more than 500,000 new cases expected for the year<br />
<strong>2012</strong>.<br />
● Lung cancer will be responsible for 17,210 new cases<br />
in men and 10,110 new cases in women.<br />
● Brazil reduced the percentage <strong>of</strong> smokers from<br />
around 32% in the 1970s to 17% in 2010s, and a<br />
further annual decrease <strong>of</strong> about 0.3% is projected by<br />
government health authorities.<br />
● Diagnosis, staging, and treatment guidelines for patients<br />
with lung cancer follow the recommendations<br />
<strong>of</strong> the Brazilian <strong>Society</strong> <strong>of</strong> <strong>Clinical</strong> <strong>Oncology</strong> and the<br />
Brazilian Cancer <strong>Society</strong>, which are consistent with<br />
the National Comprehensive Cancer Network and<br />
European <strong>Society</strong> for Medical <strong>Oncology</strong> guidelines.<br />
● As a developing country with large cultural and<br />
socioeconomic contrasts and an estimated population<br />
<strong>of</strong> over 190 million inhabitants for the year <strong>2012</strong>,<br />
disparities in patient access to proper medical care<br />
are still a major challenge for cancer control in Brazil.<br />
Classification and Prognostic Factors<br />
Lung cancer is grouped as NSCLC and small-cell lung<br />
cancer (SCLC). NSCLC accounts for more than 85% <strong>of</strong> all<br />
cases and is classified as squamous (SCC) and nonsquamous<br />
cell carcinoma (NSCC). NSCC includes adenocarcinomas,<br />
large-cell carcinomas, and other subtypes. In a retrospective<br />
analysis <strong>of</strong> 240 consecutive patients treated at an academic<br />
hospital in Sao Paulo, Brazil, between 2000 and 2006, the<br />
most common NSCLC subtype was SCC. 8 Similar results<br />
were reported in a study performed in Manaus, Brazil,<br />
between 1995 and 2002. 9 In contrast, in a recent series <strong>of</strong><br />
patients with NSCLC from the Brazilian National Cancer<br />
Institute in Rio de Janeiro, Brazil, the most common histologic<br />
subtype was adenocarcinoma. 10 This pattern is being<br />
also observed in an NSCLC series from developed countries.<br />
11 As a rule, most patients with NSCLC (80%) present<br />
with locally advanced or advanced disease (stages III and<br />
IV), while a minority (20%) <strong>of</strong> patients have early disease<br />
(stages I and II) at presentation. 8,10<br />
Favorable prognostic factors for survival in patients with<br />
NSCLC include early stage, good performance status, no<br />
weight loss, and female gender. In contrast, p53 mutations<br />
and KRAS activation may predict a less favorable prognosis.<br />
12 In another study, 114 patients with NSCLC from Porto<br />
Alegre, Brazil, underwent tumor resection for stage I disease.<br />
The 5-year survival rate was 85.5% and 46.4% for<br />
women and men, respectively (p � 0.0001). 13<br />
Pathology<br />
Tumor specimens can be obtained from fine-needle aspiration,<br />
core needle, endobronchial or transbronchial biopsy,<br />
427
Fig. 2. Total number and percent distribution <strong>of</strong> the ten most frequent types <strong>of</strong> cancers projected for the year <strong>2012</strong> in Brazil (excluding<br />
nonmelanoma skin cancer). 3<br />
Abbreviation: CNS, central nervous system.<br />
and bronchial brushings or washings. Mediastinal lymph<br />
nodes are <strong>of</strong>ten sampled for staging. The pathologist provides<br />
information about the histologic type (or subtype), the<br />
presence and extent <strong>of</strong> tumor invasion, and surgical margins.<br />
The presence <strong>of</strong> EGFR or EML4-ALK mutations is also<br />
important, as it can direct treatment toward the use <strong>of</strong><br />
specific tyrosine kinase inhibitors (TKIs). 12<br />
In a study performed by investigators at the Brazilian<br />
National Cancer Institute, 162 paraffin-embedded specimens<br />
obtained from patients with NSCLC were analyzed for<br />
EGFR (exons 18 to 21), KRAS (exon 2), and BRAF (exons 11<br />
and 15) mutations and for MET amplification and ALK<br />
rearrangement (Fig. 3). 10 EGFR analysis was successful<br />
in 150 cases, and mutations were identified in 25.3% (38<br />
patients) <strong>of</strong> patients (6, 19, 13, and 5 mutations in exons 18,<br />
19, 20, and 21, respectively; 5 cases encompassing mutations<br />
in 2 different exons). These abnormalities were observed<br />
mainly in adenocarcinomas (52.8%) but were also frequent<br />
in SCC (36.1%).<br />
In the above study, KRAS mutations were observed in 30<br />
<strong>of</strong> 148 analyzed cases (20.3%), in which 76.7% were adenocarcinomas.<br />
BRAF mutations were found in 13 <strong>of</strong> 145<br />
patients (9.0%), mainly in squamous histology (61.5%). Notably,<br />
mutations in exon 11 in men with SCC were identified,<br />
although no case <strong>of</strong> V600E mutation was identified. 10<br />
The latter observation differs from previous reports <strong>of</strong> a high<br />
prevalence <strong>of</strong> this abnormality in patients with a BRAF<br />
mutation. 14 MET gene–increased copy number (mean � 5<br />
copies/cell) was observed in 21 <strong>of</strong> 152 evaluated cases<br />
(13.8%), mostly in adenocarcinomas (73.7%). ALK rearrangement<br />
was present in four <strong>of</strong> 161 cases (2.5%), three <strong>of</strong><br />
them with adenocarcinoma and one with bronchioloalveolar<br />
histology.<br />
During the surgical procedure, the removed tissue should<br />
428<br />
provide information about the status <strong>of</strong> resection margins,<br />
involvement <strong>of</strong> regional lymph nodes and the presence <strong>of</strong><br />
incidental nodules. The information on the WHO histologic<br />
type (and subtype), disease staging, and prognostic factors<br />
are essential for treatment planning. The recognition <strong>of</strong> a<br />
bronchioloalveolar subtype <strong>of</strong> adenocarcinoma is also relevant,<br />
as EGFR-TKIs are beneficial for these patients. 15<br />
Furthermore, immunohistochemistry is a critical tool in<br />
distinguishing between primary lung cancer and other neoplasms,<br />
such as mesothelioma, metastatic cancer <strong>of</strong> another<br />
origin, or neuroendocrine tumors. 12<br />
<strong>Clinical</strong> Evaluation<br />
GILBERTO SCHWARTSMANN<br />
The initial evaluation <strong>of</strong> a patient with NSCLC includes<br />
medical history, physical examination, laboratory tests,<br />
chest x-ray, abdominal ultrasound, and/or CT scans. Bronchoscopy<br />
is recommended for the diagnosis <strong>of</strong> central lesions,<br />
while percutaneous biopsy is usually performed in<br />
peripheral lesions. It is recommended that the patient be<br />
referred to a smoking cessation program. Staging includes a<br />
chest CT with the inclusion <strong>of</strong> the upper abdomen and<br />
adrenals.<br />
Mediastinoscopy is the gold standard for the study <strong>of</strong><br />
mediastinal lymph nodes. The sampling <strong>of</strong> mediastinal<br />
lymph nodes is essential, as CT scans have limitations to<br />
rule out lymph node involvement. Endobronchial ultrasoundtransbronchial<br />
needle aspiration (EBUS-TBNA) is also being<br />
considered for the mediastinal staging <strong>of</strong> patients with<br />
NSCLC worldwide. Unfortunately, only a small number <strong>of</strong><br />
cancer centers in Brazil are currently applying this technique<br />
in the routine treatment <strong>of</strong> patients with lung cancer.<br />
The use <strong>of</strong> PET imaging is also a valuable tool in the<br />
current staging <strong>of</strong> patients with NSCLC. However, because<br />
<strong>of</strong> its high cost and limited availability through the National
LUNG CANCER IN BRAZIL<br />
Fig. 3. Characteristics <strong>of</strong> groups mutated<br />
for EGFR, KRAS, BRAF, and TP53. 10<br />
Abbreviations: EGFR, epidermal growth<br />
factor receptor; SCC, squamous cell carcinoma.<br />
Health System in Brazil, its use is usually restricted to<br />
patients with early disease who are candidates for curative<br />
surgery. In the absence <strong>of</strong> specific symptoms, there is no<br />
indication for the use <strong>of</strong> brain MRI or bone scans in patients<br />
with stage IV disease. An MRI <strong>of</strong> the brain, however, is<br />
considered by many oncologists in Brazil in the case <strong>of</strong> lung<br />
adenocarcinoma, because <strong>of</strong> the higher risk <strong>of</strong> brain metastases.<br />
Finally, surgical resection without prior invasive<br />
testing may be an option in patients with a highly suspicious<br />
solitary pulmonary nodule.<br />
In a series <strong>of</strong> 291 patients with NSCLC treated in an<br />
academic hospital in Sao Paulo, Brazil, clinical staging was<br />
based on clinical and imaging studies. PET scan was not<br />
routinely performed. Pathologic staging differed from clinical<br />
staging in 33% <strong>of</strong> cases (15% were upstaged and 18%<br />
downstaged). Sensitivity, specificity, positive and negative<br />
predictive values, and accuracy for clinical staging were<br />
78%, 69%, 82%, 64%, and 67%, respectively. 16<br />
In another study performed in Porto Alegre, Brazil, the<br />
authors looked at the indications for MRI studies in patients<br />
with NSCLC. MRI was used mainly for patients with superior<br />
sulcus tumors, suspected spinal cord canal invasion, and<br />
suspected brain metastases. 17<br />
Biomarkers<br />
EGFR exon 19 deletion or exon 21 L858R mutation result<br />
in activation <strong>of</strong> the tyrosine kinase domain and are predictive<br />
<strong>of</strong> better sensitivity to EGFR-TKIs. These mutations are<br />
found in approximately 10% to 15% <strong>of</strong> white and 30% to 40%<br />
<strong>of</strong> Asian patients with NSCLC. 10,18 As described below,<br />
these results were not confirmed in a recent study performed<br />
by investigators <strong>of</strong> the Brazilian National Cancer Institute,<br />
in which 25.3% (38 <strong>of</strong> 150 cases) <strong>of</strong> patients with NSCLC<br />
showed these mutations. EGFR mutations were more common<br />
in women (60.6%), smokers (51.4%), and in the adeno-<br />
carcinoma subtype (54.5%). These mutations were also<br />
detected in never (15.2%) and previous (27.3%) smokers, as<br />
well as in a significant percentage <strong>of</strong> patients with the<br />
squamous cell subtype (33.4%). 10<br />
Patients with NSCLC with high ERCC1 expression in the<br />
tumor were shown to have a better overall survival than<br />
those with low ERCC1 expression. High ERCC1 levels were<br />
also associated with resistance to platinum-based chemotherapies.<br />
12,19 KRAS mutation is observed in approximately<br />
25% <strong>of</strong> adenocarcinomas. It is associated with worse survival<br />
and resistance to platinum-based chemotherapy or EGFR-<br />
TKIs. 12,20 Furthermore, high RRM1 levels are associated<br />
with better survival but poor response to gemcitabine or<br />
carboplatin. 12,21<br />
Treatment<br />
Surgical resection should be attempted in all patients with<br />
stage I or II good performance status NSCLC, as it provides<br />
the best chances <strong>of</strong> cure. 12 In a study performed in a cancer<br />
hospital in Sao Paulo, Brazil, including 737 patients with<br />
NSCLC, complete tumor resection was performed in 24.6%<br />
<strong>of</strong> patients. The overall 5-year survival rate was 28%, and<br />
the median survival was 18.9 months. 22<br />
The survival advantage <strong>of</strong> adjuvant cisplatin-based chemotherapy<br />
for patients with completely resected stage I, II,<br />
or III NSCLC was demonstrated in prospective randomized<br />
trials (IALT, NCIC, and ANITA). Postoperative cisplatinbased<br />
chemotherapy produced an absolute benefit <strong>of</strong> 5% in<br />
5-year survival, especially in patients with stage II-III<br />
disease with good performance status. Adjuvant chemotherapy<br />
was also beneficial for patients with stage I disease<br />
with � 4-cm tumors. 12,23<br />
Several therapeutic options can be considered for patients<br />
with stage IIIA disease. For unresectable cases <strong>of</strong> stage IIIA<br />
or IIIB disease, chemoradiation was shown to be superior to<br />
429
adiation alone. Concurrent chemoradiation showed better<br />
results as compared with sequential therapy. Most centers<br />
in Brazil adopt a cisplatin/etoposide concurrent chemoradiation<br />
regimen in this context, although other regimens are<br />
used, such as carboplatin/paclitaxel or cisplatin/vinblastine.<br />
The question <strong>of</strong> adding additional courses <strong>of</strong> chemotherapy<br />
after chemoradiation is unsolved.<br />
In a trial performed at the Brazilian National Cancer<br />
Institute, 30 patients with clinical stages IB to IIIA NSCLC<br />
who were candidates for surgical resection received three<br />
cycles <strong>of</strong> neoadjuvant chemotherapy. Patients without evidence<br />
<strong>of</strong> progression underwent mediastinoscopy. Those<br />
with negative lymph nodes underwent resection, whereas<br />
radiation was given to those with positive nodes. Twentythree<br />
patients (77%) responded to neoadjuvant chemotherapy,<br />
and complete resection rate was achieved in 21 patients<br />
(70%). 24<br />
For patients with stage IV disease and a good performance<br />
status, chemotherapy is the treatment <strong>of</strong> choice. With the<br />
exclusion <strong>of</strong> patients with central nervous system metastasis,<br />
chemotherapy greatly increases survival compared with<br />
best supportive care. 12 This was corroborated by a study <strong>of</strong><br />
patients with stage IV NSCLC in a cancer hospital in Sao<br />
Paulo, Brazil. 25 In another trial in three cancer centers in<br />
Brazil, 564 patients with stage IV disease were evaluated.<br />
Of those, 335 (59.4%) received chemotherapy. There was a<br />
great heterogeneity in drug regimens used in the patients.<br />
Again, overall survival was better with chemotherapy compared<br />
with best supportive care. 26<br />
Several agents showed objective responses in advanced<br />
stage NSCLC, including paclitaxel, docetaxel, vinorelbine,<br />
etoposide, pemetrexed, irinotecan, and gemcitabine.<br />
Platinum-containing doublets can lead to 30% to 40% oneyear<br />
survival rates and are superior to single agents. 12<br />
Carboplatin/paclitaxel, cisplatin/vinorelbine, cisplatin/gemcitabine,<br />
carboplatin/docetaxel, or carboplatin/pemetrexate<br />
are drug regimens commonly used in the treatment <strong>of</strong><br />
patients with advanced NSCLC in Brazil.<br />
It should be emphasized that in patients with good performance<br />
status and solitary brain metastasis, surgical<br />
resection may improve survival. The results for solitary<br />
metastatic lesions in other sites are controversial. 12<br />
New targeted therapies were shown to produce tumor<br />
responses in patients with advanced lung cancer. Bevacizumab<br />
may be used in combination with paclitaxel and<br />
carboplatin in patients with advanced stage NSCLC with<br />
nonsquamous histology and no hemoptysis. In a metaanalysis<br />
published by investigators from Campinas, Brazil,<br />
bevacizumab improved response rate and progression-free<br />
survival in patients with NSCLC receiving chemotherapy,<br />
while the effect on overall survival was uncertain. 27 Erlotinib<br />
is approved as first-line therapy in patients with EGFR<br />
mutation–positive stage IV NSCLC. It can also be used in<br />
patients who progressed after one prior chemotherapy regimen.<br />
Furthermore, cetuximab showed an increase in overall<br />
survival in patients with stage IV disease when given in<br />
430<br />
combination with cisplatin/vinorelbine. 12 The continuation<br />
<strong>of</strong> a biologic agent, such as bevacizumab or cetuximab, can<br />
be an option after four to six chemotherapy courses for<br />
responding patients or stable disease. At present, the abovementioned<br />
biological agents are not provided routinely as<br />
part <strong>of</strong> the standard treatment <strong>of</strong> patients with NSCLC<br />
through the Brazilian National Health System.<br />
Pemetrexed is an alternative maintenance therapy. Two<br />
recent studies have demonstrated progression-free and overall<br />
survival benefit with the administration <strong>of</strong> switch maintenance<br />
with pemetrexed (for nonsquamous histology only),<br />
erlotinib, or docetaxel, following four to six courses <strong>of</strong><br />
platinum-containing chemotherapy. It should be emphasized,<br />
however, that there is no data to support the continuation<br />
<strong>of</strong> combination chemotherapy beyond four to six<br />
treatment courses. 28<br />
Systemic Therapy for SCLC<br />
GILBERTO SCHWARTSMANN<br />
SCLC represents approximately 10% to 15% <strong>of</strong> all lung<br />
cancers. As SCLC has a high risk <strong>of</strong> early metastatic<br />
dissemination, only a minority <strong>of</strong> cases (30%) have limited<br />
disease at presentation. Tumor responses to both chemotherapy<br />
and radiotherapy are frequent (50% to 80%), but<br />
relapses are the rule. Platinum plus etoposide chemotherapy<br />
is the cornerstone regimen in the treatment <strong>of</strong> these<br />
patients. In most centers in Brazil, platinum/etoposide chemotherapy<br />
combined with thoracic radiotherapy is the<br />
choice for patients with limited disease, while chemotherapy<br />
alone is used in patients with extensive disease. Because <strong>of</strong><br />
the high risk <strong>of</strong> central nervous system involvement, prophylactic<br />
cranial irradiation is indicated for complete responders.<br />
Five-year survival <strong>of</strong> patients presenting with<br />
limited disease varies between 10% and 25%, although it<br />
does not exceed 10% at 2 years in patients with extensive<br />
disease. Most patients relapse within the first two years,<br />
and there are few second-line treatment options for these<br />
patients.<br />
Irinotecan was shown to produce objective responses in<br />
patients with SCLC. In a phase II trial conducted in Japan,<br />
irinotecan was associated with a median survival <strong>of</strong> 13<br />
months in patients with extensive-stage disease. Subsequently,<br />
a phase III trial comparing cisplatin/irinotecan<br />
with cisplatin/etoposide demonstrated superior median<br />
1-year and 2-year survival rates for the irinotecan arm. 29<br />
This was corroborated by a meta-analysis conducted by<br />
investigators from Sao Paulo, Brazil, that included eight<br />
trials and 3,086 patients with SCLC. 30 Presently, cisplatin/<br />
etoposide, cisplatin/irinotecan, and carboplatin/irinotecan<br />
are the most popular drug regimens used for the treatment<br />
<strong>of</strong> SCLC in Brazil. Most academic hospital still favor the use<br />
<strong>of</strong> cisplatin/etoposide with concomitant irradiation as firstline<br />
therapy for patients with limited disease and this<br />
regimen alone for the initial treatment <strong>of</strong> patients with<br />
extensive disease. The cyclophosphamide/doxorubicin/vincristine<br />
regimen is now rarely used as first-line therapy for<br />
this disease in Brazil.
LUNG CANCER IN BRAZIL<br />
Author’s Disclosure <strong>of</strong> Potential Conflicts <strong>of</strong> Interest<br />
Author<br />
Gilberto Schwartsmann*<br />
*No relevant relationships to disclose.<br />
Employment or<br />
Leadership<br />
Positions<br />
Consultant or<br />
Advisory Role<br />
1. Siegel R, Naishadham D, Jemal A. Cancer statistics, CA Cancer J Clin.<br />
<strong>2012</strong>;62:10-29.<br />
2. Ferlay J, Shin HR, Bray F, et al. Estimates <strong>of</strong> worldwide burden <strong>of</strong><br />
cancer in 2008: GLOBOCAN 2008. Int J Cancer. 2010;127:2893-2917.<br />
3. Instituto Nacional de Câncer and Ministério da Saúde. Estimativa <strong>2012</strong>:<br />
Incidência de câncer no Brasil. Rio de Janeiro: INCA. http://www.inca.gov.br.<br />
Accessed January <strong>2012</strong>.<br />
4. Souza MC, Vasconcelos AGG, Cruz OG. Trends in lung cancer mortality<br />
in Brazil from the 1980s into the early 21st century: age-period-cohort<br />
analysis. Cad Saude Publica. <strong>2012</strong>;28:21-30.<br />
5. Santoro IL, Ramos RP, Franceschini J, et al. Non-small cell lung cancer<br />
in never smokers: A clinical entity to be identified. Clinics (Sao Paulo).<br />
2011;66:1873-1877.<br />
6. Monteiro CA, Cavalcante TM, Moura EC, et al. Population-based evidence<br />
<strong>of</strong> a strong decline in the prevalence <strong>of</strong> smokers in Brazil (1989-2003).<br />
Bull World Health Organ. 2007;85:527-534.<br />
7. McLoud TC. Initial results <strong>of</strong> the National Lung Cancer Screening Trial.<br />
Cancer Imaging. 2011;11:S85.<br />
8. Novaes FT, Cataneo DC, Ruiz Junior RL, et al. Lung cancer: Histology,<br />
staging, treatment and survival. J Bras Pneumol. 2008;34:595-600.<br />
9. Westphal FL, Lima LC, Andrade EO, et al. Characteristics <strong>of</strong> patients<br />
with lung cancer in the city <strong>of</strong> Manaus. Brazil J Bras Pneumol. 2009;35:157-<br />
163.<br />
10. Melo, ACD, Inada HKP, Barros M, et al. Non-small cell lung cancer<br />
(NSCLC) genotyping in a Brazilian cohort. Poster Session P2.123 presented<br />
at 14th World Conference on Lung Cancer; 2011; Amsterdam, NL.<br />
11. Janssen-Heijnen ML, Coebergh JW. Trends in incidence and prognosis<br />
<strong>of</strong> the histological subtypes <strong>of</strong> lung cancer in North America, Australia, New<br />
Zealand and Europe. Lung Cancer. 2001;31:123-137.<br />
12. Ettinger DS, Akerley W, Bepler G, et al. Non-small cell lung cancer.<br />
J Natl Compr Canc Netw. 2010;8:740-801.<br />
13. Chatkin JM, Abreu CM, Fritscher CC, et al. Is there a gender difference<br />
in non-small cell lung cancer survival? Gend Med. 2004;1:41-47.<br />
14. Paik PK, Arcila ME, Fara M, et al. <strong>Clinical</strong> characteristics <strong>of</strong> patients<br />
with lung adenocarcinomas harboring BRAF mutations. J Clin Oncol. 2011;<br />
29:2046-2051.<br />
15. Wislez M, Lavolé A, Gounant V, et al. Bronchiolar-alveolar carcinoma:<br />
From concept to innovative therapeutic strategies. Presse Med. 2011;40:389-<br />
397.<br />
16. Younes RN, Schutz FA, Gross JL. Preoperative and pathological<br />
staging <strong>of</strong> NSCLC: Retrospective analysis <strong>of</strong> 291 cases. Rev Assoc Med Bras.<br />
2010;56:237-241.<br />
17. Hochhegger B, Marchiori E, Sedlaczek O, et al. MRI in lung cancer: A<br />
pictorial essay. Br J Radiol. 2011;84:661-668.<br />
18. Ren JH, He WS, Yan GL, et al. EGFR mutations in non-small-cell lung<br />
cancer among smokers and non-smokers: A meta-analysis. Environ Mol<br />
Mutagen. <strong>2012</strong>;53:78-82.<br />
Stock<br />
Ownership Honoraria<br />
REFERENCES<br />
Research<br />
Funding<br />
Expert<br />
Testimony<br />
Other<br />
Remuneration<br />
19. Vilmar A, Sørensen JB. Excision repair cross-complementation group 1<br />
(ERCC1) in platinum-based treatment <strong>of</strong> non-small cell lung cancer with<br />
special emphasis on carboplatin: A review <strong>of</strong> current literature. Lung Cancer.<br />
2009;64:131-139.<br />
20. Gaughan EM, Costa DB. Genotype-driven therapies for non-small cell<br />
lung cancer: Focus on EGFR, KRAS and ALK gene abnormalities. Ther Adv<br />
Med Oncol. 2011;3:113-125.<br />
21. Ryu JS, Shin ES, Nam HS, et al. Differential effect <strong>of</strong> polymorphisms <strong>of</strong><br />
CMPK1 and RRM1 on survival in advanced non-small cell lung cancer<br />
patients treated with gemcitabine or taxane/cisplatinum. J Thorac Oncol.<br />
2011;6:1320-1329.<br />
22. Younes RN, Deutsch F, Badra C, et al. Nonsmall cell lung cancer:<br />
Evaluation <strong>of</strong> 737 consecutive patients in a single institution. Rev Hosp Clin<br />
Fac Med Sao Paulo. 2004;59:119-127.<br />
23. NSCLC Meta-analyses Collaborative Group, Arriagada R, Auperin A,<br />
et al. Adjuvant chemotherapy, with or without postoperative radiotherapy, in<br />
operable non-small-cell lung cancer: Two meta-analyses <strong>of</strong> individual patient<br />
data. Lancet. 2010;375:1267-1277.<br />
24. Martins RG, Dienstmann R, de Biasi P, et al. Phase II trial <strong>of</strong><br />
neoadjuvant chemotherapy using alternating doublets in non-small-cell lung<br />
cancer. Clin Lung Cancer. 2007;8:257-263.<br />
25. Anelli A, Lima CA, Younes RN, et al. Chemotherapy versus best<br />
supportive care in stage IV non-small cell lung cancer, non metastatic to the<br />
brain. Rev Hosp Clin Fac Med Sao Paulo. 2001;56:53-58.<br />
26. Naime FF, Younes RN, Kersten BG, et al. Metastatic non-small cell<br />
lung cancer in Brazil: Treatment heterogeneity in routine clinical practice.<br />
Clinics (Sao Paulo). 2007;62:397-404.<br />
27. Botrel TE, Clark O, Clark L, et al. Efficacy <strong>of</strong> bevacizumab (Bev) plus<br />
chemotherapy (CT) compared to CT alone in previously untreated locally<br />
advanced or metastatic non-small cell lung cancer (NSCLC): systematic<br />
review and meta-analysis Lung Cancer. 2011;74:89-97.<br />
28. Azzoli CG, Temin S, Aliff T, et al. 2011 Focused Update <strong>of</strong> 2009<br />
<strong>American</strong> <strong>Society</strong> <strong>of</strong> <strong>Clinical</strong> <strong>Oncology</strong> <strong>Clinical</strong> Practice Guideline Update on<br />
Chemotherapy for Stage IV Non-Small-Cell Lung Cancer. J Clin Oncol.<br />
2011;29:3825-3831.<br />
29. Planchard D, Le Péchoux C. Small cell lung cancer: New clinical<br />
recommendations and current status <strong>of</strong> biomarker assessment. Eur J Cancer.<br />
2011;47 Suppl 3:S272-S283.<br />
30. Lima JP, dos Santos LV, Sasse EC, et al. Camptothecins compared with<br />
etoposide in combination with platinum analog in extensive stage small cell<br />
lung cancer: Systematic review with meta-analysis. J Thorac Oncol. 2010;5:<br />
1986-1993.<br />
31. Brazilian Institute <strong>of</strong> Geography and Statistics - Instituto Brasileiro de<br />
Geografia e Estatística (IBGE), Brazil, 2011. www.ibge.gov.br. Accessed<br />
January <strong>2012</strong>.<br />
431
Research and Standard Care: Lung Cancer<br />
in China<br />
By Tony S. Mok, MD, Qing Zhou, MD, and Yi-Long Wu, MD<br />
Overview: China has an enormous burden from rising tobacco<br />
consumption and lung cancer incidence. Governmental<br />
intervention on lung cancer prevention is insufficient, and both<br />
incidence and mortality related to lung cancer are still on the<br />
rise. Treatment guidelines are available, but heterogeneity in<br />
the quality <strong>of</strong> care between centers, especially the disparity<br />
between urban and rural areas, have resulted in inconsistent<br />
care to patients with lung cancer. Despite knowledge on<br />
molecular-targeted therapy, only a small fraction <strong>of</strong> patients<br />
CHINA, A COUNTRY in rapid development, is confronted<br />
with a major health hazard. Success <strong>of</strong> her<br />
development comes with the price <strong>of</strong> pollution, change in<br />
dietary habits, and increase in tobacco consumption. Cancer<br />
has become the leading cause <strong>of</strong> death in China. According to<br />
the Ministry <strong>of</strong> Health in China, mortality related to cancer<br />
has risen by 80% during the last 30 years to an alarming<br />
number <strong>of</strong> 1.8 million cancer deaths in 2010. Cancer incidence<br />
rose from 74.2/100,000 in 1970 to 135.9/100,000 in<br />
2004. 1 Among the death tolls, a substantial proportion is<br />
attributed to lung cancer. Similar to other developed countries,<br />
majority <strong>of</strong> patients with lung cancer are diagnosed at<br />
advanced stage, and only a small proportion <strong>of</strong> patients have<br />
curable disease at presentation. During the last three decades,<br />
deaths related to lung cancer have increased by 465%.<br />
But different from other developed countries, the quality <strong>of</strong><br />
health care in this most populated country is highly heterogeneous.<br />
The tremendous success in economic development<br />
brings wealth to a small sector <strong>of</strong> society, whereas a large<br />
proportion <strong>of</strong> residents are still living close to poverty line.<br />
With the lack <strong>of</strong> universal health care system, many suffer<br />
from inadequate treatment as well as from disease itself. In<br />
this review, we summarize the disease burden <strong>of</strong> lung cancer<br />
in China, the current treatment standard across the country,<br />
and the impressive development in clinical/translational<br />
research on this dreadful illness.<br />
The Burden<br />
The burden starts with tobacco consumption. Being the<br />
world largest tobacco-producing country, China also consumes<br />
majority <strong>of</strong> her own production. The World Health<br />
Organization (WHO) estimated China to have 320 million<br />
smokers in 2008, and this may indirectly cause environmental<br />
tobacco exposure to an estimate <strong>of</strong> more than 500 million<br />
nonsmokers. With the lung cancer risk being six to 10 times<br />
higher in smokers and 30% to 60% higher in nonsmokers<br />
with exposure to second-hand smoke, the incidence <strong>of</strong> lung<br />
cancer in China is expected to increase. To date, there is<br />
still an absence in reversing the trend in the number <strong>of</strong><br />
smokers in China. This country generates over US$21 billion<br />
in tobacco tax annually, and some provinces such as<br />
Yunnan are basically dependent on this tax revenue. 2 It is<br />
unlikely that there will be dramatic control <strong>of</strong> tobacco<br />
production or consumption.<br />
China has signed the WHO Framework Convention on<br />
Tobacco Control in 2003. The scale <strong>of</strong> the health care burden<br />
is being recognized, but the total budget spent on tobacco<br />
432<br />
have access to routine EGFR mutation analysis. Platinumbased<br />
doublet chemotherapy remains the most commonly<br />
used regimen irrespective <strong>of</strong> mutation status. On a positive<br />
note, both clinical and translational research on lung cancer<br />
are in rapid progress. The Chinese Thoracic <strong>Oncology</strong> Group<br />
(CTONG) has already contributed substantially to the care <strong>of</strong><br />
patients with lung cancer and is expected to continue in the<br />
trend.<br />
control was estimated to be less than US$1 million, which is<br />
a very small fraction compared with the revenue generated<br />
by tobacco. In 2002, it was estimated in a national survey in<br />
30 provinces in China that smoking prevalence rate was<br />
35.8% (66% in men and 3.1% in women). 3 The greatest<br />
concern is the increasing number <strong>of</strong> young smokers. It was<br />
estimated that about 15 million young people between the<br />
ages <strong>of</strong> 13 and 18 years are regular smokers, and another 40<br />
million in this age group are occasional smokers. The median<br />
age <strong>of</strong> starting smoking is 19 years. 4 The smoking rate<br />
in men (60%) is much higher than that in women (� 5%), but<br />
the incidence <strong>of</strong> women smokers is rapidly rising.<br />
Given the grim situation in tobacco consumption, it is not<br />
surprising that lung cancer incidence has increased. More<br />
than 400,000 cancer deaths were attributed to tobacco<br />
consumption in 2005, among which lung cancer accounted<br />
for more than 240,000. 5 In 2004, the mortality rate <strong>of</strong> lung<br />
cancer was 41/100,000 in urban areas and 26/100,000 in<br />
rural areas. WHO estimated that total number <strong>of</strong> new cases<br />
<strong>of</strong> lung cancer may reach 1 million every year by 2025; thus<br />
the death toll will increase proportionally. 6<br />
The only logical way to reduce lung cancer incidence and<br />
related death is tobacco control. The China National Office<br />
for Cancer Prevention is responsible for reduction <strong>of</strong> cancer<br />
incidence and mortality, and their primary targets are lung,<br />
liver, and gastric cancers. Tobacco control is on the top <strong>of</strong><br />
their agenda, but investment in the program was relatively<br />
small comparing to the magnitude <strong>of</strong> the problem. Joining<br />
other countries, the WHO Framework Convention on Tobacco<br />
Control was signed in 2003 and ratified in 2005. The<br />
government had invested 10 million RMB on the project.<br />
The objective is to promote a smoke-free environment, build<br />
an antismoking network, and provide assistance in smoking<br />
cessation. But for the 320 million current smokers in China,<br />
this amount is highly insufficient. To date there is legislation<br />
against smoking in public places but lack <strong>of</strong> effort in<br />
execution <strong>of</strong> the law. Cigarette cost and taxes are still<br />
relatively low. Although there is a law against sales <strong>of</strong><br />
From the Department <strong>of</strong> <strong>Clinical</strong> <strong>Oncology</strong>, State Key Laboratory <strong>of</strong> South China, Hong<br />
Kong Cancer Institute, The Chinese University <strong>of</strong> Hong Kong, Hong Kong; Guangdong Lung<br />
Cancer Institute, Guangdong General Hospital and Guangdong Academy <strong>of</strong> Medical<br />
Sciences, Guangzhou, China.<br />
Authors’ disclosures <strong>of</strong> potential conflicts <strong>of</strong> interest are found at the end <strong>of</strong> this article.<br />
Address reprint requests to Tony S. Mok, MD; email: tony@clo.cuhk.edu.hk.<br />
© <strong>2012</strong> by <strong>American</strong> <strong>Society</strong> <strong>of</strong> <strong>Clinical</strong> <strong>Oncology</strong>.<br />
1092-9118/10/1-10
LUNG CANCER IN CHINA<br />
Fig 1. Chinese Guideline for Diagnosis<br />
and Treatment <strong>of</strong> Lung Cancer.<br />
Abbreviation: TNM, tumor mode metastasis.<br />
tobacco to minors, there is again no specific action on<br />
execution <strong>of</strong> this law. Unless the government has substantial<br />
change in attitude toward tobacco control, the burden<br />
related to lung cancer is unlikely to cease.<br />
The Treatment<br />
Treatment for lung cancer is quite diverse in China.<br />
Economic and technologic advantages in urban areas provide<br />
much better access to high-quality care, whereas treat-<br />
KEY POINTS<br />
● China has a large burden <strong>of</strong> lung cancer. Being the<br />
largest tobacco-producing country and a major consumer,<br />
its incidence <strong>of</strong> lung cancer is expected to<br />
increase unless major action on tobacco control is<br />
taken.<br />
● Treatment guidelines are available, but clinical practice<br />
may not always follow the guidelines.<br />
● With the exception <strong>of</strong> bevacizumab, most <strong>of</strong> the active<br />
agents for lung cancer are available in China. Drug<br />
cost is either partially reimbursed or self-financed.<br />
● Despite the high incidence <strong>of</strong> lung cancer, routine<br />
EGFR mutation analysis is available to only a small<br />
fraction <strong>of</strong> patients and is limited to major academic<br />
centers.<br />
● The Chinese Thoracic <strong>Oncology</strong> Group (CTONG) is an<br />
extremely active and well-organized study group.<br />
Within a few short years since establishment, they<br />
have already published major studies such as OPTI-<br />
MAL and INFORM.<br />
ment in rural area could be quite primitive. Unfortunately,<br />
limited information is available from rural area, and for the<br />
selected patients from rural areas who can afford the medical<br />
expenses, they would travel to urban areas for cancer<br />
care. Thus, in this review, we will focus on only the current<br />
status <strong>of</strong> lung cancer treatment in major cities.<br />
Treatment guidelines are available, although there is no<br />
legal or financial obligation for doctors to follow the guidelines.<br />
These are being used as references because the reimbursement<br />
system does not always follow the guidelines.<br />
The most commonly used guidelines include the Chinese<br />
Guideline for Diagnosis and Treatment <strong>of</strong> Lung Cancer 7 and<br />
the Chinese translation <strong>of</strong> the National Comprehensive<br />
Cancer Network (NCCN) Guideline. The former was developed<br />
by local experts, whereas the latter is a direct translation<br />
<strong>of</strong> a U.S. national guideline with minor variations.<br />
Contents <strong>of</strong> the two guidelines are very similar. The basic<br />
framework <strong>of</strong> the Chinese guideline is summarized in Figure<br />
1. Use <strong>of</strong> a multimodality approach is well recognized. The<br />
major difference is limited to the availability <strong>of</strong> specific<br />
medication in China. For example, bevacizumab has not<br />
been approved for use in lung cancer in China. The Chinese<br />
Guideline stated that “antiangiogenesis drug may be added<br />
if applicable,” whereas the Chinese translation <strong>of</strong> NCCN<br />
guidelines include endostar (a Chinese made angiotensin)<br />
and ginseng extract as an option for first-line treatment in<br />
combination with chemotherapy. Pemetrexed was not approved<br />
for first-line use until recently. Apart from the minor<br />
variation, the use <strong>of</strong> surgery for early-stage disease, chemoradiotherapy<br />
for local advanced disease, epidermal growth<br />
factor receptor tyrosine kinase inhibitor (EGFR TKI) for<br />
patients with EGFR mutation, and systemic chemotherapy<br />
for stage IV disease are well accepted. However, the guide-<br />
433
lines are not necessarily routinely translated into clinical<br />
practice.<br />
Shanghai is among the few cities that have detailed<br />
epidemiologic data on their patients with lung cancer. The<br />
Shanghai Municipal Center for Disease Control and Prevention<br />
reported a total <strong>of</strong> 25,927 male lung cancer deaths and<br />
10,468 female lung cancer deaths between 2003 and 2007.<br />
The pathologic distribution for adenocarcinoma, squamous<br />
cell carcinoma and small cell lung cancer were 58.9%, 35.4%,<br />
and 5.7%, respectively. Interestingly, the death rate has a<br />
slight trend <strong>of</strong> decline since 2001, and the 5-year survival<br />
rate has increased in the last decade. In Beijing, a total <strong>of</strong><br />
32,845 cases <strong>of</strong> lung cancer were diagnosed between 1998<br />
and 2007. 8 The crude incidence in women has increased.<br />
However, only approximately half <strong>of</strong> the patients (54.9%)<br />
had histologic confirmation <strong>of</strong> diagnosis. Histologic subtypes<br />
<strong>of</strong> squamous cell carcinoma and adenocarcinoma was 30.4%<br />
and 42.8%, respectively, in 1998, and changed to 24.1% and<br />
46.8% in 2007, respectively.<br />
The quality <strong>of</strong> thoracic surgery is generally good in urban<br />
major hospitals. Yang and colleagues 9 performed 621 anatomic<br />
lobectomies between 1996 and 2003, <strong>of</strong> which 113 were<br />
by video-assisted thorascopic surgery (VATS). Perioperative<br />
morbidity and mortality rates was 0% and 0.9%, respectively.<br />
The postoperative complication rate was 10.6%. The<br />
survival rates at 5 years for stage I, II, and III non-small cell<br />
lung cancer (NSCLC) were 79.1%, 45.2%, and 22.2%, respectively<br />
for VATS; and 81.6%, 47.2%, and 24.1%, respectively,<br />
for open lobectomy. Chen and colleagues from Peking University<br />
School <strong>of</strong> <strong>Oncology</strong> (personal communication, November<br />
2011) performed surgical resection in 660 patients<br />
and reported 5-year survival for stage I, II and III disease to<br />
be 78.3%, 52.3%, and 32.8%, respectively. These results are<br />
not different from the international standard.<br />
There is only limited information on the clinical practice<br />
in management <strong>of</strong> advanced-stage disease in China. A national<br />
survey on medical treatment <strong>of</strong> NSCLC was performed<br />
by the Sun Yat-Sen University Cancer Center in<br />
Guangzhou (personal communication, December 2011).<br />
Questionnaires were sent to 202 practicing doctors in 135<br />
centers in 12 major cities (Beijing, Shanghai, Guangzhou,<br />
Chengdu, Hangzhou, Xi’an, Jinan, Wuhan, Tianjin, Nanjing,<br />
Chongqing, and Shenyang). A total <strong>of</strong> 987 cases were<br />
recorded (381 early-stage lung cancer, 606 advanced-stage<br />
lung cancer). The majority <strong>of</strong> patients (525 <strong>of</strong> 606; 86.7%)<br />
with advanced-stage lung cancer received first-line chemo-<br />
434<br />
therapy. The most commonly used regimens included gemcitabine/platinum<br />
(27.4%), docetaxel/platinum (16.2%), and<br />
paclitaxel/platinum (13.5%). For patients with adenocarcinoma,<br />
approximately 16% would use pemetrexed/platinum.<br />
Only 4.9% used first-line EGFR TKI. The main reason is<br />
that the number <strong>of</strong> patients who underwent EGFR mutation<br />
analysis was low. Only 47 patients underwent EGFR mutation<br />
analysis, <strong>of</strong> whom 22 (47%) were positive for the<br />
mutation. Demographics <strong>of</strong> the tested population are not<br />
available, but the high mutation rate suggested that patients<br />
were clinically selected (female, nonsmoker, adenocarcinoma)<br />
for testing. Only 54 patients (9%) received secondline<br />
treatment. Interestingly, gemcitabine/platinum was<br />
still the most popular regimen, whereas single-agent docetaxel,<br />
gefitinib, and pemetrexed account for 13%, 11%, and<br />
9.3%, respectively. According to the authors, this is the first<br />
national survey in China. This reflects the true clinical<br />
practice across vast geographic areas. However, the study<br />
falls short because <strong>of</strong> the relatively small sample size for a<br />
large number <strong>of</strong> participating centers. On average, there<br />
were fewer than eight patients from each center, and these<br />
patients may not be representative <strong>of</strong> the clinical practice <strong>of</strong><br />
hosting hospital. This survey provides only a rough view <strong>of</strong><br />
current practice in management <strong>of</strong> advanced NSCLC in<br />
China.<br />
Molecular-targeted therapy has become an important part<br />
<strong>of</strong> disease management. China contributed a substantial<br />
proportion <strong>of</strong> patients to the IPASS study that established<br />
the role <strong>of</strong> gefitinib in patients with lung cancer with EGFR<br />
mutation. 10 However, the afore-described survey suggested<br />
that molecular testing is not widely practiced in China. A<br />
total <strong>of</strong> 26 hospitals in China have in-house capacity to test<br />
for EGFR mutation as a standard service, whereas another<br />
50 hospitals routinely send their specimens to other hospitals<br />
or private laboratories. In 2011, an estimated 12,000<br />
EGFR mutation analyses were performed, which is only a<br />
small fraction <strong>of</strong> patients with lung cancer in China. The<br />
cost <strong>of</strong> testing and availability <strong>of</strong> tumor sample are the<br />
major reasons for not testing; furthermore, the majority <strong>of</strong><br />
patients were not able to afford the expensive EGFR TKI.<br />
The Research<br />
MOK, ZHOU, AND WU<br />
Fig 2. Growth in Number <strong>of</strong> <strong>Clinical</strong> Trials in<br />
China.<br />
SEER, Surveillance, Epidemiology and End<br />
Results.<br />
China has become an important partner and contributor<br />
in laboratory, translational, and clinical research for lung<br />
cancer. The nation’s case number represents the largest<br />
patient resource in the world. Educated investigators are
LUNG CANCER IN CHINA<br />
well equipped and organized for clinical trials. Impressed<br />
by the large market size and potential growth, multiple<br />
pharmaceutical companies have set up headquarters and<br />
research facilities in China. The combination <strong>of</strong> pharmaceutical<br />
sponsorship, experienced researchers, and huge patient<br />
resource, positions China to be a leading country in lung<br />
cancer research.<br />
<strong>Clinical</strong> research in China took <strong>of</strong>f in 2004 on the discovery<br />
<strong>of</strong> EGFR mutation (Fig. 2). Incidence <strong>of</strong> EGFR mutation<br />
was markedly higher in the Asian population; thus, clinical<br />
research on the mutation would be most feasible in countries<br />
such as Japan and China. IPASS is the first major international<br />
study that compares gefitinib with standard chemotherapy<br />
in a clinical selected population for EGFR mutation,<br />
and the translational study has confirmed the superiority <strong>of</strong><br />
gefitinib in the mutated population. Sixteen centers from<br />
China contributed 372 patients (31%), and the study was<br />
completed in record speed and quality. On the basis <strong>of</strong> this<br />
collaboration, Pr<strong>of</strong>essor Yi Long Wu founded the Chinese<br />
Table 1. Ongoing Chinese Thoracic <strong>Oncology</strong> Group Studies<br />
Study No. NCT No. Investigational Drug(s) Title Status<br />
CTONG 0801 NCT00765687 Bisphosphonates Screening Non Small Cell Lung Cancer With Bone Metastasis and<br />
Efficacy and Safety Research <strong>of</strong> Receiving Bisphosphonates<br />
(BLEST)<br />
Ongoing<br />
CTONG 0802 NCT00874419 Erlotinib Erlotinib Versus Gemcitabine/Carboplatin in Chemo-naive Stage<br />
IIIB/IV Non-Small Cell Lung Cancer Patients With Epidermal<br />
Growth Factor Receptor (EGFR) Exon 19 or 21 Mutation(Optimal)<br />
Ongoing<br />
CTONG 0803 NCT00663689 Erlotinib Efficacy <strong>of</strong> Erlotinib for Brain Metastasis <strong>of</strong> Non-Small Cell Lung<br />
Cancer<br />
Ongoing<br />
CTONG 0804 NCT00770588 Gefitinib Assess the Efficacy, Safety and Tolerability <strong>of</strong> Gefitinib (Iressa 250<br />
mg) as Maintenance Therapy in Locally Advanced or Metastatic<br />
(Stage IIIB/IV) Non Small Cell Lung Cancer (NSCLC) (INFORM)<br />
Completed<br />
CTONG 0805 NCT00922584 Sorafenib Sorafenib Treatment in Non-Small Cell Lung Cancer After Failure <strong>of</strong><br />
Epidermal Growth Factor Receptor-Tyrosine Kinase Inhibitor<br />
Recruiting<br />
CTONG 0806 NCT00891579 Pemetrexed/gefitinib Study <strong>of</strong> Pemetrexed Versus Gefitinib in Patients With Locally<br />
Advanced or Metastatic Non Small Cell Lung Cancer Who Have<br />
Previously Received Platinum-Based Chemotherapy Without<br />
Epidermal Growth Factor Receptor (EGFR) Mutations<br />
Recruiting<br />
CTONG 0807 NCT00816868 Erlotinib/capecitabine A Study <strong>of</strong> TX Regimen as First-Line Treatment in Elderly Patients<br />
With Stage IIIB/IV Adenocarcinoma Non-Small Cell Lung Cancer<br />
Ongoing<br />
CTONG 0901 NCT01024413 Erlotinib/gefitinib Erlotinib Versus Gefitinib in Advanced Non Small Cell Lung Cancer<br />
With exon21 MutationA,306:A Randomized Trial<br />
Recruiting<br />
CTONG 0902 NCT00883779 Erlotinib A Study <strong>of</strong> Tarceva (Erlotinib) or Placebo in Combination With<br />
Platinum-Based Therapy as First Line Treatment in Patients With<br />
Advanced or Recurrent Non-Small Cell Lung Cancer<br />
Ongoing<br />
CTONG 0904 NCT01038661 Docetaxel Tax First-line Chemotherapy With Different Doses and Then<br />
Maintenance Therapy (TFINE)<br />
Recruiting<br />
CTONG 1001 NCT01319669 rhTPO <strong>Clinical</strong> Trial on the Prevention <strong>of</strong> Thrombocytopenia After First-line<br />
Chemotherapy<br />
Recruiting<br />
CTONG 1002 NCT01236716 Nab-paclitaxel/gemcitabine Nab-Paclitaxel Treatment in Advanced Squamous Cell Carcinoma <strong>of</strong><br />
Lung<br />
Not yet opening<br />
CTONG 1003 NCT01175096 Rad001 (everolimus) Safety and Tolerability Pr<strong>of</strong>ile <strong>of</strong> RAD001 Daily in Chinese Patients<br />
With Advanced Pulmonary Neuroendocrine Tumor<br />
Ongoing<br />
CTONG 1101 NCT01297101 Erlotinib A Single Arm, One Center, Phase II Study <strong>of</strong> Sequential<br />
Administration <strong>of</strong> Erlotinib in Combination with<br />
Gemcitabine/Cisplatin As Neoadjuvant Treatment in Patients with<br />
Stage IIIA NSCLC<br />
Recruiting<br />
CTONG 1102 Gefitinib Iressa Versus Chemo As Intermittent Treatment in Advanced NSCLC Not yet opening<br />
CTONG 1103 NCT 01407822 Erlotinib Erlotinib Versus Chemo As Neoadjuvant in IIIA-N2 NSCLC with<br />
EGFR Mutation in Exon 19 or 21<br />
Ongoing<br />
CTONG 1104 NCT01405079 Gefitinib Gefitinib Versus Vinorelbine/Platinum As Adjuvant Treatment in<br />
Stage II - IIIA(N1-N2) NSCLC with EGFR Mutation<br />
Ongoing<br />
Abbreviations: EGFR, epidermal growth factor receptor; NCT, National <strong>Clinical</strong> Trial; NSCLC, non-small cell lung cancer; rhTPO, recombinant human thrombopoietin;<br />
TX, treatment.<br />
Thoracic <strong>Oncology</strong> Group (CTONG) in 2007. CTONG’s mission<br />
is to design and develop multicenter clinical trials and<br />
provide a high level <strong>of</strong> evidence for clinical practice. They<br />
have now 23 active centers that participate either in<br />
CTONG studies or in global studies on behalf <strong>of</strong> CTONG.<br />
The number <strong>of</strong> lung cancer studies has been consistently<br />
increasing over the years, and CTONG is the leading group.<br />
Table 1 lists the ongoing CTONG studies. CTONG 0802<br />
(OPTIMAL study) is the first randomized study that established<br />
the role <strong>of</strong> erlotinib in patients with EGFR mutation,<br />
and its results have helped to register erlotinib as first-line<br />
therapy in China and Europe. 11 The INFORM study<br />
(CTONG 0804) is the first maintenance study on gefitinib,<br />
and the positive results were orally presented at the 2011<br />
ASCO Annual Meeting. With her relatively short history,<br />
CTONG has completed two important studies that influenced<br />
clinical management. Other ongoing studies such as a<br />
comparative study <strong>of</strong> erlotinib with gefitinib in patients with<br />
exon 21 mutation (CTONG 0901) and a comparative study <strong>of</strong><br />
435
gefitinib and chemotherapy as adjuvant therapy with patients<br />
with EGFR–positive, resectable lung cancer (CTONG<br />
1104) are extremely promising.<br />
There are multiple tumor banks in China, most <strong>of</strong> which<br />
are located in major academic centers. Investigators are<br />
very active in molecular genomic researches and have provided<br />
important translational data for novel drug development.<br />
Wu and colleagues are among the first to report in a<br />
meta-analysis from six centers in China on the relationship<br />
between EGFR mutation and clinical parameters. 12 There<br />
have been extensive translational works on EGFR mutations<br />
including EGFR heterogenecity and detection <strong>of</strong> EGFR<br />
mutation from plasma DNA. More importantly, these tumor<br />
banks started to become centralized and there is plan for<br />
unification <strong>of</strong> clinical data. More high-quality translational<br />
research is expected in the future.<br />
Authors’ Disclosures <strong>of</strong> Potential Conflicts <strong>of</strong> Interest<br />
Author<br />
Employment or<br />
Leadership<br />
Positions<br />
Consultant or<br />
Advisory Role<br />
Tony S. Mok AstraZeneca;<br />
AVEO;<br />
Boehringer<br />
Ingelheim;<br />
Bristol-Myers<br />
Squibb; Lilly;<br />
Merck Serono;<br />
Pfizer; Roche;<br />
Taiho<br />
Pharmaceutical<br />
Conclusion<br />
China is a developing country with an enormous burden<br />
from rising tobacco consumption and lung cancer incidence.<br />
Governmental intervention on lung cancer prevention is<br />
insufficient. Treatment guidelines are available, but the<br />
heterogeneity in quality <strong>of</strong> care between centers, especially<br />
the disparity between urban and rural areas, has resulted in<br />
inconsistent care for patients with lung cancer. Despite the<br />
knowledge <strong>of</strong> molecular-targeted therapy, only a small fraction<br />
<strong>of</strong> patients have access to routine EGFR mutation<br />
analysis. Platinum-based doublet chemotherapy remains<br />
the most commonly used regimen irrespective <strong>of</strong> mutation<br />
status. On a positive note, clinical and translational research<br />
on lung cancer are rapidly increasing. CTONG has<br />
already contributed substantially to the care <strong>of</strong> patients<br />
with lung cancer.<br />
Stock<br />
Ownership Honoraria<br />
AstraZeneca;<br />
AVEO;<br />
Boehringer<br />
Ingelheim; Lilly;<br />
Pfizer; Roche<br />
Qing Zhou*<br />
Yi-Long Wu AstraZeneca;<br />
Eli Lilly; Novartis;<br />
Pfizer; Roche;<br />
San<strong>of</strong>i<br />
*No relevant relationships to disclose.<br />
1. Ministry <strong>of</strong> Health <strong>of</strong> the People’s Republic <strong>of</strong> China. Report on the Third<br />
National Sampling Survey <strong>of</strong> Causes <strong>of</strong> Death. Beijing: The People’s Health<br />
Press; 2008.<br />
2. Lee AH, Liang Y. Tobacco Control in China. In Hu TW (ed). Tobacco<br />
Control Policy Analysis in China: Economics and Health. Berkeley, CA: World<br />
Scientific Publishing Company; 2007.<br />
3. Yang G, Fan L, Tan J, et al. Smoking in China: findings <strong>of</strong> the 1996<br />
National Prevalence Survey. JAMA. 2002;282:1247-1253.<br />
4. Ministry <strong>of</strong> Health <strong>of</strong> the People’s Republic <strong>of</strong> China. 2007 China Tobacco<br />
Control Report. Beijing: Ministry <strong>of</strong> Health <strong>of</strong> the People’s Republic <strong>of</strong> China;<br />
2007.<br />
5. Wang JB, Jiang Y, Wei WQ, et al. Estimation <strong>of</strong> cancer incidence and<br />
mortality attributable to smoking in China. Cancer Causes Control. 2010;21:<br />
959-965.<br />
6. Parkin DM, Bray F, Ferlay J, et al. Global cancer statistics, 2002. CA<br />
Cancer J Clin. 2005;55:74-108.<br />
7. Zhi XY, Wu YL, Bu H, et al. Chinese guideline on diagnosis and<br />
treatment <strong>of</strong> primary lung cancer. J Thorac Dis. 2011;4:88-101.<br />
436<br />
REFERENCES<br />
Research<br />
Funding<br />
AstraZeneca<br />
Expert<br />
Testimony<br />
MOK, ZHOU, AND WU<br />
Other<br />
Remuneration<br />
8. Wang N, Chen WQ, Zhu WX, et al. [Incidence trends and pathological<br />
characteristics <strong>of</strong> lung cancer in urban Beijing during period <strong>of</strong> 1998–2007].<br />
Zhonghua Yu Fang Yi Xue Za Zhi. 2011;45:249-254.<br />
9. Yang X, Wang S, Qu J. Video-assisted thoracic surgery (VATS) compares<br />
favorable with thoracotomy for the treatment <strong>of</strong> lung cancer: a five year<br />
outcome comparison. World J Surg. 2009;33:1857-1861.<br />
10. Mok TS, Wu YL, Thongprasert S, et al. Gefitinib or carboplatinpaclitaxel<br />
in pulmonary adenocarcinoma. N Engl J Med. 2009;361:947-<br />
957.<br />
11. Zhou C, Wu YL, Chen G, et al. Erlotinib versus chemotherapy as<br />
firstline treatment for patients with advanced EGFR mutation-positive nonsmall-cell<br />
lung cancer (OPTIMAL, CTONG-0802): a multicenter, open label,<br />
randomized phase III study Lancet Oncol. 2011;12:735-742.<br />
12. Wu YL, Zhong WZ, Li LY, et al. Epidermal growth factor receptor<br />
mutations and their correlation with gefitinib therapy in patients with<br />
non-small cell lung cancer: a meta-analysis based on updated individual<br />
patient data from six medical centers in mainland China. J Thorac Oncol.<br />
2007;2:430-439.
Research and Standard <strong>of</strong> Care: Lung Cancer<br />
in Romania<br />
Overview: In Romania, lung cancer is the most frequent<br />
cancer in men and fourth most frequent in women, and its<br />
incidence and mortality continue to rise. Recently, firm antitobacco<br />
policies were implemented, in agreement with the<br />
MPOWER strategies recommended by the World Health Organization<br />
(WHO). As <strong>of</strong> January <strong>2012</strong>, the recognized “<strong>of</strong>ficial”<br />
standard <strong>of</strong> care in lung cancer is still represented by the 2009<br />
edition <strong>of</strong> the European <strong>Society</strong> for Medical <strong>Oncology</strong> (ESMO)<br />
guidelines. Cancer treatment is free, as the National Program<br />
<strong>of</strong> <strong>Oncology</strong> covers the budget for all cytotoxic agents and<br />
targeted therapy. However, reimbursement for several expen-<br />
ROMANIA BECAME a member <strong>of</strong> the European Union<br />
in 2007. With 21.3 million inhabitants and a size<br />
roughly the same as Great Britain, the country inherited a<br />
predominantly “state-owned” health system, with welltrained<br />
health pr<strong>of</strong>essionals but limited resources and a<br />
centralized organization.<br />
Cancer is the second cause <strong>of</strong> death in Romania, following<br />
cardiovascular diseases, and the drugs used for its management<br />
are funded by a dedicated National Program <strong>of</strong><br />
<strong>Oncology</strong>. 1<br />
Patterns <strong>of</strong> Lung Cancer Epidemiology<br />
According to Globocan 2008, 70,300 new patients with<br />
cancer and 46,300 cancer deaths were encountered in Romania.<br />
2 Lung cancer was the most frequently diagnosed cancer<br />
in both genders. Incidence <strong>of</strong> lung cancer was 10,384 new<br />
cases (14.8% <strong>of</strong> all cancers), with an age standardized rate<br />
(ASR) <strong>of</strong> 30%. This compares favorably with nearby Hungary,<br />
with the highest rate <strong>of</strong> lung cancer in the world and<br />
an ASR <strong>of</strong> 52%. Mortality in Romania was 9,427 cases<br />
(20.4%) with an ASR <strong>of</strong> 26.8%. The 5-year prevalence was<br />
59.3%.<br />
According to the Romanian National Center for Health<br />
Statistics, the crude incidence <strong>of</strong> lung cancer rose continuously<br />
in a 25-year period, rising from 32.14% in men in 1982<br />
to 56.14% in 2007 and from 5.93% in women to 15.5%.<br />
Within 50 years, (from 1959 to 2010), the ASR for lung<br />
cancer mortality more than doubled both in men and women<br />
(men, 17.65% to 48.04%; women, 4.43% to 8.94%). Not all <strong>of</strong><br />
the eight regional Cancer Registries are yet providing accurate<br />
data because <strong>of</strong> a lack <strong>of</strong> trained personnel, insufficient<br />
funds, or methodological gaps. According to the Cluj County<br />
Registry, the cumulative lifetime risk <strong>of</strong> developing lung<br />
cancer is 7.46% for men and 1.21% for women.<br />
As compared with Western Europeans, more Romanian<br />
patients have squamous cell carcinoma and less adenocarcinoma.<br />
A study at the Pneumology Institute <strong>of</strong> Bucharest<br />
found 48% squamous cell, 29% adenocarcinomas, 7% large<br />
cell carcinomas, and 16% small cell lung cancers (SCLC)<br />
among 7,792 patients with confirmed lung carcinomas.<br />
Sixty-five percent <strong>of</strong> the patients were aged 55 to 74 years at<br />
diagnosis. 3<br />
Standards <strong>of</strong> Lung Cancer Care<br />
Until recently, the management <strong>of</strong> lung cancer was rather<br />
heterogenous among centers, although previous collabora-<br />
By Tudor E. Ciuleanu, MD, PhD<br />
sive drugs such as pemetrexed, erlotinib, and bevacizumab is<br />
individually approved by a centralized commission. All new<br />
drugs registered in Europe by the European Medicines Agency<br />
are concomitantly registered in Romania. However, no new<br />
drugs (such as gefitinib) or new indications (such as first-line<br />
tyrosine-kinase inhibitors or maintenance treatment) have<br />
been accepted for reimbursement since 2008. <strong>Clinical</strong> research<br />
is rapidly growing, and Romanian centers demonstrate<br />
a high recruitment rate in pivotal trials, despite initial delays<br />
because <strong>of</strong> a slow approval <strong>of</strong> the studies by authorities.<br />
tion efforts led to published Romanian therapeutic guidelines.<br />
4 In 2009, the National Commission <strong>of</strong> <strong>Oncology</strong><br />
<strong>of</strong>ficially adopted the European <strong>Society</strong> for Medical <strong>Oncology</strong><br />
(ESMO) guidelines, but no update followed. 5 Recently, the<br />
third Central European Cooperative <strong>Oncology</strong> Group<br />
(CECOG) Consensus on the Treatment <strong>of</strong> non-small cell<br />
lung cancer (NSCLC) was published and is expected to<br />
influence the management <strong>of</strong> lung cancer in Romania. 6<br />
Briefly, the consensus included the following recommendations:<br />
(1) early (operable) NSCLC: surgery (followed by<br />
adjuvant chemotherapy in stage II and III and in selected<br />
patients with stage IB disease; (2) locoregionally advanced<br />
NSCLC: combined chemoradiotherapy and; (3) advanced<br />
NSCLC: four to six courses <strong>of</strong> cisplatin-based chemotherapy<br />
with a third generation cytotoxic drug (pemetrexed in nonsquamous<br />
NSCLC) for first-line treatment, with bevacizumab<br />
providing modest benefit with added toxicity;<br />
epidermal growth factor receptor (EGFR) TKIs depending<br />
on EGFR-activating mutation status; pemetrexed as maintenance<br />
<strong>of</strong> response immediately following cisplatin-based<br />
chemotherapy resulting in significantly improved survival<br />
(particularly in patients with nonsquamous NSCLC). Docetaxel,<br />
pemetrexed (for nonsquamous NSCLC), or erlotinib<br />
for second-line therapy.<br />
First-line tryrosine kinase inhibitors (TKIs), as well as<br />
maintenance treatment, are not reimbursed in Romania,<br />
although the drugs are registered and available at the<br />
patient’s expense.<br />
The standard ESMO guidelines apply for SCLC, which<br />
include: (1) limited disease: combined chemoradiotherapy;<br />
(2) extended disease: chemotherapy alone (platinum and<br />
etoposide); (3) prophylactic cranial radiotherapy: recommended<br />
for responders, in both limited and extended disease<br />
and; (4) second-line chemotherapy: recommended in patients<br />
with good performance status, with drugs such as<br />
topotecan (oral or IV), ifosfamide, taxanes, or the CAV<br />
(cyclophosphamide, doxorubicin, vincristine) combination.<br />
From the Medical <strong>Oncology</strong> Department, Institute <strong>of</strong> <strong>Oncology</strong> Ion Chiricuta, Cluj-<br />
Napoca, Romania.<br />
Author’s disclosures <strong>of</strong> potential conflicts <strong>of</strong> interest are found at the end <strong>of</strong> this article.<br />
Address reprint requests Tudor Ciuleanu, MD, PhD, Institute <strong>of</strong> <strong>Oncology</strong> Ion Chiricuta,<br />
St. Republicii 34-36, Cluj-Napoca, 400015, Romania; email: tudor@iocn.ro.<br />
© <strong>2012</strong> by <strong>American</strong> <strong>Society</strong> <strong>of</strong> <strong>Clinical</strong> <strong>Oncology</strong>.<br />
1092-9118/10/1-10<br />
437
Some unsolved problems and disparities still persist in the<br />
management <strong>of</strong> patients with lung cancer because <strong>of</strong> a<br />
generally underfinanced health care system. Although a highquality<br />
disease management program can be <strong>of</strong>fered in many<br />
<strong>of</strong> the Romanian specialized cancer units, problems persist<br />
related to the access to a complete diagnostic work-up and an<br />
immediate multidisciplinary treatment. Each <strong>of</strong> the 40 counties<br />
in Romania have at least one oncology department<br />
(medical oncology with or without radiation oncology). The<br />
largest are the Cancer Institutes “Alexandru Trestioreanu”<br />
in Bucharest and “Ion Chiricuta” in Cluj-Napoca, with a<br />
third one recently inaugurated in Iasi.<br />
Fiberoptic diagnosis is performed by 150 pneumologists,<br />
but bronchoscopy departments are lacking in nine counties.<br />
The number <strong>of</strong> bronchoscopies per year raised continuously<br />
between 1990 and 2010 (2,100 vs. 10,000 procedures). This<br />
represent an overload <strong>of</strong> the main four centers (Bucuresti,<br />
Cluj, Timisoara, and Iasi), which together perform 75% <strong>of</strong> all<br />
the examinations. Lack <strong>of</strong> sufficient modern bronchoscopes<br />
or biopsy devices, lack <strong>of</strong> specialized pathology and cytology<br />
departments (decreased diagnostic accuracy), and lack <strong>of</strong><br />
sufficient intensive care and thoracic surgery departments<br />
(avoidance <strong>of</strong> the procedures with a higher risk <strong>of</strong> complications)<br />
have a negative effect on the quality <strong>of</strong> diagnosis. 3<br />
Thoracic surgery has evolved, but radical surgical treatment<br />
is confined to few specialized university centers.<br />
Twenty nine centers have radiotherapy facilities throughout<br />
the country, served by 110 radiation oncology specialists.<br />
However, a recent analysis revealed that overall, the equipment<br />
is outdated and insufficient—only 16 centers have<br />
high-energy radiotherapy machines (11 linear accelerators<br />
and 15 telecobalt machines, 6 <strong>of</strong> them outdated). It is<br />
estimated that only 27% <strong>of</strong> the patients that require radiotherapy<br />
have access to treatment, mostly because <strong>of</strong> a lack <strong>of</strong><br />
facilities. There is one megavoltage machine per 1 million<br />
inhabitants, even though the European Union standard<br />
recommends one machine per every 300,000 inhabitants. 7<br />
Therefore, concomitant chemoradiotherapy is rarely used,<br />
KEY POINTS<br />
● The <strong>of</strong>ficially recommended standard <strong>of</strong> care in Romania<br />
is in agreement with the 2009 European <strong>Society</strong><br />
for Medical <strong>Oncology</strong> guidelines.<br />
● Antitobacco policies are in place, following the<br />
MPOWER strategies recommended by the WHO.<br />
● Lung cancer treatment in Romania is free and funded<br />
by a National Program <strong>of</strong> <strong>Oncology</strong>, with the exception<br />
<strong>of</strong> pemetrexed and targeted treatments, which<br />
are individually reimbursed following the decision <strong>of</strong><br />
a centralized National Commission.<br />
● Although new treatments are promptly registered in<br />
Romania through a common European procedure, the<br />
decision for reimbursement is a multistep bureaucratic<br />
process that may take several years.<br />
● Romanian centers consistently contributed to recent<br />
research in lung cancer, and clinical research is<br />
rapidly growing, despite delays related to the initial<br />
approval by authorities.<br />
438<br />
Table 1. Smoking Prevalence during the Lifetime, Romania, 2004*<br />
Gender %<br />
Male 75.4<br />
Female<br />
Age Group (years)<br />
48.7<br />
15–24 61.8<br />
25–34 70.5<br />
35–44 65.0<br />
45–54 62.7<br />
55–64<br />
Education<br />
44.2<br />
Primary school (8 yr) 33.3<br />
Secondary school (12 yr) 65.8<br />
High school, university<br />
Income<br />
76.0<br />
Low 58.5<br />
Medium 67.6<br />
High<br />
Place <strong>of</strong> residence<br />
82.0<br />
Rural 53.9<br />
Small town 67.5<br />
Big city<br />
Geographical region<br />
68.5<br />
Moldova 62.9<br />
Muntenia 54.6<br />
Transilvania 68.2<br />
Bucharest 73.0<br />
Total 62.1<br />
* Source: Center for Health Policies and Services.<br />
with the sequential approach permitting the scheduling <strong>of</strong><br />
radiotherapy.<br />
Chemotherapy and targeted therapy are delivered by the<br />
270 specialists in medical oncology. Optimal chemotherapy<br />
can be <strong>of</strong>fered by medical oncology departments throughout<br />
the country in a timely manner. The registration <strong>of</strong> new<br />
drugs is simultaneous with the European Medicines Agency<br />
registration. The access to some expensive new molecules is,<br />
however, limited because <strong>of</strong> the reimbursement policies.<br />
Strategies for Tobacco Control<br />
TUDOR E. CIULEANU<br />
A study carried out in Bucharest, between 1999 and 2001,<br />
found 84% <strong>of</strong> 8,856 patients with lung cancer were smokers. 3<br />
Data from the Center for Health Policies and Services show<br />
a prevalence <strong>of</strong> smoking in the general population <strong>of</strong> 62.1%<br />
during the entire life. 8 The prevalence was higher in men<br />
age 25 to 34 years, with a high education level, high income,<br />
and from urban areas (Table 1). A study published in 2006<br />
ranked Romania 29 <strong>of</strong> 30 European countries in respect to<br />
the total measures taken to control tobacco smoking. The<br />
score was based on price, public place bans, public information<br />
campaign spending, advertising bans, health warnings,<br />
and treatment. 9<br />
The WHO Framework Convention on Tobacco Control<br />
(WHO FCTC) and its guidelines provide the foundation for<br />
countries to implement and manage tobacco control. The<br />
WHO introduced the MPOWER measures, intended to assist<br />
in the country-level implementation <strong>of</strong> effective interventions<br />
to reduce the demand for tobacco. Romanian<br />
tobacco control strategy follows the MPOWER measures<br />
recommended by the WHO:<br />
● Monitoring <strong>of</strong> tobacco use and prevention policies. A<br />
periodical monitoring <strong>of</strong> smoking is done every 2 years.
LUNG CANCER IN ROMANIA<br />
● Protection <strong>of</strong> people from tobacco smoke by banning<br />
smoking in public places, including bars and restaurants.<br />
The legislation has improved. From a total allowance<br />
<strong>of</strong> smoking to being permitted in public places only<br />
in separated, ventilated rooms and totally banned in<br />
medical care units and transportation.<br />
● Offering assistance for quitting tobacco use. Treatment<br />
for tobacco addiction has been freely available in the<br />
framework <strong>of</strong> a national program funded by the Ministry<br />
<strong>of</strong> Health since 2007. A toll-free quit-line is also<br />
available.<br />
● Warning about the dangers <strong>of</strong> tobacco. The health warnings<br />
are covering 30% to 40% <strong>of</strong> the main surfaces <strong>of</strong><br />
tobacco products, according to the provisions <strong>of</strong> the<br />
European Union Tobacco Directive. In 2008, Romania<br />
became the second European Union country implementing<br />
the pictorial health warnings on all tobacco products.<br />
● Enforcement <strong>of</strong> bans on tobacco advertising, promotion,<br />
and sponsorship on radio, TV, outdoor and indoor billboards,<br />
mass media, toys, nontobacco objects, and<br />
minors-intended events.<br />
● Raising taxes on tobacco, according to the European<br />
Union Tobacco Taxation requirements. A part <strong>of</strong> the<br />
funds collected finances the tobacco control program<br />
and the treatment <strong>of</strong> some smoking-related diseases.<br />
The National Tobacco Control Program implemented in<br />
2007 by the Ministry <strong>of</strong> Health includes prevention (coordinated<br />
by the National Center for Health Promotion), treatment,<br />
and monitoring activities (coordinated by the National<br />
Institute <strong>of</strong> Pneumology). Educational programs and communication<br />
campaigns are organized throughout the year,<br />
but special interventions are prepared on May 31 and<br />
November 17 (No-Tobacco Day). A toll-free quit-line is<br />
manned by psychologists. The medical treatment is free-<strong>of</strong>charge<br />
and prescribed by 60 doctors trained in smoking<br />
cessation.<br />
The State Sanitary Inspection and the National Authority<br />
for Consumer Protection are mandated by law to control the<br />
implementation <strong>of</strong> the regulations. Some public health campaigns<br />
conducted by nongovernmental organizations and<br />
pr<strong>of</strong>essional associations associate the stop-smoking component<br />
in their programs, addressing lung disease, cardiovascular,<br />
or cancer prevention programs.<br />
Media campaigns, in collaboration with international<br />
partners, such as the “HELP campaign” are periodically<br />
organized. These integrated programs use television, Internet,<br />
and mobile phones to build capacity for “a life without<br />
tobacco” for youngsters and young adults. Currently, funds<br />
available do not allow for comprehensive mass-media campaigns.<br />
Community-based campaigns are difficult to organize<br />
because <strong>of</strong> limited resources and lack <strong>of</strong> coordination<br />
between institutions.<br />
Health Care Financing in Romania<br />
The treatment <strong>of</strong> insured patients with cancer is free. A<br />
National Program <strong>of</strong> <strong>Oncology</strong> was created in 2003. The<br />
budget allocated for this program raised continuously from<br />
2003 to 2009 and then reached a plateau. All approved<br />
anticancer agents are included on a special list <strong>of</strong> drugs that<br />
are 100% reimbursed, financed by the National House <strong>of</strong><br />
Insurance. A special budget is prospectively allocated for the<br />
National Program <strong>of</strong> <strong>Oncology</strong>, which is covering all anticancer<br />
drugs. The acquisition is done following a national<br />
auction. Each oncology department is periodically buying<br />
the drugs needed according to a planned budget. Oral drugs<br />
may also be obtained from the pharmacies outside hospitals.<br />
However, for the expensive new drugs such as pemetrexed,<br />
MoAbs (i.e., bevacizumab), and TKIs (i.e., erlotinib),<br />
the reimbursement is subject to an individual approval from<br />
a National Commission. The number <strong>of</strong> patients who benefited<br />
<strong>of</strong> systemic treatments covered by the National Program<br />
<strong>of</strong> <strong>Oncology</strong>, raised from 75,000 in 2007 to 96,700 in<br />
2010 for all cancers. Approximately 3,400 patients are<br />
treated concomitantly with expensive drugs in Romania.<br />
There is a mismatch between the new published guidelines<br />
(including CECOG) 6 and the lack <strong>of</strong> reimbursement for<br />
the newly registered drugs or indications. First-line TKIs for<br />
patients with EGFR mutations are not yet reimbursed.<br />
Maintenance chemotherapy is not yet reimbursed, even<br />
though Romanian centers were among the most active in the<br />
pivotal maintenance registration trials for pemetrexed and<br />
erlotinib. 10,11 The Romanian legislation foresees the yearly<br />
update <strong>of</strong> the list <strong>of</strong> compensated drugs. However, this has<br />
not happened since 2008, and about 30 oncology drugs still<br />
wait for the reimbursement decision. The way from registration<br />
to reimbursement is a multistep bureaucratic<br />
process. Additionally, a new drug may be proposed for<br />
compensation only after a minimum <strong>of</strong> 1 year <strong>of</strong> use in at<br />
least three European countries.<br />
In an attempt to provide the new drugs to patients,<br />
different risk-sharing strategies (such as cost-volume, costresult,<br />
headroom agreements, and discounts) were initiated<br />
between authorities and the pharmaceutical companies with<br />
variable success. At the moment <strong>of</strong> this writing, a radical<br />
reform in the Romanian Health System is expected, which<br />
will probably include the privatization <strong>of</strong> many hospitals<br />
and the transfer <strong>of</strong> the financing <strong>of</strong> all medical activities<br />
(including oncology) to several private Insurance Companies,<br />
besides the public National House <strong>of</strong> Insurance.<br />
The Research Structure at the Cancer Institute<br />
Ion Chiricuta<br />
The oldest Cancer Institute in Romania was founded in<br />
1929, located in Cluj, and is called “Ion Chiricuta,” after the<br />
name <strong>of</strong> a famous Romanian cancer surgeon, who was<br />
awarded by the Heidelberg University as the pioneer <strong>of</strong> the<br />
use <strong>of</strong> the omentum in plastic surgery and who was integral<br />
to the development <strong>of</strong> the Institute. It is a comprehensive<br />
cancer center that <strong>of</strong>fers cancer diagnosis and treatment<br />
(with 550 hospital beds). The clinical departments are represented<br />
by surgery, radiation oncology (external beam<br />
radiotherapy with two linear accelerators and one cobalt<br />
unit and brachytherapy with low-dose rate and high-dose<br />
rate), medical oncology (continuous hospitalization and day<br />
hospital), and pediatric oncology. There are dedicated histology<br />
and cytology departments, a certified clinical laboratory,<br />
and an imagery plateau with computed tomography<br />
(CT) scans, ultrasound, and scintigraphy. All categories <strong>of</strong><br />
personnel are represented, but there is a work overload<br />
because <strong>of</strong> an increasing number <strong>of</strong> patients (from 727 new<br />
patients in 1955 to 7,156 new patients in 2008). The Institute<br />
has basic science/translational research departments,<br />
with cellular biology, immunology, and radiobiology platforms.<br />
It is an active member <strong>of</strong> OECI (Organization <strong>of</strong> the<br />
439
Table 2. The Contribution <strong>of</strong> the Cancer Institute “Ion Chiricuta” Cluj-Napoca, Romania (CIIC), to Multicenter Trials in Lung Cancer<br />
Setting/Study Phase Number pts CIIC/all pts % pts CIIC Reference<br />
NSCLC adjuvant/ IALT (platinum doublets versus observation) III 55/1,867 2.9% 13<br />
NSCLC 1 st line/<br />
CTNR (platinum versus non platinum doublets) IIR 40/102 39.2 14<br />
Stellar 3 (paclitaxel poliglumex/carbo versus paclitaxel/carbo) III 19/400 4.75 15<br />
NSCLC maintenance/<br />
CECOG (gemcitabine versus observation) III 24/352 6.8 18<br />
JMEN (pemetrexed versus placebo) III 70/663 10.6 10<br />
SATURN (erlotinib versus observation) III 65/889 7.3 11<br />
NSCLC rescue treatments/<br />
BR 21 (erlotinib versus placebo) III 68/731 9.3 16<br />
ISEL (gefitinib versus placebo) III 58/1,692 3.4 17<br />
TITAN (erlotinib versus docetaxel or pemetrexed) III 48/424 11.3 23<br />
SCLC prophylactic cranial irradiation (PCI) PCI 99–01 (2 doses <strong>of</strong> PCI) III 35/720 4.9% 21<br />
SCLC 1 st line JMHO (pemetrexed/carbo versus etoposide/carbo) III 20/908 2.2 19<br />
SCLC rescue treatment/406 (oral topotecan versus observation) III 13/141 9.2% 20<br />
Biogenerics/XM 02-INT 03 (biogeneric versus brand name G-CSF) III 34/240 14.2 22<br />
Total 549/9,129 6.01<br />
Abbreviations: NSCLC, non-small cell lung cancer; IALT, International Adjuvant Lung Cancer Trial; CECOG, Central European Cooperative <strong>Oncology</strong> Group; SCLC,<br />
small cell lung cancer.<br />
European Cancer Institutes) and part <strong>of</strong> the CECOG and<br />
BUON (Balkan Union <strong>of</strong> <strong>Oncology</strong>) network.<br />
Highlights <strong>of</strong> Romanian Participation in Lung<br />
Cancer Research<br />
<strong>Clinical</strong> Trials and the Specific National Environment in <strong>Oncology</strong><br />
<strong>Clinical</strong> research has been growing steadily in Romania.<br />
Although heterogeneity exists in the level <strong>of</strong> care, an increasing<br />
number <strong>of</strong> state-owned and private centers qualify<br />
for participation in clinical studies. According to the Romanian<br />
National Drug Agency, the number <strong>of</strong> phase II and<br />
phase III studies to which Romanian centers are contributing<br />
patients doubled in the past 2 years.<br />
The climate in cancer research is influenced by three<br />
factors: economic (a negative trend in 2009–<strong>2012</strong>, affected<br />
by the global crisis), political (wind shear, with the permanent<br />
need to reform the sanitary system carried out by each<br />
new government, the close parliamentary/presidential elections),<br />
and scientific (good premises with an increasing<br />
presence in international studies, high recruitment potential,<br />
and increasingly experienced investigators).<br />
Local Versus International Research Priorities<br />
There is a steep decrease in local academic research in<br />
favor <strong>of</strong> the international research, mainly industrysponsored<br />
clinical trials. Based on the European legislation<br />
(Directive 2001/20/EC), clinical trials must get ethical approval<br />
and approval from the competent authorities. However,<br />
the duration <strong>of</strong> these regulatory procedures to initiate<br />
a clinical trial is a factor determining the competitive<br />
position in clinical research.<br />
An analysis <strong>of</strong> the time interval between final protocol<br />
approval (FPA) and inclusion <strong>of</strong> the first patient into randomized<br />
clinical trials was performed for six multicenter<br />
trials in 25 CECOG study centers, including Romania. 12 The<br />
average time interval from FPA to the inclusion <strong>of</strong> the first<br />
patient was 18.4 months. Most <strong>of</strong> this time has been spent<br />
for regulatory procedures, i.e., the approval by the Ethical<br />
Review Boards (9.6 � 7.2 months) and CAs (10.0 � 6.6<br />
months). The Letters <strong>of</strong> Agreement were signed 11.5 to 9.4<br />
months after FPA. As the regulatory procedures accounted<br />
440<br />
for more than 50% <strong>of</strong> the duration <strong>of</strong> the whole ‘paper to<br />
patient process,’ optimization is necessary to make novel<br />
therapies available to patients more quickly.<br />
As <strong>of</strong> January <strong>2012</strong>, a total <strong>of</strong> 1,107 studies involved<br />
Romanian sites for all the medical specialties, according to<br />
clinicaltrials.gov. Fifty-eight involved lung cancers (compared<br />
with 48 for breast and 46 for digestive tumors).<br />
Romanian centers participated mainly in phase III trials, 34<br />
followed by phase II, 19 phase IV, 3 and phase I/II (two<br />
studies). Among these, 16 are actively recruiting. The investigational<br />
products in these studies were as follows: bevacizumab,<br />
BIBF1120, BNP7787, CS7017, darbepoetin alfa,<br />
erlotinib, ganetespib, gefitinib, IMC-11F8, ipilimumab, LY252-<br />
3355, necitumumab, obatoclax, onartuzumab, OSI-906, pemetrexed,<br />
picoplatin, ramucirumab, stimuvax, and sunitinib.<br />
Since 2004, the Cancer Institute “Ion Chiricuta” recruited<br />
549 <strong>of</strong> the total <strong>of</strong> 9,129 patients in 13 finalized multicentric<br />
trials, representing 6% <strong>of</strong> the overall population included.<br />
The participation ranged from 2.2% to 39.2% between studies<br />
(Table 2). 13-23<br />
The results lead to the United States Food and Drug<br />
Administration (FDA) and/or European Medicines Agency<br />
registration <strong>of</strong> several new molecules (i.e., erlotinib, oral<br />
topotecan), new biogenerics (filgrastim), or new indications<br />
(i.e., maintenance for pemetrexed and erlotinib). Some <strong>of</strong><br />
these trials helped to define the role <strong>of</strong> different therapeutic<br />
modalities in lung cancer such as adjuvant chemotherapy<br />
(IALT), rescue treatments (BR21, ISEL, TITAN), maintenance<br />
therapy (JMEN, SATURN), and prophylactic cranial<br />
irradiation (PCI99-01). The first ASCO International <strong>Clinical</strong><br />
Trials Workshop was held in Cluj in 2011, focusing on the<br />
best practice in the implementation <strong>of</strong> a protocol and an<br />
overview <strong>of</strong> clinical trial design.<br />
Conclusion<br />
TUDOR E. CIULEANU<br />
Lung cancer constitutes an important health problem in<br />
Romania, with a rising incidence and mortality. The implemented<br />
antitobacco strategies are expected to slow down<br />
this trend in the future. All new drugs are simultaneously<br />
registered in Europe and in Romania. However, a generally<br />
underfinanced health system does not cover an expedited
LUNG CANCER IN ROMANIA<br />
reimbursement <strong>of</strong> the new drugs for standard care. <strong>Clinical</strong><br />
research is constantly growing, and Romanian centers al-<br />
Author’s Disclosures <strong>of</strong> Potential Conflicts <strong>of</strong> Interest<br />
Author<br />
Employment or<br />
Leadership<br />
Positions<br />
Consultant or<br />
Advisory Role<br />
Tudor E. Ciuleanu Amgen; Bristol-<br />
Myers Squibb;<br />
GlaxoSmithKline;<br />
Lilly; Merck;<br />
OSIP; Pfizer;<br />
Roche<br />
1. Eniu A. International Insight: Cancer in Romania. ASCO News &<br />
Forum. July 2009:42-43.<br />
2. GLOBOCAN 2008 database (version 1.2). http://globocan.iarc.fr/. Accessed<br />
online January <strong>2012</strong>.<br />
3. Ulmeanu R. Bronchology in Romania: Where to? Pneumologia. 2006;55:<br />
147-150.<br />
4. Ciuleanu TE, Dediu M, Rusu P, et al. Lung cancer: Diagnostic and<br />
Treatment Guideline. J Radiother & Med Oncol. 2007;13:5-18.<br />
5. D’Addario G, Felip E. On behalf <strong>of</strong> the ESMO Guidelines Working<br />
Group. Non-small-cell lung cancer: ESMO <strong>Clinical</strong> Recommendations for<br />
diagnosis, treatment and follow-up. Ann Oncol. 2009;20(suppl 4):68-70.<br />
6. Brodowicz T, Ciuleanu T, Crawford J, et al. Third CECOG consensus on<br />
the systemic treatment <strong>of</strong> non-small-cell lung cancer. Ann Oncol. Epub 2011<br />
Sept 22.<br />
7. Cernea V, Nagy V, Irimie A, et al. Report Regarding the Present State <strong>of</strong><br />
Radiotherapy Laboratories in Romania, National Program for Radiotherapy<br />
2008. J Radiother & Med Oncol. 2008;15:7-24.<br />
8. Vlãdescu C, Mihãlþan F, Sanda L. On behalf <strong>of</strong> the Center for Health<br />
Policies and Services. Smoking and Public Health in Romania, 2004. http://<br />
www.stopfumat.eu/Materiale/Studiu_CPSS_04. Accessed online January<br />
<strong>2012</strong>.<br />
9. Joosens L, Raw M. The Tobacco Control Scale: A new scale to measure<br />
country activity. Tob Control. 2006;15:247-253.<br />
10. Ciuleanu T, Brodowicz T, Zielinski C, et al. Maintenance pemetrexed<br />
plus best supportive care versus placebo plus best supportive care for<br />
non-small-cell lung cancer: A randomised, double-blind, phase 3 study.<br />
Lancet. 2009;374:1432-1440.<br />
11. Cappuzzo F, Ciuleanu T, Stelmakh L, et al. Erlotinib as maintenance<br />
treatment in advanced non-small-cell lung cancer: A multicentre, randomised,<br />
placebo-controlled phase 3 study. Lancet Oncol. 2010;11:521-529.<br />
12. Brodowicz, I. Steiner, S. Beslija, et al. Time interval between final<br />
protocol approval (FPA) and inclusion <strong>of</strong> the first patient into randomized<br />
clinical trials (RCTs) performed by the Central European Cooperative <strong>Oncology</strong><br />
Group (CECOG): A 10-year experience. J Clin Oncol. 2009;27:15s (suppl;<br />
abstr 6546).<br />
13. The International Adjuvant Lung Cancer Trial Collaborative Group.<br />
Cisplatin-Based Adjuvant Chemotherapy in Patients with Completely Resected<br />
Non–Small-Cell Lung Cancer. N Engl J Med. 2004;350:351-360.<br />
14. Grigorescu Al, Ciuleanu T, Firoiu E, et al. A randomized phase II trial<br />
<strong>of</strong> sequential gemcitabine plus vinorelbine followed by gemcitabine plus<br />
ifosfamide versus gemcitabine plus cisplatin in the treatment <strong>of</strong> chemo-naive<br />
ready showed a high recruitment rate in several pivotal<br />
trials.<br />
Stock<br />
Ownership Honoraria<br />
Amgen; Bristol-<br />
Myers Squibb;<br />
GlaxoSmithKline;<br />
Lilly; Merck;<br />
Novartis; OSIP;<br />
Pfizer; Roche<br />
REFERENCES<br />
Research<br />
Funding<br />
Expert<br />
Testimony<br />
Other<br />
Remuneration<br />
patients with stages III and IV non-small cell lung cancer (NSCLC). Lung<br />
Cancer. 2007;57:168-174.<br />
15. Langer CJ, O’Byrne KJ, Socinski MA, et al. Phase III Trial Comparing<br />
Paclitaxel Poliglumex (CT-2103, PPX) in Combination with Carboplatin<br />
Versus Standard Paclitaxel and Carboplatin in the Treatment <strong>of</strong> PS 2<br />
Patients with Chemotherapy-Naive Advanced Non-small Cell Lung Cancer.<br />
J Thorac Oncol. 2008;3:623-630.<br />
16. Shepherd FA, Pereira JR, T Ciuleanu T, et al. Erlotinib in Previously<br />
Treated Non-Small-Cell Lung Cancer. N Engl J Med. 2005;353:123-132.<br />
17. Thatcher N, Chang A, Parikh P, et al. Gefitinib plus best supportive<br />
care in previously treated patients with refractory advanced non-small-cell<br />
lung cancer: Results from a randomised, placebo-controlled, multicentre<br />
study (Iressa Survival Evaluation in Lung Cancer). Lancet. 2005;366:1527-<br />
1537.<br />
18. Brodowicz T, Krzakowski M, Zwitter M, et al. Cisplatin and gemcitabine<br />
first-line chemotherapy followed by maintenance gemcitabine or best<br />
supportive care in advanced non-small cell lung cancer: A phase III trial for<br />
the Central European Cooperative <strong>Oncology</strong> Group (CECOG). Lung Cancer.<br />
2006;52:155-163.<br />
19. Socinski MA, Smit EF, Lorigan R, et al. Phase III Study <strong>of</strong> Pemetrexed<br />
Plus Carboplatin Compared With Etoposide Plus Carboplatin in<br />
Chemotherapy-Naive Patients With Extensive-Stage Small-Cell Lung Cancer.<br />
J Clin Oncol. 2009;27:4787-4792.<br />
20. O’Brien MER, Ciuleanu TE, Tsekov H, et al. Phase III Trial Comparing<br />
Supportive Care Alone With Supportive Care With Oral Topotecan in Patients<br />
With Relapsed Small-Cell Lung Cancer. J Clin Oncol. 2006;24:5441-<br />
5447.<br />
21. Le Péchoux C, Dunant A, Senan S, et al. Standard-dose versus<br />
higher-dose prophylactic cranial irradiation (PCI) in patients with limitedstage<br />
small-cell lung cancer in complete remission after chemotherapy and<br />
thoracic radiotherapy (PCI 99-01, EORTC 22003-08004, RTOG 0212, and<br />
IFCT 99-01): a randomised clinical trial. Lancet Oncol. 2009;10:467-474.<br />
22. Gatzemeier U, Ciuleanu T, Dediu M, et al. XM02, the First Biosimilar<br />
G-CSF, is Safe and Effective in Reducing the Duration <strong>of</strong> Severe Neutropenia<br />
and Incidence <strong>of</strong> Febrile Neutropenia in Patients with Small Cell or Nonsmall<br />
Cell Lung Cancer Receiving Platinum-Based Chemotherapy. J Thorac<br />
Oncol. 2009;4:736-740.<br />
23. Ciuleanu T, Stelmakh L, Cicenas S, et al. Efficacy and safety <strong>of</strong><br />
erlotinib versus chemotherapy in second-line treatment <strong>of</strong> patients with<br />
advanced, non-small-cell lung cancer with poor prognosis (TITAN): a randomised<br />
multicentre, open-label, phase 3 study. Lancet Oncol. <strong>2012</strong>; [epub<br />
ahead <strong>of</strong> print].<br />
441
LEVERAGING VIRTUAL PATIENT COMMUNITIES<br />
FOR OPTIMAL CLINICAL CARE AND RESEARCH<br />
CHAIR<br />
Howard J. West, MD<br />
Swedish Cancer Institute<br />
Seattle, WA<br />
SPEAKERS<br />
George D. Demetri, MD<br />
Dana-Farber Cancer Institute, Harvard Medical School<br />
Boston, MA<br />
Dave deBronkart, SB<br />
Boston, MA
A New Model: Physician-Patient Collaboration<br />
in Online Communities and the <strong>Clinical</strong><br />
Practice <strong>of</strong> <strong>Oncology</strong><br />
By Howard J. West, MD, Dave deBronkart, SB, and George D. Demetri, MD<br />
Overview: The practice <strong>of</strong> medicine is in the midst <strong>of</strong> a<br />
fundamental transformation based on the new availability <strong>of</strong><br />
health information through the Internet and other sources<br />
accessible by the broad lay public, as well as on the easy<br />
sharing <strong>of</strong> experiences and content through social media. This<br />
is occurring at a time when the volume <strong>of</strong> new information<br />
required for optimal medical care is exceeding that which an<br />
individual physician can feasibly follow and master. The<br />
changes in cancer care are especially acute as we experience<br />
an ongoing reclassification <strong>of</strong> many disease entities to reflect<br />
divisions by molecular variables, <strong>of</strong>ten with new clinical options<br />
now optimized for very limited patient subsets. The<br />
increasing complexity <strong>of</strong> the field, combined with the high<br />
stakes <strong>of</strong> optimizing treatment decisions and the growing<br />
THE INTERNET has transformed many industries,<br />
ranging from news media to travel, real estate, and<br />
politics, with the practice <strong>of</strong> medicine also in the midst <strong>of</strong> its<br />
own belated disruption as well. For centuries, our dominant<br />
medical model has been based on a unidirectional flow <strong>of</strong><br />
information, in which only physicians had access to the<br />
knowledge sufficient to weigh differential diagnoses and to<br />
recommend optimal treatments. Patients readily accepted<br />
physician wisdom and recommendations as the only readily<br />
available source <strong>of</strong> medical information.<br />
Among the factors that have dramatically altered the<br />
course <strong>of</strong> medical practice, and oncology in particular, are<br />
two central issues. The first is that the sheer amount <strong>of</strong> new<br />
medical information emerging over the past few decades has<br />
ballooned to the point that no single physician could possibly<br />
maintain expertise in the full breadth <strong>of</strong> topics required to<br />
care optimally for a full range <strong>of</strong> patients and clinical<br />
problems. In particular, oncology has exploded with new<br />
content that is a boon in defining a new era <strong>of</strong> mechanistic<br />
understanding <strong>of</strong> cancer pathophysiology, as well as necessary<br />
for the rational development and use <strong>of</strong> molecularly<br />
targeted therapies; however, it is simply infeasible for an<br />
oncologist to internalize the depth and breadth <strong>of</strong> new<br />
content. The second core element is that this content is now<br />
readily accessible to the lay public, including patients and<br />
caregivers, essentially without limitation and in real time—<br />
perhaps even before the clinician has become aware <strong>of</strong> it.<br />
Engaged Patients Usher in a New Model <strong>of</strong><br />
Medical Practice<br />
There is clear evidence that the amount <strong>of</strong> medical information<br />
for any one doctor to consume and digest has grown<br />
remarkably. PubMed (www.ncbi.nlm.nih.gov/pubmed/) currently<br />
categorizes more than 21 million citations, with new<br />
publications being added at a rate <strong>of</strong> approximately one per<br />
minute—a rate that has more than doubled over the past<br />
20 years. 1 The number <strong>of</strong> medical journals has also increased,<br />
and this is especially true for oncology, where there<br />
are now approximately 180 journals covering just this one<br />
area <strong>of</strong> clinical subspecialty practice. 2 As new research<br />
availability <strong>of</strong> a wide range <strong>of</strong> information in the public domain,<br />
make oncology an area in which patients and caregivers are<br />
most motivated to become active seekers <strong>of</strong> medical information<br />
and participants in their care decisions.<br />
The credibility <strong>of</strong> the available online information in such a<br />
situation has emerged as a critical issue, but physicians have<br />
historically been reluctant to create content or interact with<br />
the lay public in online patient communities. Here we will<br />
highlight several examples <strong>of</strong> collaborative engagement between<br />
health care pr<strong>of</strong>essionals and motivated patients in an<br />
online environment that illustrate how a new bidirectional or<br />
even networked model that is a product <strong>of</strong> the Internet age can<br />
accelerate clinical research and improve delivery <strong>of</strong> cancer<br />
care.<br />
results are released early online and new trial data are <strong>of</strong>ten<br />
released via a press release long before being presented<br />
and/or published in a peer-reviewed setting, physicians only<br />
have more sources <strong>of</strong> practice-changing information to interpret<br />
and incorporate into their management recommendations.<br />
This broader distribution <strong>of</strong> medical information is<br />
also occurring as physicians are typically pressed to see<br />
more patients in less time.<br />
Concurrent with these changes, patients and caregivers<br />
are seeking both support and medical information online<br />
at a rapidly escalating pace—from 25% <strong>of</strong> U.S. adults in<br />
2,000% to 61% by 2010 3 —with this trend rapidly accelerating<br />
as fast Internet connections become more readily available.<br />
People also have become increasingly comfortable with<br />
seeking and expecting to find relevant content online free<br />
<strong>of</strong> charge. Among health care topics, those that are lifethreatening<br />
and/or chronic lend themselves best to online<br />
patient communities and to Internet-based searches for<br />
information.<br />
This is unquestionably a mixed blessing for patients and<br />
physicians alike. Internet searches can instantaneously return<br />
reliable and timely information but also might deliver<br />
content from unreliable sources that instead prey on the<br />
desperation <strong>of</strong> people by peddling the false hope <strong>of</strong> a “Miracle<br />
Cancer Cure” (www.theCancerCureMiracle.com) (Fig. 1).<br />
Beyond charlatans misleading patients and caregivers for<br />
pr<strong>of</strong>it, online communities provide a wide range <strong>of</strong> recommendations<br />
from people who might be knowledgeable and<br />
well intended but who might definitively promote treatment<br />
ideas that are not proven to be superior to others or are<br />
possibly even detrimental. Such information may directly<br />
compete with thoughtful recommendations from qualified<br />
medical and scientific pr<strong>of</strong>essionals. The fact that the Inter-<br />
From the Department <strong>of</strong> Thoracic <strong>Oncology</strong>, Swedish Cancer Institute, Seattle, WA.<br />
Authors’ disclosures <strong>of</strong> potential conflicts <strong>of</strong> interest are found at the end <strong>of</strong> this article.<br />
Address reprint requests to Howard West, MD, Swedish Cancer Institute, 1221 Madison<br />
St., Suite 1020, Seattle, WA 98104; email: howard.west@swedish.org.<br />
© <strong>2012</strong> by <strong>American</strong> <strong>Society</strong> <strong>of</strong> <strong>Clinical</strong> <strong>Oncology</strong>.<br />
1092-9118/10/1-10<br />
443
net makes it easier than ever to share ideas and information<br />
means that the difference between this being a net benefit or<br />
harm depends entirely on the quality <strong>of</strong> the information<br />
available.<br />
Practicing physicians have historically been reluctant to<br />
create content or to participate in online communities for the<br />
lay public. The most frequently cited reasons for physician<br />
reluctance to engage online are wariness about liability,<br />
patient privacy concerns, and a lack <strong>of</strong> time or reimbursement<br />
for these efforts (Fig. 2). 4 Although there are a growing<br />
number <strong>of</strong> online resources with pr<strong>of</strong>essionally vetted information<br />
for patients and caregivers, these are few and far<br />
between compared with the volume <strong>of</strong> content from nonpr<strong>of</strong>essional<br />
sources. In addition, although a growing proportion<br />
<strong>of</strong> physicians participate in online communities for<br />
physicians, these are generally completely segregated from<br />
patient-oriented communities, so that there are very few<br />
online settings in which physicians and patients interact<br />
together online.<br />
Yet, the potential benefits for online physician engagement<br />
are very substantial, not only for patients but also for<br />
physicians and for the global practice <strong>of</strong> medicine. Patients<br />
KEY POINTS<br />
● Medicine in general, and oncology in particular, is<br />
now experiencing a rapid growth in new content and<br />
a reclassification <strong>of</strong> many cancers into smaller, molecularly<br />
defined subgroups that has made it infeasible<br />
for an individual physician to maintain sufficient<br />
expertise to remain the sole source <strong>of</strong> treatment<br />
information.<br />
● Patients and caregivers are increasingly turning to<br />
online sources to supplement what they learn from<br />
their own medical care teams, and this can be a<br />
beneficial or detrimental change depending entirely<br />
on the quality <strong>of</strong> the medical information available<br />
online.<br />
● Physicians have historically remained wary to engage<br />
in online communities or to provide information<br />
via the Internet, largely because <strong>of</strong> concerns about<br />
legal liability, patient privacy, and time limitations.<br />
● Despite these challenges, there is a wide range <strong>of</strong><br />
examples in which physicians have produced credible<br />
online content and/or partnered with communities <strong>of</strong><br />
motivated patients and caregivers facilitated by<br />
Internet-based platforms to distribute highly relevant<br />
and high-quality information and to conduct<br />
valued clinical research.<br />
● Reliance on Internet-based information will only escalate<br />
as more patients and caregivers become connected,<br />
making it more compelling for physicians to<br />
embrace the constructive possibilities <strong>of</strong> collaborative<br />
engagement and interaction between patients and<br />
caregivers online rather than passively cede the ability<br />
to influence these online communities in a constructive<br />
way.<br />
444<br />
WEST, DEBRONKART, AND DEMETRI<br />
Fig. 1. Medical advice from unqualified sources is plentiful online<br />
(theCancerCureMiracle.com).<br />
rank a physician’s input as the most influential factor in<br />
shaping their care decisions (see Table 1), 5 and timely and<br />
credible information from experts can help overcome the gulf<br />
between the prohibitive expanse <strong>of</strong> information for a patient’s<br />
care and for the amount that any single physician can<br />
review and retain. With the expanse <strong>of</strong> new content, physicians<br />
are now apt to become a bottleneck and to limit care<br />
options if medical practice follows an outdated unidirectional<br />
model. Increasingly, patients with access to a network<br />
<strong>of</strong> committed advocates, potentially including many other<br />
fellow patients, in the context <strong>of</strong> highly accessible medical<br />
information can collaborate with their own local medical<br />
team to shape a more bidirectional or even a networked<br />
model in which patients conduct their own research and<br />
discuss their own views regarding the most appropriate or<br />
preferred options with their physicians. This respects the<br />
physician’s role as an expert and caring provider, while also<br />
recognizing that no single physician can be expected to be<br />
the sole source <strong>of</strong> all medical knowledge. In this new model,<br />
the physician becomes the pivotal individual to provide<br />
context and to shape recommendations for a patient who is<br />
increasingly engaged in key decisions.<br />
Provision <strong>of</strong> Readily Available, Vetted Content<br />
through Physician Engagement<br />
The substantial benefits from this new model are all<br />
predicated on high-quality information being available online<br />
from identifiable, credible sources to counter the abundance<br />
<strong>of</strong> less reliable information, or even misinformation,<br />
being promulgated by less qualified or less competent<br />
sources. Practically speaking, physicians can engage by<br />
<strong>of</strong>fering “pushed” online content, in which knowledge is<br />
<strong>of</strong>fered and can be consumed by a limitless number <strong>of</strong> people<br />
who can view this digital content without any additional<br />
effort from its producer. Examples <strong>of</strong> this type <strong>of</strong> content<br />
include blog posts, audio and video podcasts, or information<br />
found on “micro-blogging” platforms such as Twitter. The
PHYSICIAN ENGAGEMENT IN ONLINE PATIENT COMMUNITIES<br />
Fig. 2. Top concerns <strong>of</strong> physicians regarding interaction with patients online (respondents choose up to three). 4<br />
alternative, “pulled” content—patient solicitation <strong>of</strong> answers<br />
on an interactive online community or participation<br />
in discussions via Facebook or Twitter—<strong>of</strong>fer the benefit <strong>of</strong><br />
great appeal to the lay audience because <strong>of</strong> potentially more<br />
individualized and, therefore, valuable answers. These efforts<br />
create challenges, however, by being more time consuming<br />
and not scalable like pushed content. In addition,<br />
pulled, interactive content that involves individualized<br />
case discussions could potentially entail an implied medical<br />
recommendation in the absence <strong>of</strong> broad disclaimers and<br />
careful wording. Additionally, there is a risk <strong>of</strong> oversimplification<br />
because the optimal practice <strong>of</strong> medicine always<br />
depends on the totality <strong>of</strong> details and accuracy <strong>of</strong> the<br />
information. In the most egregious example, patients with a<br />
mistaken diagnosis might not even have the disease that<br />
they think they have.<br />
Patients can clearly benefit from becoming far more informed<br />
about appropriate treatment options and may learn<br />
about standard or clinical research-based options about<br />
which their own physician might not be aware. They can<br />
also be comforted by the ability to play an active role in<br />
achieving a consensus about the optimal treatment for a<br />
complex situation. However, benefits are also readily available<br />
for physicians who are willing to invest the time to<br />
impart their knowledge into global online discussions. Aside<br />
from the very substantial intrinsic value <strong>of</strong> enabling more<br />
patients to participate directly in their own care while<br />
armed with vetted information, physicians who engage with<br />
patients online may need to spend less time covering these<br />
same topics repeatedly in individual discussions, may be<br />
Table 1. Physicians Remain Most Valued Source <strong>of</strong><br />
Health Information 5<br />
Who is more helpful when you need ...<br />
Pr<strong>of</strong>essional<br />
sources, e.g.,<br />
MD and RN (%)<br />
Fellow patients,<br />
friends, and<br />
family (%)<br />
Both<br />
equally<br />
(%)<br />
An accurate medical diagnosis 91 5 2<br />
Information about prescription drugs 85 9 3<br />
Information about alternative treatments 63 24 5<br />
A recommendation for a doctor or specialist 62 27 6<br />
A recommendation for a hospital or other<br />
medical facility<br />
62 27 6<br />
better trusted by their patients, and may attract more new<br />
patients based on their online outreach.<br />
The medical community can potentially curate highquality<br />
content, such as by creating an online resource akin<br />
to a Wikipedia or Khan Academy <strong>of</strong> medical information<br />
accessible to physicians and to the lay public alike. In<br />
many ways, this was the impetus for ASCO to develop the<br />
pr<strong>of</strong>essionally vetted Web resource CancerNet (www.cancer.<br />
net) to facilitate these new pr<strong>of</strong>essional roles and responsibilities<br />
with the public. Effective pr<strong>of</strong>essional activities in<br />
this regard will engender two critical and increasingly<br />
necessary fundamental changes. First, physicians will no<br />
longer need to hold an unmanageable capacity <strong>of</strong> specialized<br />
knowledge but can work with patients to access and interpret<br />
the best data to create an optimal management strategy.<br />
By necessity, this change will also be accompanied by<br />
an increased physician acceptance <strong>of</strong> outside sources <strong>of</strong> new<br />
information as relevant and potentially valid. Secondly,<br />
rather than have the same content recreated and recapitulated<br />
for thousands <strong>of</strong> patients thousands <strong>of</strong> times individually,<br />
this information can be shared communally, freeing<br />
time otherwise spent in reduplicated efforts.<br />
One example <strong>of</strong> successful engagement <strong>of</strong> multiple oncologists<br />
providing timely content to a patient and caregiver<br />
population is provided by the Global Resource for Advancing<br />
Cancer Education (GRACE, CancerGRACE.org). GRACE is<br />
a nonpr<strong>of</strong>it organization comprised primarily <strong>of</strong> an expert<br />
physician-mediated online forum that started with a focus<br />
on lung cancer and recently expanding into other cancer<br />
subtypes, in which one <strong>of</strong> the authors (Dr. West) along with<br />
other oncology experts distill the latest trial results and<br />
summarize current best practices, along with personal perspectives,<br />
in accessible language and multiple formats.<br />
These formats include blog posts, video and audio podcasts,<br />
and a very popular interactive discussion forum in which<br />
patients and caregivers can ask experts, and each other,<br />
questions about the best current treatments and new research<br />
concepts just emerging. Although requiring time and<br />
effort to develop and maintain, this resource <strong>of</strong>fering both<br />
pushed and interactive content efficiently delivers expertquality<br />
information, enabling tens <strong>of</strong> thousands <strong>of</strong> highly<br />
motivated people every month from all over the world to<br />
445
ecome extremely sophisticated about the leading treatments<br />
and promising new trial-based options.<br />
Other important new online vehicles are being developed<br />
by oncology pr<strong>of</strong>essionals along these lines as well. Cancer<br />
Commons (cancercommons.org) is a nonpr<strong>of</strong>it, open-science<br />
initiative that focuses on several forms <strong>of</strong> cancers for which<br />
the molecular understanding is rapidly evolving, with important<br />
implications for current clinical research and care<br />
options. Using highly interactive online models as teaching<br />
tools, this site aims to illuminate complex scientific pathways<br />
and bring them into the clinical context in a meaningful<br />
way for patients, caregivers, and health care<br />
pr<strong>of</strong>essionals.<br />
Online Medical Information: Facilitating Research in<br />
Small and Geographically Distributed Subgroups<br />
Centralized online content provided by medical experts<br />
can help overcome an emerging new challenge that is a<br />
by-product <strong>of</strong> the recognition that patients with any given<br />
cancer diagnosis viewed as a single monolithic group are<br />
actually far more heterogeneous, comprising many, much<br />
smaller subgroups. The rapid evolution <strong>of</strong> molecular oncology<br />
and molecular diagnostics over the past few years has<br />
led the erosion <strong>of</strong> large populations <strong>of</strong> stage-specific “lung<br />
cancer” or “breast cancer” into groups defined by clinically<br />
relevant molecular markers that redefine diagnostic groups,<br />
natural history, prognosis, optimal treatments and appropriate<br />
trial populations such as ALK-positive non–small cell<br />
lung cancer or triple-negative breast cancer. In fact, it is<br />
widely expected that “cancer” as a term will cede to the more<br />
accurate representation <strong>of</strong> subtypes <strong>of</strong> cancers as a multidimensional<br />
group <strong>of</strong> many different types <strong>of</strong> rare diseases.<br />
The implications <strong>of</strong> this shift for clinical research and care<br />
are clear and significant. When an eligible trial population<br />
transitions from an easily accessible pool <strong>of</strong> more than<br />
tens <strong>of</strong> thousands <strong>of</strong> potentially eligible patients to a much<br />
smaller subgroup <strong>of</strong> perhaps less than 1,000 potential candidates<br />
who are broadly geographically distributed with a<br />
very low density, researchers are no longer able to pursue<br />
the classical model in which hundreds <strong>of</strong> centers <strong>of</strong>fer the<br />
same clinical trial (e.g., standard chemotherapy with or<br />
without novel agent X).<br />
Instead, research will be facilitated if these much smaller<br />
yet geographically diverse populations <strong>of</strong> rare subsets can be<br />
identified (or ideally, self-identify) and, increasingly, if these<br />
patients are enabled to actively seek out treatment at a<br />
very limited number <strong>of</strong> highly specialized centers <strong>of</strong>fering a<br />
menu <strong>of</strong> unique clinical trials for this limited subgroup. To<br />
borrow a sartorial phrase, this could be considered as<br />
“bespoke clinical research.” Although inconvenient and expensive,<br />
this model has been proven to be effective by the<br />
development <strong>of</strong> agents such as crizotinib for non–small cell<br />
lung carcinoma driven by genomic aberrancies in the ALK<br />
kinase, as well as vemurafenib for V600E BRAF mutationdriven<br />
melanoma—settings in which the anticipated benefits<br />
far exceed those we typically see for broad, untargeted<br />
populations.<br />
Patients with rare cancers, acquired resistance to effective<br />
therapies, and other narrowly defined subgroups have even<br />
self-aggregated online to facilitate clinical research, an effort<br />
that has been extremely fruitful when these coordinated<br />
efforts by motivated patients are encouraged and facilitated<br />
446<br />
WEST, DEBRONKART, AND DEMETRI<br />
by medical pr<strong>of</strong>essionals. More than a decade ago, at least<br />
two different self-organized collections <strong>of</strong> patients with gastrointestinal<br />
stromal tumors (GISTs) and their caregivers<br />
were formed, including the Life Raft Group and GIST<br />
Support International. Similar groups subsequently formed<br />
internationally, such as Das Lebenhaus in Germany and<br />
Ensemble Contre le GIST in France. These online groups<br />
began sharing and aggregating their experiences with this<br />
rare disease, which had been inconsistently identified or<br />
diagnosed before the year 2000, brought together by the<br />
breakthrough results demonstrated with the tyrosine kinase<br />
inhibitor, imatinib. This patient-driven online community<br />
derived, in part, from prior online collaborations with one<br />
the authors (Dr. Demetri), who was participating actively in<br />
patient-moderated online discussions in the late 1990s under<br />
the auspices <strong>of</strong> the nonpr<strong>of</strong>it organization known as<br />
the Association <strong>of</strong> Online Cancer Resources (acor.org). This<br />
organization was founded by a private philanthropic individual,<br />
Gilles Frydman, who had been frustrated by the<br />
difficulties he encountered in finding up-to-date information<br />
about cancer. The GIST online patient communities were<br />
very helpful in driving awareness <strong>of</strong> this new diagnosis in a<br />
relatively rare group <strong>of</strong> patients, and this awareness helped<br />
stimulate interest and participation in critically important<br />
clinical trials and clinical research initiatives linking investigators<br />
directly with the patient population that would be<br />
the subject and ultimate beneficiary <strong>of</strong> this research. Interestingly,<br />
certain groups were facilitated in their operations<br />
by support from the pharmaceutical firms that were developing<br />
and marketing imatinib and other targeted therapies,<br />
representing a complex intersection <strong>of</strong> corporate-derived<br />
philanthropic support to patient support groups with<br />
corporate-sponsored outreach to a target population <strong>of</strong> consumers<br />
and patients. These patient communities have been<br />
highly visible, promulgating viewpoints and even making<br />
public presentations <strong>of</strong> the collated, self-reported patient<br />
experiences from their database including a qualitative<br />
patient-developed scale for rating severity <strong>of</strong> toxicity. 6 Although<br />
the reliability and rigor <strong>of</strong> this methodology might be<br />
questionable, it is certainly a testament to the engagement<br />
<strong>of</strong> this very dedicated community <strong>of</strong> patients and caregivers<br />
with a devotion to this disease.<br />
The website PatientsLikeMe (PatientsLikeMe.com) has<br />
now enrolled more than 100,000 patients with a wide range<br />
<strong>of</strong> medical conditions, as a means to share experiences,<br />
provide mutual support, and also facilitate clinical research<br />
on the aggregated patient populations. Its own representatives<br />
have even conducted an observational study <strong>of</strong> a<br />
relatively uniform population <strong>of</strong> patients with amyotrophic<br />
lateral sclerosis who were on lithium carbonate treatment.<br />
Representatives also collected and recently published efficacy<br />
and toxicity data on participating patients who were<br />
compared statistically with matched controls. 7<br />
Finally, the accelerating ability <strong>of</strong> Internet-based patient<br />
recruitment for clinical research on rare and globally distributed<br />
populations is illustrated by other examples. A<br />
relatively large study aiming to accrue 1,000 people who<br />
have been diagnosed with any form <strong>of</strong> sarcoma is in progress<br />
under the sponsorship <strong>of</strong> 23andMe (www.23andMe.com), a<br />
company that aims to link genome-wide screening analyses<br />
with elements <strong>of</strong> online community and social networking. 8<br />
This study may subsequently expand to evaluate first- and
PHYSICIAN ENGAGEMENT IN ONLINE PATIENT COMMUNITIES<br />
second-generation relatives to assess risk factors that may<br />
be related to why people may be at risk <strong>of</strong> developing such<br />
rare diseases. Another example is a small trial that focused<br />
on the rare clinical problem <strong>of</strong> spontaneous coronary artery<br />
dissection. This was conducted by investigators at the Mayo<br />
Clinic and was able to enroll its target <strong>of</strong> 12 patients within<br />
just one week <strong>of</strong> being granted institutional review board<br />
approval. 9 Although a small and limited trial outside <strong>of</strong><br />
the cancer setting, it is clear that the partnership <strong>of</strong> engaged<br />
medical pr<strong>of</strong>essionals with a coordinated and motivated<br />
online patient population can facilitate research otherwise<br />
infeasible without Internet-based participation.<br />
It is critical to emphasize that each <strong>of</strong> these novel mechanisms<br />
<strong>of</strong> identifying appropriate trials for patients requires<br />
the input and guidance <strong>of</strong> a patient’s primary oncologist to<br />
provide context and guidance to review the range <strong>of</strong> options<br />
and facilitate participation in an optimal choice, if one<br />
exists. These network-furnished opportunities alter but do<br />
not obviate the relationship between the patient and the<br />
local oncologist.<br />
Conclusions: Picking Up the Gauntlet<br />
The Internet and related information technology has ushered<br />
in disruptive changes in the practice <strong>of</strong> medicine. The<br />
volume <strong>of</strong> new information has grown to a point where<br />
individual physicians increasingly find themselves unable to<br />
feasibly master the range <strong>of</strong> material required by any but<br />
the most specialized clinics, especially with increasingly<br />
clinic and paperwork demands. Most significantly, this information<br />
is no longer exclusively available to physicians;<br />
rather, it is now also available to motivated patients who<br />
seek relevant content in hopes <strong>of</strong> better understanding their<br />
condition and participating more actively in their own care<br />
decisions. These trends are especially true in cancer care,<br />
where the chronicity and <strong>of</strong>ten life-threatening nature <strong>of</strong> the<br />
disease leads patients to be exceptionally motivated to learn<br />
about a field that has become exponentially more complex<br />
and yet also more mechanistic with this new era <strong>of</strong> molecular<br />
oncology.<br />
As the proportion <strong>of</strong> the general public seeking health care<br />
information online grows steadily, the quality <strong>of</strong> the content<br />
they encounter is a concerning variable. Although patientcentered<br />
online communities <strong>of</strong>ten feature very knowledgeable<br />
members <strong>of</strong> the lay public who have become extremely<br />
sophisticated, such communities are also a potential source<br />
for misinformation that can be detrimental to good care, or<br />
at least will compete for patient attention with more accurate,<br />
constructive educational material. The need for openminded<br />
and expert pr<strong>of</strong>essional input is critical despite the<br />
fact that physicians have historically been wary about engaging<br />
in such public dissemination <strong>of</strong> information by producing<br />
vetted online content.<br />
Although the investment <strong>of</strong> time and effort for such<br />
activities is significant, the potential for practical benefits<br />
cannot be overstated. Patients and caregivers will seek<br />
assistance from online sources in greater numbers regardless<br />
<strong>of</strong> whether health care pr<strong>of</strong>essionals provide content<br />
that is pr<strong>of</strong>essionally reviewed for quality, accuracy, methodology,<br />
fair balance, and reliability. The physician’s ability<br />
to engage and largely direct the conversation, as well as<br />
ensure the quality <strong>of</strong> online content, is likely to translate to<br />
the difference between whether the newly defined relationship<br />
between patient and physician will become more oppositional<br />
or more collaborative. As we move toward an era <strong>of</strong><br />
bidirectional rather than unidirectional flow <strong>of</strong> health care<br />
information, physicians have the potential to leverage the<br />
efficiency <strong>of</strong> digital content to convey the most current<br />
expert information broadly, opening up the possibilities <strong>of</strong><br />
new strategies for molecularly-guided clinical research that<br />
will capitalize on the ability <strong>of</strong> the Internet to connect small<br />
groups <strong>of</strong> geographically distributed people, along with the<br />
motivation and communication within patient online communities.<br />
The role <strong>of</strong> the oncologist, as for nearly all other physicians,<br />
is being altered in real time, and the opportunity (if<br />
not responsibility) to provide high-quality, vetted medical<br />
information remains crucial. There are enough examples<br />
now <strong>of</strong> the realized potential <strong>of</strong> these efforts that more<br />
physicians should feel compelled to engage in the pr<strong>of</strong>essional<br />
discourse with patients online.<br />
Importantly, no public online source will have the details<br />
<strong>of</strong> a patient’s case that can inform care decisions, and even<br />
knowledgeable sources may <strong>of</strong>fer a range <strong>of</strong> perspectives on<br />
questions with no absolute correct answer. It is therefore<br />
critical to underscore that, even as the role <strong>of</strong> serving as a<br />
patient’s primary oncologist evolves to increasingly integrate<br />
a plurality <strong>of</strong> sources <strong>of</strong> knowledge, it remains the<br />
pivotal mechanism for vetting and prioritizing content and<br />
its applicability for a particular patient’s context.<br />
447
Authors’ Disclosures <strong>of</strong> Potential Conflicts <strong>of</strong> Interest<br />
Employment or<br />
Leadership<br />
Positions<br />
Consultant or<br />
Advisory Role<br />
Author<br />
Howard J. West*<br />
Dave deBronkart*<br />
George D. Demetri Amgen; ARIAD;<br />
Daiichi Sankyo;<br />
Genentech;<br />
GlaxoSmithKline;<br />
Idera<br />
Pharmaceuticals;<br />
Infinity; Johnson<br />
& Johnson;<br />
Kolltan<br />
Pharmaceuticals;<br />
Merck Serono;<br />
Momenta<br />
Pharmaceuticals;<br />
Novartis; Pfizer;<br />
Plexxikon;<br />
Ziopharm<br />
<strong>Oncology</strong><br />
*No relevant relationships to disclose.<br />
Stock<br />
Ownership Honoraria<br />
Champions<br />
Biotechnology;<br />
EmergingMed;<br />
Kolltan<br />
Pharmaceuticals;<br />
Plexxikon<br />
1. U.S. National Library <strong>of</strong> Medicine, Institutes <strong>of</strong> Health Web site.<br />
http://www.nlm.nih.gov/bsd/medline_cit_counts_yr_pub.html. Accessed February<br />
23, <strong>2012</strong>.<br />
2. Cancer Journals Web site. http://www.cancerindex.org/clinks9.htm. Accessed<br />
February 23, <strong>2012</strong>.<br />
3. Fox S, Jones S. The Social Life <strong>of</strong> Health Information. Pew Internet and<br />
<strong>American</strong> Life Project, 2009 Web site. http://www.pewinternet.org/Reports/<br />
2009/8-The-Social-Life-<strong>of</strong>-Health-Information/02-A-Shifting-Landscape/2-61<strong>of</strong>-adults-in-the-US-gather-health-information-online.aspx.<br />
Accessed February<br />
23, <strong>2012</strong>.<br />
4. QuantiaMD web site. www.quantiamd.com. Accessed February 23, <strong>2012</strong>.<br />
4. Modahl M, Tompsett L, Moorhead T. Doctors, Patients & Social Media:<br />
QuantiaMD, 2011. http://www.quantiamd.com/q-qcp/DoctorsPatientSocial<br />
Media.pdf. Accessed February 23, <strong>2012</strong>.<br />
5. Fox S. Peer-to-Peer Healthcare: Pew Internet and <strong>American</strong> Life Project,<br />
448<br />
REFERENCES<br />
Research<br />
Funding<br />
Novartis; Pfizer Amgen; ARIAD;<br />
Bristol-Myers<br />
Squibb; Daiichi<br />
Sankyo;<br />
Genentech;<br />
Infinity; Johnson<br />
& Johnson;<br />
Novartis; Pfizer;<br />
PharmaMar<br />
WEST, DEBRONKART, AND DEMETRI<br />
Expert<br />
Testimony<br />
ARIAD (U);<br />
Infinity (U);<br />
Johnson &<br />
Johnson (U);<br />
Novartis (U);<br />
Pfizer (U);<br />
PharmaMar (U)<br />
Other<br />
Remuneration<br />
2011. Available at http://www.pewinternet.org/�/media//Files/Reports/2011/<br />
Pew_P2PHealthcare_2011.pdf, accessed 2/10/12.<br />
6. Call J, Scherzer NJ, Josephy PD, et al. Evaluation <strong>of</strong> self-reported<br />
progression and correlation <strong>of</strong> imatinib dose to survival in metastatic gastrointestinal<br />
stromal tumors: An open cohort study. J Gastrointest Cancer.<br />
2010;41(1):60-70.<br />
7. Wicks P, Vaughan TE, Massigli MP, et al. Accelerated clinical discovery<br />
using self-reported patient data collected online and a patient-matching<br />
algorithm. Nat Biotechnol. 2011;29(5):411-414.<br />
8. 23andMe Sarcoma Community: A Patient-Driven Revolution in Sarcoma<br />
Research. https://www.23andme.com/sarcoma/. Accessed February 23,<br />
<strong>2012</strong>.<br />
9. Tweet MS, Gulati R, Aase LA, et al. Spontaneous coronary artery<br />
dissection: A disease-specific, social networking community-based study.<br />
Mayo Clinic Proc. 2011;86(9):845-850.
LUNG CANCER SCREENING 101<br />
CHAIR<br />
Christine D. Berg, MD<br />
National Cancer Institute<br />
Bethesda, MD<br />
SPEAKERS<br />
Denise R. Aberle, MD<br />
University <strong>of</strong> California, Los Angeles David Geffen School <strong>of</strong> Medicine<br />
Los Angeles, CA<br />
Douglas E. Wood, MD<br />
University <strong>of</strong> Washington<br />
Seattle, WA
Lung Cancer Screening: Promise and Pitfalls<br />
By Christine D. Berg, MD, Denise R. Aberle, MD, and Douglas E. Wood, MD<br />
OVERVIEW: The results <strong>of</strong> the National Lung Screening Trial<br />
(NLST) have provided the medical community and <strong>American</strong><br />
public with considerable optimism about the potential to<br />
reduce lung cancer mortality with imaging-based screening.<br />
Designed as a randomized trial, the NLST has provided the<br />
first evidence <strong>of</strong> screening benefit by showing a 20% reduction<br />
in lung cancer mortality and a 6.7% reduction in all-cause<br />
mortality with low dose helical computed tomography (LDCT)<br />
screening relative to chest X-ray. The major harms <strong>of</strong> LDCT<br />
screening include the potential for radiation-induced carcinogenesis;<br />
high false-positivity rates in individuals without lung<br />
cancer, and overdiagnosis. Following the results <strong>of</strong> the NLST,<br />
the National Comprehensive Cancer Network (NCCN) published<br />
the first <strong>of</strong> multiple lung cancer screening guidelines<br />
under development by major medical organizations. These<br />
FOR DECADES, the early detection <strong>of</strong> common cancers<br />
has been advocated in an attempt to improve the chance<br />
for long-term survival and cure. Breast, colon, and prostate<br />
cancer all have established screening programs that are<br />
covered by insurers, embraced by physicians and the public,<br />
endorsed by pr<strong>of</strong>essional societies and policy-makers, and<br />
touted as critical public-health measures as the U.S. health<br />
care system strives to prevent rather than treat disease.<br />
Lung cancer, the leading cause <strong>of</strong> cancer death in the United<br />
States and the world has lagged behind. There are many<br />
reasons, including that patients with lung cancer may suffer<br />
from the public’s and policy-makers’ perception <strong>of</strong> lung<br />
cancer as a “self-inflicted” disease. Also, the poor long term<br />
survivorships <strong>of</strong> lung cancer patients has compromised advocacy<br />
efforts. However, we are now at the beginning <strong>of</strong> a<br />
new and exciting era for patients with lung cancer. Revolutions<br />
in molecular genetics and modern technology have<br />
begun to have an effect on the course <strong>of</strong> this here-to-for<br />
highly lethal disease. For adenocarcinomas, in particular,<br />
several targeted therapies, such as the use <strong>of</strong> erlotinib, have<br />
emerged. Because <strong>of</strong> the National Lung Cancer Screening<br />
Trial (NLST), medical imaging advances have recently assessed<br />
the utilization <strong>of</strong> computerized tomographic scanning<br />
<strong>of</strong> the lung with a low radiation dose technique and provided<br />
the medical community and patients with optimism. 1<br />
Evidence for Benefit<br />
Previously, trials that tested lung cancer screening with<br />
chest x-ray (CXR) showed disappointing results. Four randomized<br />
trials included one study in the Czech Republic and<br />
three National Cancer Institute (NCI)-sponsored trials that<br />
included sputum cytology in two trials that demonstrated no<br />
effect on lung-cancer specific mortality. 2-55 The Mayo Lung<br />
Project (MLP), in particular, demonstrated a known problem<br />
with screening, i.e., <strong>of</strong> overdiagnosis—detecting lesions so<br />
indolent that they are not medically significant, with 17%<br />
more lung cancers detected in the screened arm, an excess<br />
that persisted for at least 20 years. 6 (Interestingly, a recent<br />
reanalysis <strong>of</strong> the MLP and the Johns Hopkins Lung Project<br />
shows some possible evidence <strong>of</strong> a very small beneficial<br />
mortality effect with sputum analysis. 7 ) However, as a<br />
consequence <strong>of</strong> these studies, no major medical group recommended<br />
lung cancer screening until recently.<br />
450<br />
recommendations amalgamated screening cohorts, practices,<br />
interpretations, and diagnostic follow-up based on the NLST<br />
and other published studies to provide guidance for the<br />
implementation <strong>of</strong> LDCT screening. There are major areas <strong>of</strong><br />
opportunity to optimize implementation. These include standardizing<br />
practices in the screening setting, optimizing risk<br />
pr<strong>of</strong>iles for screening and for managing diagnostic evaluation<br />
in individuals with indeterminate nodules, developing interdisciplinary<br />
screening programs in conjunction with smoking<br />
cessation, and approaching all stakeholders systematically to<br />
ensure the broadest education and dissemination <strong>of</strong> screening<br />
benefits relative to risks. The incorporation <strong>of</strong> validated biomarkers<br />
<strong>of</strong> risk and preclinical lung cancer can substantially<br />
enhance the effectiveness screening programs.<br />
The Prostate, Lung, Colorectal, and Ovarian Cancer<br />
screening trial (PLCO) was launched in 1993 as a multimodal<br />
screening trial with ambitious goals. One was to<br />
assess with a large sample size the effect <strong>of</strong> CXR (posteroanterior<br />
only) screening (three rounds for nonsmokers and<br />
four rounds for current or former smokers) on lung cancer<br />
mortality. The trial enrolled 154, 901 individuals, 10% <strong>of</strong><br />
whom were current smokers and 41.5% former smokers. The<br />
result unfortunately confirmed that this approach did not<br />
affect lung cancer mortality. 8 There was perhaps some<br />
evidence <strong>of</strong> overdiagnosis—but not <strong>of</strong> the magnitude seen<br />
with the Mayo Lung Project.<br />
Computed tomography (CT) has been clinically available<br />
in the United States for decades; however, image-acquisition<br />
time was slow until the advent <strong>of</strong> the helical CT. Also, until<br />
it was demonstrated that a low-dose technique could reliably<br />
image the lung parenchyma, valid concerns about radiation<br />
dose and subsequent carcinogenesis discouraged its use for<br />
screening a healthy population. 9 With the advent <strong>of</strong> low-dose<br />
helical computed tomography (LDCT), several groups undertook<br />
screening <strong>of</strong> at-risk individuals and reported promising<br />
results. Among these was the Early Lung Cancer<br />
Action Program (ELCAP), in which 1000 individuals at risk<br />
<strong>of</strong> lung cancer underwent combined chest-x-ray and LDCT<br />
screening. A high proportion <strong>of</strong> early stage lung cancers<br />
were observed using LDCT, and the ELCAP was among the<br />
major studies to firmly introduce LDCT screening into the<br />
<strong>American</strong> consciousness. A number <strong>of</strong> downsides emerged<br />
from these various studies, including high false positivity<br />
rates and the challenges <strong>of</strong> distinguishing true mortality<br />
benefit from the well-known biases <strong>of</strong> lead-time, length, and<br />
overdiagnosis that arise from single arm screening studies.<br />
From the Early Detection Research Group, Division <strong>of</strong> Cancer Prevention, National<br />
Cancer Institute, Bethesda, MD; Radiological Sciences, David Geffen School <strong>of</strong> Medicine at<br />
UCLA, Los Angeles, CA; Division <strong>of</strong> Cardiothoracic Surgery, University <strong>of</strong> Washington,<br />
Seattle, WA.<br />
Authors’ disclosures <strong>of</strong> potential conflicts <strong>of</strong> interest are found at the end <strong>of</strong> this article.<br />
Address reprint requests to Christine Berg, MD, National Cancer Institute, Early<br />
Detection Research Group, Division <strong>of</strong> Cancer Prevention, Executive Plaza North, Room<br />
3112, 6130 Executive Boulevard, MSC 7346, Bethesda, MD 20892; email: bergc@<br />
mail.nih.gov.<br />
© <strong>2012</strong> by <strong>American</strong> <strong>Society</strong> <strong>of</strong> <strong>Clinical</strong> <strong>Oncology</strong>.<br />
1092-9118/10/1-10
LUNG CANCER SCREENING<br />
The concern in much <strong>of</strong> the scientific community was that<br />
given these limitations, as well as the highly aggressive and<br />
heterogeneous biology <strong>of</strong> lung cancer, the validity <strong>of</strong> LDCT<br />
screening would require demonstrate <strong>of</strong> a mortality benefit<br />
in order to have great public health importance. Therefore,<br />
NCI decided that a randomized, controlled clinical trial<br />
should be designed and conducted to determine reliably the<br />
effect <strong>of</strong> LDCT on lung cancer–specific mortality.<br />
The NLST was designed to answer one question in the<br />
shortest, most definitive manner: does screening with LDCT<br />
in an at-risk population lower lung cancer–specific mortality?<br />
A 20% mortality reduction was judged to be clinically<br />
important and feasible to assess. The entry criteria for the<br />
trial were set to bring a high-risk group into screening.<br />
These criteria included current or former smokers 55 to74<br />
years, 30 pack-years <strong>of</strong> smoking, if former smokers having<br />
quit within 15 years. The participants had to be asymptomatic<br />
and healthy enough to withstand surgery, and could not<br />
have had prior invasive cancers, and also not have had a CT<br />
within the last18 months <strong>of</strong> enrollment. The trial was to<br />
compare three rounds <strong>of</strong> screening at 12 month intervals<br />
with LDCT compared to postero-anterior CXR. Consideration<br />
was given to having a third arm <strong>of</strong> no screening, but, as<br />
the PLCO was ongoing, it was decided to compare the NLST<br />
results to a matched cohort in the PLCO. 11 Also, an analysis<br />
was done to assess whether or not adding rounds <strong>of</strong> screening<br />
would have any major, additional benefit; results showed<br />
that they did not. With these parameters, the needed sample<br />
size was 50,000 with 90% power and an � <strong>of</strong> 5% to determine<br />
a 20% mortality reduction (p � 0.05). The NCI launched the<br />
trial, which was managed by both the Lung Screening<br />
Study, under contract to the Division <strong>of</strong> Cancer Prevention,<br />
and the <strong>American</strong> College <strong>of</strong> Radiology Imaging Network, a<br />
cooperative group through the Division <strong>of</strong> Cancer Treatment<br />
and Diagnosis.<br />
Accrual was rapid. Over 20 months from August 2002 to<br />
April 2004, 53,454 individuals were enrolled. Overall, the<br />
trial participants were younger, <strong>of</strong> higher educational status,<br />
and more likely to be former smokers than those who<br />
also matched the NLST entry criteria in the U. S. population.<br />
12 Screening was accomplished with a high degree <strong>of</strong><br />
KEY POINTS<br />
● Screening with low dose computed tomography<br />
(LDCT) has been shown to reduce lung cancer by 20%<br />
relative to chest X-ray.<br />
● Complication rates from screening and downstream<br />
diagnostic procedures are low in individuals with<br />
positive screens.<br />
● The major harms <strong>of</strong> screening relate to high false<br />
positivity rates, the potential for radiation-induced<br />
carcinogenesis, and overdiagnosis.<br />
● LDCT screening implementation should be interdisciplinary<br />
and integrated with smoking cessation programs<br />
to derive maximum benefit.<br />
● Successful dissemination <strong>of</strong> LDCT screening will require<br />
a systematically approach to address the<br />
unique challenges <strong>of</strong> the screening centers, primary<br />
care environment, and population at risk.<br />
compliance in both arms—95% in LDCT and 93% in CXR.<br />
Follow-up continued after screening was completed. Regular<br />
assessments were done through questionnaires to participants<br />
and searches through tumor registries, National<br />
Death Index, and other sources to determine whether or not<br />
a participant developed lung cancer, whether he or she died,<br />
and the cause <strong>of</strong> death. A detailed endpoint verification<br />
process was undertaken to ensure the highest degree <strong>of</strong><br />
accuracy and consistency in determining cause <strong>of</strong> death,<br />
particularly for lung cancer and deaths possibly from complications<br />
related to procedures done to evaluate for lung<br />
cancer. In October 2010, the Data and Safety Monitoring<br />
Board observed that a stopping boundary had been crossed<br />
and recommended that the trial cease. In November 2010,<br />
the initial findings from the NLST were released. On June<br />
29, 2011, the primary results were published online in the<br />
New England Journal <strong>of</strong> Medicine and appeared in the print<br />
issue on August 4, 2011. 1<br />
Screening with LDCT resulted in a 20% decrease in<br />
lung-cancer specific mortality. A subset <strong>of</strong> the PLCO that<br />
matched the NLST was analyzed. There was no evidence<br />
that when compared with community care there was any<br />
mortality reduction with CXR, and the lung cancer–specific<br />
mortality in this cohort in the PLCO was the same as in the<br />
CXR arm <strong>of</strong> the NLST. The NLST was the first ever report<br />
from a randomized clinical trial documenting that lung<br />
cancer mortality could be reduced with a screening modality.<br />
Overall mortality was also lowered. However, when lung<br />
cancer deaths were removed, this difference was no longer<br />
statistically significant (p � 0.05). Compared with CXR,<br />
LDCT screening was associated with a stage shift towards<br />
earlier stages for all histologies <strong>of</strong> non-small cell lung<br />
cancer. There was no stage shift with small cell lung cancers.<br />
Unfortunately, the detection <strong>of</strong> limited small cell carcinoma<br />
was not enhanced with LDCT compared with CXR.<br />
During screening in the LDCT arm, 649 cases <strong>of</strong> lung cancer<br />
were diagnosed after a positive screen and 44 cases as<br />
interval cancer, whereas in the CXR arm, 279 cases were<br />
diagnosed after a positive screen, with 137 as interval<br />
cancer. A total <strong>of</strong> 1060 cases <strong>of</strong> lung cancer occurred in the<br />
LDCT arm compared with 941 in the CXR arm. Therefore,<br />
129 additional cases <strong>of</strong> lung cancer were diagnosed in the<br />
LDCT arm than in the CXR arm; this absolute number is not<br />
an estimate <strong>of</strong> the amount <strong>of</strong> overdiagnosis. More follow-up<br />
would be helpful to determine precisely the numbers <strong>of</strong><br />
excess cancers detected with LDCT compared with CXR that<br />
would not come to clinical detection during a participant’s<br />
life time. Alternatively, modeling can be done to better<br />
estimate the amount <strong>of</strong> overdiagnosis. Several approaches to<br />
determining overdiagnosis exist and further work is planned<br />
on this topic.<br />
A positive result <strong>of</strong> suspected lung cancer was defined as<br />
a nodule � 4 mm, or other findings potentially related to<br />
lung cancer. The average percentage <strong>of</strong> positive screens was<br />
high: 24.2% <strong>of</strong> LDCTs and 6.9% <strong>of</strong> CXRs. The chance for a<br />
participant overall after three screens to have one positive<br />
result was 39.1% in the LDCT arm and 16.0% in the CXR<br />
arm. Other significant abnormalities were also found more<br />
frequently in the LDCT arm than the CXR arm (7.5%<br />
compared with 2.1%, respectively). During the trial, there<br />
were guidelines for the evaluation <strong>of</strong> a positive screen both<br />
in the LDCT arm and the CXR arm. These were not<br />
mandatory, as it was judged important to leave follow-up<br />
451
in the hands <strong>of</strong> the radiologists and physicians caring for<br />
the individual patients, taking into account regional differences<br />
and patient-specific preferences. Fortunately, much <strong>of</strong><br />
the evaluation <strong>of</strong> abnormalities could be conducted noninvasively.<br />
In general, a clinical evaluation and a diagnostic CT<br />
were performed. For a diagnosis <strong>of</strong> malignancy, invasive<br />
procedures were performed. The number <strong>of</strong> invasive procedures<br />
per malignancy diagnosed was relatively low. Complications<br />
were few (1.4% in the LDCT arm and 1.6% in the<br />
CXR arm). Major complications were primarily seen in<br />
individuals with underlying lung cancer, i.e., in the LDCT<br />
arm the rate <strong>of</strong> major complications in those with lung<br />
cancer was 11.2% compared to 0.06% in those without. 1<br />
Reader variability studies done in the NLST reported that<br />
radiologists have a low level <strong>of</strong> agreement in detecting<br />
nodules and in measuring the growth <strong>of</strong> nodules. 13,14 A<br />
panel <strong>of</strong> radiologists reviewed NLST baseline studies and<br />
reported the total number <strong>of</strong> abnormalities detected and<br />
classified as pulmonary nodules. There was up to a two-fold<br />
difference among radiologists. For cases classified as positive,<br />
consistency among recommendations for follow-up<br />
was poor. A similar study was done to assess changes in<br />
nodule morphology between two annual scans. Out <strong>of</strong> 95<br />
nodules originally interpretend as present on both scans, 19<br />
were judged by at least one <strong>of</strong> nine independent reviewers<br />
not to be present initially. 14<br />
Another potential harm from an imaging test with ionizing<br />
radiation is that <strong>of</strong> radiation carcinogenesis. The medical<br />
physics group involved with the trial was diligent in setting<br />
image acquisition parameters to keep the dose low and to<br />
keep image quality high. 15,16 Effective doses were estimated<br />
using volume CT dose index (CTDI) for the 97 scanners, and<br />
a whole-body mean effective dose was calculated. This was<br />
1.4 millisievert (mSv). This compares with the average<br />
whole-body effective dose <strong>of</strong> 7 mSv from diagnostic CT. 17 An<br />
estimate <strong>of</strong> radiation-induced cancers in an individual<br />
screened at 55, 56 and 57 years with the NLST LDCT<br />
technique is 1 to 3 lung cancer deaths per 10,0000, and<br />
breast cancers induced is 0.3 per 10,000. This compares with<br />
30 lung cancer deaths prevented per 10,000 screened three<br />
times. Of note, the cancers caused by radiation would occur<br />
many years after the screen whereas deaths prevented occur<br />
within a few years. 18<br />
A detailed assessment <strong>of</strong> cost-effectiveness is planned<br />
utilizing data from the NLST. This will take into account not<br />
only screening and diagnostic evaluation costs for positive<br />
screens but those evaluations undertaken in those screened<br />
who had other abnormalities or entered the medical system<br />
as a consequence <strong>of</strong> the screen. A preliminary report indicated<br />
that the incremental cost per year <strong>of</strong> life gained in the<br />
NLST was $38,000. 19 Additionally, work with the Cancer<br />
Intervention and Surveillance Network (CISNET) is ongoing.<br />
The CISNET groups will validate and improve as<br />
needed their models using the NLST results. Other questions<br />
then that are very important when considering implementation<br />
<strong>of</strong> screening in the population, such as age at<br />
which to start screening, other smoking intensities, as well<br />
as other frequencies and durations <strong>of</strong> screening, will be<br />
addressed.<br />
Concerns for Implementation<br />
Clearly this is a major advance for patients at risk for lung<br />
cancer and will mean a major policy change by payers,<br />
452<br />
BERG, ABERLE, AND WOOD<br />
policy-makers, guideline groups, and patient advocates.<br />
However, this enthusiasm, although deserved, must be<br />
tempered with caution. Another view <strong>of</strong> the NLST data<br />
reveals that it is necessary to screen 320 individuals every<br />
12 months for three rounds for each lung cancer death<br />
avoided. Many patients will be exposed to the emotional<br />
and physical risks <strong>of</strong> lung-cancer screening to achieve<br />
the desired benefit. A careful, measured approach is<br />
important for the institution <strong>of</strong> lung cancer screening nationwide.<br />
For the radiology community, the following considerations<br />
apply: the screening process itself should be standardsdriven.<br />
Image-acquisition protocols must be consistent to<br />
ensure adequate image quality at the lowest reasonable<br />
radiation exposure 17,20 ; this is particularly important if<br />
computer-aided diagnosis (CAD) is incorporated into routine<br />
nodule detection and characterization. 21,22 Imagers experienced<br />
in the management <strong>of</strong> lung nodules should provide<br />
interpretations and use consistent follow-up guidelines. 23,24<br />
Viable commercial solutions to track patients and nodules<br />
do not currently exist but would have a major beneficial<br />
effect on screening effectiveness and efficiencies; this critical<br />
need should be the basis for developing partnerships between<br />
screening centers and industry to understand how<br />
s<strong>of</strong>tware technologies can facilitate workflow. There are<br />
several questions that remain to be addressed by the imaging<br />
community. Among them are the following:<br />
Interpretation guidelines. Screening interpretations in<br />
the NLST were largely dichotomous, based on considerations<br />
<strong>of</strong> nodule size and morphology. The NELSON trial<br />
being conducted in the Netherlands and Denmark uses a<br />
two-tiered interpretation paradigm in which nodules falling<br />
between certain size thresholds are considered “indeterminate,”<br />
which mandate a 3-month follow-up LDCT to determine<br />
whether the screen is negative or positive. 23 Using this<br />
algorithm, the positive predictive value <strong>of</strong> LDCT was substantially<br />
improved. Although it may be argued that medical<br />
resource utilization is not significantly different between the<br />
two interpretation paradigms, the implications <strong>of</strong> the<br />
screening result using the NELSON model more closely<br />
approximate lung cancer “risk” in individuals with indeterminate<br />
nodules, which has significant implications for both<br />
the individual patient and her/his provider.<br />
Results communication. Screening centers should not<br />
only communicate results to the one being screened and his<br />
or her provider but have the necessary resources to follow up<br />
individuals with indeterminate nodules while keeping the<br />
primary provider fully informed.<br />
Role <strong>of</strong> image analysis. Screening interpretation in the<br />
NLST was based on visual assessment. The European<br />
screening trials have predicated results interpretation on<br />
quantitative nodule volumetry. The incorporation <strong>of</strong> quantitative<br />
s<strong>of</strong>tware into the screening process will impose<br />
modifications to workflow in imaging practice and will<br />
probably result in the expansion <strong>of</strong> trained allied personnel<br />
who can oversee s<strong>of</strong>tware analysis before formal radiologist<br />
review.<br />
The implications <strong>of</strong> lung cancer screening in the primarycare<br />
setting are substantial. Primary-care providers will<br />
need to be convinced that LDCT can be effective in reducing<br />
lung-cancer mortality and that the benefits <strong>of</strong> LDCT screening<br />
outweigh the potential harms <strong>of</strong> radiation exposure, high<br />
false positivity rates, and potential overdiagnosis. The in-
LUNG CANCER SCREENING<br />
Fig. 1. Guidelines published by the National Comprehensive Cancer Network for the diagnostic evaluation <strong>of</strong> positive screens in which solid<br />
or part-solid nodules are detected. Abbreviations: LDCT, low-dose helical computed tomography; PET/CT, positron emission tomography/<br />
computed tomography. Reproduced with permission from the NCCN <strong>Clinical</strong> Practice Guidelines in <strong>Oncology</strong> (NCCN Guidelines for Lung Cancer<br />
Screening V.1.<strong>2012</strong>). © <strong>2012</strong> National Comprehensive Cancer Network, Inc. All rights reserved. The NCCN Guidelines and illustrations herein<br />
may not be reproduced in any form for any purpose without the express written permission <strong>of</strong> the NCCN. To view the most recent and complete<br />
version <strong>of</strong> the NCCN Guidelines, go online to NCCN.org. NATIONAL COMPREHENSIVE CANCER NETWORK ® , NCCN ® , NCCN GUIDELINES ® , and all<br />
other NCCN Content are trademarks owned by the National Comprehensive Cancer Network, Inc.<br />
corporation <strong>of</strong> lung cancer screening adds additional complexity<br />
to a busy out-patient clinic, erodes already limited<br />
clinical time and resources, and disrupts workflow. The<br />
intent <strong>of</strong> screening and its importance must be communicated<br />
unambiguously to patients, screening exams must be<br />
scheduled, patients with significant findings on screens<br />
must be referred for additional testing, and patients counseled<br />
about likely outcomes and their implications. Finally,<br />
screening implementation will compete for attention with<br />
other validated and less costly health care measures like<br />
breast cancer or colon cancer screening, smoking cessation,<br />
weight loss, and exercise. 25,26<br />
There are formidable challenges to the implementation <strong>of</strong><br />
lung cancer screening from the community perspective.<br />
Successful implementation will require the diffusion <strong>of</strong><br />
screening across all socioeconomic strata. Community engagement<br />
will be an important element <strong>of</strong> implementation<br />
in many cultural settings and requires trust and a successful<br />
dialogue in which community members can be informed<br />
about the health consequences <strong>of</strong> smoking, lung-cancer risk,<br />
and the balance <strong>of</strong> benefits versus risks <strong>of</strong> early detection,<br />
while also educating the medical pr<strong>of</strong>ession regarding community<br />
priorities. The diffusion <strong>of</strong> screening and preventive<br />
services is particularly important in underserved and minority<br />
populations because these communities are disproportionately<br />
adversely affected by lung cancer; they are<br />
commonly diagnosed at advanced stages, less commonly<br />
undergo surgical resections, and, in particular, black men<br />
have a lower overall survival from lung cancer. If lungcancer<br />
screening is to be equitably administered to all<br />
individuals at risk, the following barriers must be addressed:<br />
lack <strong>of</strong> awareness and low prioritization <strong>of</strong> lungcancer<br />
prevention and early detection; cultural concerns <strong>of</strong><br />
trust, fatalism, and stigmatization; financial constraints;<br />
and geographical barriers to access. 27,28<br />
The NLST enrolled only patients at high risk <strong>of</strong> lung<br />
cancer. Although it would be naïve and narrow-minded to<br />
not recognize that additional patients, outside <strong>of</strong> the NLST<br />
criteria, may have substantial risk <strong>of</strong> lung cancer that<br />
warrants screening, one must be very cautious in extrapo-<br />
453
lating the NLST results to other patient populations and<br />
recognize the unintended consequences <strong>of</strong> screening these<br />
patients. Second, the NLST centers included only programs<br />
with substantial experience and resources in radiology,<br />
pulmonary medicine, thoracic surgery, and pathology, along<br />
with a disciplined and multidisciplinary approach to nodule<br />
management. Ninety-six percent <strong>of</strong> lung nodules found were<br />
ultimately determined to be “false positives,” and 39% <strong>of</strong><br />
patients had at least one positive result during the study. A<br />
highly organized and disciplined approach to management<br />
is the only way to mitigate the potential harms caused to<br />
patients by excessive and unnecessary testing and the<br />
morbidity <strong>of</strong> invasive procedures. The margin between net<br />
benefit and net harm in lung-cancer screening is likely<br />
small, and the benefit to patients could easily be lost if a<br />
higher percentage <strong>of</strong> the patients with false positive findings<br />
undergo unnecessary work-up and invasive testing. Successful<br />
implementation <strong>of</strong> lung-cancer screening will require the<br />
following: (1) pragmatic and thoughtful guidelines that<br />
define patients eligible for screening (not limited to NLST<br />
criteria yet reasonably narrow in scope); (2) experienced<br />
radiologists to interpret screening studies and minimize<br />
false positives; (3) a protocolized approach for the management<br />
<strong>of</strong> screen detected nodules; (4) diagnostic and therapeutic<br />
surgical procedures performed by board-certified<br />
thoracic surgeons in order to optimize staging and minimize<br />
morbidity; and (5) experienced multidisciplinary oncology<br />
management with thoracic surgery, medical, and radiation<br />
oncology to optimize oncology treatment and outcomes.<br />
Guideline Development<br />
The development <strong>of</strong> lung-cancer screening guidelines is<br />
underway, with the first being published by the National<br />
Comprehensive Cancer Network (NCCN) in October<br />
2011. 29 The NCCN has a strong history and experience<br />
in the development <strong>of</strong> cancer guidelines. The group assembled<br />
a panel <strong>of</strong> 26 pr<strong>of</strong>essionals, representing thoracic surgery,<br />
radiology, pulmonary medicine, medical oncology,<br />
epidemiology, pathology, internal medicine, and patient advocacy,<br />
and worked together to produce the first lung cancer<br />
screening guidelines developed after the NLST. Most notable<br />
in the NCCN guidelines is the extrapolation <strong>of</strong> high-risk<br />
patients beyond the inclusion criteria <strong>of</strong> the NLST, to<br />
include patients 50 to 54 years (NLST included only 55 to 74<br />
years), and patients with � 20 pack per year smoking<br />
history (NLST required � 30 pack-years) if the patient had<br />
another lung-cancer risk factor as well (chronic obstructive<br />
pulmonary disease, pulmonary fibrosis, radon or occupational<br />
exposure, cancer history, or family history). Another<br />
extrapolation was to recommend that screening continue<br />
annually until the individual reached 74 years. The NCCN<br />
also recommended a highly protocolized approach to the<br />
follow-up, work-up, and invasive testing <strong>of</strong> positive findings,<br />
similar to those recommended by the Fleischner <strong>Society</strong> 24<br />
and others.<br />
The biggest challenge in lung-cancer screening is the<br />
thoughtful management <strong>of</strong> screen- detected nodules, the<br />
majority <strong>of</strong> which are benign. There are several variations<br />
on the management <strong>of</strong> screen-detected lung nodules, proposed<br />
by the Fleischner <strong>Society</strong>, 24 the International Early<br />
Lung Cancer Action Program (I-ELCAP), 30 the NLST, 1 the<br />
NELSON Trial, 23 and specific recommendations regarding<br />
nonsolid nodules by Godoy and colleagues. 31 The NCCN<br />
454<br />
BERG, ABERLE, AND WOOD<br />
Lung Screening Panel has amalgamated these recommendations<br />
into a pragmatic algorithm for nodule management<br />
(Fig. 1 and Fig. 2). 29 The NCCN recommendations are less<br />
aggressive than the I-ELCAP for the work-up <strong>of</strong> baseline,<br />
new solid, and part solid nodules � 6 mm. The NCCN<br />
recommendations are also slightly different in recommending<br />
a contrast enhanced CT or positron emission tomography<br />
(PET) in the evaluation <strong>of</strong> solid or part solid nodules<br />
� 8 mm. Finally, the NCCN defined nodule growth as either<br />
an increase in the mean diameter <strong>of</strong> 2 mm or more for<br />
nodules � 15 mm or in the solid portion <strong>of</strong> a part-solid<br />
nodule, or an increase <strong>of</strong> 15% or more in the mean diameter<br />
for nodules � 15 mm. This definition <strong>of</strong> nodule growth is<br />
simplified compared with I-ELCAP and should result in<br />
fewer false positive results than seen in the NLST. Of note,<br />
surveys <strong>of</strong> compliance with the Fleischner <strong>Society</strong> guidelines<br />
have shown only 35% to 60% compliance by members <strong>of</strong> the<br />
Radiological <strong>Society</strong> <strong>of</strong> North America, 32 and 27% compliance<br />
by members <strong>of</strong> the <strong>Society</strong> <strong>of</strong> Thoracic Radiology, 33 with<br />
an overall trend toward over-management. It will be important<br />
for the successful application <strong>of</strong> screening programs to<br />
assure an algorithmic and disciplined approach to nodule<br />
work-up and follow-up in order to minimize the serious<br />
potential harms from excessive and invasive testing in these<br />
patients undergoing screening.<br />
Once a nodule has been identified, the involvement <strong>of</strong> an<br />
experienced thoracic surgeon will help the multidisciplinary<br />
team refine a strategy for further work-up, including biopsy<br />
and/or resection. The Lung Cancer Early Detection and<br />
Prevention Clinic at the University <strong>of</strong> Washington incorporates<br />
a “Nodule Board,” consisting <strong>of</strong> specialists from thoracic<br />
radiology, pulmonary medicine, and thoracic surgery.<br />
This group reviews clinical details and imaging and develops<br />
a management plan based on a treatment algorithm<br />
and informed by the combined expertise <strong>of</strong> the involved<br />
specialists. The Non–Small Cell Lung Cancer Panel <strong>of</strong> the<br />
NCCN now recommends assessment and management <strong>of</strong><br />
presumed or proven lung cancer by “board certified thoracic<br />
surgeons who perform lung cancer surgery as a prominent<br />
part <strong>of</strong> their practice.” 34 This recommendation is based on<br />
the data that as much as 50% <strong>of</strong> lung cancer surgery in the<br />
United States continues to be performed by general surgeons<br />
and that surgical outcomes (morbidity and mortality),<br />
as well as oncology outcomes (correct staging, extent <strong>of</strong><br />
resection, and cancer survival) are better when performed<br />
by specialists in thoracic surgery. 35,36 There are multiple<br />
potential adverse consequences <strong>of</strong> nonspecialist surgery,<br />
which are even more pr<strong>of</strong>ound for recipients <strong>of</strong> lung-cancer<br />
screening: unnecessary surgery in cases where follow-up or<br />
other diagnostic testing may have been preferred, inadequate<br />
staging before and/or during lung cancer surgery,<br />
underutilization <strong>of</strong> minimally invasive surgery for both<br />
diagnostic and resection procedures, and a lack <strong>of</strong> advanced<br />
techniques (segmentectomy or sleeve resection) to minimize<br />
the extent <strong>of</strong> pulmonary resection. Specialist thoracic<br />
surgeons, working with a multidisciplinary lung cancer<br />
team, are best equipped to help maximize the benefit <strong>of</strong> early<br />
detection. They are an important part <strong>of</strong> avoiding the adverse<br />
consequences <strong>of</strong> unnecessary procedures or substandard<br />
cancer outcomes that potentially could result in more<br />
harm than good from lung-cancer screening programs applied<br />
without adherence to guidelines and necessary pr<strong>of</strong>essional<br />
expertise.
LUNG CANCER SCREENING<br />
Fig. 2. Guidelines published by the National Comprehensive Cancer Network for the diagnostic evaluation <strong>of</strong> positive screens in which<br />
non-solid (ground glass) nodules are detected. Abbreviations: LDCT, low-dose helical computed tomography. Reproduced with permission from<br />
the NCCN <strong>Clinical</strong> Practice Guidelines in <strong>Oncology</strong> (NCCN Guidelines for Lung Cancer Screening V.1.<strong>2012</strong>). © <strong>2012</strong> National Comprehensive<br />
Cancer Network, Inc. All rights reserved. The NCCN Guidelines and illustrations herein may not be reproduced in any form for any purpose<br />
without the express written permission <strong>of</strong> the NCCN. To view the most recent and complete version <strong>of</strong> the NCCN Guidelines, go online to<br />
NCCN.org. NATIONAL COMPREHENSIVE CANCER NETWORK ® , NCCN ® , NCCN GUIDELINES ® , and all other NCCN Content are trademarks owned<br />
by the National Comprehensive Cancer Network, Inc.<br />
Opportunities<br />
There are major opportunities to be gained in the process<br />
<strong>of</strong> screening implementation. The NLST successfully addressed<br />
the critical endpoint <strong>of</strong> differential lung cancer<br />
mortality, and secondary analyses will inform the costeffectiveness<br />
<strong>of</strong> LDCT in older, heavy smokers. However,<br />
broad-scale implementation <strong>of</strong> LDCT screening is predicated<br />
on several variables that the NLST does not directly address<br />
and for which further research is crucial. Among these are<br />
considerations <strong>of</strong> the optimal risk pr<strong>of</strong>ile <strong>of</strong> those who are<br />
screened, and how risk pr<strong>of</strong>iles might be used to guide<br />
diagnostic strategies.<br />
Morphologic features <strong>of</strong> indeterminate lung nodules on<br />
CT have been studied as potential predictors <strong>of</strong> lung cancer.<br />
Most analyses have relied on subjective visual assessment<br />
<strong>of</strong> nodule features such as: size (diameter), consistency<br />
(ground glass, part-solid, or solid), border definition, and<br />
internal features, such as reticulation, air bronchograms<br />
and bubble-like lucencies. 37 Quantitative analysis <strong>of</strong> lung<br />
nodules using CAD seeks to characterize nodules by math-<br />
ematical feature descriptors. We are at the cusp <strong>of</strong> validating<br />
analytic s<strong>of</strong>tware that can reproducibly characterize<br />
lung nodules across a range <strong>of</strong> nodule types. 38,39 Such<br />
nodule characterization could become standard in the diagnostic<br />
stratification <strong>of</strong> individuals with indeterminate nodules.<br />
Between 80% and 90% <strong>of</strong> lung cancers occur in tobacco<br />
smokers, yet only 10% to 15% <strong>of</strong> chronic smokers develop<br />
lung cancer. Prospective studies have also shown that approximately<br />
25% <strong>of</strong> smokers develop COPD as defined by<br />
spirometry, whereas 50% to 80% <strong>of</strong> patients with lung<br />
cancer have COPD. 40,41 Relative to smokers with normal<br />
lung function, those with COPD have up to a six-fold<br />
increased risk <strong>of</strong> lung cancer, making COPD by far the<br />
greatest risk factor for lung cancer in ever smokers. 41 These<br />
observations suggest an inherently greater risk <strong>of</strong> lung<br />
cancer among smokers with COPD than smokers with<br />
normal lung function. Although COPD and lung cancer have<br />
in common smoking exposure, several lines <strong>of</strong> evidence now<br />
support underlying shared genetic susceptibility that acts in<br />
455
concert with the shared risk <strong>of</strong> smoking-related genetic and<br />
epigenetic effects. Genome-association studies have identified<br />
several heritable susceptibility or protective loci thought<br />
to affect both COPD and lung cancer development: single<br />
nucleotide polymorphisms on loci 15q25 that regulate cholinergic<br />
nicotine receptors (CHRNA3/5); several haplotypes<br />
involved in the xenobiotic metabolism <strong>of</strong> tobacco lung carcinogens,<br />
and; genes involved in cell-cycle control, apoptosis,<br />
airway inflammation, and repair. 42,43<br />
Emphysema has recently been found to be associated<br />
with lung cancer, independent <strong>of</strong> airflow obstruction on<br />
spirometry. Emphysema can be directly quantified on LDCT<br />
with high reproducibility, and commercial s<strong>of</strong>tware is also<br />
available that can objectively quantify the severity <strong>of</strong><br />
smoking-related airway remodeling. 44 In patients with indeterminate<br />
nodules, such characterization could factor into<br />
diagnostic algorithms and may ultimately inform the determination<br />
<strong>of</strong> screening frequency at the individual patient<br />
level.<br />
Finally, the peripheral blood serves as a repository <strong>of</strong><br />
lung cancer-associated cytokines, soluble proteins, and<br />
microRNAs that derive from the tumor microenvironment<br />
and that exhibit molecular signatures similar to those<br />
in tumor tissues. 45,46 Similarly, samples <strong>of</strong> airway epithelium<br />
obtained through bronchoscopy, sputum expectoration,<br />
or nasal cellular brushings express aberrant methylation<br />
and microRNA patterns observed in lung cancers. 47,48<br />
Authors’ Disclosures <strong>of</strong> Potential Conflicts <strong>of</strong> Interest<br />
Author<br />
Christine D. Berg*<br />
Denise R. Aberle*<br />
Douglas E. Wood*<br />
*No relevant relationships to disclose.<br />
Employment or<br />
Leadership<br />
Positions<br />
Consultant or<br />
Advisory Role<br />
1. The National Lung Screening Trial Research Team. Reduced lungcancer<br />
mortality with low-dose computed tomographic screening. N Engl<br />
J Med. 2011;365:395-409.<br />
2. Melamed MR, Flehinger BJ, Zaman MB, et al. Screening for early lung<br />
cancer: Results <strong>of</strong> the Memorial Sloan-Kettering study in New York. Chest.<br />
1984;86:44-53.<br />
3. Tockman MS, Levin ML, Frost JK, et al. Screening and detection <strong>of</strong> lung<br />
cancer. In Aisner J (ed). Lung Cancer. New York: Churchill Livingstone,1985;<br />
25-26.<br />
4. Fontana RS, Sanderson DR, Woolner LB, et al. Lung cancer screening:<br />
The Mayo program. J Occup Med. 1986;28:746-750.<br />
5. Kubik A, Parkin DM, Khlat M, et al. Lack <strong>of</strong> benefit from semi-annual<br />
screening for cancer <strong>of</strong> the lung: Follow-up report <strong>of</strong> a randomized controlled<br />
trial on a population <strong>of</strong> high-risk males in Czechoslovakia. Int J Cancer.<br />
1990;45:26-33.<br />
6. Marcus PM, Bergstralh EJ, Fagerstrom RM, et al. Lung cancer mortality<br />
in the Mayo Lung Project: Impact <strong>of</strong> extended follow-up. J Natl Cancer Inst.<br />
2000;92:1308-1316.<br />
7. Doria-Rose VP, Marcus PM, Szabo E, et al. Randomized controlled trials<br />
<strong>of</strong> the efficacy <strong>of</strong> lung cancer screening by sputum cytology revisited:a combined<br />
mortality analysis from the Johns Hopkins Lung Project and the<br />
Memorial Sloan-Kettering Lung Study. Cancer. 2009;115:5007-5017.<br />
8. Oken MM, Hocking WG, Kvale PA, et al. Screening by chest radiograph<br />
and lung cancer mortality: The Prostate, Lung, Colorectal, and Ovarian<br />
(PLCO) randomized trial. JAMA. 2011;306:1865-1873.<br />
9. Naidich DP, Marshall CH, Gribbin C, et al. Low-dose CT <strong>of</strong> the lungs:<br />
preliminary observations. Radiology. 1990;175:729-731.<br />
10. The National Lung Screening Trial Research Team. The National<br />
456<br />
If validated, these lung cancer-specific molecular signatures<br />
fromeasily accessible tissues will enable their<br />
translation into clinical practice and will substantially alter<br />
how we define lung-cancer risk and screening in the<br />
future. At 15 <strong>of</strong> the NLST centers, sponsored by the <strong>American</strong><br />
College <strong>of</strong> Radiology Imaging Network, participants<br />
volunteered to provide serial blood, sputum, and urine<br />
specimens. Lung cancer and other tissue specimens were<br />
collected across the trial and used to construct tissuemicroarrays.<br />
These specimens, when combined with the<br />
voluminous data from the study, may be useful in enhancing<br />
this molecular-signature research. The biospecimens are<br />
available to the research community through a peerreviewed<br />
process. 49<br />
As we begin to more systematically define lung-cancer risk<br />
through combinations <strong>of</strong> clinical, phenotypic, and molecular<br />
pr<strong>of</strong>iling, we will be better positioned to distinguish between<br />
individuals who have lung cancer versus no cancer. Such<br />
discrimination can significantly lower the harms <strong>of</strong> screening<br />
by reducing unnecessary interventions, minimizing<br />
anxiety, and lowering costs while promoting early diagnosis<br />
and intervention. Finally, the integration <strong>of</strong> biologic and<br />
imaging-based biomarkers to define risk provides significant<br />
opportunity to stimulate the motivational tension to stop<br />
smoking, which is most important in the prevention <strong>of</strong> lung<br />
cancer and all smoking-related diseases. The goal is to bring<br />
this epidemic <strong>of</strong> smoking-related disorders to an end.<br />
Stock<br />
Ownership Honoraria<br />
REFERENCES<br />
Research<br />
Funding<br />
BERG, ABERLE, AND WOOD<br />
Expert<br />
Testimony<br />
Other<br />
Remuneration<br />
Lung Screening Trial: Overview and study design. Radiology. 2011;258:243-<br />
253.<br />
11. Church TR, National Lung Screening Trial Executive Committee.<br />
Chest radiography as the comparison for spiral CT in the National Lung<br />
Screening Trial. Acad Radiol. 2003;10:713-715.<br />
12. The National Lung Screening Trial Research Team. Baseline characteristics<br />
<strong>of</strong> participants in the randomized National Lung Screening Trial.<br />
J Natl Cancer Inst. 2010;102:1771-1779.<br />
13. Gierada DS, Pilgrim TK, Ford M, et al. Lung cancer: Interobserver<br />
agreement on interprestion <strong>of</strong> pulmonary findings at low-dose CR screening.<br />
Radiology. 2008;246:265-272.<br />
14. Singh S, Pinsky P, Fineberg NS, et al. Evaluation <strong>of</strong> reader variability<br />
in the interpretations <strong>of</strong> follow-up CT Scans at lung cancer screening.<br />
Radiology. 2011;259:263-270.<br />
15. Cagnon CH, Cody DD, McNitt-Gray MF, et al. Description and implementation<br />
<strong>of</strong> a quality control program in an imaging-based clinical trial.<br />
Acad Radiol. 2006;13:1431-1441.<br />
16. Gierada DS, Garg K, Nath H, et al. CT quality assurance in the<br />
lung screening study component <strong>of</strong> the National Lung Screening Trial:<br />
Implications for multicenter imaging trials. AJR Am J Roentgenol. 2009;193:<br />
419-424.<br />
17. Larke FJ, Kruger RL, Cagnon CH, et al. Estimated radiation dose<br />
associated with low-dose chest CT <strong>of</strong> average-size participants in the National<br />
Lung Screening Trial. AJR Am J Roentgenol. 2011;197:1165-1169.<br />
18. Berrington de González A, Kim KP, Berg CD. Low-dose lung computed<br />
tomography screening before age 55: estimates <strong>of</strong> the mortality reduction<br />
required to outweigh the radiation-induced cancer risk. J Med Screen.<br />
2008;15:153-8.
LUNG CANCER SCREENING<br />
19. Montes RP. The cost <strong>of</strong> CT screening: year gained via lung cancer<br />
screening could cost $38,000. The Cancer Letter. 2011;37:15-16.<br />
20. CT Technique Chart. National Lung Screening Trial-<strong>American</strong> College<br />
<strong>of</strong> Radiology Imaging Network Protocol 6654. http://www.acrin.org/Default.<br />
aspx?tabid�282. Accessed February 19, 2011.<br />
21. Daqing M, Linning E. Volumetric measurement pulmonary groundglass<br />
opacity nodules with multi-detector CT: Effect <strong>of</strong> various tube current<br />
on measurement accuracy—a chest CT phantom study. AJR. 2009;16:934-<br />
939.<br />
22. Park CM, Goo JM, Lee HJ, et al. Persistent pure ground-glass nodules<br />
in the lung: Interscan variability <strong>of</strong> semiautomated volume and attenuation<br />
measurements. AJR. 2010;195:W408-414.<br />
23. Van Klaveren RJ, Oudkerk M, Prokop M, et al. Management <strong>of</strong><br />
lung nodules detected by volume CT scanning. N Engl J Med. 2009;361:2221-<br />
2229.<br />
24. MacMahon H, Austin JHM, Gamsu G, et al. Guidelines for management<br />
<strong>of</strong> small pulmonary nodules detected on CT scans: a statement from the<br />
Fleischner <strong>Society</strong>. Radiol. 2005;237:395-400.<br />
25. Raupach T, Merker J, Hasenfuss G, et al. Knowledge gaps about<br />
smoking cessation in hospitalized patients and their doctors. Eur J Cardiovasc<br />
Prev Rehabil. 2011;18:334-341.<br />
26. Tong EK, Strouse R, Hall J, et al.. National survey <strong>of</strong> US health<br />
pr<strong>of</strong>essionals’ smoking prevalence, cessation practices, and beliefs. Nicotine<br />
Tobac Res. 2010;12:724-733.<br />
27. Stuber J, Galea S, Link BG. Smoking and the emergence <strong>of</strong> a stigmatized<br />
social status. Soc Science Med. 2008;67:420-430.<br />
28. Niederdeppe J, Levy AG. Fatalistic beliefs about cancer prevention and<br />
three prevention behaviors. Cancer Epidemiol Biomarkers Prev. 2007;16:998-<br />
1003.<br />
29. Wood DE, Eapen GA, Ettinger DS, et al. Lung cancer screening. J Natl<br />
Compr Canc Netw. <strong>2012</strong>;10:24-65.<br />
30. International Early Lung Cancer Action Program: Enrollment and<br />
Screening Protocol. http://www.ielcap.org/pr<strong>of</strong>essionals/docs/ielcap.pdf. Accessed<br />
March 27, <strong>2012</strong>.<br />
31. Godoy MC, Naidich DP. Subsolid pulmonary nodules and the spectrum<br />
<strong>of</strong> peripheral adenocarcinomas <strong>of</strong> the lung: Recommended interim guidelines<br />
for assessment and management. Radiology. 2009;253:606-622.<br />
32. Eisenberg RL, Bankier AA, Boiselle PM. Compliance with Fleischner<br />
<strong>Society</strong> guidelines for management <strong>of</strong> small lung nodules: A survey <strong>of</strong> 834<br />
radiologists. Radiology. 2010;255:218-224.<br />
33. Esmaili A, Munden RF, Mohammed TLH. Small pulmonary nodule<br />
management: A survey <strong>of</strong> the members <strong>of</strong> the <strong>Society</strong> <strong>of</strong> Thoracic Radiology<br />
with comparison to the Fleischner <strong>Society</strong> guidelines. J Thorac Imaging.<br />
2011;26:27-31.<br />
34. Ettinger DS, Akerley W, Bepler G, et al. Non-small cell lung cancer.<br />
J Natl Compr Canc Netw. 2010;8:740-801.<br />
35. Wood DE, Farjah F. Surgeon Specialty is Associated with Better<br />
Outcomes: The Facts Speak for Themselves. Ann Thorac Surg. 2009;88:1393-<br />
1395.<br />
36. Farjah F, Flum DR, Varghese TK, et al. Surgeon specialty and longterm<br />
survival following resection for lung cancer. Ann Thorac Surg. 2009;87:<br />
995-1006.<br />
37. Travis WD, Brambilla E, Noguchi M, et al. International Association<br />
for the Study <strong>of</strong> Lung Cancer/<strong>American</strong> Thoracic <strong>Society</strong>/European Respiratory<br />
<strong>Society</strong> international multidisciplinary classification <strong>of</strong> lung adenocarcinoma.<br />
J Thorac Oncol. 2011;6:244-285.<br />
38. Iwano S, Nakamura T, Kamioka Y, et al. Computer-aided differentiation<br />
<strong>of</strong> malignant from benign solitary pulmonary nodules imaged by highresolution<br />
CT. Comput Med Imag Graph. 2008;32:416-422.<br />
39. Way TW, Sahiner B, Chan H-P, et al. Computer-aided diagnosis <strong>of</strong><br />
pulmonary nodules on CT scans: Improvement <strong>of</strong> classification performance<br />
with nodule surface features. Med Phys. 2009;36:3086-3098.<br />
40. Lokke A, Lang P, Scharling H, et al. Developing COPD: A 25-year<br />
follow up study <strong>of</strong> the general population. Thorax. 2006;61:935-939.<br />
41. Young RP, Hopkins RJ, Christmas T, et al. COPD prevalence is<br />
increased in lung cancer, independent <strong>of</strong> age, sex, and smoking history. Eur<br />
Respir J. 2009;34:380-386.<br />
42. Adcock IM, Caramori G, Barnes PJ. Choric obstructive pulmonary disease<br />
and lung cancer: New molecular insights. Respiration. 2011;81:265-284.<br />
43. Young RP, Hopkins RJ, Whittington CF, et al. Individual and cumulative<br />
effects <strong>of</strong> GWAS susceptibility loci in lung cancer: Associations after<br />
sub-phenotyping for COPD. PLos ONE. 2011;6:e16476.<br />
44. Hasegawa M, Nasuhara Y, Onodera Y, et al. Airflow limitations and<br />
airway dimensions in chronic obstructive pulmonary disease. Am J Respir<br />
Crit Care Med. 2006;173:1309-1315.<br />
45. Punturieri A, Szabo E, Croxton T, et al. Lung cancer and chronic<br />
obstructive pulmonary disease: Needs and opportunities for integrated research.<br />
J Natl Cancer Inst. 2009;101:554-559.<br />
46. Taylor DD, Gercel-Taylor C. MicroRNA signatures <strong>of</strong> tumor-derived<br />
exosomes as diagnostic biomarkers <strong>of</strong> ovarian cancer. Gynecol Oncol. 2008;<br />
110:13-21.<br />
47. Mitchell PS, Parkin RK, Kroh EM, et al. Circulating microRNAs as<br />
stable blood-based markers for cancer detection. Proc Natl Acad Sci USA.<br />
2008;105:10513-10518.<br />
48. Palmisano WA, Divine KK, Saccomanno G, et al. Predicting lung cancer<br />
by detecting aberrant promoter methylation in sputum. Cancer Res. 2000;60:<br />
5954-5958.<br />
49. NLST-ACRIN Biorepository. http://www.acrin.org/RESEARCHERS/<br />
POLICIES/NLSTACRINBIOREPOSITORY.aspx. Accessed March 28, <strong>2012</strong>.<br />
457
PERSONALIZED MEDICINE IN LUNG<br />
CANCER IN <strong>2012</strong><br />
CHAIR<br />
Pasi A. Janne, MD, PhD<br />
Dana-Farber Cancer Institute<br />
Boston, MA<br />
SPEAKERS<br />
Ignacio I. Wistuba, MD<br />
University <strong>of</strong> Texas M. D. Anderson Cancer Center<br />
Houston, TX<br />
Tetsuya Mitsudomi, MD, PhD<br />
Aichi Cancer Center Hospital<br />
Nagoya, Japan
Molecular Testing <strong>of</strong> Non–Small Cell Lung<br />
Carcinoma Biopsy and Cytology Specimens<br />
Overview: During the past decade, substantial progress has<br />
been made in the characterization <strong>of</strong> molecular abnormalities<br />
in non–small cell carcinoma (NSCLC) tumors that are being<br />
used as molecular targets and predictive biomarkers for<br />
selection <strong>of</strong> targeted therapy. These recent advances in<br />
NSCLC targeted therapy require the analysis <strong>of</strong> a panel <strong>of</strong><br />
molecular abnormalities in tumor specimens, including gene<br />
mutations (e.g., EGFR, KRAS, BRAF, DDR2), gene amplifications<br />
(e.g., MET, FGFR1), and fusions (e.g., EML4-ALK) by<br />
applying different methods to tumor tissue (biopsy) and cell<br />
(cytology) samples. However, the biopsy and cytology samples<br />
LUNG CANCER continues to be the most common and<br />
deadly malignant tumor worldwide. 1 The main challenge<br />
to improve the poor survival rate (5-year survival <strong>of</strong><br />
approximately 15%) <strong>of</strong> this disease is to develop novel<br />
strategies to better stratify high-risk populations for early<br />
diagnosis and to select the adequate treatment for different<br />
lung cancer subsets. 2<br />
NSCLC represents more than 80% <strong>of</strong> lung cancers. 3 Adenocarcinoma<br />
(40%) and squamous cell carcinoma (30%) are<br />
the most frequent histologic subtypes, but there are also<br />
less frequent types, including large cell, adenosquamous,<br />
and sarcomatoid carcinomas. Although most NSCLCs are<br />
associated with smoking, some <strong>of</strong> them (approximately<br />
15%), mostly adenocarcinoma, also occur in never-smoker<br />
patients.<br />
Lung tumors are the result <strong>of</strong> a multistep process in<br />
which normal lung cells accumulate multiple genetic and<br />
epigenetic abnormalities and evolve into cells with malignant<br />
biological capabilities. 4 Recent advances in understanding<br />
the complex biology <strong>of</strong> NSCLC, particularly the<br />
activation <strong>of</strong> oncogenes by mutation, translocation, and<br />
amplification, have provided new treatment targets<br />
and allowed the identification <strong>of</strong> subsets <strong>of</strong> tumors with<br />
unique molecular pr<strong>of</strong>iles that can predict response to<br />
therapy in this disease. 5 The identification <strong>of</strong> specific genetic<br />
and molecular abnormalities in tumor tissue specimens and<br />
the administration <strong>of</strong> specific inhibitors to those targets are<br />
the basis <strong>of</strong> personalized cancer treatment. 5<br />
The successful development <strong>of</strong> personalized therapy depends<br />
on the identification <strong>of</strong> a specific molecular target that<br />
drives cancer growth, subsequent validation <strong>of</strong> a clinically<br />
applicable biomarker, and development <strong>of</strong> a clinically sound<br />
and rational endpoint, coupled with understanding <strong>of</strong> the<br />
molecular mechanisms associated with the tumor’s resistance.<br />
In this process, the role <strong>of</strong> the pathologist in the<br />
analysis <strong>of</strong> molecular changes in lung cancer tumor tissue<br />
specimens is becoming increasingly important. These<br />
changes in the paradigms <strong>of</strong> lung cancer diagnosis and<br />
treatment have posed multiple new challenges for pathologists<br />
to adequately integrate both routine histopathologic<br />
assessment and molecular testing into the clinical pathology<br />
for tumor diagnosis and subsequent selection <strong>of</strong> the most<br />
appropriate therapy.<br />
By Ignacio I. Wistuba, MD<br />
available for molecular testing in advanced metastatic NSCLC<br />
tumors are likely to be small specimens, including core needle<br />
biopsies and/or fine needle aspiration, which may limit the<br />
molecular and genomic analysis with currently available methods<br />
and technologies. In this process, the role <strong>of</strong> the pathologist<br />
is becoming increasingly important to adequately<br />
integrate both routine histopathologic assessment and molecular<br />
testing into the clinical pathology for proper tumor<br />
diagnosis and subsequent selection <strong>of</strong> the most appropriate<br />
therapy.<br />
Molecular Abnormalities <strong>of</strong> NSCLC<br />
During the past decade, substantial progress has been<br />
made in the characterization <strong>of</strong> molecular abnormalities in<br />
NSCLC tumors that are being used as molecular targets and<br />
predictive biomarkers for selection <strong>of</strong> targeted therapy (Table<br />
1). In lung adenocarcinoma, at least two different major<br />
pathways have been identified in its pathogenesis: a<br />
smoking-associated activation <strong>of</strong> KRAS signaling and a<br />
non–smoking-associated activation <strong>of</strong> EGFR signaling. 6<br />
Lung adenocarcinomas arising in never or light smokers are<br />
characterized by markedly higher frequencies <strong>of</strong> a series <strong>of</strong><br />
targetable oncogenes abnormalities, 5 including EGFR and<br />
HER2 tyrosine kinase (TK) domain–activating mutations 6<br />
and EML4-ALK (2;5)(p23q35) translocation. 7 Recently, an<br />
additional potentially targetable gene translocation, KIF5B-<br />
RET (10p;11q)(p11.22; q11-21), has been identified in lung<br />
adenocarcinoma from never and ever smokers. 8-11<br />
Squamous cell carcinoma <strong>of</strong> the lung has been less histologically<br />
and molecularly studied than adenocarcinoma.<br />
Squamous cell carcinoma also harbors genetic abnormalities,<br />
resulting in activation <strong>of</strong> oncogenes, including EGFRvIII<br />
(deletion <strong>of</strong> exons 2–7) and DDR2 mutations and FGFR1<br />
(8p12) gene amplification (Table 1). 12,13 Other potentially<br />
targetable genetic abnormalities have been detected in both<br />
major NSCLC histologic subtypes, including, among others,<br />
PIK3CA mutation and amplification, MET amplification<br />
(7q21-q31), and AKT1 and MAP2K1 mutations. 5,14<br />
The most frequent clinically relevant driver gene abnormalities<br />
that define new molecular subsets <strong>of</strong> NSCLC are<br />
reviewed below.<br />
EGFR Mutation<br />
Epidermal growth factor receptor (EGFR) molecular abnormalities<br />
are common events in NSCLC and include<br />
gene-activating mutations, gene amplification, and overex-<br />
From the Departments <strong>of</strong> Pathology and Thoracic/Head and Neck Medical <strong>Oncology</strong>,<br />
University <strong>of</strong> Texas M. D. Anderson Cancer Center, Houston, TX.<br />
Author’s disclosures <strong>of</strong> potential conflicts <strong>of</strong> interest are found at the end <strong>of</strong> this article.<br />
Address reprint requests to Ignacio I. Wistuba, MD, Department <strong>of</strong> Pathology and<br />
Thoracic/Head and Neck Medical <strong>Oncology</strong>, Unit 85, University <strong>of</strong> Texas M. D. Anderson<br />
Cancer Center, 1515 Holcombe Blvd., Houston, TX 77030; email: iiwistuba@mdanderson.org.<br />
© <strong>2012</strong> by <strong>American</strong> <strong>Society</strong> <strong>of</strong> <strong>Clinical</strong> <strong>Oncology</strong>.<br />
1092-9118/10/1-10<br />
459
Table 1. Summary <strong>of</strong> Molecular Abnormalities Associated<br />
with the Lung Adenocarcinoma and Squamous Cell<br />
Carcinoma Histologies<br />
Gene Molecular Change Adenocarcinoma<br />
pression <strong>of</strong> the protein and its ligands. 4 Mutations <strong>of</strong> EGFR<br />
occur in approximately 24% <strong>of</strong> adenocarcinomas and up to<br />
60% in tumors from never smokers. The mutations are<br />
limited to the first four exons <strong>of</strong> the TK domain (exons 18 to<br />
21). 15,16 The most frequent mutations are in-frame deletions<br />
in exon 19 (44% <strong>of</strong> all mutations) and missense mutations in<br />
exon 21 (41% <strong>of</strong> all mutations). 17 In addition, in-frame<br />
duplications or insertions occurring in exon 20 have been<br />
described in approximately 5% <strong>of</strong> the mutant cases, and rare<br />
missense mutations occur in multiple sites. 17 EGFR mutations<br />
occur predominantly in adenocarcinoma (approximately<br />
20% to 48% vs. approximately 2% for other NSCLC<br />
histologic subtypes) and are more frequent in never smokers<br />
(54% vs. 16% in ever smokers) and female patients (49% vs.<br />
19% in male patients). 17 EGFR mutation is the most important<br />
criterion used to select patients for EGFR TK inhibitor<br />
therapy in lung cancer. 18<br />
HER2 Mutation<br />
In lung cancer, HER2/Neu abnormalities include gene<br />
mutation, gene amplification, and overexpression <strong>of</strong> the<br />
protein. 19 HER2 gene mutations have been detected in 2%<br />
KEY POINTS<br />
Squamous cell<br />
Carcinoma<br />
EGFR Mutation 10–40% Very rare<br />
Amplification/CNG 15% 30%<br />
IHC overexpression 15–40% 60%<br />
HER2 Mutation 2% Very rare<br />
Amplification 4% 2%<br />
EML4-ALK Translocation 7% Very rare<br />
KIF5B-RET Translocation 2% Not reported<br />
KRAS Mutation 10–30% Very rare<br />
BRAF Mutation 1–3% Very rare<br />
FGFR1 Amplification Not reported 20%<br />
DDR2 Mutation Not reported 4%<br />
PIK3CA Amplification/CNG 2–6% 30%<br />
Mutation 2% 2%<br />
Abbreviations: CNG, copy number gain; IHC, immunohistochemistry.<br />
● Non-small cell lung carcinoma (NSCLC) can be molecularly<br />
classified in multiple subtypes for selection<br />
<strong>of</strong> targeted therapy.<br />
● Currently, a panel <strong>of</strong> gene abnormalities (mutations,<br />
amplifications, and translocations) can be tested in<br />
NSCLC tumor biopsy and cytology specimens.<br />
● The role <strong>of</strong> the pathologist is crucial to integrate<br />
histology diagnosis and molecular testing <strong>of</strong> small<br />
tumor samples.<br />
● Quality control <strong>of</strong> tumor tissue and cell specimens for<br />
adequacy is extremely important for successful molecular<br />
testing.<br />
● Novel methodologies, including multiplex mutation<br />
detection platforms and next-generation sequencing,<br />
are useful to test multiple genetic aberrations in<br />
small tumor specimens.<br />
460<br />
<strong>of</strong> adenocarcinomas and occur mostly in exon 20. 19 HER2<br />
mutations have been described predominantly in patients<br />
with lung cancer with East-Asian ethnic background and a<br />
history <strong>of</strong> never smoking. HER2 amplification has been<br />
reported in 2 to 4% <strong>of</strong> NSCLCs and is more frequent in<br />
adenocarcinoma (4%). 20<br />
ALK Fusion Genes<br />
In lung cancer, aberrant ALK expression has been identified<br />
in a subset <strong>of</strong> adenocarcinomas, and this abnormality<br />
consists <strong>of</strong> the formation <strong>of</strong> a fusion transcript with celltransforming<br />
activity, which is the product <strong>of</strong> a inverted<br />
translocation <strong>of</strong> EML4 gene located at chromosome 2p21 and<br />
the ALK gene located at 2p23. 7 EML4-ALK translocations<br />
have multiple distinct is<strong>of</strong>orms (up to 9) with demonstrated<br />
transforming activity. EML4-ALK translocation has been<br />
detected in 7% <strong>of</strong> lung adenocarcinomas, particularly in<br />
patients with a history <strong>of</strong> never or light smoking, and is<br />
associated with early onset <strong>of</strong> tumor. 21 Histologically,<br />
EML4-ALK–rearranged adenocarcinomas have been described<br />
to have a predominantly solid pattern with signet<br />
ring cells, but combined acinar and cribriform patterns<br />
also have been described in these tumors. 21 The standard<br />
method to assess EML4-ALK fusion in lung cancer tumors is<br />
fluorescence in situ hybridization (FISH) using a “breakapart”<br />
probe, and samples are considered to have positive<br />
FISH results for EML4-ALK fusion if more than 15% <strong>of</strong><br />
scored tumor cells have split ALK 5� and 3� probe signals or<br />
have isolated 3� signals. 22 There are some reports that<br />
suggest that ALK protein expression assessment by immunohistochemistry<br />
(IHC) correlates with the presence <strong>of</strong><br />
EML4-ALK fusion, and there are ongoing studies testing<br />
ALK protein expression as a screening method for ALK<br />
fusions. 23<br />
BRAF and KRAS Mutations<br />
BRAF oncogene can be activated in NSCLC, particularly<br />
adenocarcinoma (1% to 3%), by gene point mutations. 5<br />
Contrary to melanoma, most BRAF mutations detected in<br />
lung cancer are non-Val600Glu mutations that affect exons<br />
11 and 15, and they are mutually exclusive to EGFR and<br />
KRAS mutations.<br />
KRAS mutations are more common in lung adenocarcinoma<br />
than other NSCLC histologic types and are more<br />
frequently found in tumors from patients with a smoking<br />
history (approximately 30%). 6 In lung cancer, KRAS mutations<br />
are found in codons 12, 13, and 61 (42% <strong>of</strong> all mutations),<br />
which are mainly GGT to TGT transversions that<br />
produce glycine to cysteine amino acid changes. 6 KRAS<br />
mutations are rarely detected in EGFR-mutant tumors.<br />
RAS is considered <strong>of</strong> untargetable molecule; therefore, recent<br />
studies have evaluated the RAS downstream pathway,<br />
RAS/RAF/MEK, as a potential target for therapy in lung<br />
cancer.<br />
FGFR1 Amplification<br />
IGNACIO I. WISTUBA<br />
FGFR1 is a transmembrane TK and member <strong>of</strong> the<br />
fibroblast growth factor receptor (FGFR) TK family that<br />
comprises four kinases (FGFR-1 to -4). 13 In lung cancer,<br />
amplification <strong>of</strong> FGFR1 (chromosome 8p11-12) is a driver<br />
event in NSCLC and is predominantly detected in squamous<br />
cell carcinomas (approximately 20%) compared with adenocarcinomas<br />
(1% to 3%). 13 Currently, the preferred method to
MOLECULAR TESTING AND NSCLC<br />
Fig 1. Left, Photomicrographs <strong>of</strong> representative examples<br />
<strong>of</strong> core needle biopsy (CNB) and fine needle<br />
aspiration (FNA) specimens frequently available for histologic<br />
diagnosis <strong>of</strong> advanced metastatic non-small cell<br />
lung cancer (NSCLC). The insert in the CNB photomicrograph<br />
corresponds to a tissue paraffin block. Right, In<br />
NSCLC, diagnosis <strong>of</strong> the histologic subtype (adenocarcinoma<br />
or squamous cell carcinoma) is the first step. In<br />
tumors with poorly differentiated histologic subtypes,<br />
the diagnosis <strong>of</strong> NSCLC not otherwise specified is frequently<br />
performed; however, a more specific histologic<br />
diagnosis should be reached by using a limited immunohistochemistry<br />
panel: thyroid transcription factor1 is<br />
marker <strong>of</strong> adenocarcinoma, and p63 is a marker <strong>of</strong><br />
squamous cell carcinoma. After assessment <strong>of</strong> tissue<br />
quality for molecular testing, the sample should be submitted<br />
for a panel <strong>of</strong> tests, including gene fusions and<br />
amplification analyses by fluorescent in situ hybridization<br />
and DNA extraction for gene mutation analysis. It<br />
has been recommended that all lung adenocarcinomas<br />
be tested for EML4-ALK fusion and EGFR mutation, while<br />
squamous cell carcinomas should be tested for other<br />
gene abnormalities (DDR2 mutation and FGFR1 amplification).<br />
However, the utilization <strong>of</strong> multiplex platforms<br />
to test mutations in tumor samples allows testing <strong>of</strong> all<br />
NSCLC histologies for a panel <strong>of</strong> mutations and other<br />
gene abnormalities regardless <strong>of</strong> their histology.<br />
asses FGFR1 copy number is FISH, but the definitions <strong>of</strong><br />
copy number gain and gene amplification still must be<br />
determined.<br />
DDR2 Mutations<br />
Mutations <strong>of</strong> this TK have been described in 4% <strong>of</strong> lung<br />
squamous cell carcinomas. 12 Mutations were found in both<br />
the kinase domain and other regions <strong>of</strong> the protein sequence<br />
without hotspots, which makes the analysis <strong>of</strong> mutations<br />
<strong>of</strong> this gene challenging. Tumors established from a DDR2mutant<br />
cell line were sensitive to dasatinib, and a patient<br />
with squamous cell carcinoma that responded to dasatinib<br />
and erlotinib treatment harbored a DDR2 kinase domain<br />
mutation. 12<br />
PIK3CA Mutations and Amplification<br />
In NSCLC, copy number gain (� 3 copies per cell) <strong>of</strong><br />
PIK3CA is a common abnormality, predominantly in squamous<br />
cell carcinomas (33% to 35%) compared with adenocarcinomas<br />
(2% to 6%). 14 Mutations in the helical or kinase<br />
domain <strong>of</strong> PIK3CA have been reported in very low frequencies<br />
(2%) in NSCLC. 14 PI3K and its downstream effectors<br />
PTEN, mTOR, and AKT are potential therapeutic targets<br />
for NSCLC therapy and are being evaluated in clinical trials<br />
for lung cancer.<br />
Molecular Testing <strong>of</strong> NSCLC Tissue and<br />
Cell Specimens<br />
The recent advances in NSCLC targeted therapy require<br />
the analysis <strong>of</strong> a panel <strong>of</strong> molecular abnormalities in tumor<br />
specimens (Fig. 1), including gene mutations, amplifications,<br />
and fusions, by applying different methods to tumor tissue<br />
specimens. 24 However, the tissue (biopsy) and cell (cytology)<br />
samples available for molecular testing in advanced metastatic<br />
tumors are likely to be small specimens, including core<br />
needle biopsies (CNB) and/or fine needle aspiration (FNA)<br />
(Fig. 1), which may limit molecular and genomic analysis<br />
with currently available methods and technologies. There is<br />
a need to adapt and incorporate the current and new<br />
NSCLC Small Biopsy<br />
and Cytology<br />
CNB<br />
FNA<br />
Biopsy or<br />
Cytology<br />
NSCLC – NOS<br />
Adenocarcinoma<br />
IHC:<br />
TTF-1<br />
p63<br />
Squamous Cell Ca<br />
FISH:<br />
EML4-ALK Fusion<br />
KIF5B-RET Fusion<br />
FGFR1 Amp<br />
MET Amp<br />
Mutation:<br />
EGFR<br />
DDR2<br />
PI3KCA<br />
BRAF<br />
HER2<br />
K-, N-, H-RAS<br />
MEK<br />
AKT1, etc<br />
emerging technologies to the molecular analysis <strong>of</strong> small<br />
tissue specimens, such as CNB and FNA, obtained from<br />
NSCLC patients.<br />
There are several scientific and methodologic challenges,<br />
as well as practical barriers, to the use <strong>of</strong> widespread<br />
molecular testing using lung biopsy and cytology specimens.<br />
The ideal specimens for molecular testing would be freshly<br />
obtained tumor tissues followed by immediate snap freezing.<br />
However, these samples are usually available only for research<br />
purposes in academic centers and are typically used<br />
for discovery purposes. 2<br />
In pathology laboratories, the diagnostic clinical tumor<br />
tissue specimens (e.g., CNB, bronchoscopy samples, surgical<br />
resections) are fixed in formalin and embedded in paraffin<br />
for the histologic process. Both formalin fixation and paraffin<br />
embedding compromise the integrity <strong>of</strong> protein and<br />
nuclei acids (RNA, DNA) for molecular testing, particularly<br />
when nonbuffered formalin is used and the specimens are<br />
fixed in formalin for greater than 24 hours. The cytology<br />
specimens (e.g., bronchial brushes, bronchoalveolar lavages,<br />
pleural fluids, and FNAs) are usually fixed in alcohol, which<br />
is optimal for preservation <strong>of</strong> nucleic acids. When the cytology<br />
specimen has abundant material, the sample can be<br />
fixed in formalin and processed as a tissue specimen (cell<br />
block) to obtain histologic sections. Although tissue specimens<br />
are preferable for molecular testing, cytology samples<br />
with abundant malignant cells can be successfully used for<br />
molecular testing.<br />
Among others, there are two important aspects that must<br />
be addressed when small lung cancer biopsy or cytology<br />
specimens are received for diagnosis in a pathology laboratory.<br />
First, the handling <strong>of</strong> the biopsy and cytology specimen<br />
for histologic analysis and subsequent molecular testing<br />
requires thoughtful prioritization <strong>of</strong> the utilization <strong>of</strong> the<br />
sample to prevent the use <strong>of</strong> tissue in less important analysis<br />
than the definitive molecular testing required for selection<br />
<strong>of</strong> therapy. Second, the pathologist should determine<br />
whether the amount <strong>of</strong> malignant cells available in the<br />
specimen is adequate for nuclei acid extractions and also for<br />
461
histologic section–based molecular tests (e.g., FISH and<br />
IHC).<br />
On the other hand, our growing understanding <strong>of</strong> the<br />
cancer biology <strong>of</strong> NSCLC, particularly the molecular evolution<br />
<strong>of</strong> tumors during local progression and metastasis and<br />
the identification <strong>of</strong> molecular abnormalities contributing to<br />
resistance to TK inhibitor therapies, emphasizes the importance<br />
<strong>of</strong> characterizing the molecular abnormalities <strong>of</strong> the<br />
disease at every stage <strong>of</strong> its evolution. For molecular testing<br />
<strong>of</strong> advanced metastatic NSCLC, it is important to sample<br />
and analyze the tumors’ sample at each time point <strong>of</strong> clinical<br />
decision making.<br />
Relevant Molecular Pathologic Analysis Methods<br />
Currently, most <strong>of</strong> the predictive molecular markers available<br />
for therapy selection in NSCLC are oncogene mutations<br />
and amplifications. However, other molecular and genetic<br />
changes modulate the sensitivity <strong>of</strong> tumor to targeted therapies,<br />
including protein overexpression, gene methylation,<br />
and gene expression abnormalities. The need for analysis <strong>of</strong><br />
multiple molecular and genetic changes in small, clinically<br />
relevant biopsy and cytology specimens has driven the<br />
scientific community and the molecular pathology laboratories<br />
to develop multiplex approaches for molecular testing<br />
<strong>of</strong> small tumors samples, particularly for gene mutation<br />
analysis.<br />
Mutation Analysis<br />
Currently, direct nucleic acid sequencing with previous<br />
polymerase chain reaction (PCR) amplification <strong>of</strong> extracted<br />
DNA is the most commonly used technique for gene mutation<br />
analysis <strong>of</strong> tumor biopsy and cytology samples to detect<br />
mutations <strong>of</strong> clinically relevant genes. Several sequencing<br />
methods are available for mutation analysis <strong>of</strong> DNA extracted<br />
from tumor tissue and cell specimens, especially<br />
for formalin-fixed paraffin-embedded (FFPE) samples. The<br />
current PCR-based sequencing mutation analysis methods<br />
can be divided into uniplex (e.g., Sanger sequencing and<br />
pyrosequencing) and multiplex (e.g., matrix-assisted laser<br />
desorption ionization time-<strong>of</strong>-light mass spectrometry and<br />
primer extension assay) methods. In the uniplex method,<br />
one hotspot sequence is examined at a time, while the<br />
multiplex technique multiple hotspot mutations are examined<br />
simultaneously. The multiplex approaches are clearly<br />
the preferred methods since they allow more efficient mutation<br />
analysis <strong>of</strong> small amounts <strong>of</strong> DNA for multiple hotspots<br />
(approximately 100 to 200) from a panel <strong>of</strong> genes (approximately<br />
10 to 30).<br />
Sanger Sequencing<br />
In solid tumors, including lung, Sanger sequencing is the<br />
most commonly used sequencing method to detect hotspot<br />
mutations <strong>of</strong> oncogenes (e.g., EGFR, KRAS, BRAF). This<br />
type <strong>of</strong> sequencing can detect essentially all base substitutions,<br />
small insertions, and deletions. Its main disadvantage<br />
is the relatively low sensitivity (approximately 20%) for the<br />
detection <strong>of</strong> mutant alleles in the DNA sample extracted<br />
from tumor specimens. These DNA samples are usually a<br />
mixture <strong>of</strong> mutant and wild-type alleles as a result <strong>of</strong> the<br />
presence <strong>of</strong> malignant and nonmalignant (from adjacent<br />
normal or tumor stroma) cells in the tumor tissue specimens.<br />
25<br />
462<br />
Pyrosequencing<br />
Pyrosequencing uses sequencing by a synthesis method<br />
to sequence nucleic acids and relies on the detection <strong>of</strong><br />
pyrophosphate release on nucleotide incorporation. Pyrosequencing<br />
is considered to be more sensitive than Sanger<br />
sequencing and detects approximately 5% <strong>of</strong> mutant compared<br />
with wild-type alleles. 26<br />
Matrix-Assisted Laser Desorption Ionization Time-<strong>of</strong>-Light<br />
Mass Spectrometry<br />
Matrix-assisted laser desorption ionization time-<strong>of</strong>-light<br />
mass spectrometry (Sequenom) utilizes a high-throughput<br />
PCR-based sequencing assay to detect multiple hotspot<br />
mutations simultaneously using small amounts <strong>of</strong> DNA<br />
obtained from biopsy and cytology specimens. 27 It has a high<br />
level <strong>of</strong> sensitivity (approximately 5% <strong>of</strong> the mutant alleles)<br />
and allows quantification <strong>of</strong> the percentage <strong>of</strong> mutant<br />
DNA. 27 The Sequenom method is also useful to assess gene<br />
amplification.<br />
Primer Extension Assay<br />
Primer extension assay is a primer extension–based<br />
method that allows simultaneous analysis <strong>of</strong> up to 10 different<br />
mutations. 28 It is a sensitive, low-cost, and rapid method<br />
to screen for mutations and to analyze methylation. This<br />
assay uses the SNaPshot Multiplex Kit (Applied Biosystems<br />
Inc), which contains a reaction mix <strong>of</strong> four differentially<br />
fluorescently labeled ddNTPs, allowing the interrogation <strong>of</strong><br />
each base at a mutation site. 28<br />
Translocation and Gene Copy Number Analyses<br />
FISH is a cytogenetic technique that uses fluorescentlabeled<br />
probes to hybridize specific sequences <strong>of</strong> DNA on<br />
chromosomes. 29 It is applied to visualize chromosome deletions,<br />
amplification, and structural rearrangements. The<br />
main advantage <strong>of</strong> this technique is that it allows in situ localization<br />
<strong>of</strong> the specific sequences and the simultaneous<br />
detection <strong>of</strong> multiple sites by using hybridization probes<br />
labeled with different fluorophores. The main disadvantage<br />
<strong>of</strong> FISH is the need <strong>of</strong> additional equipment for analysis,<br />
such as dark-field microscopy and multiband fluorescent<br />
filters. Chromogenic in situ hybridization is a variant <strong>of</strong> the<br />
in situ hybridization technique that visualizes the specific<br />
DNA sequence by a peroxidase reaction and allows the<br />
visualization in a light microscope. The main disadvantage<br />
<strong>of</strong> this method is that it limits the use <strong>of</strong> different label<br />
probes to target multiple sites; however, recent advances in<br />
methodologies permit the use <strong>of</strong> dual colors to target two<br />
sequencing sites and enable the visualization <strong>of</strong> tissue<br />
architecture and cytomorphologic analysis. Although there<br />
are other methods available for gene copy number and<br />
fusion genes analyses, such as DNA quantitative PCR assay<br />
for copy number assessment and messenger RNA quantitative<br />
PCR assay for gene fusion analysis, FISH continues to<br />
be the preferred method for gene copy number (e.g., MET,<br />
FGFR1) and gene fusion (e.g., EML4-ALK) in lung cancer.<br />
Protein Expression Analysis<br />
IGNACIO I. WISTUBA<br />
IHC is a widely used technique in pathology laboratories<br />
to detect the presence and levels <strong>of</strong> expression <strong>of</strong> a specific<br />
protein in FFPE tumor cytology and cytology specimens. In<br />
lung cancer, the use <strong>of</strong> IHC is currently used for histopatho-
MOLECULAR TESTING AND NSCLC<br />
logic diagnosis and classification <strong>of</strong> tumors, particularly<br />
when small tissue specimens are examined. Currently, IHC<br />
markers are frequently used by pathologists to clinically<br />
subtype NSCLC; for example, cytokeratin 7 and thyroid<br />
transcription factor 1 are positive in most adenocarcinomas,<br />
whereas p63, p40, and cytokeratin 5/6 are positive in most<br />
squamous cell carcinomas. 24 Despite being readily available,<br />
there are no validated molecular markers based on protein<br />
expression by IHC being used to predict response to therapy<br />
in lung cancer. One <strong>of</strong> the most important advantaged <strong>of</strong><br />
IHC is that it allows the identification <strong>of</strong> the protein expression<br />
in specific types <strong>of</strong> cells as well as distinct subcellular<br />
localization.<br />
Conclusion<br />
Most <strong>of</strong> the current biomarkers discovered and now used<br />
in clinical applications to date consist <strong>of</strong> a single genetic<br />
mutation, gene amplification, or translocation, but these<br />
aberrations are rare and not sufficient to select the majority<br />
<strong>of</strong> patients for targeted therapies. It is also known that, in<br />
many cases, multiple changes in tumor cells, rather than a<br />
single modification, lead to activation <strong>of</strong> selective and <strong>of</strong>ten<br />
interactive molecular pathways promoting tumor growth<br />
and survival. In addition, various targeted treatment regimens<br />
have been shown to result in the activation <strong>of</strong> alternative,<br />
compensatory molecular pathways that continue to<br />
promote cancer cell survival.<br />
The development <strong>of</strong> new technologies, such as highthroughput<br />
arrays, has allowed researchers to screen the<br />
whole genome, proteome, and transcriptome for new biomarkers<br />
in tumor tissue, serum, plasma, or other human<br />
body fluids and develop genomic and proteomic pr<strong>of</strong>iles, or<br />
“signatures,” to better reflect the complex molecular aberrations<br />
within a single tumor.<br />
The rapid development <strong>of</strong> technologies for large-scale<br />
Author’s Disclosures <strong>of</strong> Potential Conflicts <strong>of</strong> Interest<br />
Author<br />
Ignacio I. Wistuba*<br />
*No relevant relationships to disclose.<br />
Employment or<br />
Leadership<br />
Positions<br />
Consultant or<br />
Advisory Role<br />
1. Jemal A, Siegel R, Xu J, et al. Cancer statistics, 2010. CA Cancer J Clin.<br />
2010;60:277-300.<br />
2. Wistuba II, Gelovani JG, Jacoby JJ, et al. Methodological and practical<br />
challenges for personalized cancer therapies. Nat Rev Clin Oncol. 2011;8:135-<br />
141.<br />
3. Travis WD, Brambilla E, Muller-Hermelink HK, et al. Tumours <strong>of</strong><br />
the lung. In: Travis WD, Brambilla E, Muller-Hermelink HK, et al<br />
(eds). Pathology and Genetics: Tumours <strong>of</strong> the Lung, Pleura, Thymus<br />
and Heart. World Health Organization Classification <strong>of</strong> Tumours. Pathology<br />
& Genetics. Lyon, France: International Agency for Research on Cancer;<br />
2004:9-124.<br />
4. Herbst RS, Heymach JV, Lippman SM. Lung cancer. N Engl J Med.<br />
2008;359:1367-1380.<br />
5. Pao W, Girard N. New driver mutations in non-small cell lung cancer.<br />
Lancet Oncol. 2011;12:175-180.<br />
6. Sun S, Schiller JH, Gazdar AF. Lung cancer in never smokers—a<br />
different disease. Nat Rev Cancer. 2007;7:778-790.<br />
7. Soda M, Choi YL, Enomoto M, et al. Identification <strong>of</strong> the transforming<br />
EML4-ALK fusion gene in non-small-cell lung cancer. Nature. 2007;448:561-<br />
566.<br />
sequencing (next-generation sequencing [NGS]) <strong>of</strong> DNA<br />
and RNA has facilitated high-throughput molecular analysis,<br />
holding various advantages over traditional sequencing.<br />
These new technologies provide capabilities to fully<br />
sequence large numbers <strong>of</strong> genes in a single test and<br />
simultaneously detect deletions, insertions, copy number<br />
alterations, translocations, and exome-wide base substitutions<br />
(including known hotspot mutations) in all known<br />
cancer-related genes. The amount <strong>of</strong> starting material (DNA<br />
or RNA) needed for the newest NGS applications is getting<br />
smaller, and currently, the analysis <strong>of</strong> a panel (approximately<br />
200 to 300) <strong>of</strong> gene mutations and fusions can be<br />
performed even in DNA extracted from FFPE tumor tissue<br />
specimens. However, one <strong>of</strong> the potential difficulties in this<br />
process is the large computing capacities needed to manage<br />
the billions <strong>of</strong> small sequence readouts generated and to<br />
assemble those with large databases to interpret the raw<br />
data. Another challenge for the NGS is the identification <strong>of</strong><br />
meaningful driver mutations and the separation <strong>of</strong> “true”<br />
mutations among a background <strong>of</strong> intrinsic sequence variations.<br />
30 In addition, verifying and validating the discovered<br />
“driver” mutations will require experimental and detailed<br />
classic molecular pathology studies to bring NGS into clinical<br />
context.<br />
In summary, the recent advances in NSCLC targeted<br />
therapy require the analysis <strong>of</strong> a panel <strong>of</strong> molecular abnormalities<br />
<strong>of</strong> tumor specimens, including gene mutations,<br />
amplifications, and fusions, by applying different methods<br />
to the samples. In this new era <strong>of</strong> personalized therapy<br />
in lung cancer using targeted agents, the role <strong>of</strong> the<br />
pathologist in the analysis <strong>of</strong> molecular changes in lung<br />
cancer tumor tissue specimens is becoming increasingly<br />
important to properly integrate routine histopathologic diagnosis<br />
and molecular testing into the clinical practice.<br />
Stock<br />
Ownership Honoraria<br />
REFERENCES<br />
Research<br />
Funding<br />
Expert<br />
Testimony<br />
Other<br />
Remuneration<br />
8. Ju YS, Lee WC, Shin JY, et al. A transforming KIF5B and RET gene<br />
fusion in lung adenocarcinoma revealed from whole-genome and transcriptome<br />
sequencing. Genome Res. <strong>2012</strong>;22:436-445.<br />
9. Lipson D, Capelletti M, Yelensky R, et al. Identification <strong>of</strong> new ALK and<br />
RET gene fusions from colorectal and lung cancer biopsies. Nat Med. Epub<br />
<strong>2012</strong> Feb 12.<br />
10. Kohno T, Ichikawa H, Totoki Y, et al. KIF5B-RET fusions in lung<br />
adenocarcinoma. Nat Med. Epub <strong>2012</strong> Feb 12.<br />
11. Takeuchi K, Soda M, Togashi Y, et al. RET, ROS1 and ALK fusions in<br />
lung cancer. Nat Med. Epub <strong>2012</strong> Feb 12.<br />
12. Hammerman PS, Sos ML, Ramos AH, et al. Mutations in the DDR2<br />
kinase gene identify a novel therapeutic target in squamous cell lung cancer.<br />
Cancer Discov. 2011;1:78-89.<br />
13. Weiss J, Sos ML, Seidel D, et al. Frequent and focal FGFR1 amplification<br />
associates with therapeutically tractable FGFR1 dependency in squamous<br />
cell lung cancer. Sci Transl Med. 2010;2:62ra93.<br />
14. Yamamoto H, Shigematsu H, Nomura M, et al. PIK3CA mutations<br />
and copy number gains in human lung cancers. Cancer Res. 2008;68:6913-<br />
6921.<br />
15. Paez JG, Janne PA, Lee JC, et al. EGFR mutations in lung cancer:<br />
463
correlation with clinical response to gefitinib therapy. Science. 2004;304:1497-<br />
1500.<br />
16. Lynch TJ, Bell DW, Sordella R, et al. Activating mutations in the<br />
epidermal growth factor receptor underlying responsiveness <strong>of</strong> non-small-cell<br />
lung cancer to gefitinib. N Engl J Med. 2004;350:2129-2139.<br />
17. Shigematsu H, Gazdar AF. Somatic mutations <strong>of</strong> epidermal growth<br />
factor receptor signaling pathway in lung cancers. Int J Cancer. 2006;118:<br />
257-262.<br />
18. Mok TS, Wu YL, Thongprasert S, et al. Gefitinib or carboplatinpaclitaxel<br />
in pulmonary adenocarcinoma. N Engl J Med. 2009;361:947-957.<br />
19. Shigematsu H, Takahashi T, Nomura M, et al. Somatic mutations <strong>of</strong><br />
the HER2 kinase domain in lung adenocarcinomas. Cancer Res. 2005;65:<br />
1642-1646.<br />
20. Hirsch FR, Varella-Garcia M, Franklin WA, et al. Evaluation <strong>of</strong><br />
HER-2/neu gene amplification and protein expression in non-small cell lung<br />
carcinomas. Br J Cancer. 2002;86:1449-1456.<br />
21. Inamura K, Takeuchi K, Togashi Y, et al. EML4-ALK lung cancers are<br />
characterized by rare other mutations, a TTF-1 cell lineage, an acinar<br />
histology, and young onset. Mod Pathol. 2009;22:508-515.<br />
22. Kwak EL, Bang YJ, Camidge DR, et al. Anaplastic lymphoma kinase<br />
inhibition in non-small-cell lung cancer. N Engl J Med. 2010;363:1693-<br />
1703.<br />
23. Mino-Kenudson M, Chirieac LR, Law K, et al. A novel, highly sensitive<br />
464<br />
IGNACIO I. WISTUBA<br />
antibody allows for the routine detection <strong>of</strong> ALK-rearranged lung adenocarcinomas<br />
by standard immunohistochemistry. Clin Cancer Res. 2010;16:1561-<br />
1571.<br />
24. Kerr KM. Personalized medicine for lung cancer: new challenges for<br />
pathology. Histopathology. <strong>2012</strong>;60:531-546.<br />
25. Anderson SM. Laboratory methods for KRAS mutation analysis. Expert<br />
Rev Mol Diagn. 2011;11:635-642.<br />
26. Tsiatis AC, Norris-Kirby A, Rich RG, et al. Comparison <strong>of</strong> Sanger<br />
sequencing, pyrosequencing, and melting curve analysis for the detection <strong>of</strong><br />
KRAS mutations: diagnostic and clinical implications. J Mol Diagn. 2010;12:<br />
425-432.<br />
27. Fumagalli D, Gavin PG, Taniyama Y, et al. A rapid, sensitive, reproducible<br />
and cost-effective method for mutation pr<strong>of</strong>iling <strong>of</strong> colon cancer and<br />
metastatic lymph nodes. BMC Cancer. 2010;10:101.<br />
28. Su Z, Dias-Santagata D, Duke M, et al. A platform for rapid detection<br />
<strong>of</strong> multiple oncogenic mutations with relevance to targeted therapy in<br />
non-small-cell lung cancer. J Mol Diagn. 2011;13:74-84.<br />
29. Varella-Garcia M. Chromosomal and genomic changes in lung cancer.<br />
Cell Adh Migr. 2010;4:100-106.<br />
30. Cronin M, Ross JS. Comprehensive next-generation cancer genome<br />
sequencing in the era <strong>of</strong> targeted therapy and personalized oncology. Biomark<br />
Med. 2011;5:293-305.
THYMOMA AND THYMIC CARCINOMA:<br />
UPDATE ON MANAGEMENT<br />
CHAIR<br />
Gregory J. Riely, MD, PhD<br />
Memorial Sloan-Kettering Cancer Center<br />
New York, NY<br />
SPEAKERS<br />
Nicolas Girard, MD, MSc<br />
Hopital Louis Pradel<br />
Bron, France<br />
Frank C. Detterbeck, MD<br />
Yale School <strong>of</strong> Medicine<br />
New Haven, CT
Multidisciplinary Management <strong>of</strong><br />
Thymic Carcinoma<br />
By Gregory J. Riely, MD, PhD, James Huang, MD, and Andreas Rimner, MD, PhD<br />
Overview: Thymic carcinomas represent approximately 10%<br />
<strong>of</strong> thymic tumors. In our approach to patients with thymic<br />
carcinoma, we emphasize multimodality treatment with close<br />
communication between the pathologist, thoracic surgeon,<br />
medical oncologist, and radiation oncologist. Given the paucity<br />
<strong>of</strong> high-quality clinical research data, treatment decisions<br />
are guided by a small amount <strong>of</strong> prospective trial data,<br />
retrospective reports, and clinical experience. Surgical management<br />
<strong>of</strong> thymic carcinoma must account for the more<br />
aggressive biology, higher degree <strong>of</strong> local invasion <strong>of</strong> neighboring<br />
structures, greater propensity for nodal metastasis,<br />
and higher risk <strong>of</strong> distant metastatic disease. Although surgi-<br />
THYMIC CARCINOMAS represent approximately 10%<br />
<strong>of</strong> thymic tumors, with thymoma and thymic carcinoids<br />
encompassing the remaining 90%. In our approach to patients<br />
with thymic carcinoma, we emphasize multimodality<br />
treatment with close communication between the pathologist,<br />
thoracic surgeon, medical oncologist, and radiation<br />
oncologist. Given the paucity <strong>of</strong> data with which to evaluate<br />
approaches to therapy, treatment decisions are guided by<br />
a small amount <strong>of</strong> prospective trial data, retrospective<br />
reports, and institutional experience. For patients who present<br />
with localized or locally advanced disease for which<br />
surgical resection is feasible, we routinely recommend preoperative<br />
chemotherapy with regimens evaluated in the<br />
advanced disease setting followed by complete surgical resection<br />
with RT dependent on the findings at the time <strong>of</strong><br />
resection. For patients with unresectable localized disease,<br />
we use an approach combining chemotherapy and RT. Patients<br />
who have evidence <strong>of</strong> distant metastatic disease or<br />
recurrent disease after prior surgery or RT, palliative chemotherapy<br />
is the mainstay <strong>of</strong> treatment.<br />
Pathology<br />
Thymic carcinomas are malignant epithelial tumors with<br />
overt cytologic atypia. Thymic carcinoma cells are thought<br />
to be derived from thymic epithelial cells. Although they<br />
may have different histologic features (similar to neuroendocrine,<br />
mucoepidermoid, and lymphoepithelioma), the most<br />
common is squamous differentiation with large polyhedral<br />
cells, sometimes with keratinization. Thymic carcinomas<br />
are routinely found to be immunoreactive to antibodies to<br />
CD5 and CD117 (c-Kit). Given the squamous histology,<br />
histology review alone cannot distinguish squamous cell<br />
carcinoma <strong>of</strong> the lung from thymic carcinoma, so clinical<br />
correlation is required for this diagnosis. Although not<br />
clinically used, copy number data show different patterns<br />
and frequencies <strong>of</strong> chromosomal gains and losses for thymic<br />
carcinoma tumors as compared with squamous cell carcinomas<br />
<strong>of</strong> the lung, emphasizing that they are different disease<br />
entities. 1 Of note, it is rare for patients with thymic carcinoma<br />
to have associated paraneoplastic syndromes such as<br />
myasthenia gravis and pure red cell aplasia, which are<br />
relatively common in patients with thymoma.<br />
To better understand thymic carcinoma and its association<br />
with thymomas, a number <strong>of</strong> groups have evaluated<br />
466<br />
cal resection remains the most important component in the<br />
management <strong>of</strong> localized thymic tumors, radiation therapy<br />
(RT) may be used as adjuvant therapy after surgical resection<br />
or as the definitive treatment modality in patients who are<br />
deemed unresectable because <strong>of</strong> medical comorbidities or<br />
technical reasons. Systemic therapy for thymic carcinoma is<br />
used in two clinical scenarios: preoperative treatment and<br />
palliative therapy. First-line, platinum-based chemotherapy<br />
regimens are associated with response rates between 22%<br />
and 75%. Recent data from targeted therapy trials do not<br />
reveal a clear role for targeted therapies for patients with<br />
thymic carcinoma.<br />
specific molecular features <strong>of</strong> thymic carcinoma. Further<br />
characterization <strong>of</strong> thymic carcinomas has found that the<br />
epidermal growth factor receptor is expressed at high levels<br />
in virtually all thymic carcinomas (when detected by immunohistochemistry).<br />
1 The insulin-like growth factor receptor<br />
(IGF1-R) is also expressed at high levels in most, but not<br />
all, thymic carcinomas. 2 With targeted mutational analyses,<br />
KIT mutations have been identified as well as a limited<br />
number <strong>of</strong> KRAS mutations). 1 Unfortunately, there are no<br />
chromosomal gains or losses that have been uniformly<br />
identified in all thymic carcinomas. 3<br />
Surgical Management <strong>of</strong> Thymic Carcinoma<br />
Surgery is considered the mainstay <strong>of</strong> treatment for most<br />
thymic tumors. However, thymic carcinoma presents challenges<br />
for management given its more aggressive behavior.<br />
The literature regarding the surgical treatment <strong>of</strong> thymic<br />
carcinoma is sparse because <strong>of</strong> its rarity and is limited to<br />
small retrospective case series with heterogeneous treatments.<br />
Furthermore, many series have grouped thymic<br />
carcinoma with thymoma in their published experiences, not<br />
surprisingly since thymic carcinoma was labeled as type C<br />
thymoma by the World Health Organization up until 2004,<br />
when they recognized thymic carcinoma as a separate distinct<br />
entity. Similarly, some series <strong>of</strong> thymic carcinomas<br />
have also included thymic neuroendocrine tumors, such as<br />
carcinoids.<br />
Surgical considerations must account for the more aggressive<br />
biology exhibited by thymic carcinoma, which is manifested<br />
by a higher degree <strong>of</strong> local invasion <strong>of</strong> neighboring<br />
structures, greater propensity for nodal metastasis, and<br />
higher risk <strong>of</strong> distant metastatic disease. Taken together,<br />
these factors result in a lower resectability rate, higher<br />
From the Thoracic <strong>Oncology</strong> Service, Division <strong>of</strong> Solid Tumor <strong>Oncology</strong>, Department <strong>of</strong><br />
Medicine, Memorial Sloan-Kettering Cancer Center, Weill Cornell Medical College, New<br />
York, NY; Thoracic Surgery Service, Department <strong>of</strong> Surgery, Memorial Sloan-Kettering<br />
Cancer Center New York, NY; Department <strong>of</strong> Radiation <strong>Oncology</strong>, Memorial Sloan-<br />
Kettering Cancer Center, New York, NY.<br />
Authors’ disclosures <strong>of</strong> potential conflicts <strong>of</strong> interest are found at the end <strong>of</strong> this article.<br />
Address reprint requests to Gregory J. Riely, MD, PhD, Department <strong>of</strong> Medicine,<br />
Memorial Sloan-Kettering Cancer Center, Weill Cornell Medical College, Room 1261, 300<br />
East 66 th St., New York, NY; email: rielyg@mskcc.org.<br />
© <strong>2012</strong> by <strong>American</strong> <strong>Society</strong> <strong>of</strong> <strong>Clinical</strong> <strong>Oncology</strong>.<br />
1092-9118/10/1-10
MANAGEMENT OF THYMIC CARCINOMA<br />
recurrence rate, and shorter survival relative to patients<br />
with thymoma. Thymic carcinoma rarely presents at an<br />
early stage. A small, well-encapsulated thymic carcinoma is<br />
an uncommon occurrence and would likely only be diagnosed<br />
after being resected. More typically, thymic carcinoma presents<br />
in a locally advanced setting, with evidence <strong>of</strong> local<br />
invasion, adenopathy, or pleural or pericardial involvement.<br />
In a recent series at our institution, 75% <strong>of</strong> thymic carcinomas<br />
presented as Masaoka stage III or IV. 4<br />
In patients who present with locally advanced disease,<br />
achieving a complete resection is <strong>of</strong>ten difficult. Within the<br />
anatomic confines <strong>of</strong> the mediastinum, the ability to obtain<br />
a clear surgical margin is frequently limited by the great<br />
vessels <strong>of</strong> the heart. Published complete resection rates vary<br />
from 20% to 60%, and our institutional complete resection<br />
rate <strong>of</strong> 52% is consistent with these results. 4-6 Extended<br />
resections involving resection <strong>of</strong> the innominate vein or<br />
superior vena cava can be done in appropriate settings. In<br />
one series, one-third <strong>of</strong> the patients required partial or<br />
complete resection <strong>of</strong> the vena cava. 7 Circumferential resection<br />
<strong>of</strong> the vena cava also necessitates resection <strong>of</strong> the<br />
phrenic nerve, and careful assessment <strong>of</strong> pulmonary function<br />
in the preoperative planning is mandatory, with consideration<br />
given to diaphragm plication at the time <strong>of</strong> surgery<br />
if indicated. It is debatable whether a tumor that invades<br />
the aorta, pulmonary trunk, or a cardiac chamber can still<br />
be resected. Although these situations may be technically<br />
feasible with the use <strong>of</strong> cardiopulmonary bypass, there is no<br />
data to suggest that the outcomes would be better than the<br />
alternative <strong>of</strong> debulking and R1 resection followed by radiotherapy<br />
for the residual disease. 7 There is insufficient evidence<br />
<strong>of</strong> benefit to support the added morbidity and risk <strong>of</strong><br />
such measures.<br />
There is no consensus regarding the role <strong>of</strong> mediastinal<br />
lymph node dissection for patients with thymic carcinoma.<br />
Patients with thymic carcinoma much more frequently<br />
have nodal involvement than those with thymoma. 8 Any<br />
clinically involved nodes must be resected en bloc with the<br />
specimen, and a careful search <strong>of</strong> the mediastinum should<br />
entail the prevascular, aortopulmonary, internal mammary,<br />
and cervical stations. Whether a systematic mediastinal<br />
lymphadenectomy—including subcarinal and paratracheal<br />
stations—is warranted is unclear and must be individualized<br />
to the patient. 9<br />
Local invasion <strong>of</strong> the lung is common and can be fre-<br />
KEY POINTS<br />
● Thymic carcinomas are an uncommon subset <strong>of</strong> thymic<br />
tumors.<br />
● There are limited prospective data regarding treatment.<br />
● The outcomes <strong>of</strong> patients with thymic carcinoma are<br />
worse than that seen with thymoma.<br />
● In localized thymic carcinoma, complete surgical resection,<br />
if possible, should be performed.<br />
● Radiation therapy may be used as adjuvant therapy<br />
after surgical resection or as the definitive treatment<br />
modality in patients with localized thymic carcinoma<br />
that is deemed unresectable.<br />
quently managed with wedge resections or lobectomy if<br />
necessary. Individual pleural metastases can be easily resected<br />
in some circumstances. However, pleural involvement<br />
that presents as innumerable miliary metastases, or<br />
conversely, as bulky confluent disease will preclude complete<br />
resection. Extrapleural pneumonectomy has been advocated<br />
for stage IVa thymoma by several institutions<br />
including our own, but the poorer prognosis for patients with<br />
thymic carcinoma and the morbidity associated with the<br />
procedure engenders less enthusiasm for this aggressive<br />
approach. 7,10<br />
RT<br />
Although surgical resection remains the most important<br />
component in the management <strong>of</strong> all thymic tumors, RT<br />
may be used as adjuvant therapy after surgical resection or<br />
as the definitive treatment modality in patients who are<br />
deemed unresectable because <strong>of</strong> medical comorbidities or<br />
technical reasons. No prospective trials have established a<br />
clear role for RT, but a series <strong>of</strong> retrospective studies have<br />
demonstrated that excellent local control rates can be<br />
achieved when surgical resection and RT are combined.<br />
The largest study was reported by Kondo et al. 8 Onehundred<br />
and eighty-six patients with thymic carcinoma at<br />
multiple institutions had surgically excision and treatment<br />
with adjuvant chemotherapy, RT, both, or no adjuvant<br />
therapy. Fifty-one percent <strong>of</strong> patients had undergone a<br />
complete resection. The most important prognostic factor<br />
for overall survival was a complete resection. Nevertheless,<br />
51% <strong>of</strong> patients developed a recurrence. Although RT improved<br />
the results in incompletely resected tumors, there<br />
was no clear benefit in patients who had undergone a<br />
complete resection.<br />
Similarly, a retrospective study <strong>of</strong> 40 cases with long-term<br />
follow-up (median follow-up time <strong>of</strong> 87 months for surviving<br />
patients) found that a complete resection was the most<br />
important prognostic factor on multivariate analysis. 11 All<br />
patients in this study were treated with either adjuvant RT<br />
in resectable patients or definitive RT in unresectable cases.<br />
They achieved 100% in-field local control in 16 patients who<br />
underwent a complete resection followed by adjuvant RT<br />
with a median dose <strong>of</strong> 50 Gy. Other significant prognostic<br />
factors included a KPS <strong>of</strong> 70% or greater and low-grade<br />
histology. The 5- and 10-year overall survival rates for the<br />
whole patient group was 38% and 28%, respectively.<br />
Hsu et al reported their experience <strong>of</strong> 26 patients with a<br />
minimum follow-up <strong>of</strong> 40 months, all <strong>of</strong> which were treated<br />
with adjuvant RT at a median dose <strong>of</strong> 60 Gy after a total<br />
or subtotal resection. 12 Excellent 5-year local control rates<br />
<strong>of</strong> 92% after a complete surgical resection and adjuvant RT<br />
were obtained. Even after an incomplete resection and<br />
adjuvant RT, 5-year local control rates <strong>of</strong> 88% were<br />
achieved. The 5-year overall survival in this study was 77%.<br />
RT techniques<br />
Most published studies to date have reported outcomes on<br />
patients who were treated over a long period <strong>of</strong> time with<br />
<strong>of</strong>ten times old two-dimensional RT techniques using opposed<br />
fields. However, the development <strong>of</strong> more conformal<br />
RT (CRT) techniques—such as three-dimensional CRT in<br />
the early 1990s and more recently intensity-modulated RT<br />
(IMRT) or proton therapy—has allowed significantly im-<br />
467
Reference Therapy<br />
proved sparing <strong>of</strong> normal organs surrounding the tumor/<br />
tumor bed. In addition, four-dimensional treatment<br />
planning now allows a quantitative determination <strong>of</strong> tumor<br />
motion caused by respiration at the time <strong>of</strong> simulation and<br />
provides further accuracy and confidence in our treatment<br />
delivery. Expert guidelines recommend that these techniques<br />
be used for thymic malignancies where available. 13 It<br />
needs to be noted that the increased conformality <strong>of</strong> these<br />
techniques harbors the risk <strong>of</strong> underdosing the tumor, which<br />
may lead to marginal recurrences. Therefore, a clear set <strong>of</strong><br />
definitions and guidelines for the simulation, treatment<br />
planning, and delivery using CRT techniques is necessary.<br />
Recently, the International Thymic Malignancy Interest<br />
Group (ITMIG) has published its first set <strong>of</strong> guidelines to<br />
increase consistency in how RT is delivered to patients with<br />
thymic malignancies. 14 These may allow a more systematic<br />
approach to determine the best use <strong>of</strong> RT in thymic tumors.<br />
RT dose<br />
The optimal RT dose has not yet been determined. Most<br />
published studies used between 40 and 70 Gy in 1.8 to 2.0 Gy<br />
daily fractions. One study on thymic malignancies, which<br />
included six patients with thymic carcinoma, found that an<br />
RT dose <strong>of</strong> 50 Gy or greater was associated with improved<br />
local control. 15 However, this has not been confirmed in<br />
other studies. 12 In general, doses <strong>of</strong> 45 to 50 Gy for negative/<br />
close margins, 50 to 54 Gy for microscopically positive<br />
margins, and 60 Gy for gross residual disease in conventional<br />
daily fractions <strong>of</strong> 1.8 to 2.0 Gy are recommended<br />
(National Comprehensive Cancer Network guidelines<br />
v2.<strong>2012</strong>).<br />
Organs at Risk<br />
The organs at risk for short- and long-term toxicities from<br />
RT for thymic carcinomas include the heart, lungs, esophagus,<br />
and spinal cord. More modern RT techniques, such as<br />
three-dimensional CRT and IMRT, allow better sparing <strong>of</strong><br />
these organs at risk without compromising target coverage.<br />
General dosimetric guidelines include a mean heart dose <strong>of</strong><br />
less than 26 Gy, mean lung dose 20 Gy or less and V20 at<br />
30% to 35% or less, mean esophageal dose less than 34 Gy,<br />
and Dmax to the spinal cord less than 45 Gy. 14<br />
Systemic Therapy<br />
Table 1. Selected Prospective and Larger Retrospective Series <strong>of</strong> Patients with Thymic Carcinoma Treated with<br />
Cytotoxic Chemotherapy<br />
Systemic therapy for thymic carcinoma is used in two<br />
clinical scenarios, preoperative treatment and palliative<br />
Prospective or<br />
Retrospective Patients RR (%)<br />
Median PFS/TTP<br />
(months)<br />
Median OS<br />
(months) Comment<br />
18 ADOC/ADOCb R 34 50% 21<br />
19 Cisplatin/irinotecan R 9 56% 8 34<br />
20 CODE R 12 42% 6 46<br />
21 Carboplatin, paclitaxel P 23 22% 5 20 Included type B3 thymoma<br />
22 Carboplatin, paclitaxel P 11 36% 8 23<br />
23 Etoposide, ifosfamide, cisplatin R 4 25%<br />
24 Etoposide, Ifosfamide, cisplatin P 8 25%<br />
25 Doxorubicin, cyclophosphamide, cisplatin, vincristine P 8 75% 19<br />
16 Pemetrexed P 11 0 5<br />
Abbreviations: RR, recurrence rate; PFS, progression-free survival; TTP, time to progression; OS, overall survival; ADOC, doxorubicin, cyclophosphamide, vincristine,<br />
cisplatin; ADOCb, doxorubicin, cyclophosphamide, vincristine, carboplatin; CODE, cisplatin, vincristine, doxorubicin, etoposide.<br />
468<br />
RIELY, HUANG, AND RIMNER<br />
therapy. Given the significance <strong>of</strong> a complete surgical resection—if<br />
significant tumor response occurs—the probability<br />
<strong>of</strong> complete surgical resection can be improved. In comparison<br />
with more common diseases, identifying the appropriate<br />
chemotherapy regimen to choose for initial or subsequent<br />
therapy for patients with thymic carcinoma is challenging.<br />
As with surgical and RT data, prior reports <strong>of</strong> thymic tumor<br />
management have <strong>of</strong>ten combined patients with thymoma<br />
and thymic carcinoma into small studies. However, there<br />
are relatively large retrospective series and subsets from<br />
prospective series that provide some data to guide therapy.<br />
Cytotoxic Chemotherapy<br />
All chemotherapy regimens evaluated in thymic carcinoma<br />
have been studied either as part <strong>of</strong> series evaluating<br />
thymoma or used chemotherapy regimens developed for<br />
thymoma. In general, the combination chemotherapy regimens<br />
most widely evaluated have combined platinum analogs,<br />
anthracyclines, along with other agents (Table 1). For<br />
first-line, platinum-based chemotherapy regimens, the response<br />
rates have ranged between 22% and 75%; however,<br />
the varying chemotherapy regimens, different ways data<br />
were collected, and the small sample sizes limit the conclusions<br />
that can be drawn about individual chemotherapy<br />
regimens. Despite our inability to differentiate individual<br />
regimens, it seems clear that the most commonly used<br />
treatments are cisplatin-based chemotherapy, sometimes in<br />
combination with an anthracycline.<br />
Despite the relatively short median overall survival times<br />
reported, there is a frustrating lack <strong>of</strong> data to guide use <strong>of</strong><br />
second-line cytotoxic therapies for treatment <strong>of</strong> patient with<br />
thymic carcinoma. The exception is pemetrexed in patients<br />
with previously treated advanced thymic carcinoma; unfortunately,<br />
there were no radiographic responses observed,<br />
but there was a median time to progression <strong>of</strong> 5 months. 16<br />
This summary <strong>of</strong> data underscores the need for more prospective<br />
evaluation <strong>of</strong> the cytotoxic chemotherapies <strong>of</strong>ten<br />
used to treat patients with thymic carcinoma.<br />
“Targeted” Systemic Therapies<br />
The absence <strong>of</strong> high rates <strong>of</strong> radiographic response and<br />
short progression-free survival associated with cytotoxic<br />
chemotherapy has blunted enthusiasm for further evaluation<br />
<strong>of</strong> such drugs and more recent studies have focused on<br />
newer targeted therapies, <strong>of</strong>ten with preclinical rationales<br />
supporting evaluation <strong>of</strong> a given drug (Table 2). With two
MANAGEMENT OF THYMIC CARCINOMA<br />
Reference Therapy<br />
Table 2. Selected Prospective and Larger Retrospective Series <strong>of</strong> Patients with Thymic Carcinoma Treated with<br />
“Targeted” Therapies<br />
exceptions, testing <strong>of</strong> targeted therapies in thymic carcinoma<br />
has met with no significant success. Several prospective<br />
trials <strong>of</strong> imatinib were launched after initial case<br />
reports identified mutated KIT in patients with thymic<br />
carcinoma and subsequent significant response to multitargeted<br />
tyrosine kinase inhibitors. These studies included<br />
patients with thymic carcinoma, sometimes requiring KIT<br />
(or platelet-derived growth factor expression), but none<br />
required the presence <strong>of</strong> KIT mutations for enrollment. In<br />
the absence <strong>of</strong> patients with documented KIT mutations, no<br />
patients had radiographic responses. Because <strong>of</strong> high epidermal<br />
growth factor receptor (EGFR) expression in thymic<br />
carcinoma, both gefitinib and erlotinib (with bevacizumab)<br />
have been prospectively evaluated in a total <strong>of</strong> 14 patients<br />
without radiographic response. No evidence <strong>of</strong> EGFR dependence<br />
in thymic tumors has been reported (no EGFR mutations<br />
and no significant focal EGFR gene amplification). One<br />
targeted therapy has shown initial success in patients with<br />
thymic carcinoma who were enrolled in a phase I trial. Two<br />
<strong>of</strong> three patients with thymic carcinoma treated with PHA-<br />
848125AC—a dual cyclin-dependent kinase 2/thropomyosin<br />
receptor kinase A inhibitor—had radiographic responses<br />
and long responses to therapy. This initial success has led to<br />
Authors’ Disclosures <strong>of</strong> Potential Conflicts <strong>of</strong> Interest<br />
Author<br />
Employment or<br />
Leadership<br />
Positions<br />
Prospective or<br />
Retrospective Patients RR (%)<br />
Consultant or<br />
Advisory Role<br />
Gregory J. Riely ARIAD;<br />
Boehringer<br />
Ingelheim;<br />
Chugai Pharma;<br />
Daiichi Sankyo;<br />
Novartis;<br />
Tragara<br />
a multicenter phase II study <strong>of</strong> this drug in patients with<br />
thymic carcinoma (clinicaltrials.gov NCT01011439).<br />
Future Directions<br />
The rarity <strong>of</strong> all thymic tumors, especially thymic carcinomas,<br />
complicates prospective clinical research in these<br />
diseases. As noted above, at sites with expertise in this<br />
disease, the largest series <strong>of</strong> patients prospectively evaluated<br />
was 23 and none <strong>of</strong> these studies has led to a breakthrough<br />
in therapy for this disease. Given the uncommon<br />
nature <strong>of</strong> thymoma and thymic carcinoma, empiric approaches<br />
to drug evaluation conducted at a limited number<br />
<strong>of</strong> centers are likely to be <strong>of</strong> little benefit. Under the<br />
guidance <strong>of</strong> the ITMIG, researchers are developing a large<br />
retrospective and prospective database to identify potential<br />
improvements in care that can be derived from current<br />
practices. In addition, ITMIG has put forth standards for<br />
clinical research to allow comparison <strong>of</strong> clinical trial results<br />
across groups. 17 Finally, the thymic cancer research community<br />
needs to harness recent advances in sequencing technology<br />
to vastly improve our understanding <strong>of</strong> the biology<br />
<strong>of</strong> this disease to generate testable clinical research hypotheses.<br />
Stock<br />
Ownership Honoraria<br />
Research<br />
Funding<br />
Bristol-Myers<br />
Squibb<br />
James Huang Bristol-Myers<br />
Squibb<br />
Andreas Rimner*<br />
*No relevant relationships to disclose.<br />
1. Girard N, Shen R, Guo T, et al. Comprehensive genomic analysis reveals<br />
clinically relevant molecular distinctions between thymic carcinomas and<br />
thymomas. Clin Cancer Res. 2009;15:6790-6799.<br />
2. Girard N, Teruya-Feldstein J, Payabyab EC, et al. Insulin-like growth<br />
factor-1 receptor expression in thymic malignancies. J Thorac Oncol. 2010;5:<br />
1439-1446.<br />
3. Zettl A, Strobel P, Wagner K, et al. Recurrent genetic aberrations in<br />
thymoma and thymic carcinoma. Am J Pathol. 2000;157:257-266.<br />
4. Huang J, Rizk NP, Travis WD, et al. Comparison <strong>of</strong> patterns <strong>of</strong> relapse<br />
REFERENCES<br />
Median PFS/TTP<br />
(months)<br />
Median OS<br />
(months) Comment<br />
26 Imatinib P 11 0 All patients were KIT� or PDGF� by IHC<br />
27 Imatinib P 5 0 No patients had KIT mutations<br />
28 Imatinib P 3 0 No patients had KIT mutations<br />
29 Erlotinib, bevacizumab P 7 0<br />
30 Gefitinib P 7 0<br />
31 Belinostat P 16 0 3 12<br />
32 Octreotide P 5 0 5 23<br />
33 PHA 848125 P 3 67% Part <strong>of</strong> a phase I trial<br />
Abbreviations: RR, recurrence rate; PFS, progression-free survival; TTP, time to progression; OS, overall survival; PDGF, platelet-derived growth factor receptor; IHC,<br />
immunohistochemical.<br />
Expert<br />
Testimony<br />
Other<br />
Remuneration<br />
in thymic carcinoma and thymoma. J Thorac Cardiovasc Surg. 2009;138:26-<br />
31.<br />
5. Chalabreysse L, Etienne-Mastroianni B, Adeleine P, et al. Thymic<br />
carcinoma: A clinicopathological and immunohistological study <strong>of</strong> 19 cases.<br />
Histopathology. 2004;44:367-374.<br />
6. Takeda S, Sawabata N, Inoue M, et al. Thymic carcinoma. <strong>Clinical</strong><br />
institutional experience with 15 patients. Eur J Cardiothorac Surg. 2004;26:<br />
401-406.<br />
7. Huang J, Rizk NP, Travis WD, et al. Feasibility <strong>of</strong> multimodality<br />
469
therapy including extended resections in stage IVA thymoma. J Thorac<br />
Cardiovasc Surg. 2007;134:1477-1483, discussion 1483-1484.<br />
8. Kondo K, Monden Y. Therapy for thymic epithelial tumors: A clinical<br />
study <strong>of</strong> 1,320 patients from Japan. Ann Thorac Surg. 2003;76:878-884.<br />
9. Detterbeck FC, Moran C, Huang J, et al. Which way is up? Policies and<br />
procedures for surgeons and pathologists regarding resection specimens <strong>of</strong><br />
thymic malignancy. J Thorac Oncol. 2011;6:S1730-S1738.<br />
10. Wright CD. Stage IVA thymoma: Patterns <strong>of</strong> spread and surgical<br />
management. Thorac Surg Clin. 2011;21:93-97, vii.<br />
11. Ogawa K, Toita T, Uno T, et al. Treatment and prognosis <strong>of</strong> thymic<br />
carcinoma a retrospective: Analysis <strong>of</strong> 40 cases. Cancer. 2002;94:3115-3119.<br />
12. Hsu HC, Huang EY, Wang CJ, et al. Postoperative radiotherapy in<br />
thymic carcinoma: Treatment results and prognostic factors. Int J Radiat<br />
Oncol Biol Phys. 2002;52:801-805.<br />
13. Gomez D, Komaki R. Technical advances <strong>of</strong> radiation therapy for<br />
thymic malignancies. J Thorac Oncol. 2010;5:S336-S343.<br />
14. Gomez D, Komaki R, Yu J, et al. Radiation therapy definitions and<br />
reporting guidelines for thymic malignancies. J Thorac Oncol. 2011;6:S1743-<br />
S1748.<br />
15. Mayer R, Beham-Schmid C, Groell R, et al. Radiotherapy for invasive<br />
thymoma and thymic carcinoma: Clinicopathological review. Strahlenther<br />
Onkol. 1999;175:271-278.<br />
16. Loehrer PJ, Yiannoutsos CT, Dropcho S, et al. A phase II trial <strong>of</strong><br />
pemetrexed in patients with recurrent thymoma or thymic carcinoma. J Clin<br />
Oncol. 2006;24:7079.<br />
17. Girard N, Lal R, Wakelee H, et al. Chemotherapy definitions and<br />
policies for thymic malignancies. J Thorac Oncol. 2011;6:S1749-S1755.<br />
18. Agatsuma T, Koizumi T, Kanda S, et al. Combination chemotherapy<br />
with doxorubicin, vincristine, cyclophosphamide, and platinum compounds<br />
for advanced thymic carcinoma. J Thorac Oncol. 2011;6:2130-2134.<br />
19. Okuma Y, Hosomi Y, Takagi Y, et al. Cisplatin and irinotecan combination<br />
chemotherapy for advanced thymic carcinoma: Evaluation <strong>of</strong> efficacy<br />
and toxicity. Lung Cancer. 2011;74:492-496.<br />
20. Yoh K, Goto K, Ishii G, et al. Weekly chemotherapy with cisplatin,<br />
vincristine, doxorubicin, and etoposide is an effective treatment for advanced<br />
thymic carcinoma. Cancer. 2003;98:926-931.<br />
21. Lemma GL, Lee JW, Aisner SC, et al. Phase II study <strong>of</strong> carboplatin<br />
470<br />
RIELY, HUANG, AND RIMNER<br />
and paclitaxel in advanced thymoma and thymic carcinoma. J Clin Oncol.<br />
2011;29:2060-2065.<br />
22. Igawa S, Murakami H, Takahashi T, et al. Efficacy <strong>of</strong> chemotherapy<br />
with carboplatin and paclitaxel for unresectable thymic carcinoma. Lung<br />
Cancer. 2010;67:194-197.<br />
23. Grassin F, Paleiron N, Andre M, et al. Combined etoposide, ifosfamide,<br />
and cisplatin in the treatment <strong>of</strong> patients with advanced thymoma and<br />
thymic carcinoma. A French experience. J Thorac Oncol. 2010;5:893-897.<br />
24. Loehrer PJ Sr., Jiroutek M, Aisner S, et al. Combined etoposide,<br />
ifosfamide, and cisplatin in the treatment <strong>of</strong> patients with advanced thymoma<br />
and thymic carcinoma: An intergroup trial. Cancer. 2001;91:2010-2015.<br />
25. Koizumi T, Takabayashi Y, Yamagishi S, et al. Chemotherapy for<br />
advanced thymic carcinoma: <strong>Clinical</strong> response to cisplatin, doxorubicin,<br />
vincristine, and cyclophosphamide (ADOC chemotherapy). Am J Clin Oncol.<br />
2002;25:266-268.<br />
26. Salter JT, Lewis D, Yiannoutsos CT, et al. Imatinib for the treatment <strong>of</strong><br />
thymic carcinoma. J Clin Oncol. 2008;26 (suppl; abstr 8116).<br />
27. Giaccone G, Rajan A, Ruijter R, et al. Imatinib mesylate in patients<br />
with WHO B3 thymomas and thymic carcinomas. J Thorac Oncol. 2009;4:<br />
1270-1273.<br />
28. Palmieri G, Marino M, Buonerba C, et al. Imatinib mesylate in thymic<br />
epithelial malignancies. Cancer Chemother Pharmacol. <strong>2012</strong>;69:309-315.<br />
29. Bedano PM, Perkins S, Burns M, et al. A phase II trial <strong>of</strong> erlotinib plus<br />
bevacizumab in patients with recurrent thymoma or thymic carcinoma. J Clin<br />
Oncol. 2008;26 (suppl; abstr 19087).<br />
30. Kurup A, Burns M, Dropcho S, et al. Phase II study <strong>of</strong> gefitinib<br />
treatment in advanced thymic malignancies. J Clin Oncol. 2005;23:7068.<br />
31. Giaccone G, Rajan A, Berman A, et al. Phase II study <strong>of</strong> belinostat in<br />
patients with recurrent or refractory advanced thymic epithelial tumors.<br />
J Clin Oncol. 2011;29:2052-2059.<br />
32. Loehrer PJ Sr, Wang W, Johnson DH, et al. Octreotide alone or with<br />
prednisone in patients with advanced thymoma and thymic carcinoma: An<br />
Eastern Cooperative <strong>Oncology</strong> Group Phase II Trial. J Clin Oncol. 2004;22:<br />
293-299.<br />
33. Weiss GJ, Hidalgo M, Borad MJ, et al. Phase I study <strong>of</strong> the safety,<br />
tolerability and pharmacokinetics <strong>of</strong> PHA-848125AC, a dual tropomyosin<br />
receptor kinase A and cyclin-dependent kinase inhibitor, in patients with<br />
advanced solid malignancies. Invest New Drugs. Epub 2011 Dec 9.
The Creation <strong>of</strong> the International Thymic<br />
Malignancies Interest Group as a Model for<br />
Rare Diseases<br />
Overview: Similar to other orphan diseases, little progress<br />
has been made in the past decades in thymic malignancies. A<br />
determination to make a difference, despite the challenges<br />
facing a rare disease, led to the formation <strong>of</strong> the International<br />
Thymic Malignancies Interest Group (ITMIG) in 2010. This<br />
organization has brought together the majority <strong>of</strong> those focused<br />
on the management <strong>of</strong> thymic malignancies and has<br />
built a foundation for scientific collaboration, including con-<br />
THYMOMA IS a relatively rare malignancy. The agestandardized<br />
incidence has been reported to be 2.5 and<br />
2.8 per million in Denmark and Iceland, respectively. 1,2<br />
Studies in the United Kingdom and United States have<br />
reported incidence rates <strong>of</strong> 0.72 to 1.5 per million, but these<br />
series may have missed many smaller thymomas (i.e., those<br />
previously thought to be benign). 3,4 A study <strong>of</strong> the Surveillance<br />
Epidemiology and End Results (SEER) database in the<br />
United States from 1973 to 2006 found a modest and<br />
consistent increase in incidence (Fig. 1). 5 This was true for<br />
all subtypes and stages, suggesting that the increased incidence<br />
was unlikely to be an artifact (e.g., because <strong>of</strong> a higher<br />
prevalence <strong>of</strong> computed tomography imaging).<br />
A recent comparative analysis <strong>of</strong> SEER data from 1988 to<br />
2003 found that there was no improvement in survival<br />
during this period (Fig. 2). 6 For each stage <strong>of</strong> disease, there<br />
was no consistent evidence <strong>of</strong> even a trend toward better<br />
survival in more recent years, despite potential advances in<br />
medicine and surgery.<br />
Why has no progress in outcomes been seen over the past<br />
2 decades? A major factor is certainly that the disease is<br />
relatively rare and physicians have largely worked independently.<br />
The treatment approach has been primarily empiric,<br />
based on individual judgment with little supporting data.<br />
In addition, most published studies are retrospective series<br />
spanning many decades during which many changes have<br />
occurred and provide only a vague idea <strong>of</strong> what can be<br />
learned from this experience (Fig. 3). <strong>Clinical</strong> trials have<br />
been rare, involving only limited numbers <strong>of</strong> patients in<br />
phase II studies. Also, as it is with any rare disease, research<br />
funding mechanisms and health care structures make it<br />
difficult to establish a scientific basis for approaching the<br />
disease. Thus funding is not available because there is no<br />
scientific basis to build on, and there is no scientific basis<br />
because there is no funding. 7<br />
In the case <strong>of</strong> thymoma, several other issues have hampered<br />
progress. First, a common misconception is that many<br />
thymomas are benign. The data do not support this, and this<br />
misnomer should be abandoned. 8 However, this misconception<br />
together with the view that the thymus (in adults) is an<br />
involuted functionless organ contribute to a lack <strong>of</strong> interest<br />
in and focus on thymic malignancy. This is worsened by the<br />
fact that cardiac surgeons see the thymus every day as<br />
inconsequential tissue and frequently are willing to remove<br />
or debulk a thymic malignancy with little understanding <strong>of</strong><br />
the disease process itself.<br />
Another factor has been the inconsistency with which<br />
By Frank C. Detterbeck, MD<br />
sistent use <strong>of</strong> terms, an international database, and multidisciplinary<br />
engagement <strong>of</strong> clinicians and researchers from<br />
around the world. ITMIG has embarked on the development <strong>of</strong><br />
novel approaches to research particularly suited to a rare<br />
condition. ITMIG has gained substantial recognition for the<br />
rapid progress that has been made and serves as a model for<br />
the advancement <strong>of</strong> knowledge in a rare disease.<br />
terms have been interpreted. This makes published data<br />
almost impossible to compare. Yet for a rare disease, collaboration<br />
is absolutely essential. For example, published literature<br />
contains a variety <strong>of</strong> outcome measures; because <strong>of</strong><br />
the nature <strong>of</strong> thymic malignancies, associated diseases and<br />
causes <strong>of</strong> death, these yield dramatically different outcome<br />
results (Fig. 4). 9 These distinctions have rarely been appreciated,<br />
and no consistency in reporting results had emerged<br />
until the advent <strong>of</strong> the ITMIG.<br />
The Development <strong>of</strong> ITMIG<br />
ITMIG was inaugurated as a formal not-for-pr<strong>of</strong>it organization<br />
in May 2010. ITMIG is an academic organization,<br />
whose mission is to promote the advancement <strong>of</strong> science<br />
related to thymic malignancies and other mediastinal conditions<br />
to achieve better outcomes for patients. The goal <strong>of</strong><br />
ITMIG is to develop an infrastructure that facilitates collaboration<br />
and to create innovative approaches that maximize<br />
the progress that can be made.<br />
The catalyst for the development <strong>of</strong> ITMIG came from the<br />
Foundation for Thymic Cancer Research, an organization<br />
formed by patients and family members who were frustrated<br />
about having to search for prolonged periods before finding a<br />
physician who was truly knowledgeable about thymic malignancy.<br />
This group held two conferences in 2007 and 2008<br />
to which physicians active in this disease were invited. In<br />
addition to stimulating discussion and some collaborative<br />
projects, it became clear that real progress in a rare disease<br />
such as thymic malignancy would require creating a scientific<br />
infrastructure to foster collaborative research. At a<br />
third meeting held in 2009 at the National Institutes <strong>of</strong><br />
Health in Bethesda, Maryland, a provisional structure<br />
was created and tasked with the formal development <strong>of</strong><br />
ITMIG. 10<br />
Many pr<strong>of</strong>essional organizations have come forward to<br />
support the creation <strong>of</strong> ITMIG, including the <strong>American</strong><br />
Association for Thoracic Surgery, the European Association<br />
<strong>of</strong> Cardiothoracic Surgeons, the European <strong>Society</strong> for Myasthenia<br />
Gravis, the European <strong>Society</strong> <strong>of</strong> Thoracic Surgeons,<br />
From the Yale University School <strong>of</strong> Medicine, Thoracic Surgery, New Haven, CT.<br />
Author’s disclosures <strong>of</strong> potential conflicts <strong>of</strong> interest are found at the end <strong>of</strong> this article.<br />
Address reprint requests to Frank C. Detterbeck, MD, Yale University School <strong>of</strong> Medicine,<br />
330 Cedar St., BB205, New Haven, CT 06520-8062; email: frank.detterbeck@yale.edu.<br />
© <strong>2012</strong> by <strong>American</strong> <strong>Society</strong> <strong>of</strong> <strong>Clinical</strong> <strong>Oncology</strong>.<br />
1092-9118/10/1–10<br />
471
Fig. 1. Changing incidence <strong>of</strong> thymoma and thymic cancer. Data<br />
taken from the Surveillance Epidemiology and End Results (SEER)<br />
database. 5<br />
the Fleischner <strong>Society</strong>, the General Thoracic Surgical Club,<br />
the International Association for the Study <strong>of</strong> Lung Cancer<br />
(IASLC), the Japanese Association <strong>of</strong> Chest Surgeons, the<br />
Japanese Association for Research on the Thymus, the<br />
Japanese Association for Thoracic Surgery, the mediastinal<br />
workgroup <strong>of</strong> the European <strong>Society</strong> <strong>of</strong> Pathologists, the<br />
Myasthenia Gravis Foundation <strong>of</strong> America, the <strong>Society</strong> <strong>of</strong><br />
Thoracic Radiologists, the <strong>Society</strong> <strong>of</strong> Thoracic Surgeons, and<br />
the Thoracic <strong>Oncology</strong> Network <strong>of</strong> the <strong>American</strong> College <strong>of</strong><br />
Chest Physicians.<br />
As <strong>of</strong> January <strong>2012</strong>, ITMIG has nearly 400 members from<br />
all continents (except Antarctica). ITMIG is a multispecialty<br />
organization, involving thoracic surgeons, medical and radiation<br />
oncologists, pulmonologists, radiologists, pathologists,<br />
neurologists, and basic science researchers. Perhaps the<br />
most important early achievement <strong>of</strong> ITMIG has been the<br />
enthusiastic engagement <strong>of</strong> nearly all <strong>of</strong> those around the<br />
world who have been active in studying thymic malignancies.<br />
The ITMIG annual meetings in 2010 (in New York) and<br />
in 2011 (in Amsterdam) have been well attended with many<br />
submitted abstracts and posters.<br />
ITMIG Projects and Accomplishments<br />
A prerequisite for collaboration is the ability to speak the<br />
same language. Differences in the interpretation <strong>of</strong> terms<br />
were surprisingly wide in this field and largely unrecog-<br />
KEY POINTS<br />
● Progress in a rare disease cannot occur without global<br />
collaboration.<br />
● The consistent use <strong>of</strong> terms across a field <strong>of</strong> study is<br />
necessary for collaboration.<br />
● Details <strong>of</strong> how outcomes are reported for thymic<br />
malignancies are critically important when comparing<br />
results.<br />
● The recognition <strong>of</strong> the limitations inherent in data<br />
from small patient cohorts is important to appropriately<br />
interpret available results.<br />
● Compared with traditional approaches, innovative<br />
approaches to statistics and clinical science provide a<br />
research strategy better suited to a rare disease.<br />
472<br />
FRANK C. DETTERBECK<br />
Fig. 2. Trends in 5-year survival for thymoma and thymic cancer<br />
by tumor extent. Data taken from the Surveillance Epidemiology and<br />
End Results (SEER) database. 6<br />
nized. The ITMIG community organized a broad process to<br />
clarify critical terms. Multiple workgroups were assembled,<br />
and core members drafted initial proposals, which were<br />
vetted with workgroup members. At a 2-day workshop at<br />
Yale University with broad international representation,<br />
these definitions were discussed and revised so that they<br />
would be aligned with one another. These were then further<br />
refined by the workgroups with input from the entire ITMIG<br />
membership. The final documents were approved by the<br />
ITMIG membership for use in all ongoing research and<br />
publications. The level <strong>of</strong> engagement and broad consensus<br />
in this process was in itself a major accomplishment <strong>of</strong><br />
ITMIG. These definitions were published in a supplement to<br />
the Journal <strong>of</strong> Thoracic <strong>Oncology</strong> (JTO) in July 2011 and are<br />
openly available for download from the ITMIG or JTO<br />
websites (itmig.org and jto.org).<br />
A fundamental aspect <strong>of</strong> cancer research is stage classification,<br />
but there is no formal classification system for<br />
thymic malignancies. ITMIG partnered with IASLC, which<br />
has extensive experience in developing the revision <strong>of</strong> the<br />
staging system for lung cancer, to develop an <strong>of</strong>ficial thymic<br />
stage classification system for the next (eighth) edition <strong>of</strong> the<br />
stage classification <strong>of</strong> tumors in 2017. This project is conducted<br />
under the auspices <strong>of</strong> the Union for International<br />
Cancer Control and <strong>American</strong> Joint Committee on Cancer,<br />
the entities that determine the <strong>of</strong>ficial classification systems<br />
for all tumors.<br />
The histologic classification <strong>of</strong> thymic malignancies has<br />
also been a source <strong>of</strong> confusion and controversy. There is<br />
Fig. 3. Overview <strong>of</strong> published literature results from a PubMed<br />
search from 1989 to 2009 for thymoma, grouped by type <strong>of</strong> paper<br />
and size <strong>of</strong> patient cohort.
ITMIG AS A MODEL FOR RARE DISEASES<br />
Fig. 4. Ten-year outcomes for different outcome measures in the<br />
same cohort <strong>of</strong> resected patients with stage III thymoma. 9<br />
variability in how the current World Health Organization<br />
classification system is interpreted and applied, and its<br />
prognostic value is inconsistent. ITMIG has conducted two<br />
international workshops to identify the sources <strong>of</strong> the difficulties<br />
and to develop a strategy for defining a better system.<br />
<strong>Clinical</strong> science depends heavily on statistics to separate<br />
what we know from perceptions or beliefs. However, a rare<br />
disease presents many statistical challenges. The limited<br />
number <strong>of</strong> patients magnifies misperceptions caused by<br />
common practices about how clearly something has been<br />
demonstrated. ITMIG invested effort in developing an understanding<br />
<strong>of</strong> the limitations and techniques to minimize or<br />
at least evaluate the level <strong>of</strong> uncertainty (Fig. 5). As examples,<br />
confidence intervals around survival curves provide a<br />
clearer picture <strong>of</strong> the findings, and identification <strong>of</strong> prognostic<br />
factors may carry a risk <strong>of</strong> false-positive or false-negative<br />
findings that should be acknowledged. A description <strong>of</strong> such<br />
issues has also been published in the JTO supplement. 11<br />
A sophisticated, detailed international ITMIG database<br />
has been built on the HUBzero platform, thus benefiting<br />
from the engineering expertise <strong>of</strong> Purdue University, over a<br />
million users <strong>of</strong> this platform for multiple major initiatives,<br />
and more than a decade <strong>of</strong> research and experience in<br />
developing platforms for international collaboration involving<br />
multiple disparate types <strong>of</strong> data and analyses. Furthermore,<br />
the developers <strong>of</strong> this platform are focused on<br />
providing tools to facilitate research rather than on promoting<br />
a commercial product. The database is linked to a virtual<br />
tissue bank that is actively accruing samples.<br />
Strategy for the Future<br />
Making progress in a rare disease is more challenging<br />
than in a common condition. Merely trying to duplicate the<br />
measures taken in common will leave thymic malignancies<br />
still far behind other areas, despite facilitating some degree<br />
<strong>of</strong> progress. Therefore a specific focus <strong>of</strong> ITMIG is to seek out<br />
novel and innovative approaches that allow ITMIG to leapfrog<br />
ahead. Each annual meeting has several lectures chosen<br />
specifically to promote insight and thinking in areas<br />
that are not widely known but may have implications for<br />
thymic malignancies. Innovative strategies and research<br />
approaches are being explored as well.<br />
A search for a novel strategy that would maximize the rate<br />
<strong>of</strong> progress led ITMIG to a collaborative project at Simon<br />
Cancer Center at Indiana University and the Purdue Department<br />
<strong>of</strong> Engineering. This project, known as Cancer<br />
Care Engineering (CCE), applies techniques <strong>of</strong> complex<br />
modeling to cancer research. Developing an adaptive model<br />
allows new insights to be quickly assessed in a virtual<br />
manner and allows for more strategic planning <strong>of</strong> how to<br />
prioritize and how best to attempt to validate early findings.<br />
Such an approach has been successful in other types <strong>of</strong><br />
cancer, and the people involved in the CCE project had<br />
independently come to the idea that the adaptation <strong>of</strong> this<br />
approach would be particularly useful in a rare disease—<br />
coincident with ITMIG’s initiative to find novel approaches<br />
that would maximize progress. The initial accomplishments<br />
<strong>of</strong> ITMIG provide a good foundation on which to build this<br />
effort; it is now time to begin actual development <strong>of</strong> this<br />
approach.<br />
The traditional clinical research approach relies on providing<br />
clear pro<strong>of</strong> <strong>of</strong> one approach over another through the<br />
use <strong>of</strong> randomized clinical trials. Although this is part <strong>of</strong> the<br />
scope <strong>of</strong> ITMIG’s research plans, this strategy is also associated<br />
with major challenges, especially in a rare disease.<br />
ITMIG therefore is also including other approaches, particularly<br />
the use <strong>of</strong> Bayesian statistics. These do not seek to<br />
prove superiority or exclude expectations based on prior<br />
data (i.e., “biases”) as with the traditional frequentist approach.<br />
Instead, the Bayesian approach makes use <strong>of</strong> prior<br />
knowledge and quantifies the possibility that one treatment<br />
is or is not better. Bayesian analyses have the advantage in<br />
a rare disease <strong>of</strong> being applicable no matter how many<br />
patients are available for inclusion and <strong>of</strong> refining predictions<br />
based on each observation as it happens, instead <strong>of</strong><br />
blinding for years until the data are mature.<br />
Traditional research approaches also play a role, where<br />
applicable, in the ITMIG approach. However, in a rare<br />
disease this requires global collaboration. It is hard enough<br />
negotiating the regulatory hurdles for a multi-institutional<br />
study in one country much less meeting expectations across<br />
many countries. ITMIG has partnered with the International<br />
Rare Diseases Initiative, which is a collaboration<br />
between relevant organizations in the United States, the<br />
United Kingdom, and Europe, to manage such issues.<br />
A problem that is magnified for a rare disease is funding<br />
for the infrastructure necessary to perform collaborative<br />
Fig. 5. Precision in estimating. The vertical bars are 95% confidence<br />
intervals for a 5-year study, based on a standard model <strong>of</strong><br />
exponentially decreasing survival, a constant rate <strong>of</strong> accrual, and no<br />
loss to follow-up. 9<br />
Abbreviation: MST, median survival time.<br />
473
work and to accomplish the work itself. Traditional mechanisms<br />
such as research grants are difficult to qualify for—<br />
projects without a large clinical effect or with limited data<br />
existing have a low chance <strong>of</strong> competing for funding. Industry<br />
generally sees the market niche as small with little<br />
return on investment. ITMIG has struggled with these<br />
issues but has managed to stay afloat. The work performed<br />
by the ITMIG members, <strong>of</strong> course, is purely donated time by<br />
physicians, researchers, and other health care pr<strong>of</strong>essionals<br />
who feel that the opportunity to move forward is simply<br />
something they have to support. Several industry sponsors<br />
have donated unrestricted gifts in what represents primarily<br />
an altruistic gesture. Many related pr<strong>of</strong>essional organizations<br />
have not only <strong>of</strong>ficially endorsed ITMIG but also<br />
have been willing to provide start-up money. The bulk <strong>of</strong><br />
funding, however, comes from patients and their families<br />
and friends, most notably from the Foundation for Thymic<br />
Cancer Research.<br />
However, one advantage for a rare disease is that a<br />
relatively small amount <strong>of</strong> funding can have a substantial<br />
effect precisely because it is a rare disease. This limits the<br />
Author’s Disclosures <strong>of</strong> Potential Conflicts <strong>of</strong> Interest<br />
size and cost <strong>of</strong> the infrastructure needed and promotes<br />
willingness on a broad front for people to volunteer a<br />
manageable amount <strong>of</strong> time. Maintaining such willingness<br />
broadly and consistently remains a challenge. ITMIG members<br />
are engaged and committed, in part because it is easier<br />
to have a feeling <strong>of</strong> ownership in a smaller group and<br />
because the effort seems to produce real progress.<br />
Conclusion<br />
Employment or<br />
Leadership Consultant or Stock<br />
Author<br />
Positions Advisory Role Ownership Honoraria<br />
Frank C. Detterbeck I-Flow<br />
1. Mariusdottir E, Nikulasson S, Bjornsson J, et al. Thymic epithelial<br />
tumours in Iceland: incidence and histopathology, a population-based study.<br />
APMIS. 2010;118:927-933.<br />
2. Engel P, Marx A, Müller-Hermelink HK. Thymic tumours in Denmark.<br />
A retrospective study <strong>of</strong> 213 cases from 1970-1993. Pathol Res Pract. 1999;<br />
195:565-570.<br />
3. Engels EA, Pfeiffer RM. Malignant thymoma in the United States:<br />
demographic patterns in incidence and associations with subsequent malignancies.<br />
Int J Cancer. 2003;105:546-551.<br />
4. dos Santos Silva I, Swerdlow AJ. Thymus cancer epidemiology in<br />
England and Wales. Br J Cancer. 1990;61:899-902.<br />
5. Schwartz A, Kostun L, Henson D. Thymoma and thymic carcinomas: an<br />
analysis <strong>of</strong> 2,189 cases from the SEER database. Out 4. International Thymic<br />
Malignancy Interest Group (ITMIG) 2 nd Annual Conference. Amsterdam, The<br />
Netherlands. July 7-8, 2011.<br />
474<br />
REFERENCES<br />
Thymic malignancies and other mediastinal tumors represent<br />
rare diseases, in which there has been little progress<br />
over many decades. For a rare disease, it is clear that<br />
progress is only possible if international collaboration can be<br />
achieved. ITMIG represents an organization that is devoted<br />
to making progress in the management <strong>of</strong> rare diseases<br />
through international collaboration. ITMIG has built an<br />
infrastructure, has engaged a broad multidisciplinary group<br />
<strong>of</strong> people in a global initiative, and sought novel approaches<br />
to maximize the progress that can be made in improving<br />
outcomes for patients with these orphan diseases.<br />
Research<br />
Funding<br />
Expert<br />
Testimony<br />
FRANK C. DETTERBECK<br />
Other<br />
Remuneration<br />
6. Detterbeck F, Morgensztern D. Progress in outcomes for patients with<br />
Thymoma International Thymic Malignancy Interest Group (ITMIG) 2 nd<br />
Annual Conference. Amsterdam, The Netherlands: July 7-8, 2011 (abstr<br />
02.01).<br />
7. Lilford RJ, Thornton JG, Braunholtz D. <strong>Clinical</strong> trials and rare diseases:<br />
A way out <strong>of</strong> a conundrum. BMJ. 1995;311:1621-1625.<br />
8. Detterbeck FC, Parsons AM. Thymic tumors. Ann Thorac Surg. 2004;<br />
77:1860-1869.<br />
9. Huang J, Wang Z, Loehrer P, et al. Standard Outcome Measures for<br />
Thymic Malignancies. J Thorac Oncol. 2010;5:2017-2023.<br />
10. Detterbeck F. ITMIG: A Way Forward. J Thorac Oncol. 2010;5:s365s370<br />
(suppl 4).<br />
11. Gonen M. Bias, Biostatistics & Prognostic Factors. J Thorac Oncol.<br />
2011;6:s1705-s1709 (suppl 3).
Thymoma: From Chemotherapy to<br />
Targeted Therapy<br />
Overview: Thymic malignancies are rare epithelial tumors<br />
that may be aggressive and difficult to treat. Thymomas are<br />
frequently eligible for upfront surgical resection. However,<br />
nearly 30% <strong>of</strong> patients present with locally advanced tumor at<br />
time <strong>of</strong> diagnosis, and chemotherapy is then used to reduce<br />
the tumor burden—possibly allowing subsequent surgery<br />
THYMIC MALIGNANCIES represent a wide range <strong>of</strong><br />
clinical, histologic, and radiologic entities, which may<br />
be aggressive and difficult to treat. 1-3 The current histopathologic<br />
classification distinguishes thymomas from thymic<br />
carcinomas; thymomas are further subdivided into<br />
different types (types A, AB, B1, B2, and B3), according to<br />
the morphology <strong>of</strong> epithelial tumor cells (with an increasing<br />
degree <strong>of</strong> atypia from type A to type B3), the relative<br />
proportion <strong>of</strong> the nontumoral lymphocytic component (decreasing<br />
from types B1 to B3), and the resemblance to<br />
normal thymic architecture. 2 More than 25% <strong>of</strong> thymomas<br />
actually exhibit morphologic heterogeneity, whereas 10% to<br />
15% combine different histologic types. Tumor invasiveness,<br />
as evaluated by the Masaoka staging system, is a major<br />
predictor <strong>of</strong> outcome. 4,5<br />
Surgery is the mainstay <strong>of</strong> the curative-intent treatment<br />
<strong>of</strong> thymic tumors, 1 and complete resection represents the<br />
most noteworthy favorable prognostic factor. 1,5 Contrary to<br />
thymic carcinomas, recurrences after complete surgical<br />
resection <strong>of</strong> thymomas are rare and mostly occur locoregionally,<br />
5 which limits the rationale for postoperative chemotherapy.<br />
Nearly 30% <strong>of</strong> patients present with locally<br />
advanced tumor at time <strong>of</strong> diagnosis, with invasion <strong>of</strong><br />
intrathoracic neighboring structures, and/or dissemination<br />
to the pleura and the pericardium. In such cases, primary<br />
chemotherapy has been used both to reduce the tumor<br />
burden—possibly allowing subsequent surgery and/or radiotherapy—and<br />
to achieve prolonged disease control. 1,6<br />
Finally, chemotherapy is also the palliative-intent treatment<br />
for unresectable, metastatic, and recurrent thymic<br />
tumors. 1,6<br />
As a consequence <strong>of</strong> the rarity <strong>of</strong> thymomas, our knowledge<br />
regarding chemotherapy in this setting has mainly<br />
been based on retrospective series, most <strong>of</strong> which were<br />
published years ago. Only a few prospective trials have been<br />
conducted. Most studies included both thymomas and thymic<br />
carcinomas, and did not report detailed results by<br />
histologic type. Taken together, these studies, despite recruiting<br />
limited numbers <strong>of</strong> patients over extended period <strong>of</strong><br />
times and being heterogeneous with regard to patient selection<br />
criteria, therapeutic sequence, and intent <strong>of</strong> the treatment,<br />
demonstrated the chemosensitivity <strong>of</strong> thymoma to<br />
various cytotoxic agent combinations. 1,6 Interpretation <strong>of</strong><br />
outcomes data, especially overall survival and response<br />
rates, should integrate 1) the fact that only half <strong>of</strong> patients<br />
with thymoma actually die as a result <strong>of</strong> tumor progression,<br />
whereas 25% <strong>of</strong> deaths are related to thymoma-associated<br />
immunologic manifestations, including myasthenia 7 ;2)the<br />
correlation between histology and stage, considering that<br />
type A to B1 thymomas more frequently present as stage I to<br />
By Nicolas Girard, MD, PhD<br />
and/or radiotherapy. Metastatic and recurrent thymic malignancies<br />
may be similarly treated with chemotherapy. More<br />
recently, the molecular characterization <strong>of</strong> thymoma led to the<br />
identification <strong>of</strong> potentially druggable targets, laying the foundation<br />
to implement personalized medicine for patients.<br />
II disease, whereas type B2 to B3 thymomas are usually<br />
diagnosed as stage III to IV disease 8 ; this specific feature<br />
may potentially confound the prognostic or predictive value<br />
<strong>of</strong> these variables; and 3) the potential effect on lymphocytic<br />
thymomas (types AB, B1, and B2) <strong>of</strong> corticosteroids, which<br />
are usually delivered concurrently with chemotherapy, and<br />
may produce a substantial reduction <strong>of</strong> lesion size at imaging<br />
studies through lymphocytic depletion, with no antitumor<br />
effect. 9<br />
Novel strategies are still needed, especially for type B3<br />
thymomas, which, similarly to thymic carcinomas, carry<br />
a poor prognosis despite multimodal treatment. 5 In the<br />
past, insights in the biology <strong>of</strong> thymic tumors were originally<br />
made after anecdotal clinical responses to targeted<br />
therapies, 3 and substantial efforts were subsequently conducted<br />
to dissect the molecular pathways involved in carcinogenesis.<br />
3,10-13 Research is hampered by the rarity <strong>of</strong><br />
these tumors, evolution <strong>of</strong> histopathologic concepts, and a<br />
lack <strong>of</strong> established cell lines and animal models. However,<br />
these studies led to the identification <strong>of</strong> potentially druggable<br />
targets, laying the foundations to implement personalized<br />
medicine in the field.<br />
Chemotherapy for Thymoma<br />
Curative-Intent, Preoperative Chemotherapy<br />
In locally advanced thymic malignancies (i.e., unresectable<br />
Masaoka stage III and IVA disease at time <strong>of</strong> diagnosis),<br />
chemotherapy aims at making feasible subsequent R0<br />
resection to achieve long-term survival. 1,6 Several chemotherapy<br />
regimens have been used in this setting, mostly<br />
consisting <strong>of</strong> doxorubicin- and/or platinum-based multiagent<br />
combinations (Table 1). Usually two to four cycles <strong>of</strong> chemotherapy<br />
are administered before imaging reassessment. In<br />
this setting, response rates to chemotherapy ranged from<br />
70% to 80% in the largest studies (Table 1). Patients for<br />
whom R0 resection was thought to be feasible underwent<br />
surgery, and complete resection was achieved in approximately<br />
50% <strong>of</strong> cases (Table 1).<br />
In those studies, when the patient was not deemed to be a<br />
From the Department <strong>of</strong> Respiratory Medicine, Pilot Unit for the Management <strong>of</strong> Rare<br />
Intrathoracic Tumors, National Expert Center for Thymic Malignancies, Louis Pradel<br />
Hospital, Hospices Civils de Lyon; and UMR 754 “Retrovirus and Compared Pathology,”<br />
Claude Bernard University, Lyon, France.<br />
Author’s disclosures <strong>of</strong> potential conflicts <strong>of</strong> interest are found at the end <strong>of</strong> this article.<br />
Address reprint requests to Nicolas Girard, Service de Pneumologie, Hôpital Louis<br />
Pradel, 28, Avenue Doyen Lépine, 69677 Lyon (Bron), France; email: nicolas.girard@chulyon.fr.<br />
© <strong>2012</strong> by <strong>American</strong> <strong>Society</strong> <strong>of</strong> <strong>Clinical</strong> <strong>Oncology</strong>.<br />
1092-9118/10/1-10<br />
475
Table 1. Selected Studies Reporting on Preoperative Chemotherapy or Chemoradiation for Locally Advanced Thymic Tumors<br />
Study<br />
Primary<br />
Chemotherapy<br />
Regimen<br />
No. <strong>of</strong><br />
Patients<br />
Tumor<br />
Type Stage<br />
surgical candidate—either because R0 resection was not<br />
thought to be achievable or because <strong>of</strong> poor performance<br />
status or coexistent medical condition—curative-intent, definitive<br />
radiotherapy was delivered. Loehrer and colleagues<br />
reported a phase II trial including 23 patients with thymoma<br />
and treated with cyclophosphamide, doxorubicin, and<br />
cisplatin (CAP) chemotherapy followed by radiotherapy to a<br />
total dose <strong>of</strong> 54 Gy. 25 Five-year survival was 53%.<br />
Of note, two studies including patients with locally advanced<br />
thymic tumors reported on the use <strong>of</strong> primary<br />
chemotherapy associated with sequential or concurrent ra-<br />
Design<br />
Response<br />
Rate (%)<br />
Any<br />
Surgery<br />
diotherapy in a preoperative intent (Table 1). Available<br />
retrospective data do not provide interpretable results comparing<br />
chemotherapy with chemoradiotherapy in the preoperative<br />
setting. A randomized phase II trial is currently<br />
ongoing, evaluating cisplatin and etoposide chemotherapy<br />
combined with conformal or intensity-modulated radiotherapy<br />
as primary treatment <strong>of</strong> locally advanced thymoma and<br />
thymic carcinoma (clinicaltrials.gov ID: NCT00387868).<br />
After primary chemotherapy, if radiotherapy was not<br />
feasible because <strong>of</strong> a large tumor burden that precluded safe<br />
delivery <strong>of</strong> appropriate doses or because <strong>of</strong> comorbidities<br />
increasing the risks <strong>of</strong> radiation-induced toxicity, treatment<br />
consisted <strong>of</strong> chemotherapy alone in a strategy that may<br />
ultimately be considered palliative. In the reported literature,<br />
approximately 15% to 20% <strong>of</strong> patients with locally<br />
advanced thymic tumors receiving upfront chemotherapy<br />
did not receive either surgery or radiation therapy or other<br />
local treatment (Table 1). Survival <strong>of</strong> these patients was<br />
limited.<br />
Consolidation Chemotherapy<br />
Surgery<br />
Subsequent Treatment (%)<br />
Complete<br />
Resection<br />
Radiotherapy None<br />
Chemotherapy<br />
Macchiarini et al 1991 14 CEE 7 T/TC III Phase II 100 100 57 0 0<br />
Berruti et al 1993 15 ADOC 6 T III-IVA Phase II 83 NA 17 NA NA<br />
Rea et al 1993 16 ADOC 16 T III-IVA Retrosp 100 100 69 0 0<br />
Berruti et al 1999 17 ADOC 16 T III-IVA Phase II 81 56 56 31 13<br />
Venuta et al 2003 18 CEE 15 T/TC III Retrosp 66 100 NA NA NS<br />
Bretti et al 2004 19 ADOC/PE 25 T/TC III-IVA Retrosp 72 68 44 NA NA<br />
Kim et al 2004 20 CAPP 22 T Phase II 77 100 72 0 0<br />
Lucchi et al 2005 21 CEE 36 T/TC III-IVA Retrosp 67 69 78 19 3<br />
Jacot et al 2005 22 CAP 5 T/TC III-IVA Retrosp 75 38 25 50 12<br />
Yokoi et al 2007 23 CAMP 14 T/TC III, IV Retrosp 93 64 14 14 21<br />
Kunitoh et al 2009 24 CODE 21 T III Phase II 62 62 43 24 14<br />
Chemoradiation<br />
Loehrer et al 199725 CAP/54 Gy 23 T/TC III-IVA Phase II 70 0 0 0 100<br />
Wright et al 200826 PE, ADOC, CAP,<br />
CEE/45–60 Gy<br />
10 T/TC III-IVA Retrosp 40 100 80 0 0<br />
Abbreviations: T, thymoma; TC, thymic carcinoma; Retrosp, retrospective; CAP, cisplatin (50 mg/m 2 /3 weeks), doxorubicin (50 mg/m 2 /3 weeks), cyclophosphamide<br />
(500 mg/m 2 /3 weeks); ADOC, doxorubicin (40 mg/m 2 /3 weeks), cisplatin (50 mg/m 2 /3 weeks), vincristine (0.6 mg/m 2 /3 weeks), cyclophosphamide (700 mg/m 2 /3 weeks);<br />
PE, cisplatin (60 mg/m 2 /3 weeks), etoposide (120 mg/m 2 � 3/3 weeks); CODE, cisplatin (25 mg/m 2 /week), vincristine (1 mg/m 2 /week), doxorubicin (40 mg/m 2 /week),<br />
etoposide (80 mg/m 2 � 3 days/week); CEE, cisplatin (75 mg/m 2 /3 weeks), epirubicin (100 mg/m 2 /3 weeks), etoposide (120 mg/m 2 � 3 days/3weeks); CAMP, CAP plus<br />
prednisolone (1000 mg/m 2 � 4 days, 500 mg/m 2 � 2 days/3 weeks); NA, not available.<br />
KEY POINTS<br />
● Surgery is the mainstay <strong>of</strong> the treatment <strong>of</strong> thymic<br />
malignancies. Chemotherapy is recommended in unresectable,<br />
locally advanced, and metastatic tumors.<br />
● Thymomas are chemosensitive. Major regimens are<br />
based on doxorubicin and cisplatin. No randomized<br />
trial is available that compares different cytotoxic<br />
agent combinations.<br />
● Preoperative chemotherapy is recommended for unresectable,<br />
locally advanced thymomas, aiming at<br />
allowing subsequent R0 resection, which is the major<br />
predictor <strong>of</strong> long-term survival.<br />
● Adjuvant chemotherapy is not recommended for thymomas,<br />
but may be used for thymic carcinomas.<br />
● Targeted therapies have been developed empirically<br />
in refractory thymomas, with limited rationale and<br />
poor patient selection. The use <strong>of</strong> targeted agents in<br />
thymic tumors is currently investigational, not a<br />
routine practice, because other options may exist for<br />
refractory tumors.<br />
476<br />
NICOLAS GIRARD<br />
Consolidation chemotherapy refers to chemotherapy delivered<br />
after multimodal, curative-intent treatment, aiming<br />
at treating possible residual microscopic disease after surgery.<br />
This strategy has been reported by investigators<br />
from the M. D. Anderson Cancer Center (Houston, TX). 20<br />
Patients with stage III to IV thymoma received upfront<br />
chemotherapy with three cycles <strong>of</strong> CAPP (CAP plus prednisone),<br />
followed by surgery and adjuvant radiotherapy, followed<br />
by consolidation chemotherapy with three cycles <strong>of</strong><br />
CAPP. Contrary to adjuvant chemoradiation, which usually<br />
consists <strong>of</strong> chemotherapy followed by radiotherapy, consolidation<br />
chemotherapy was delivered after adjuvant radiotherapy.<br />
The role <strong>of</strong> consolidation chemotherapy within such<br />
multimodal strategy has not been specifically evaluated.
SYSTEMIC TREATMENT OF THYMOMA<br />
Study<br />
Table 2. Landmark Studies Reporting on Palliative Chemotherapy Regimens in Advanced Thymic Malignancies<br />
No. <strong>of</strong><br />
Patients<br />
Period <strong>of</strong><br />
Accrual<br />
(years)<br />
Palliative-Intent, Definitive Chemotherapy<br />
Palliative chemotherapy is administered as the sole treatment<br />
modality for thymic tumors, with no plan for surgery<br />
or radiotherapy (e.g., in patients with metastatic or recurrent<br />
disease, or with locally advanced tumors that did not<br />
sufficiently shrink after primary chemotherapy to be eligible<br />
for subsequent focal treatment). 6 The main objectives <strong>of</strong><br />
palliative-intent chemotherapy are to improve potential<br />
tumor-related symptoms and to achieve tumor response.<br />
Prolonged disease control is possible, but tumor eradication<br />
is not expected. Several studies—both prospective and retrospective—described<br />
several regimens for definitive chemotherapy<br />
(Table 2), but because there are no randomized<br />
studies, it is unclear which cytotoxic agents are best; multiagent<br />
combination regimens and anthracycline-based regimens<br />
appear to have improved response rates compared<br />
with others. In general, a combination regimen is recommended,<br />
for at least three and no more than six cycles.<br />
Overall, response rates are 20% to 40%, lower than those<br />
observed in the preoperative setting.<br />
In the palliative-intent setting, several consecutive lines<br />
<strong>of</strong> chemotherapy may be administered when patients present<br />
with tumor progression. Delay <strong>of</strong> progression ranges<br />
from 4 to 80 months in the literature. It is estimated that<br />
50% to 70% <strong>of</strong> patients with thymoma recurrence receive<br />
chemotherapy, while some recurrences may be eligible for<br />
surgery and/or radiotherapy. 7 Strategy may consist <strong>of</strong> the<br />
readministration <strong>of</strong> a previously effective regimen, 37 as well<br />
as the use <strong>of</strong> less toxic agents, including paclitaxel or<br />
pemetrexed (Table 2). The repeated use <strong>of</strong> anthracyclines is<br />
limited by potential cardiac toxicity, which may be even<br />
more relevant in the recurrence setting, in case <strong>of</strong> combination<br />
with radiotherapy and surgery, and given the possible<br />
development <strong>of</strong> paraneoplastic myocarditis. Specific to recur-<br />
Tumor<br />
Type Design Regimen Agents Doses<br />
Single-agent chemotherapy<br />
Bonomi et al 1993 27 21 4 T/TC Phase II Cisplatin 50 mg/m 2 /3 weeks 10<br />
Highley et al 1999 28 15 12 T/TC Retrosp Ifosfamide 1.5g/m 2 � 5 days/3 weeks 46<br />
Loehrer et al 2006 29 27 1 T/TC Phase II Pemetrexed 500 mg/m 2 /3 weeks 17<br />
Combination chemotherapy<br />
Fornasiero et al 1990 30 32 11 T Retrosp ADOC Doxorubicin 40 mg/m 2 /3 weeks 91<br />
Cisplatin 50 mg/m 2 /3 weeks<br />
Vincristin 0.6 mg/m 2 /3 weeks<br />
Cyclophosphamide 700 mg/m 2 /3 weeks<br />
Loehrer et al 1994 31 30 9 T/TC Phase II CAP Cisplatin 50 mg/m 2 /3 weeks 51<br />
Doxorubicin 50 mg/m 2 /3 weeks<br />
Cyclophosphamide 500 mg/m 2 /3 weeks<br />
Giaccone et al 1996 32 16 6 T Phase II PE Cisplatin 60 mg/m 2 /3 weeks 56<br />
Etoposide 120 mg/m 2 � 3/3 weeks<br />
Loehrer et al 2001 33 34 2 T/TC Phase II VIP Etoposide 75 mg/m 2 � 4 days/3 weeks 32<br />
Ifosfamide 1.2 g/m 2 � 4 days/3 weeks<br />
Cisplatin 20 mg/m 2 � 4 days/3 weeks<br />
Lemma et al 2011 34 46 7 T/TC Phase II Carbo-Px Carboplatin AUC 5/3 weeks 43<br />
Paclitaxel 225 mg/m 2 /3 weeks<br />
Palmieri et al 2011 35 15 3 T/TC Phase II CAP-GEM Capecitabine 650 mg/m 2 bid � 14 days/3 weeks 40<br />
Gemcitabine 1000 mg/m 2 � 2 days/3 weeks<br />
Okuma et al 2011 36 9 8 TC Retrosp Cisplatin-Irinotecan Cisplatin 80 mg/m 2 /4 weeks 56<br />
Irinotecan 60 mg/m 2 � 3 days/4 weeks<br />
Abbreviations: T, thymoma; TC, thymic carcinoma; Retrosp, retrospective.<br />
rent thymic tumors after first course <strong>of</strong> treatment with definitive<br />
chemotherapy is the use <strong>of</strong> octreotide, which as single<br />
agent produced objective tumor responses rates ranging from<br />
10% to 37% in tumors showing increased uptake at OctreoScan<br />
(Indium-111 pentetreotide; Covidien; Dublin, Ireland) scintigraphy.<br />
38 Of note, objective responses to octreotide have<br />
been reported only in thymomas, not in thymic carcinomas.<br />
Overall, given the modest results <strong>of</strong> chemotherapy in the<br />
palliative-intent setting, novel strategies are needed. In this<br />
way, amrubicin, a new generation anthracycline, is currently<br />
investigated in refractory thymic tumors (clinicaltrials.gov<br />
ID: NCT01364727).<br />
Targeted Therapy for Thymic Malignancies<br />
Response<br />
Rate (%)<br />
Despite the rarity <strong>of</strong> thymic tumors, substantial efforts<br />
have been made to dissect the molecular pathways involved<br />
in thymus carcinogenesis. 3,10-13 The main signaling pathways<br />
with potential druggable targets that have been explored<br />
in thymic tumors are the epidermal growth factor<br />
receptor (EGFR), the KIT/mast/stem-cell growth factor receptor<br />
(KIT/SCFR), and the insulin-like growth factor-1<br />
receptor (IGF-1R) pathways. Antiangiogenic agents may<br />
also be <strong>of</strong> interest.<br />
EGFR Pathway Inhibitors<br />
Overall, EGFR is overexpressed in approximately 70% <strong>of</strong><br />
thymomas and 50% <strong>of</strong> thymic carcinomas. 3,10,11 EGFR expression,<br />
as well as EGFR gene amplification, was shown to<br />
be higher in stage III to IV tumors. However, EGFR mutations<br />
are rare in thymic malignancies. 3,10 Thus far, only two<br />
cases <strong>of</strong> thymomas harboring EGFR mutations have been<br />
found <strong>of</strong> a total <strong>of</strong> 158 tumors collectively analyzed. The<br />
mutations were L858R in one case and G863D in the other<br />
case, both <strong>of</strong> which are associated with response to EGFR<br />
477
tyrosine kinase inhibitors in lung cancer. 3,39 Besides EGFR,<br />
no mutation has been identified in other genes <strong>of</strong> the<br />
pathway: PIK3CA, AKT1, ERBB2, MEK1, and PTEN. 10 In<br />
the Memorial Sloan-Kettering Cancer Center (New York,<br />
NY) series, RAS mutations were observed in two thymomas:<br />
one G12A KRAS mutation and one G13V HRAS mutation. 10<br />
These mutations are associated with primary resistance to<br />
EGFR inhibitors.<br />
From a clinical standpoint, the low frequency <strong>of</strong> EGFRactivating<br />
mutations in thymic tumors may explain why<br />
responses to EGFR tyrosine kinase inhibitors have rarely<br />
been observed. 3,40,41 On the contrary, several observations<br />
<strong>of</strong> heavily pretreated recurrent thymoma exhibiting partial<br />
response to cetuximab have been reported. 42,43 All tumors<br />
harbored strong EGFR expression by immunohistochemistry.<br />
A phase II trial that evaluates the feasibility <strong>of</strong> delivering<br />
cetuximab in combination with CAP in unresectable<br />
locally advanced thymomas (clinicaltrails.gov ID: NCT-<br />
01025089) is ongoing.<br />
KIT and Angiogenesis Inhibitors<br />
KIT is a transmembrane growth factor with tyrosine<br />
kinase activity. A recent pooled analysis <strong>of</strong> data reported in<br />
the literature indicates that collectively, KIT is overexpressed<br />
in 2% <strong>of</strong> thymomas and 79% <strong>of</strong> thymic carcinomas. 3<br />
KIT gene mutations have been found exclusively in thymic<br />
carcinomas, predicting the efficacy <strong>of</strong> specific inhibitors,<br />
including imatinib, sunitinib, sorafenib, and dasatinib. 3 Interestingly,<br />
most <strong>of</strong> these agents also potently inhibit other<br />
kinases, including vascular endothelial growth factor<br />
(VEGF) receptors (VEGFRs), platelet-derived growth factor<br />
receptor (PDGFR), and Src. The effect <strong>of</strong> these drugs, especially<br />
in KIT-wild-type thymic carcinomas and in thymomas,<br />
may then be partially related to <strong>of</strong>f-target effects, especially<br />
angiogenesis inhibition, 44-46 because VEGF-A and VEGFR-1<br />
and -2 are overexpressed by epithelial tumor cells. 47<br />
However, only sparse data are available regarding the use<br />
<strong>of</strong> angiogenesis inhibitors in thymic malignancies. In a<br />
phase II trial, bevacizumab was tested in combination with<br />
erlotinib in 11 thymomas and seven thymic carcinomas. 41<br />
No tumor response was observed, but stable disease rate<br />
was 60%. Motesanib diphosphate was reported to induce<br />
partial response in a refractory thymoma. 46 In a phase I<br />
study combining docetaxel with aflibercept, one patient with<br />
thymoma experienced partial response. 48 Interestingly, de-<br />
Author’s Disclosures <strong>of</strong> Potential Conflicts <strong>of</strong> Interest<br />
Author<br />
Nicolas Girard*<br />
*No relevant relationships to disclose.<br />
Employment or<br />
Leadership<br />
Positions<br />
Consultant or<br />
Advisory Role<br />
1. Girard N, Mornex F, Van Houtte P, et al. Thymoma: a focus on current<br />
therapeutic management. J Thorac Oncol. 2009;4:119-126.<br />
2. WHO histological classification <strong>of</strong> tumours <strong>of</strong> the thymus. In: Travis WB,<br />
Brambilla A, Muller-Hermelinck HK, et al. World Health Organization<br />
Classification <strong>of</strong> Tumour:. Pathology and Genetics <strong>of</strong> Tumours <strong>of</strong> the Lung,<br />
Pleura, Thymus and Heart. Lyon: IARC Press, 2004;146.<br />
3. Girard N. Thymic tumors: relevant molecular data in the clinic. J Thorac<br />
Oncol. 2010;5:S291-S295.<br />
4. Detterbeck FC, Nicholson AG, Kondo K, et al. The Masaoka-Koga stage<br />
478<br />
spite the large tumor burden <strong>of</strong> thymic tumors and the<br />
frequent abutment to mediastinal vascular structures, no<br />
hemorrhagic adverse effect has been reported with the use <strong>of</strong><br />
these drugs in these observations.<br />
IGF-1R Pathway Inhibitors<br />
IGF-1R is a transmembrane receptor with tyrosine kinase<br />
activity that was reported by immunohistochemistry to be<br />
overexpressed in thymic tumors. 3,11,13 Moderate to high<br />
IGF-1R expression was more frequent in thymic carcinomas<br />
than in thymomas (70% vs. 21%, respectively, p � 0.001).<br />
<strong>Clinical</strong>ly, figitumumab, an anti-IGF1-R antibody, showed<br />
clinical activity in a patient with refractory thymoma. 49 A<br />
phase II trial is ongoing evaluating IMC-A12, another anti-<br />
IGF-1R antibody, in advanced and refractory thymomas and<br />
thymic carcinomas (clinicaltrials.gov ID: NCT00965250). In<br />
vitro data in a thymoma cell line suggest that IGF-1R<br />
overexpression may also be targeted by heat-shock protein<br />
90 chaperone inhibitors. 13<br />
Other Targeted Therapies<br />
Belinostat, a histone deacetylase inhibitor, was evaluated<br />
in thymic malignancies in a recently completed phase II trial<br />
enrolling 41 patients (25 thymomas and 16 thymic carcinomas).<br />
50 Response and 2-year survival rates were 8% and<br />
77%, respectively in thymomas, but the drug had no effect<br />
in thymic carcinomas. A subsequent phase II trial is evaluating<br />
combination <strong>of</strong> belinostat with CAP as first-line<br />
treatment for unresectable tumors (clinicaltrials.gov ID:<br />
NCT01100944).<br />
Cyclin-dependent kinase (CDK) proteins, which control<br />
the cell cycle G1-S transition, are altered through p16INK4<br />
loss in thymic tumors, related to a gene methylation mechanism.<br />
51 A phase II trial with a CDK inhibitor, PHA-<br />
848125AC, in advanced thymoma (clinicaltrials.gov IDs:<br />
NCT01301391) is currently ongoing.<br />
For routine practice, the use <strong>of</strong> targeted agents in thymoma<br />
is currently investigational because other options<br />
may exist for refractory tumors. Research efforts are currently<br />
being conducted to dissect the molecular biology <strong>of</strong><br />
thymic malignancies, and to better understand and predict<br />
the efficacy <strong>of</strong> chemotherapy and targeted agents. Promising<br />
results are emerging. Given the rarity <strong>of</strong> these tumors,<br />
translation <strong>of</strong> preclinical findings to the clinic may be quick<br />
and represents one <strong>of</strong> the most promising therapeutic approaches<br />
for advanced-stage thymoma.<br />
Stock<br />
Ownership Honoraria<br />
REFERENCES<br />
Research<br />
Funding<br />
Expert<br />
Testimony<br />
NICOLAS GIRARD<br />
Other<br />
Remuneration<br />
classification for thymic malignancies: clarification and definition <strong>of</strong> terms.<br />
J Thorac Oncol. 2011;6:S1710-1716.<br />
5. Kondo K, Monden Y. Therapy for thymic epithelial tumors: a clinical<br />
study <strong>of</strong> 1,320 patients from Japan. Ann Thorac Surg. 2003;76:878-884.<br />
6. Girard N, Lal R, Wakelee H, et al. Chemotherapy definitions and policies<br />
for thymic malignancies. J Thorac Oncol. 2011;6:S1749-S1755.<br />
7. Huang J, Detterbeck FC, Wang Z, et al. Standard outcome measures for<br />
thymic malignancies. J Thorac Oncol. 2011;6:S1691-S1697.<br />
8. Okumura M, Miyoshi S, Fujii Y, et al. <strong>Clinical</strong> and functional signifi-
SYSTEMIC TREATMENT OF THYMOMA<br />
cance <strong>of</strong> WHO classification on human thymic epithelial neoplasms: a study <strong>of</strong><br />
146 consecutive tumors. Am J Surg Pathol. 2001;25:103-110.<br />
9. Kirkove C, Berghmans J, Noel H, et al. Dramatic response <strong>of</strong> recurrent<br />
invasive thymoma to high doses <strong>of</strong> corticosteroids. Clin Oncol. 1992;4:64-66.<br />
10. Girard N, Shen R, Guo T, et al. Comprehensive genomic analysis<br />
reveals clinically relevant molecular distinctions between thymic carcinomas<br />
and thymomas. Clin Cancer Res. 2009;15:6790-6799.<br />
11. Girard N, Teruya-Feldstein J, Payabyab EC, et al. Insulin-like growth<br />
factor-1 receptor expression in thymic malignancies. J Thorac Oncol. 2010;5:<br />
1439-1446.<br />
12. Petrini I, Wang Y, Pineda M, et al. Complete genome sequencing <strong>of</strong> a<br />
human B3 thymoma. J Thorac Oncol. 2011;6:S452 (suppl).<br />
13. Breinig M, Mayer P, Harjung A, et al. Heat shock protein 90-sheltered<br />
overexpression <strong>of</strong> insulin-like growth factor 1 receptor contributes to malignancy<br />
<strong>of</strong> thymic epithelial tumors. Clin Cancer Res. 2011;17:2237-2249.<br />
14. Macchiarini P, Chella A, Ducci F, et al. Neoadjuvant chemotherapy,<br />
surgery, and postoperative radiation therapy for invasive thymoma. Cancer.<br />
1991;68:706-713.<br />
15. Berruti A, Borasio P, Roncari A, et al. Neoadjuvant chemotherapy with<br />
adriamycin, cisplatin, vincristine and cyclophosphamide (ADOC) in invasive<br />
thymomas: results in six patients. Ann Oncol. 1993;4:429-431.<br />
16. Rea F, Sartori F, Loy M, et al. Chemotherapy and operation for invasive<br />
thymoma. J Thorac Cardiovasc Surg. 1993;106:543-549.<br />
17. Berruti A, Borasio P, Gerbino A, et al. Primary chemotherapy with<br />
adriamycin, cisplatin, vincristine and cyclophosphamide in locally advanced<br />
thymomas: a single institution experience. Br J Cancer. 1999;81:841-845.<br />
18. Venuta F, Rendina EA, Longo F, et al. Long-term outcome after<br />
multimodality treatment for stage III thymic tumors. Ann Thorac Surg.<br />
2003;76:1866-1872.<br />
19. Bretti S, Berruti A, Loddo C, et al. Multimodal management <strong>of</strong> stages<br />
III-IVa malignant thymoma. Lung Cancer. 2004;44:69-77.<br />
20. Kim ES, Putnam JB, Komaki R, et al. Phase II study <strong>of</strong> a multidisciplinary<br />
approach with induction chemotherapy, followed by surgical resection,<br />
radiation therapy, and consolidation chemotherapy for unresectable<br />
malignant thymomas: final report. Lung Cancer. 2004;44:369-379.<br />
21. Lucchi M, Ambrogi MC, Duranti L, et al. Advanced stage thymomas<br />
and thymic carcinomas: results <strong>of</strong> multimodality treatments. Ann Thorac<br />
Surg. 2005;79:1840-1844.<br />
22. Jacot W, Quantin X, Valette S, et al. Multimodality treatment program<br />
in invasive thymic epithelial tumor. Am J Clin Oncol. 2005;28:5-7.<br />
23. Yokoi K, Matsuguma H, Nakahara R, et al. Multidisciplinary treatment<br />
for advanced invasive thymoma with cisplatin, doxorubicin, and methylprednisolone.<br />
J Thorac Oncol. 2007;2:73-78.<br />
24. Kunitoh H, Tamura T, Shibata T, et al. A phase II trial <strong>of</strong> dose-dense<br />
chemotherapy, followed by surgical resection and/or thoracic radiotherapy, in<br />
locally advanced thymoma: report <strong>of</strong> a Japan <strong>Clinical</strong> <strong>Oncology</strong> Group trial<br />
(JCOG 9606). Br J Cancer. 2010;103:6-11.<br />
25. Loehrer PJ Sr, Chen M, Kim K, et al. Cisplatin, doxorubicin, and<br />
cyclophosphamide plus thoracic radiation therapy for limited-stage unresectable<br />
thymoma: an intergroup trial. J Clin Oncol. 1997;15:3093-3099.<br />
26. Wright CD, Choi NC, Wain JC, et al. Induction chemoradiotherapy<br />
followed by resection for locally advanced Masaoka stage III and IVA thymic<br />
tumors. Ann Thorac Surg. 2008;85:385-389.<br />
27. Bonomi PD, Finkelstein D, Aisner S, et al. EST 2582 phase II trial <strong>of</strong><br />
cisplatin in metastatic or recurrent thymoma. Am J Clin Oncol. 1993;16:342-<br />
345.<br />
28. Highley MS, Underhill CR, Parnis FX, et al. Treatment <strong>of</strong> invasive<br />
thymoma with single-agent ifosfamide. J Clin Oncol. 1999;17:2737-2744.<br />
29. Loehrer PJ, Yiannoutsos CT, Dropcho S, et al. A phase II trial <strong>of</strong><br />
pemetrexed in patients with recurrent thymoma or thymic carcinoma. J Clin<br />
Oncol. 2006;24:7079.<br />
30. Fornasiero A, Daniele O, Ghiotto C, et al. Chemotherapy <strong>of</strong> invasive<br />
thymoma. J Clin Oncol. 1990;8:1419-1423.<br />
31. Loehrer PJ Sr, Kim K, Aisner SC, et al. Cisplatin plus doxorubicin plus<br />
cyclophosphamide in metastatic or recurrent thymoma: final results <strong>of</strong> an<br />
intergroup trial. The Eastern Cooperative <strong>Oncology</strong> Group, Southwest Oncol-<br />
ogy Group, and Southeastern Cancer Study Group. J Clin Oncol. 1994;12:<br />
1164-1168.<br />
32. Giaccone G, Ardizzoni A, Kirkpatrick A, et al. Cisplatin and etoposide<br />
combination chemotherapy for locally advanced or metastatic thymoma. A<br />
phase II study <strong>of</strong> the European Organization for Research and Treatment <strong>of</strong><br />
Cancer Lung Cancer Cooperative Group. J Clin Oncol. 1996;14:814-820.<br />
33. Loehrer PJ Sr, Jiroutek M, Aisner S, et al. Combined etoposide,<br />
ifosfamide, and cisplatin in the treatment <strong>of</strong> patients with advanced thymoma<br />
and thymic carcinoma: an intergroup trial. Cancer. 2001;91:2010-2015.<br />
34. Lemma GL, Lee JW, Aisner SC, et al. Phase II study <strong>of</strong> carboplatin and<br />
paclitaxel in advanced thymoma and thymic carcinoma. J Clin Oncol.<br />
2011;29:2060-2065.<br />
35. Palmieri G, Merola G, Federico P, et al. Preliminary results <strong>of</strong> phase II<br />
study <strong>of</strong> capecitabine and gemcitabine (CAP-GEM) in patients with metastatic<br />
pretreated thymic epithelial tumors (TETs). Ann Oncol. 2010;21:1168-<br />
1172.<br />
36. Okuma Y, Hosomi Y, Takagi Y, et al. Cisplatin and irinotecan combination<br />
chemotherapy for advanced thymic carcinoma: evaluation <strong>of</strong> efficacy<br />
and toxicity. Lung Cancer. 2011;74:492-496.<br />
37. Lara PN Jr, Bonomi PD, Faber LP. Retreatment <strong>of</strong> recurrent invasive<br />
thymoma with platinum, doxorubicin, and cyclophosphamide. Chest. 1996;<br />
110:1115-1117.<br />
38. Loehrer PJ Sr, Wang W, Johnson DH, et al. Octreotide alone or with<br />
prednisone in patients with advanced thymoma and thymic carcinoma: an<br />
Eastern Cooperative <strong>Oncology</strong> Group phase II trial. J Clin Oncol. 2004;22:<br />
293-299.<br />
39. Yoh K, Nishiwaki Y, Ishii G, et al. Mutational status <strong>of</strong> EGFR and KIT<br />
in thymoma and thymic carcinoma. Lung Cancer. 2008;62:316-320.<br />
40. Kurup A, Burns M, Dropcho S, et al. Phase II study <strong>of</strong> gefitinib<br />
treatment in advanced thymic malignancies. J Clin Oncol. 2005;23 (suppl;<br />
abstr 7068).<br />
41. Bedano PM, Perkins S, Burns M, et al. A phase II trial <strong>of</strong> erlotinib plus<br />
bevacizumab in patients with recurrent thymoma or thymic carcinoma. J Clin<br />
Oncol. 2008:26 (suppl; abstr 19087).<br />
42. Farina G, Garassino MC, Gambacorta M, et al. Response <strong>of</strong> thymoma to<br />
cetuximab. Lancet Oncol. 2007;8:449-450.<br />
43. Palmieri G, Marino M, Salvatore M, et al. Cetuximab is an active<br />
treatment <strong>of</strong> metastatic and chemorefractory thymoma. Front Biosci. 2007;<br />
12:757-761.<br />
44. Fiedler W, Giaccone G, Lasch P, et al. Phase I trial <strong>of</strong> SU14813 in<br />
patients with advanced solid malignancies. Ann Oncol. 2011;22:195-201.<br />
45. Ströbel P, Hohenberger P, Marx A. Thymoma and thymic carcinoma:<br />
molecular pathology and targeted therapy. J Thorac Oncol. 2010;5:S286-<br />
S290.<br />
46. Azad A, Herbertson RA, Pook D, et al. Motesanib diphosphate (AMG<br />
706), an oral angiogenesis inhibitor, demonstrates clinical efficacy in advanced<br />
thymoma. Acta Oncol. 2009;48:619-621.<br />
47. Cimpean AM, Raica M, Encica S, et al. Immunohistochemical expression<br />
<strong>of</strong> vascular endothelial growth factor A (VEGF), and its receptors<br />
(VEGFR1, 2) in normal and pathologic conditions <strong>of</strong> the human thymus. Ann<br />
Anat. 2008;190:238-245.<br />
48. Isambert N, Freyer G, Zanetta S, et al. A phase I dose escalation and<br />
pharmacokinetic (PK) study <strong>of</strong> intravenous aflibercept (VEGF trap) plus<br />
docetaxel (D) in patients (pts) with advanced solid tumors: preliminary<br />
results. J Clin Oncol. 2008;26:177s (suppl; abstr 3599).<br />
49. Haluska P, Shaw HM, Batzel GN, et al. Phase I dose escalation study<br />
<strong>of</strong> the anti insulin-like growth factor-I receptor monoclonal antibody CP-<br />
751,871 in patients with refractory solid tumors. Clin Cancer Res. 2007;13:<br />
5834-5840.<br />
50. Giaccone G, Rajan A, Berman A, et al. Phase II study <strong>of</strong> belinostat in<br />
patients with recurrent or refractory advanced thymic epithelial tumors.<br />
J Clin Oncol. 2011;29:2052-2059.<br />
51. Hirabayashi H, Fujii Y, Sakaguchi M, et al. p16INK4, pRB, p53 and<br />
cyclin D1 expression and hypermethylation <strong>of</strong> CDKN2 gene in thymoma and<br />
thymic carcinoma. Int J Cancer. 1997;73:639-644.<br />
479
CONTROVERSIES IN INDOLENT LYMPHOMA<br />
CHAIR<br />
Sonali M. Smith, MD<br />
The University <strong>of</strong> Chicago<br />
Chicago, IL<br />
SPEAKERS<br />
Gilles Salles, MD, PhD<br />
Hospices Civils de Lyon and Universite de Lyon<br />
Lyon, France<br />
Ajay K. Gopal, MD<br />
University <strong>of</strong> Washington<br />
Seattle, WA
What Is the Best Strategy for Incorporating<br />
New Agents into the Current Treatment <strong>of</strong><br />
Follicular Lymphoma?<br />
Overview: Although there is increasing knowledge about the<br />
pathobiology <strong>of</strong> follicular lymphoma (FL), the incorporation <strong>of</strong><br />
new agents is challenged by the long clinical course and<br />
inherent heterogeneity <strong>of</strong> the disease. Furthermore, a longstanding<br />
concept in FL is that although most patients have an<br />
indolent initial phase <strong>of</strong> disease, this is typically followed by<br />
sequentially shorter remission durations and justifies the<br />
continued intense search for new rationally designed agents.<br />
Ideally, there would be personalized prognostic tools, preemptive<br />
target identification, and means to predict response<br />
in individual patients. Short <strong>of</strong> having these tools, one conceptual<br />
approach is to consider FL as a series <strong>of</strong> clinical<br />
disease states divided between treatment-naïve (low tumor<br />
AS THE prototype <strong>of</strong> indolent lymphomas, follicular<br />
lymphoma (FL) mandates a longitudinal perspective<br />
when considering clinical management options. The median<br />
life expectancy has improved, with approximately 40% <strong>of</strong><br />
patients surviving in excess <strong>of</strong> 10 years. 1,2 During this time,<br />
patients can have variable clinical courses and, although<br />
clinical prognostic indices are helpful, they are <strong>of</strong>ten difficult<br />
to apply to individual patients outside <strong>of</strong> a research setting.<br />
Furthermore, treatment decisions made early in the disease<br />
may influence future options, and the optimal sequencing <strong>of</strong><br />
treatments is poorly defined. The traditional paradigm has<br />
been that with each successive therapy, the duration <strong>of</strong><br />
response shortens until a blatantly chemoresistant picture<br />
emerges. However, today a clearer understanding <strong>of</strong> FL<br />
biology has led to the emergence <strong>of</strong> a new repertoire <strong>of</strong><br />
agents with more rational and targeted mechanisms <strong>of</strong><br />
action. These agents challenge the notion <strong>of</strong> inevitably<br />
shorter response durations, and <strong>of</strong>fer hope <strong>of</strong> improved<br />
clinical outcomes compared with traditional sequential cytotoxic<br />
therapy.<br />
A major challenge to integrating new agents or new<br />
approaches is the inherent clinical and biologic heterogeneity<br />
<strong>of</strong> follicular lymphoma. Ideally, there would be personalized<br />
prognostic tools, preemptive target identification, and<br />
means to predict response in individual patients, followed by<br />
randomized clinical trials validating each agent. Despite<br />
major gains in dissecting the heterogeneity <strong>of</strong> FL (Table 1),<br />
there are few existing ways <strong>of</strong> predicting an individual<br />
patient’s overall disease course. One approach is to consider<br />
FL not simply as one disease but as sequential clinical<br />
disease states: treatment-naïve (low tumor burden and high<br />
tumor burden), relapsed (chemoimmunotherapy-sensitive),<br />
and multiply relapsed/chemoimmunotherapy resistant disease<br />
(Fig. 1). At each point, the best approach to incorporating<br />
new agents into the current treatment <strong>of</strong> FL requires an<br />
understanding <strong>of</strong> pathogenetic mechanisms, a consideration<br />
<strong>of</strong> the current and most promising agents, and conscientious<br />
trial design in the context <strong>of</strong> desired patient outcomes. This<br />
review will present a conceptual framework <strong>of</strong> clinical disease<br />
states in FL, providing examples along the way <strong>of</strong> how<br />
biologic insights and new agents are currently being integrated<br />
into existing treatment paradigms.<br />
By Sonali M. Smith, MD<br />
burden and high tumor burden), relapsed (typically still<br />
chemoimmunotherapy-sensitive), and multiply relapsed (usually<br />
chemoimmunotherapy-resistant) disease. By applying<br />
what is known about the biology <strong>of</strong> FL along with the available<br />
agents, new treatment options can be better defined and<br />
tested within these clinical contexts. During the last few<br />
years, novel chemotherapeutics, biologic agents, monoclonal<br />
antibodies, antibody drug conjugates, and maintenance strategies<br />
are all either replacing or adding onto existing strategies.<br />
These new agents and approaches challenge the notion<br />
<strong>of</strong> inevitably shorter response durations, and <strong>of</strong>fer hope <strong>of</strong><br />
improved clinical outcomes compared with traditional sequential<br />
cytotoxic therapy.<br />
Low–Tumor Burden FL (Treatment Naïve)<br />
The vast majority <strong>of</strong> patients with FL have an asymptomatic<br />
presentation, with either palpable or radiographically<br />
visible adenopathy discovered incidentally. At initial diagnosis,<br />
the pace <strong>of</strong> disease remains unknown for individual<br />
patients, and there are no tools to accurately predict disease<br />
behavior. Tumor grade is a rough measure <strong>of</strong> disease aggressiveness,<br />
with FL grade 3b behaving akin to diffuse large<br />
B-cell lymphoma. But for patients with FL grade 1 to 2,<br />
comprising the bulk <strong>of</strong> patients, grade itself is a poor<br />
predictor <strong>of</strong> outcome. The FL International Prognostic Indices<br />
(FLIPI 1 and 2) are helpful, but it is important to<br />
remember that all patients included in these retrospective<br />
analyses were already being considered for therapy and<br />
therefore do not adequately reflect a newly diagnosed,<br />
asymptomatic patient with low tumor burden. 3,4 Newer<br />
biologic markers may help identify patients at high risk for<br />
aggressive disease behavior, but are far from routine application.<br />
Despite these hurdles, low–tumor burden FL is an ideal<br />
setting in which to evaluate novel agents or novel application<br />
<strong>of</strong> existing agents. The National Lymphocare Study<br />
found that nearly one-fifth <strong>of</strong> patients undergo a “watch and<br />
wait” strategy, reflecting that no therapeutic intervention to<br />
date has shown a survival benefit. 5 Strong support for<br />
“watch and wait” can be derived from randomized trials that<br />
had shown no advantage to early therapy. 6,7 Notably, these<br />
studies were performed before the advent <strong>of</strong> monoclonal<br />
antibodies, and the recent randomized trial <strong>of</strong> “watch and<br />
wait” compared with a short course <strong>of</strong> the anti-CD20 monoclonal<br />
antibody rituximab followed by a maintenance strategy<br />
has reinvigorated the discussion regarding the timing <strong>of</strong><br />
treatment initiation in newly diagnosed, low–tumor burden<br />
From the Section <strong>of</strong> Hematology/<strong>Oncology</strong>, Lymphoma Program, The University <strong>of</strong><br />
Chicago, Chicago, IL.<br />
Author’s disclosures <strong>of</strong> potential conflicts <strong>of</strong> interest are found at the end <strong>of</strong> this article.<br />
Address reprint requests to Sonali M. Smith, MD, Section <strong>of</strong> Hematology/<strong>Oncology</strong>,<br />
Lymphoma Program, The University <strong>of</strong> Chicago, 5841 S. Maryland Ave., MC2115, Chicago,<br />
IL 60637; email: smsmith@medicine.bsd.uchicago.edu.<br />
© <strong>2012</strong> by <strong>American</strong> <strong>Society</strong> <strong>of</strong> <strong>Clinical</strong> <strong>Oncology</strong>.<br />
1092-9118/10/1-10<br />
481
KEY POINTS<br />
Table 1. <strong>Clinical</strong>, Biologic, and Genetic Factors with Prognostic Value in Follicular Lymphoma*<br />
Factor <strong>Clinical</strong> Impact Comment<br />
Histopathologic features<br />
Tumor grade Grade 1–2 is generally more indolent than grade 3 Reproducibility amongst pathologists can be low. Grade 3A is<br />
treated similarly to grade 1–2. Grade 3B is treated similarly<br />
to diffuse large B-cell lymphoma<br />
Diffuse architecture Controversial Increased areas <strong>of</strong> diffuse architecture may increase risk <strong>of</strong><br />
transformation<br />
Proliferation rate<br />
<strong>Clinical</strong> indices<br />
Controversial<br />
IPI Worsening survival with high scores Unable to distinguish prognosis amongst vast majority <strong>of</strong><br />
patients who have low IPI scores<br />
FLIPI 1 Worsening survival with high scores Analysis was done in a pre-rituximab population. Primary end<br />
point was overall survival.<br />
FLIPI 2<br />
Blood markers<br />
Worsening PFS with high scores Analysis was done in patients receiving immunochemotherapy.<br />
Primary end point was PFS.<br />
Lactate dehydrogenase Decreased survival with elevated values Part <strong>of</strong> IPI and FLIPI-1<br />
Beta-2 microglobulin Shorter FFP and decreased survival with elevated<br />
values<br />
Part <strong>of</strong> FLIPI-2<br />
VEGF, FGF, TNF, endostatin<br />
Individual proteins or RNA<br />
Shorter FL Small numbers <strong>of</strong> FL patients in most <strong>of</strong> these studies<br />
BCL6, CD10, PU.1 Favorable<br />
MCL1, MUM1, SOCS3, YY.1, BCLXL Gene expression pr<strong>of</strong>iling<br />
Unfavorable<br />
Immune response-1 and immune response-2 Immune response-1 has 9-fold more favorable<br />
survival<br />
81-gene predictor model<br />
Microenvironment<br />
Model was able to predict disease aggressiveness and disease<br />
progression but was done in pre-rituximab era.<br />
Macrophages Decreased survival<br />
CD4� or CD8� T cells Controversial Conflicting reports in literature<br />
T-regulatory cells<br />
Molecular genetic biomarkers<br />
Controversial Conflicting reports in literature<br />
BCL2 Typically diagnostic and not prognostic. There may be differing prognostic implications <strong>of</strong> translocation<br />
versus somatic mutations as source <strong>of</strong> BCL aberration.<br />
MYC Worse prognosis Usually associated with transformation to high grade lymphoma.<br />
TP53 Worse prognosis Associated with histologic transformation.<br />
CCNB1 Improved survival<br />
Abbreviations: FGF, fibroblast growth factor; FL, follicular lymphoma; FLIPI, Follicular Lymphoma International Prognostic Index; IPI, International Prognostic Index;<br />
PFS, progression-free survival; TNF, tumor necrosis factor; VEGF, vascular endothelial growth factor.<br />
* There is little consensus on the prognostic applicability <strong>of</strong> these biomarkers on an individual basis. Data abstracted from Relander T, Johnson NA, Farinha P, et al.<br />
Prognostic factors in follicular lymphoma. J Clin Oncol. 2010;28:2902–2913.<br />
● Follicular lymphoma is clinically and biologically<br />
heterogeneous.<br />
● There are currently few validated predictive or prognostic<br />
tools available to guide treatment at an individual<br />
level.<br />
● <strong>Clinical</strong>ly, follicular lymphoma can be conceptualized<br />
as a series <strong>of</strong> disease states with unique treatment<br />
approaches and goals <strong>of</strong> treatment.<br />
● Advances in the pathobiology <strong>of</strong> follicular lymphoma<br />
have led to a new generation <strong>of</strong> agents that are now in<br />
clinical trials.<br />
● The optimal incorporation <strong>of</strong> these new agents will<br />
likely build on existing treatment paradigms within<br />
clinically distinct groups: treatment-naïve (low tumor<br />
burden and high tumor burden), chemoimmunotherapy-sensitive,<br />
and chemoimmunotherapy-resistant<br />
disease.<br />
482<br />
SONALI M. SMITH<br />
FL. 8 Patients in the maintenance rituximab arm were<br />
significantly less likely to require cytotoxic therapy at 3<br />
years (91% vs. 48%; hazard ratio [HR] � 0.37; p � 0.001);<br />
there was no improvement in overall survival, which was<br />
95% in all arms. The Eastern Cooperative <strong>Oncology</strong><br />
Group (ECOG) RESORT (“Rituximab Extended Schedule<br />
or Re-Treatment Trial”) trial sought to determine whether<br />
a maintenance rituximab strategy following induction<br />
rituximab could improve time to treatment failure compared<br />
with a rituximab re-treatment strategy. 9 Although there<br />
was no difference in the primary end point <strong>of</strong> time to<br />
treatment failure (3.6 vs. 3.9 years for rituximab retreatment<br />
vs. maintenance arms, respectively), significantly<br />
fewer patients in the maintenance arm had not yet required<br />
cytotoxic therapy at 3 years (HR � 2.5; p � 0.027). Should<br />
rituximab induction plus maintenance replace “watch and<br />
wait” as the initial management approach for low–tumor<br />
burden and asymptomatic patients? This is clearly a matter<br />
<strong>of</strong> debate, but highlights the interest and need to integrate<br />
new agents in this clinical setting in which delayed time to<br />
cytotoxic therapy could be a reasonable goal for some patients.
TREATMENT OF FOLLICULAR LYMPHOMA<br />
Fig 1. Conceptual clinical approach to<br />
follicular lymphoma. Although the overall<br />
goals <strong>of</strong> disease control and improved<br />
survival apply to all phases <strong>of</strong> disease, there<br />
are unique clinical considerations and treatment<br />
options according to tumor burden<br />
and known sensitivity versus resistance to<br />
chemotherapy and immunotherapy.<br />
Abbreviations: Allo HCT, allogeneic hematopoietic<br />
cell transplantation; MR, maintenance<br />
rituximab; R-chemo, rituximab plus<br />
chemotherapy; RIT, radioimmunotherapy.<br />
For many patients with low–tumor burden FL who either<br />
need or desire treatment, several groups have tested relatively<br />
low-intensity therapies with variable objectives.<br />
Among these, rituximab plus lenalidomide is emerging as an<br />
extremely promising biologic doublet that could be a novel<br />
therapeutic backbone. In contrast to most antineoplastic<br />
agents in FL, lenalidomide appears to have pleiotropic<br />
effects on both the malignant and nonmalignant components<br />
<strong>of</strong> lymphoma (reviewed in Press et al 22 ). The importance <strong>of</strong><br />
the microenvironment in FL is underscored by studies<br />
showing a nine-fold survival difference on the basis <strong>of</strong> the<br />
gene expression pr<strong>of</strong>ile <strong>of</strong> background, nonmalignant<br />
“immune-response” cells. 23,24 As a single agent in relapsed<br />
FL, lenalidomide has a response rate <strong>of</strong> approximately 30%.<br />
When combined with rituximab, there appears to be clinical<br />
synergy, prompting this doublet to be tested in front-line<br />
settings. The Cancer and Leukemia Group B (CALGB)/<br />
Alliance recently completed accrual to a front-line study <strong>of</strong><br />
lenalidomide and rituximab (NCT01145495), with promising<br />
preliminary results. The M. D. Anderson Cancer Center<br />
also tested lenalidomide and rituximab, and reported an<br />
impressive 79% complete remission rate in 30 patients with<br />
FL. 10 The Swiss cooperative group is testing rituximab with<br />
or without lenalidomide in a randomized phase II setting<br />
(NCT01307605). Clearly, this is a regimen <strong>of</strong> great interest,<br />
and has the potential to replace chemotherapy (which is<br />
being prospectively tested in patients with high tumor<br />
burden by the Groupe d’Etude des Lymphomes de l’Adulte<br />
[GELA]).<br />
Targeted agents, including monoclonal antibodies and<br />
radioimmunoconjugates have also been tested in treatmentnaïve<br />
patients with low tumor burden. Single agent I 131<br />
tositumumab results in a 75% complete remission rate and<br />
median progression-free survival (PFS) <strong>of</strong> 6.1 years in newly<br />
diagnosed patients. 11 Similarly, 90 Y-ibritumomab tiuxetan<br />
appears most efficacious when used in earlier disease<br />
states. 12 To date, despite this encouraging data, radioimmunotherapy<br />
as a single agent has not been approved for newly<br />
diagnosed FL, and there are no comparative studies against<br />
single-agent rituximab. More recently, a new generation <strong>of</strong><br />
monoclonal antibodies with possible advantages over ritux-<br />
imab are being tested in front-line settings, including <strong>of</strong>atumumab<br />
and GA101.<br />
There are several quite promising agents being tested in<br />
the relapsed setting (discussed later herein) that should<br />
make their way to this population, but the biggest challenge<br />
to incorporating these and other new agents in the front-line<br />
setting for patients with asymptomatic or low–tumor burden<br />
FL is agreeing on the best clinical and trial end points.<br />
Overall survival benefit requires years, perhaps decades, <strong>of</strong><br />
observation, and progression-free survival does not reflect<br />
the fact that many low tumor burden patients do not require<br />
therapy even when there is radiographic or clinical pro<strong>of</strong> <strong>of</strong><br />
mild progression. Time to cytotoxic intervention, as selected<br />
in the United Kingdom trial and the ECOG RESORT trial,<br />
may be more reflective <strong>of</strong> real-world dilemmas, but the<br />
definition <strong>of</strong> this end point needs to be refined and agreed on<br />
in the broader research community.<br />
High–Tumor Burden FL (Treatment Naïve)<br />
For patients with symptoms, or for those meeting Groupe<br />
d’Etude des Lymphomes Folliculaires (GELF) criteria, effective<br />
treatment has immediate benefit in terms <strong>of</strong> symptom<br />
relief and reduction in tumor burden. The most common<br />
approach has been chemoimmunotherapy induction with or<br />
without rituximab maintenance or radioimmunotherapy<br />
consolidation. There are at least three important broad<br />
questions for patients in which new agents could make an<br />
impact in this clinical disease state: 1) What is the best<br />
induction chemoimmunotherapy combination; 2) What is<br />
the best strategy for prolonging remission; and 3) Can a<br />
non–chemotherapy-based regimen supplant classic cytotoxics?<br />
The recent success <strong>of</strong> bendamustine is an instructive<br />
example <strong>of</strong> how an agent, even if chemotherapy based, can<br />
dramatically affect the standard <strong>of</strong> care, and perhaps serve<br />
as a backbone for other agents. Bendamustine is a unique<br />
chemotherapeutic agent with bifunctional properties <strong>of</strong> both<br />
alkylating agents and purine analogs and strong singleagent<br />
activity despite prior alkylator exposure. 13-15 When<br />
combined with rituximab, again in the relapsed setting, it<br />
showed high response rates with promising durability. 16<br />
483
On the basis <strong>of</strong> encouraging data in relapsed patients, the<br />
German Low-Grade Lymphoma Study Group compared bendamustine<br />
and rituximab (BR) against rituximab, cyclophosphamide,<br />
doxorubicin, vincristine, and prednisone (R-<br />
CHOP) in 549 treatment-naïve indolent lymphomas, with<br />
FL comprising more than half <strong>of</strong> patients. All patients<br />
required therapy as per predefined criteria. Initially designed<br />
as a noninferiority study, BR instead showed superior<br />
complete response rates (40.1% vs. 30.8%; p � 0.03) and<br />
PFS (54.8 vs. 34.8 months; p � 0.0002); with a median<br />
follow-up time <strong>of</strong> approximately 3 years, there is a trend<br />
toward improved overall survival. 17 The option to avoid<br />
anthracyclines is appealing, particularly in a disease in<br />
which the median age is in the sixth decade <strong>of</strong> life and<br />
comorbidities <strong>of</strong>ten influence therapeutic choices. A confirmatory<br />
trial <strong>of</strong> BR compared with rituximab, cyclophosphamide,<br />
vincristine, prednisone (R-CVP) or R-CHOP recently<br />
completed accrual (NCT00877006). For now, although it is<br />
not clear that BR should be the front-line standard for all<br />
patients in this disease state, it clearly provides a new<br />
standard in terms <strong>of</strong> efficacy and tolerability for patients<br />
with high tumor burden/symptomatic FL and is a platform<br />
for new regimens. An Eastern Cooperative <strong>Oncology</strong> Group<br />
(ECOG)-led study is evaluating BR with or without bortezomib<br />
followed by rituximab with or without lenalidomide<br />
as one example (NCT01216683).<br />
Despite the high response rates to most chemoimmunotherapy<br />
regimens, relapse is inevitable and has prompted<br />
evaluation <strong>of</strong> maintenance or consolidation strategies. There<br />
are currently two approved postremission strategies: maintenance<br />
rituximab (on the basis <strong>of</strong> the PRIMA trial) and<br />
consolidative radioimmunotherapy (on the basis <strong>of</strong> the FIT<br />
trial). The PRIMA (Primary Rituximab and Maintenance)<br />
trial provided clear evidence that maintenance rituximab,<br />
even after a rituximab-containing chemotherapy induction,<br />
could improve PFS (74.9% vs. 57.6%; p � 0.0001). However,<br />
there was no improvement in overall survival. The FIT<br />
(First-Line Indolent) trial delivered induction chemotherapy<br />
(with or without rituximab) and then randomly assigned<br />
patients to either 90 Y-ibritumomab tiuxetan or observation.<br />
Patients in the consolidation arm had a significant conversion<br />
<strong>of</strong> partial to complete remissions and a near-tripling <strong>of</strong><br />
PFS (13.3 vs. 36.5 months; HR � 0.465; p � 0.0001). 18 A<br />
recent update <strong>of</strong> the FIT trial showed that PFS advantage<br />
persists with 66-month median follow-up. 19 However, only<br />
15% <strong>of</strong> patients received rituximab as part <strong>of</strong> their initial<br />
induction. Two phase II studies support the addition <strong>of</strong><br />
consolidative 90 Y-ibritumumab tiuxetan after chemoimmunotherapy,<br />
but a definitive phase III trial is lacking. 20,21 I 131<br />
tositumumab consolidation has also been tested, with<br />
clearly promising results after either CHOP or CVP<br />
chemotherapy. 22-24 Mature follow-up <strong>of</strong> Southwest <strong>Oncology</strong><br />
Group (SWOG) protocol S9911 shows a 5-year PFS and<br />
overall survival <strong>of</strong> 67% and 87%, respectively, which significantly<br />
exceeds results for patients receiving treatment<br />
without immunotherapy in prior SWOG studies. This promising<br />
phase II trial prompted a randomized phase III trial <strong>of</strong><br />
CHOP with either rituximab or consolidative I 131 tositumumab,<br />
which showed equivalence <strong>of</strong> both arms and overall<br />
excellent outcomes. 25<br />
Should all patients receive maintenance or consolidation<br />
after chemoimmunotherapy? An unplanned but intriguing<br />
subanalysis <strong>of</strong> the PRIMA trial evaluated the role <strong>of</strong> func-<br />
484<br />
tional imaging via fluorodeoxyglucose positron emission<br />
tomography (FDG-PET) as a reflection <strong>of</strong> minimal residual<br />
disease (MRD). 26 Thirty-two (26%) <strong>of</strong> 122 patients remained<br />
PET positive at the end <strong>of</strong> induction. PET positivity was<br />
strongly predictive <strong>of</strong> PFS (20 months vs. not reached; p �<br />
0.001) and was more sensitive than standard computed<br />
tomography (CT) criteria at identifying patients with an<br />
inferior outcome. As a result <strong>of</strong> small numbers, it was not<br />
possible to determine whether maintenance rituximab was<br />
able to overcome post-induction PET positivity. However,<br />
the use <strong>of</strong> PET in the PRIMA study reflects the general<br />
interest in determining MRD status in FL as a predictor <strong>of</strong><br />
adverse outcome, and this might identify a population in<br />
which new agents could be evaluated.<br />
Several considerations regarding maintenance and consolidation<br />
should be raised. First, if a front-line regimen has<br />
greater relative efficacy, it is more difficult to demonstrate<br />
the impact <strong>of</strong> a maintenance or consolidation strategy. As an<br />
example, to date, there are no data showing that maintenance<br />
or consolidation after BR is superior to observation<br />
alone. Second, there are currently no data guiding the<br />
selection <strong>of</strong> maintenance rituximab compared with consolidative<br />
radioimmunotherapy. In addition, there are other<br />
agents, such as lenalidomide or others (discussed later<br />
herein), that are orally bioavailable, have encouraging<br />
safety pr<strong>of</strong>iles, and show preliminary activity that should be<br />
tested in this setting.<br />
First and Second Relapse <strong>of</strong> FL<br />
SONALI M. SMITH<br />
Perhaps one <strong>of</strong> the most heterogeneous time points <strong>of</strong> FL<br />
is the time <strong>of</strong> first or second relapse. At this point, disease<br />
can be rituximab sensitive or resistant, and chemotherapy<br />
sensitive or resistant. Patients with either an asymptomatic<br />
or low-volume relapse generally receive treatment similar to<br />
that <strong>of</strong> patients who are newly diagnosed, using chemoimmunotherapy<br />
or biologic agents, whereas patients with a<br />
quick relapse or aggressive course receive treatment similar<br />
to that <strong>of</strong> patients with multiple relapses, as discussed in the<br />
following pages. Similar to the front-line setting, there are<br />
no readily available clinical or biologic markers to predict<br />
outcome or select therapeutic regimens.<br />
For many patients, the first or second relapse, particularly<br />
if the duration <strong>of</strong> response to previous regimens has been<br />
short, is the ideal time to consider autologous or allogeneic<br />
hematopoietic stem-cell transplantation. Although this is<br />
reviewed in detail elsewhere, it is important to note that<br />
dose intensification with autologous stem-cell transplantation<br />
has been a helpful tool for many patients in first or<br />
second relapse. However, it is equally critical to note that<br />
the vast majority <strong>of</strong> data were generated before the advent <strong>of</strong><br />
rituximab and that there are no randomized trials in the<br />
modern era comparing autologous stem-cell transplantion<br />
with chemoimmunotherapy regimens. Allogeneic stem-cell<br />
transplantation is the only known curative modality for<br />
relapsed advanced-stage FL, and a careful discussion <strong>of</strong><br />
potential risks and benefits with an experienced transplant<br />
center should be considered for patients at early relapse,<br />
particularly if there has been limited benefit to chemoimmunotherapeutic<br />
regimens.<br />
Similar to front-line settings, bendamustine is increasingly<br />
used as a therapeutic backbone to which new agents<br />
are added. One example is the addition <strong>of</strong> the proteasome<br />
inhibitor bortezomib. The ubiquitin-proteasome pathway
TREATMENT OF FOLLICULAR LYMPHOMA<br />
mediates the ubiquitation and degradation <strong>of</strong> proteins, and<br />
is frequently deregulated in lymphomas, including FL. Although<br />
suppression <strong>of</strong> nuclear factor-kappaB (NF�B) appears<br />
to be the major mechanism <strong>of</strong> action <strong>of</strong> bortezomib,<br />
there is also upregulation <strong>of</strong> proapoptotic factors such as<br />
NOXA and downregulation <strong>of</strong> antiapoptotic factors. The<br />
increased activity <strong>of</strong> proapoptotic factors in a disease that<br />
is characterized by universal bcl2 overexpression may be<br />
mechanistically important. The VERTICAL trial (Velcade<br />
in Relapsed or Refractory Follicular Lymphoma) added<br />
four weekly doses <strong>of</strong> bortezomib to a 35-day cycle <strong>of</strong><br />
escalating doses <strong>of</strong> bendamustine plus rituximab for five<br />
consecutive cycles. 27 Among 63 patients receiving the<br />
highest bendamustine dose level, the overall response<br />
rate was 88% and complete response rate was 53%. Another<br />
trial testing this combination had a slightly different<br />
treatment schedule that used the more standard bortezomib<br />
scheduling <strong>of</strong> twice-weekly doses. 28 Again, there was an<br />
impressive overall response rate <strong>of</strong> 83%, and this regimen<br />
is currently undergoing further testing in both relapsed<br />
and front-line settings to determine the incremental benefit<br />
<strong>of</strong> bortezomib to a bendamustine and rituximab backbone.<br />
Given high patient heterogeneity, it is clear that a subset<br />
<strong>of</strong> patients stand to benefit a great deal from new agents<br />
used at this point in the overall disease course. However,<br />
short <strong>of</strong> randomized multiarm trials, the challenge will be to<br />
standardize patients in some way to allow fair comparisons<br />
<strong>of</strong> different regimens. It is not sufficient to use the number <strong>of</strong><br />
prior regimens, because, as discussed earlier, the front-line<br />
regimens vary greatly, and time to relapse and determination<br />
<strong>of</strong> relative rituximab sensitivity compared with resistance<br />
all influence the likelihood <strong>of</strong> response to subsequent<br />
therapy. A second-line FLIPI or other biologic tool is greatly<br />
needed. Until these tools are available, this heterogeneous<br />
disease state will remain the most challenging setting in<br />
which to assess relative efficacy <strong>of</strong> a new therapy, and<br />
incorporation <strong>of</strong> new agents will be difficult to apply in a<br />
controlled fashion.<br />
Multiply Relapsed and/or Refractory FL: Promising<br />
Agents and Approaches for Chemoresistant and<br />
Rituximab-Resistant Disease<br />
Despite the major positive impact <strong>of</strong> chemoimmunotherapy,<br />
nearly all patients eventually demonstrate both<br />
chemotherapy and rituximab resistance, making multiply<br />
relapsed FL a disease state <strong>of</strong> great unmet need. An agent<br />
capable <strong>of</strong> demonstrating activity in this clinical setting<br />
would quickly have impact and could be moved earlier into<br />
the disease course, as reflected by lenalidomide and bendamustine,<br />
both <strong>of</strong> which first showed activity in relapsed/<br />
refractory FL. This is currently the disease state in which<br />
gains in knowledge regarding FL pathogenesis have spurred<br />
new and targeted agents with promising clinical validation<br />
ongoing.<br />
Among B-cell malignancies, few potential targets are as<br />
ubiquitous as the B-cell receptor (BCR) and its downstream<br />
signaling cascade. BCR normally responds to antigens by<br />
triggering an internal signal leading to gene transcription<br />
and B-cell activation and proliferation. 29 Malignant B-cell<br />
transformation usually retains the need for an intact and<br />
tonically active BCR, and agents that block its signaling<br />
have shown promising preclinical effects. 30-33 The Tec-<br />
kinase Bruton’s tyrosine kinase (BTK) appears to be required<br />
to form immunoglobulin and to allow B-cell survival<br />
as part <strong>of</strong> BCR signaling. 34 PCI-32765 is a novel, orally<br />
available, irreversible covalent inhibitor <strong>of</strong> BTK. Preclinical<br />
studies confirm its selectivity for its target, the ability to<br />
completely halt BCR signaling, and potent activity in B-cell<br />
lymphomas, chronic lymphocytic leukemia models, and autoimmune<br />
models. 35-38 A phase I trial in B-cell non-Hodgkin<br />
lymphoma showed an impressive response rate <strong>of</strong> 54% in an<br />
intent-to-treat population that included both aggressive and<br />
indolent histologies. 39 Among 16 patients with FL, one-third<br />
had an objective response including three complete responses.<br />
Early results are promising, but it is also clear that<br />
single-agent PCI-32765 primarily leads to partial responses<br />
in FL and its impact on response durability is yet to be<br />
determined. Using this agent in earlier disease states is<br />
worthy <strong>of</strong> investigation.<br />
Just downstream <strong>of</strong> BCR signaling is the phosphoinositide<br />
3-kinase (PI3K)/Akt/mammalian target <strong>of</strong> rapamycin<br />
(mTOR; PAM) axis, which is also emerging as a major<br />
pathogenetic mechanism in FL. The natural function <strong>of</strong><br />
PI3K is to transduce external growth signals and modulate<br />
downstream targets that control cellular proliferation,<br />
motility, metabolism and cell growth vs. survival<br />
(reviewed in Courtney, Corcoran, and Engelman 40 ). There<br />
are four PI3K is<strong>of</strong>orms—alpha, beta, gamma, delta—with<br />
the delta is<strong>of</strong>orm having restricted expression in human<br />
leukocytes. CAL-101 is an orally bioavailable PI3K<br />
inhibitor that is highly selective for the p110 delta is<strong>of</strong>orm. 41<br />
In vitro models show that CAL-101 can block tonic PI3K<br />
signaling with decreased activation <strong>of</strong> downstream targets,<br />
including Akt. Preliminary data from phase I studies<br />
show promising activity <strong>of</strong> CAL-101 in B-cell malignancies,<br />
including FL. 42-44 Among 30 patients with indolent<br />
lymphomas, including 17 patients with FL, the single-agent<br />
overall response rate was 63%, with a median PFS exceeding<br />
1 year. The most common grade 3 or 4 event was<br />
transient increase <strong>of</strong> hepatic enzymes, which occurred in<br />
27% <strong>of</strong> patients. On the basis <strong>of</strong> the safety pr<strong>of</strong>ile and<br />
efficacy, CAL-101 has been added to the backbone <strong>of</strong><br />
bendamustine and rituximab in a phase I combination<br />
study (NCT01088048). 45 Preliminary results show excellent<br />
tolerability and an overall response rate <strong>of</strong> more than 65%<br />
in a group <strong>of</strong> heavily pretreated patients with indolent<br />
lymphomas.<br />
An important downstream substrate <strong>of</strong> PI3K and Akt is<br />
the serine/threonine kinase mTOR. The natural functions <strong>of</strong><br />
mTOR are to integrate growth signals and nutrient availability<br />
and influence cell growth via control over mRNA<br />
translation. Several preclinical investigations support<br />
the central role <strong>of</strong> mTOR in FL (reviewed in Smith 46 ).<br />
Single-agent temsirolimus, a rapamycin analog, was tested<br />
in 39 patients with relapsed or refractory FL; more than<br />
half <strong>of</strong> patients had an objective response, including 25%<br />
complete responses. 47 Nonhematologic toxicities included<br />
metabolic abnormalities (hyperglycemia, hypertriglyceridemia,<br />
hypercholesterolemia), stomatitis, and rash. Combination<br />
studies <strong>of</strong> temsirolimus plus lenalidomide<br />
(NCT01076543) and everolimus plus lenalidomide<br />
(NCT01075321) are ongoing.<br />
A near-universal feature <strong>of</strong> FL is BCL2 overexpression,<br />
primarily as a result <strong>of</strong> the hallmark translocation, t(14;18).<br />
The constitutive expression <strong>of</strong> the antiapoptotic BCL2 pro-<br />
485
tein is not only diagnostic, but also probably underlies the<br />
characteristic disease persistence after standard therapy.<br />
Downregulation <strong>of</strong> BCL2 in preclinical models improves<br />
chemosensitivity to other agents, and could be an important<br />
therapeutic adjunct. G3139, an antisense oligonucleotide<br />
against BCL2 mRNA, showed preliminary activity in indolent<br />
lymphomas but is no longer in development. 48,49 Currently,<br />
several small-molecule inhibitors are in early phase<br />
trials, including navitoclax (ABT-263) and obatoclax mesylate<br />
(GX15–070).<br />
Finally, unique B-cell surface molecules provide opportunity<br />
for targeted agents to have a major impact. Novel<br />
monoclonal antibodies and antibody-drug conjugates targeting<br />
CD20 and non-CD20 surface molecules are under active<br />
investigation including <strong>of</strong>atumumab, GA101, veltuzumab,<br />
Author’s Disclosures <strong>of</strong> Potential Conflicts <strong>of</strong> Interest<br />
Author<br />
Employment or<br />
Leadership<br />
Positions<br />
Consultant or<br />
Advisory Role<br />
Sonali M. Smith Biogen Idec;<br />
Celgene;<br />
Cephalon;<br />
Genentech;<br />
GlaxoSmithKline;<br />
Spectrum<br />
Pharmaceuticals<br />
1. Swenson WT, Wooldridge JE, Lynch CF, et al. Improved survival <strong>of</strong><br />
follicular lymphoma patients in the United States. J Clin Oncol. 2005;23:<br />
5019-5026.<br />
2. Fisher RI, LeBlanc M, Press OW, et al. New treatment options have<br />
changed the survival <strong>of</strong> patients with follicular lymphoma. J Clin Oncol.<br />
2005;23:8447-8452.<br />
3. Federico M, Bellei M, Marcheselli L, et al. Follicular lymphoma international<br />
prognostic index 2: a new prognostic index for follicular lymphoma<br />
developed by the international follicular lymphoma prognostic factor project.<br />
J Clin Oncol. 2009;27:4555-4562.<br />
4. Solal-Céligny P, Roy P, Colombat P, et al. Follicular lymphoma international<br />
prognostic index. Blood. 2004;104:1258-1265.<br />
5. Friedberg JW, Taylor MD, Cerhan JR, et al. Follicular lymphoma in the<br />
United States: first report <strong>of</strong> the national LymphoCare study. J Clin Oncol.<br />
2009;27:1202-128.<br />
6. Ardeshna KM, Smith P, Norton A, et al. Long-term effect <strong>of</strong> a watch and<br />
wait policy versus immediate systemic treatment for asymptomatic advancedstage<br />
non-Hodgkin lymphoma: a randomised controlled trial. Lancet. 2003;<br />
362:516-522.<br />
7. Brice P, Bastion Y, Lepage E, et al. Comparison in low-tumor-burden<br />
follicular lymphomas between an initial no-treatment policy, prednimustine,<br />
or interferon alfa: a randomized study from the Groupe d’Etude des Lymphomes<br />
Folliculaires. Groupe d’Etude des Lymphomes de l’Adulte. J Clin<br />
Oncol. 1997;15:1110-1117.<br />
8. Ardeshna KM, Qian W, Smith P, et al. An Intergroup randomised trial<br />
<strong>of</strong> rituximab versus a watch and wait strategy in patients with stage II, III,<br />
IV, asymptomatic, non-bulky follicular lymphoma (grades 1, 2 and 3a): a<br />
preliminary analysis. Blood. ASH Annual Meeting Abstracts. 2010;116:<br />
abstract 6.<br />
9. Kahl BS, Hong F, Williams ME, et al. Results <strong>of</strong> Eastern Cooperative<br />
<strong>Oncology</strong> Group Protocol E4402 (RESORT): a randomized phase III study<br />
comparing two different rituximab dosing strategies for low tumor burden<br />
follicular lymphoma. Blood. ASH Annual Meeting Abstracts. 2011;118; LBA<br />
6.<br />
10. Samaniego F, Hagemeister F, McLaughlin P, et al. High response rates<br />
with lenalidomide plus rituximab for untreated indolent B-cell non-Hodgkin<br />
lymphoma, including those meeting GELF criteria. J Clin Oncol. 2011;29<br />
(suppl; abtr 8030).<br />
11. Kaminski MS, Tuck M, Estes J, et al. 131I-tositumomab therapy as<br />
initial treatment for follicular lymphoma. N Engl J Med. 2005;352:441-449.<br />
12. Scholz CW, Pinto A, Linkesch W, et al. 90Yttrium ibritumomab<br />
486<br />
AME-133, inotuzumab ozogamicin,<br />
SAR3419 and others (reviewed in Leonard and Martin 50 ).<br />
Conclusion<br />
90 Y-epratuzumab,<br />
There is a plethora <strong>of</strong> new agents, and the optimal<br />
incorporation <strong>of</strong> these agents is highly dependent on the<br />
clinical setting and goals <strong>of</strong> therapy. The ideal agent should<br />
reflect the underlying biology, with preclinically validated<br />
targets and better predictive tools to help individualize<br />
therapy. Many <strong>of</strong> these new agents are being added to or are<br />
supplanting standard cytotoxics as a result <strong>of</strong> their favorable<br />
toxicity pr<strong>of</strong>ile and promising efficacy. Although we are<br />
short <strong>of</strong> a cure for most patients, it is clear that incremental<br />
benefit has been achieved and collaborative efforts should<br />
continue.<br />
Stock<br />
Ownership Honoraria<br />
Biogen Idec;<br />
Celgene;<br />
Genentech<br />
REFERENCES<br />
Research<br />
Funding<br />
Expert<br />
Testimony<br />
SONALI M. SMITH<br />
Other<br />
Remuneration<br />
tiuxetan as first line treatment for follicular lymphoma: first results from an<br />
international phase II clinical trial. Blood. 2010;116:593A.<br />
13. Cheson BD, Rummel MJ. Bendamustine: rebirth <strong>of</strong> an old drug. J Clin<br />
Oncol. 2009;27:1492-1501.<br />
14. Kahl BS, Bartlett NL, Leonard JP, et al. Bendamustine is effective<br />
therapy in patients with rituximab-refractory, indolent B-cell non-Hodgkin<br />
lymphoma: results from a Multicenter Study. Cancer. 2010;116:106-114.<br />
15. Friedberg JW, Cohen P, Chen L, et al. Bendamustine in patients with<br />
rituximab-refractory indolent and transformed non-Hodgkin’s lymphoma:<br />
results from a phase II multicenter, single-agent study. J Clin Oncol.<br />
2008;26:204-210.<br />
16. Rummel MJ, Al-Batran SE, Kim SZ, et al. Bendamustine plus rituximab<br />
is effective and has a favorable toxicity pr<strong>of</strong>ile in the treatment <strong>of</strong> mantle<br />
cell and low-grade non-Hodgkin’s lymphoma. J Clin Oncol. 2005;23:3383-<br />
3389.<br />
17. Rummel MJ, Niederle N, Maschmeyer G, et al. Bendamustine plus<br />
rituximab is superior in respect <strong>of</strong> progression free survival and CR rate when<br />
compared to CHOP plus rituximab as first-line treatment <strong>of</strong> patients with<br />
advanced follicular, indolent, and mantle cell lymphomas: final results <strong>of</strong> a<br />
randomized phase III study <strong>of</strong> the StiL (Study Group Indolent Lymphomas,<br />
Germany). Blood. 2009;114:405A.<br />
18. Morschhauser F, Radford J, Van Ho<strong>of</strong> A, et al. Phase III trial <strong>of</strong><br />
consolidation therapy with yttrium-90-ibritumomab tiuxetan compared with<br />
no additional therapy after first remission in advanced follicular lymphoma.<br />
J Clin Oncol. 2008;26:5156-5164.<br />
19. Hagenbeek A, Radford J, Van Ho<strong>of</strong> A, et al. 90Y-ibritumomab tiuxetan<br />
(Zevalin®) consolidation <strong>of</strong> first remission in advanced-stage follicular non-<br />
Hodgkin’s lymphoma: updated results after a median follow-up <strong>of</strong> 66.2<br />
months from the international, randomized, phase III First-Line Indolent<br />
Trial (FIT) in 414 patients. Blood. 2010;116:594A.<br />
20. Hainsworth JD, Spigel DR, Markus TM, et al. Rituximab plus shortduration<br />
chemotherapy followed by Yttrium-90 Ibritumomab tiuxetan as<br />
first-line treatment for patients with follicular non-Hodgkin lymphoma: a<br />
phase II trial <strong>of</strong> the Sarah Cannon <strong>Oncology</strong> Research Consortium. Clin<br />
Lymphoma Myeloma. 2009;9:223-228.<br />
21. Jacobs SA, Swerdlow SH, Kant J, et al. Phase II trial <strong>of</strong> short-course<br />
CHOP-R followed by 90Y-ibritumomab tiuxetan and extended rituximab in<br />
previously untreated follicular lymphoma. Clin Cancer Res. 2008;14:7088-<br />
7094.<br />
22. Press OW, Unger JM, Braziel RM, et al. Phase II trial <strong>of</strong> CHOP<br />
chemotherapy followed by tositumomab/iodine I-131 tositumomab for previ-
TREATMENT OF FOLLICULAR LYMPHOMA<br />
ously untreated follicular non-Hodgkin’s lymphoma: five-year follow-up <strong>of</strong><br />
Southwest <strong>Oncology</strong> Group Protocol S9911. J Clin Oncol. 2006;24:4143-4149.<br />
23. Press OW, Unger JM, Braziel RM, et al. A phase 2 trial <strong>of</strong> CHOP<br />
chemotherapy followed by tositumomab/iodine I 131 tositumomab for previously<br />
untreated follicular non-Hodgkin lymphoma: Southwest <strong>Oncology</strong><br />
Group Protocol S9911. Blood. 2003;102:1606-1612.<br />
24. Link BK, Martin P, Kaminski MS, et al. Cyclophosphamide, vincristine,<br />
and prednisone followed by tositumomab and iodine-131-tositumomab in<br />
patients with untreated low-grade follicular lymphoma: eight-year follow-up<br />
<strong>of</strong> a multicenter phase II study. J Clin Oncol. 2010;28:3035-3041.<br />
25. Press OW, Unger JM, Rimsza L, et al. A phase III randomized<br />
intergroup trial (SWOG S0016) <strong>of</strong> CHOP chemotherapy plus rituximab vs.<br />
CHOP chemotherapy plus iodine-131-tositumomab for the treatment <strong>of</strong> newly<br />
diagnosed follicular non-Hodgkin’s lymphoma. Blood. 2011;118:98A.<br />
26. Trotman J, Fournier M, Lamy T, et al. Positron emission tomographycomputed<br />
tomography (PET-CT) after induction therapy is highly predictive<br />
<strong>of</strong> patient outcome in follicular lymphoma: analysis <strong>of</strong> PET-CT in a subset <strong>of</strong><br />
PRIMA trial participants. J Clin Oncol. 2011;29:3194-3200.<br />
27. Fowler N, Kahl BS, Lee P, et al. Bortezomib, bendamustine, and<br />
rituximab in patients with relapsed or refractory follicular lymphoma: the<br />
phase II VERTICAL study. J Clin Oncol. 2011;29:3389-3395.<br />
28. Friedberg JW, Vose JM, Kelly JL, et al. The combination <strong>of</strong> bendamustine,<br />
bortezomib, and rituximab for patients with relapsed/refractory indolent<br />
and mantle cell non-Hodgkin lymphoma. Blood. 2011;117:2807-2812.<br />
29. Lenz G, Staudt LM. Aggressive lymphomas. N Engl J Med. 2010;362:<br />
1417-1429.<br />
30. Turner M, Schweigh<strong>of</strong>fer E, Colucci F, et al. Tyrosine kinase SYK:<br />
essential functions for immunoreceptor signalling. Immunol Today. 2000;21:<br />
148-154.<br />
31. Chen L, Monti S, Juszczynski P, et al. SYK-dependent tonic B-cell<br />
receptor signaling is a rational treatment target in diffuse large B-cell<br />
lymphoma. Blood. 2008;111:2230-2237.<br />
32. Gururajan M, Jennings CD, Bondada S. Cutting edge: constitutive B<br />
cell receptor signaling is critical for basal growth <strong>of</strong> B lymphoma. J Immunol.<br />
2006;176:5715-5719.<br />
33. Friedberg JW, Sharman J, Sweetenham J, et al. Inhibition <strong>of</strong> Syk with<br />
fostamatinib disodium has significant clinical activity in non-Hodgkin lymphoma<br />
and chronic lymphocytic leukemia. Blood. 2008;115:2578-2585.<br />
34. Küppers R. Mechanisms <strong>of</strong> B-cell lymphoma pathogenesis. Nat Rev<br />
Cancer. 2005;5:251-262.<br />
35. Ponader S, Chen SS, Buggy JJ, et al. Bruton’s tyrosine kinase inhibitor<br />
PCI-32765 thwarts chronic lymphocytic leukemia cell survival and tissue<br />
homing in vitro and in vivo. Blood. <strong>2012</strong>;119:1182-1189.<br />
36. Chang BY, Huang MM, Francesco M, et al. The Bruton tyrosine kinase<br />
inhibitor PCI-32765 ameliorates autoimmune arthritis by inhibition <strong>of</strong> multiple<br />
effector cells. Arthritis Res Ther. 2011;13:R115.<br />
37. Honigberg LA, Smith AM, Sirisawad M, et al. The Bruton tyrosine<br />
kinase inhibitor PCI-32765 blocks B-cell activation and is efficacious in<br />
models <strong>of</strong> autoimmune disease and B-cell malignancy. Proc Natl Acad Sci U<br />
SA. 2010;107:13075-13080<br />
38. Herman SE, Gordon AL, Hertlein E, et al. Bruton tyrosine kinase<br />
represents a promising therapeutic target for treatment <strong>of</strong> chronic lymphocytic<br />
leukemia and is effectively targeted by PCI-32765. Blood. 2011117:6287-<br />
6296.<br />
39. Fowler N, Sharman J, Smith SM, et al. The Btk inhibitor, PCI-32765,<br />
induces durable responses with minimal toxicity in patients with relapsed/<br />
refractory B-cell malignancies: results from a phase I study. Blood. 2010;116:<br />
964A.<br />
40. Courtney KD, Corcoran RB, Engelman JA. The PI3K pathway as drug<br />
target in human cancer. J Clin Oncol. 2010;28:1075-1083.<br />
41. Lannutti BJ, Meadows SA, Herman SE, et al. CAL-101, a p110delta<br />
selective phosphatidylinositol-3-kinase inhibitor for the treatment <strong>of</strong> B-cell<br />
malignancies, inhibits PI3K signaling and cellular viability. Blood. 2011;117:<br />
591-594.<br />
42. Furman R, Byrd JC, Brown JR, et al. CAL-101, An is<strong>of</strong>orm-selective<br />
inhibitor <strong>of</strong> phosphatidylinositol 3-kinase p110, demonstrates clinical activity<br />
and pharmacodynamic effects in patients with relapsed or refractory chronic<br />
lymphocytic leukemia. Blood. 2011;116:55A.<br />
43. Kahl BS, Byrd JC, Flinn IW, et al. <strong>Clinical</strong> safety and activity in a<br />
phase 1 study <strong>of</strong> CAL-101, an is<strong>of</strong>orm-selective inhibitor <strong>of</strong> phosphatidylinositol<br />
3-kinase P110, in patients with relapsed or refractory non-Hodgkin<br />
lymphoma. Blood. 2010;116:1777A.<br />
44. Webb HK, Chen H, Yu AS, et al. <strong>Clinical</strong> pharmacokinetics <strong>of</strong> CAL-101,<br />
a p110 is<strong>of</strong>orm-selective PI3K inhibitor, following single- and multiple-dose<br />
administration in healthy volunteers and patients with hematological malignancies.<br />
Blood. 2010;116:1774A.<br />
45. De Vos S, Schreeder MT, Flinn IW, et al. A phase 1 study <strong>of</strong> the<br />
selective phosphatidylinositol 3-kinase-delta (PI3K) inhibitor, Cal-101 (GS-<br />
1101), in combination with rituximab and/or bendamustine in patients with<br />
previously treated, indolent non-Hodgkin lymphoma (iNHL). Blood. 2011;<br />
118:2699A.<br />
46. Smith SM. <strong>Clinical</strong> development <strong>of</strong> mTOR inhibitors: a focus on<br />
lymphoma. Rev Recent Clin Trials. 2007;2:103-110.<br />
47. Smith SM, van Besien K, Karrison T, et al. Temsirolimus has activity<br />
in non-mantle cell non-Hodgkin’s lymphoma subtypes: The University <strong>of</strong><br />
Chicago phase II consortium. J Clin Oncol. 2010;28:4740-4746.<br />
48. Pro B, Leber B, Smith M, et al. Phase II multicenter study <strong>of</strong><br />
oblimersen sodium, a Bcl-2 antisense oligonucleotide, in combination with<br />
rituximab in patients with recurrent B-cell non-Hodgkin lymphoma. Br J<br />
Haematol. 2008;143:355-360.<br />
49. Waters JS, Webb A, Cunningham D, et al. Phase I clinical and<br />
pharmacokinetic study <strong>of</strong> bcl-2 antisense oligonucleotide therapy in patients<br />
with non-Hodgkin’s lymphoma. J Clin Oncol. 2000;18:1812-1823.<br />
50. Leonard JP, Martin P. Novel agents for follicular lymphoma. Hematology<br />
Am Soc Hematol Educ Program. 2010;2010:259-264.<br />
487
What Is the Best First-Line Treatment Strategy<br />
for Patients with Indolent Lymphomas?<br />
By Gilles Salles, MD, PhD, Hervé Ghesquières, MD, and Emmanuel Bachy, MD<br />
Overview: Although advanced follicular lymphoma is considered<br />
incurable, patient outcomes have improved over the last<br />
decade with the use <strong>of</strong> anti-CD20 monoclonal antibodies.<br />
Multiple treatment options are available and their use depends<br />
on clinical presentation (i.e., Ann Arbor stage, tumor burden,<br />
symptoms) and patient condition and age. Radiation therapy<br />
for patients with limited stage disease remains useful, although<br />
its use in the era <strong>of</strong> anti-CD20 antibodies should be<br />
re-evaluated. Single-agent rituximab has been tested in multiple<br />
studies with patients with low tumor burden. Short<br />
treatment duration provides a response lasting 2 to 3 years,<br />
although the benefit <strong>of</strong> maintenance therapy with rituximab<br />
after induction therapy with rituximab remains unproven.<br />
When watchful waiting is not an option, a combination <strong>of</strong><br />
THE CLINICAL outcome <strong>of</strong> patients with indolent lymphoma<br />
has markedly improved over the last decade.<br />
Recent epidemiologic data have estimated that the 5- and<br />
10-year overall survival (OS) rates for patients with indolent<br />
lymphoma older than age 60 are close to 85% and 73%,<br />
respectively. 1 For patients with follicular lymphoma, several<br />
comparisons from single-center and cooperative-group studies<br />
indicate that the median OS has increased from 8 to 10<br />
years to 12 to 15 years 2,3<br />
This progress has been achieved, at least partly, because<br />
<strong>of</strong> the introduction <strong>of</strong> anti-CD20 monoclonal antibodies.<br />
However, the lack <strong>of</strong> clinical or biologic criteria either to plan<br />
the optimal time to initiate therapy or to select between<br />
using anti-CD20 monoclonal antibodies as single agents or<br />
in combination with chemotherapy likely explain the heterogeneity<br />
<strong>of</strong> first-line treatment decisions observed in the<br />
LymphoCare study. 4 Furthermore, several anti-CD20 antibodies<br />
have been developed, including naked antibodies,<br />
such as rituximab and <strong>of</strong>atumumab, and radiolabeled antibodies,<br />
such as tositumomab and ibritumomab tiutexan,<br />
yet the optimal chemotherapy regimen remains undefined.<br />
With these multiple treatment options, it is worth examining<br />
the recent and follow-up results <strong>of</strong> studies performed<br />
to help guide clinical decisions in the management <strong>of</strong> patients<br />
with follicular lymphoma. Because indolent lymphomas<br />
remain incurable and most patients experience disease<br />
recurrence, patient quality <strong>of</strong> life, the ability to deliver<br />
subsequent treatments, and potential long-term adverse<br />
effects also need to be taken into account when considering<br />
first-line treatment strategies.<br />
Although treatment algorithms used for patients with<br />
disseminated forms <strong>of</strong> mucosa-associated lymphoid tissue<br />
(MALT) or nonsplenic marginal zone lymphomas can be<br />
similar to those used for patients with follicular lymphoma,<br />
first-line management <strong>of</strong> localized MALT 5 and patients with<br />
lymphocytic and lymphoplamasmacytic lymphomas 6 are<br />
quite distinct and will not be addressed in this manuscript.<br />
We will consider several questions on first-line treatment for<br />
patients with follicular lymphoma in light <strong>of</strong> the most recent<br />
studies.<br />
488<br />
rituximab with chemotherapy is the standard <strong>of</strong> care: alkylating<br />
agents with anthracycline or bendamustine appear to be<br />
the most widely used regimens, but alkylating agents alone<br />
may still be used in selected patients subgroups. The toxicity<br />
<strong>of</strong> regimens containing fludarabine appears to limit their<br />
indication as first-line treatment. In patients responding to<br />
one <strong>of</strong> these combinations, consolidation therapy with rituximab<br />
maintenance has been shown to prolong progressionfree<br />
survival with acceptable toxicity. The benefit <strong>of</strong> radioimmunotherapy<br />
in first-line treatment is still uncertain. With<br />
patients surviving for many years, the therapeutic strategy <strong>of</strong><br />
first-line management should weigh the quality and duration <strong>of</strong><br />
response against the risk <strong>of</strong> long-term toxicities.<br />
Is Radiation Therapy for Patients with Limited-Stage<br />
Follicular Lymphoma Still an Option in <strong>2012</strong>?<br />
Radiation therapy has long been considered the treatment<br />
<strong>of</strong> choice for patients with follicular (or indolent) lymphoma<br />
with Ann Arbor stage I or stage II disease. This option has<br />
been promoted as potentially curative, although the disease<br />
might recur in areas outside the radiation fields in most<br />
patients. 7,8 Given the potential toxicity associated with<br />
radiation therapy in specific areas, a watchful waiting approach<br />
has been also proposed. 9 A large epidemiologic study<br />
supports the use <strong>of</strong> radiation therapy in patients with<br />
follicular lymphoma, with a significant improvement in<br />
long-term OS for patients with stages I and II follicular<br />
lymphoma treated with radiation therapy compared with<br />
those not receiving radiation therapy (p � 0.0001). 10 Yet,<br />
this study has several limitations, including the lack <strong>of</strong><br />
details about certain prognostic factors (e.g., lactate dehydrogenase<br />
[LDH], tumor bulk) and the observation period<br />
(1973 to 2004) when monoclonal antibodies where not an<br />
option for treatment. Another recent retrospective study 11<br />
including patients with stage I disease reported no difference<br />
in the progression-free survival (PFS) for patients<br />
treated with radiation therapy compared with those that<br />
were untreated (i.e., watchful waiting), but chemotherapy<br />
with rituximab or combined modalities were found to provide<br />
the best outcome. This suggests that radiation therapy<br />
use should be limited to those patients that have localized<br />
follicular lymphoma (i.e., stage I or confluent stage II)<br />
without adverse features (e.g., grade 3, tumor bulk, elevated<br />
LDH). 12,13 For localized follicular lymphoma, a dose <strong>of</strong> 24 Gy<br />
appears sufficient. 14<br />
From the Hospices Civils de Lyon & Université Lyon 1, Pierre-Bénite, France; Centre Léon<br />
Bérard, Lyon, France.<br />
Authors’ disclosures <strong>of</strong> potential conflicts <strong>of</strong> interest are found at the end <strong>of</strong> this article.<br />
Address reprint requests to Gilles Salles, MD, PhD, Centre Hospitalier Lyon-Sud, 69495<br />
Pierre Bénite, France; email: gilles.salles@chu-lyon.fr.<br />
© <strong>2012</strong> by <strong>American</strong> <strong>Society</strong> <strong>of</strong> <strong>Clinical</strong> <strong>Oncology</strong>.<br />
1092-9118/10/1-10
TREATMENT OPTIONS IN INDOLENT LYMPHOMA<br />
Is Early Intervention Preferable to Watchful Waiting<br />
in <strong>2012</strong>?<br />
The safety <strong>of</strong> delaying the start <strong>of</strong> systemic therapy in<br />
patients with advanced disease but a limited tumor burden<br />
(based on Groupe d’Etude des Lymphomes Folliculaires<br />
[GELF] or British National Lymphoma Investigation<br />
[BNLI] criteria, Table 1) was well documented in the 1990s,<br />
and three randomized trials reported comparable survival<br />
outcomes for patients treated immediately or followed<br />
through watchful waiting. 15-17<br />
The low toxicity pr<strong>of</strong>ile <strong>of</strong> rituximab has prompted the<br />
investigation <strong>of</strong> short treatments <strong>of</strong> this antibody as a<br />
single-agent treatment for patients with follicular lymphoma<br />
without aggressive features. 18-20 In phase II studies,<br />
median times to disease progression have ranged between 2<br />
and 3 years. The long-term follow-up in one study where<br />
patients received 4 weekly infusions followed by 4 additional<br />
infusions administered every 2 months suggested potential<br />
disease control in 45% <strong>of</strong> patients at 8 years. 21<br />
These promising results prompted the design <strong>of</strong> a large<br />
phase III study in the United Kingdom that compared<br />
rituximab with watchful waiting. 22 At 3 years, compared<br />
with the control arm, patients receiving 4 weekly courses <strong>of</strong><br />
rituximab induction therapy followed by 2 years <strong>of</strong> rituximab<br />
maintenance therapy every 2 months showed significant<br />
improvement regarding the time to initiation <strong>of</strong> the<br />
next treatment (primary endpoint, 91% compared with 48%,<br />
p � 0.001) and for time to progression (81% compared with<br />
33%, p � 0.001). 22 However, no major difference in quality <strong>of</strong><br />
life was observed and OS was identical in each study arm. 23<br />
Of note, treatment with single-agent rituximab was not<br />
available for those patients (even in the control arm) who<br />
then received cytotoxic therapy when their disease pro-<br />
KEY POINTS<br />
● Decisions to initiate treatment and to opt for a<br />
specific therapeutic strategy should take into account<br />
tumor burden, symptoms, and patient wishes and<br />
condition as well as the potential long-term effects <strong>of</strong><br />
the treatment.<br />
● New studies should investigate the role <strong>of</strong> radiation<br />
therapy for localized disease in the context <strong>of</strong> anti-<br />
CD20 antibodies.<br />
● Watchful waiting remains an option for selected patients<br />
with low tumor burden; the sustained benefit<br />
<strong>of</strong> prolonged rituximab maintenance therapy after a<br />
short-term induction is not established in these patients.<br />
● In patients in need <strong>of</strong> cytotoxic therapy, the different<br />
chemotherapy regimens available in combination<br />
with rituximab may have distinct patterns <strong>of</strong> efficacy,<br />
in term <strong>of</strong> response and progression-free survival,<br />
and short- and long-term toxicities.<br />
● In patients responding to first-line immunochemotherapy,<br />
administration <strong>of</strong> maintenance therapy with<br />
rituximab improves progression-free survival, but the<br />
role <strong>of</strong> radioimmunotherapy in this setting remains<br />
uncertain.<br />
Table 1. Tumor Burden Criteria Used to Initiate a Cytotoxic<br />
Treatment for Patients with Follicular Lymphoma<br />
GELF criteria (FL2000 and PRIMA studies)<br />
(any one <strong>of</strong> these criteria) (29)<br />
BNLI criteria<br />
(any one <strong>of</strong> these criteria) (17)<br />
High tumor bulk defined by either: Rapid generalized disease progression in<br />
- a tumor � 7cm<br />
the preceding 3 months<br />
- 3 nodes in 3 distinct areas each Life threatening organ involvement<br />
� 3cm<br />
Renal or macroscopic liver infiltration<br />
- symptomatic splenic enlargement<br />
- organ compression<br />
- ascites or pleural effusion<br />
Bone lesions<br />
Presence <strong>of</strong> systemic symptoms<br />
ECOG Performance status � 1<br />
Presence <strong>of</strong> systemic symptoms or pruritus<br />
Serum LDH or beta2-microglobulin<br />
above normal values<br />
Hemoglobin � 10 g/dL or<br />
WBC � 3.0 � 10 9 /L or<br />
platelet counts � 100 � 10 9 /L;<br />
related to marrow involvement<br />
Abbreviations: BNLI, British National Lymphoma Investigation; ECOG, Eastern<br />
Cooperative <strong>Oncology</strong> Group; GELF, Groupe d’Etude des Lymphomes Folliculaires;<br />
LDH, lactate dehydrogenase; WBC, white blood cells.<br />
gressed. Results from the recent ECOG E4402 RESORT<br />
study have also challenged the benefit <strong>of</strong> prolonged rituximab<br />
treatment in a similar group <strong>of</strong> patients. 24 In this trial,<br />
the administration <strong>of</strong> maintenance therapy with rituximab<br />
(one infusion every 3 months) for patients responding to 4<br />
weekly rituximab infusions was compared to rituximab<br />
retreatment at the time <strong>of</strong> progression. The primary endpoint,<br />
defined as time to treatment failure (i.e., initiation <strong>of</strong><br />
cytotoxic therapy or resistance to rituximab) was not different<br />
in the two study arms, although the costs and toxicities<br />
associated with rituximab maintenance therapy appear to<br />
be higher. Investigators in the United Kingdom study prematurely<br />
closed an experimental arm with only the 4 weekly<br />
infusions, and the 84 patients receiving this limited intervention<br />
appeared to have an intermediate outcome (80% not<br />
requiring treatment at 3 years, and 60% PFS) between those<br />
<strong>of</strong> the observation arm and those receiving rituximab maintenance<br />
therapy. 22<br />
From the most current data, the strategy <strong>of</strong> delaying the<br />
initiation <strong>of</strong> systemic therapy in patients with follicular<br />
lymphoma still remains a justified option in <strong>2012</strong>. Close<br />
follow-up both allows the identification <strong>of</strong> patients that are<br />
experiencing symptoms or rapid disease progression, and<br />
the ability to start delivering a combination <strong>of</strong> anti-CD20<br />
and chemotherapy, which has been shown to improve OS in<br />
patients requiring immediate intervention. If physicians<br />
and patients choose an earlier intervention with rituximab<br />
as a single agent, the duration <strong>of</strong> treatment should be<br />
limited (e.g., 4 weekly infusions, which could be eventually<br />
completed with 4 additional infusions).<br />
Other forms <strong>of</strong> short-treatment interventions, such as<br />
short courses <strong>of</strong> chemotherapy with rituximab combined<br />
with radioimmunotherapy, have also been investigated in<br />
phase II studies. 25-27 These studies included a variable<br />
proportion <strong>of</strong> patients with characteristics similar to those<br />
usually selected for watchful waiting. Given the heterogeneity<br />
<strong>of</strong> the patient population, the noncomparative nature <strong>of</strong><br />
these studies, and the lack <strong>of</strong> information about long-term<br />
toxicities (see below), such abbreviated treatment approaches<br />
cannot be regarded as standard first-line therapy<br />
today.<br />
489
Reference<br />
Table 2. Studies Combining Chemotherapy and Monoclonal Antibodies in Patients with Follicular Lymphoma<br />
Chemotherapy<br />
regimen N<br />
Median age,<br />
y<br />
Is There a Superior Chemotherapy Induction Regimen<br />
for Patients Requiring Immediate Intervention?<br />
In patients requiring treatment because <strong>of</strong> disease characteristics<br />
at diagnosis or because <strong>of</strong> disease progression<br />
after watchful waiting, the combination <strong>of</strong> rituximab plus<br />
chemotherapy has become the usual standard <strong>of</strong> care (Table<br />
2). Indeed, several randomized studies have demonstrated<br />
that this combination improves the OS <strong>of</strong> patients with<br />
follicular lymphoma, and a meta-analysis has indicated a<br />
significant reduction in the risk <strong>of</strong> death (HR for mortality<br />
0.63; 95% CI, 0.51 to 0.79). Different chemotherapy regimens<br />
(e.g., CVP [cyclophosphamide, vincristine, prednisone],<br />
CHOP [cyclophosphamide, doxorubicin, vincristine,<br />
prednisone], fludarabine combinations, bendamustine) associated<br />
with rituximab are available (e.g., R-CVP, R-CHOP,<br />
R-bendamustine, etc.) and the debate about their potential<br />
benefit remains unsettled.<br />
The CVP regimen has resulted in lower response rates<br />
(around 85%) and shorter times to disease progression, 28<br />
even when maintenance therapy with rituximab is delivered<br />
to patients that have responded to treatment. 29 The only<br />
randomized study comparing these regimens did not show a<br />
difference in OS between R-CVP and R-CHOP. 30 But in<br />
patients with adverse features (such as a high score on the<br />
Follicular Lymphoma International Prognostic Index<br />
[FLIPI]), the median times to progression were rather short<br />
(26 months), 28 and recent survey data suggest that patients<br />
with high-risk disease receiving R-CVP might have reduced<br />
survival compared with those receiving R-CHOP. 31 However,<br />
given the potential long-term cardiac toxicity <strong>of</strong> anthracyclines<br />
and their eventual benefit as second-line<br />
treatment, some physicians might defer the use <strong>of</strong> R-CHOP<br />
in first-line therapy.<br />
Although developed in the last 15 years, a regimen containing<br />
fludarabine in combination with mitoxantrone was<br />
recently compared with R-CVP and R-CHOP in a randomized<br />
study. 30 If PFS appeared favorable, lower OS was<br />
observed in the fludarabine arm, which was in line with data<br />
from the PRIMA study with R-FCM (rituximab plus fludarabine,<br />
cyclophosphamide, and mitoxantrone). 32 Of note, this<br />
lower survival appeared to be partly related to the toxicities<br />
<strong>of</strong> regimens containing fludarabine, underscoring the need<br />
to consider short- and long-term adverse events.<br />
Regarding bendamustine, one randomized study demonstrated<br />
a better response rate and prolonged time to progression<br />
with the use <strong>of</strong> R-bendamustine as compared to<br />
R-CHOP (HR 0.63, p � 0.028), but OS was not significantly<br />
different between groups. 33 In this study, the short-term<br />
toxicity pr<strong>of</strong>ile <strong>of</strong> R-bendamustine was found to be much<br />
better than that observed with R-CHOP. However, detailed<br />
and long-term data <strong>of</strong> this study are still unavailable.<br />
Median follow-up,<br />
months PFS<br />
Marcus, et al. (28) R-CVP 162 52 53 34 mo (median)<br />
Salles, et al. (29) R-CVP 268 57.5 42 53% a (3.5-y from registration)<br />
R-CHOP 881 56 42 66.5% a (3.5-y after randomization)<br />
Rummel, et al. (33) R-CHOP 140 60 34 48 mo (median) �55% (estimated PFS at 3 y)<br />
R-Bendamustine 139 60 34 median not reached �68% (estimated PFS at 3 y)<br />
Abbreviations: mo, months; PFS, progression-free survival; R-bendamustine, regimen <strong>of</strong> rituximab/bendamustine; R-CHOP, regimen <strong>of</strong> rituximab/cyclophosphamide/<br />
doxorubicin/vincristine/prednisone; R-CVP, regimen <strong>of</strong> rituximab/cyclophosphamide/vincristine/prednisone.<br />
a Only responders to induction therapy were randomized.<br />
490<br />
SALLES, GHESQUIÈRES, AND BACHY<br />
Another randomized study (NCT00877006) comparing<br />
R-bendamustine to R-CHOP or R-CVP has recently completed<br />
accrual.<br />
When considering the choice <strong>of</strong> regimen, both short- and<br />
long-term toxicities should be considered, as well as patient<br />
age and comorbidities. The incidence <strong>of</strong> patients with<br />
changes in histology at progression may also be worth<br />
investigating. In patients with a median survival exceeding<br />
10 years, it becomes essential to develop comparative studies<br />
examining long-term outcome, to ensure that the use <strong>of</strong> a<br />
given regimen does not preclude or increase toxicities associated<br />
with efficient second-line therapy.<br />
What Is the Role <strong>of</strong> Anti-CD20 Maintenance Therapy<br />
after Induction Chemotherapy?<br />
Because most patients experience disease recurrence<br />
within 3 to 5 years after chemotherapy with rituximab,<br />
attempts to prolong efficacy have been investigated. The<br />
PRIMA studies evaluated the addition <strong>of</strong> 2 years <strong>of</strong> rituximab<br />
maintenance therapy (i.e., one infusion every 2<br />
months) in patients with disease that responded to R-CVP,<br />
R-CHOP, or R-FCM. 29 The interim analysis demonstrated a<br />
significant reduction in the risk <strong>of</strong> lymphoma progression for<br />
those patients receiving maintenance therapy (HR � 0.5;<br />
95% CI, 0.39 to 0.64, p � 0.0001). At 4 years, PFS was still<br />
significantly better in the experimental arm than in the<br />
control arm (68% vs. 50%, respectively, p � 0.001) (Salles G,<br />
personal results, Table 3). The benefit <strong>of</strong> rituximab maintenance<br />
therapy was observed regardless <strong>of</strong> patient age,<br />
the quality <strong>of</strong> response after induction therapy (complete<br />
response/unconfirmed complete response compared with<br />
partial response), or FLIPI category (Fig. 1). During maintenance<br />
therapy with rituximab, patients experienced more<br />
frequent adverse events, especially grade 2 infections, but<br />
very few patients withdrew from the study for treatmentrelated<br />
toxicities. No differences in OS were observed between<br />
the two study arms. 29 A meta-analysis including<br />
2,586 patients from nine trials indicated that the use <strong>of</strong><br />
rituximab maintenance therapy could improve OS, although<br />
this benefit was not significant for patients receiving rituximab<br />
maintenance therapy after first-line treatment (HR �<br />
0.86; 95% CI, 0.60 to 1.25). 34 A large study is currently<br />
comparing the optimal duration <strong>of</strong> rituximab maintenance<br />
(2 years compared with 4 years) administered after<br />
R-bendamustine (NCT00877214).<br />
What Is the Role <strong>of</strong> Radioimmunotherapy<br />
in the First-Line Treatment <strong>of</strong> Patients<br />
with Follicular Lymphoma?<br />
Radioimmunotherapy has been investigated as singleagent<br />
and consolidation therapy after chemotherapy. Tosi-
TREATMENT OPTIONS IN INDOLENT LYMPHOMA<br />
Table 3. Studies Investigating an Anti-CD20 Antibody Consolidation or Maintenance Therapy after Chemotherapy with Rituximab<br />
Reference Induction txa n per arm Median<br />
PFS<br />
Intervention (control/intervention) follow-up HR (95% CI) Control Intervention<br />
Hagenbeek et al (37) Chemotherapyb 90Yibritumomab tiutexan 202/207 66 mo 0.51 (0.4–0,66) 15 mo (median) 49 mo (median)<br />
Salles et al (29) R-CVP rituximab maintenance<br />
113/109 48 mo 0.71 (0.48–1.04) 47% (4-y) 60% (4-y)<br />
Salles et al (29) R-CHOP (12 infusions/2 y)<br />
386/382 48 mo 0.49 (0.39–0,62) 50% (4-y) 70% (4-y)<br />
Press et al (38) R-CHOP Observation<br />
265/267 4.9 y NA 76% (2-y) 80% (2-y)<br />
CHOP<br />
131I tositumomab<br />
Abbreviations: CHOP, regimen <strong>of</strong> cyclophosphamide/doxorubicin/vincristine/prednisone CI, confidence interval; HR, hazard ratio; mo, months; NA, not available; PFS,<br />
progression-free survival; R-CHOP, regimen <strong>of</strong> rituximab/cyclophosphamide/doxorubicin/vincristine/prednisone; R-CVP, regimen <strong>of</strong> rituximab/cyclophosphamide/<br />
vincristine/prednisone; tx, treatment; y, year.<br />
a<br />
Before randomization.<br />
b<br />
Chemotherapy included anthracycline in 43% <strong>of</strong> patients and rituximab in 15%.<br />
tumomab administered as single-agent therapy in 76<br />
patients provided a high response rate (95%) and the 5-year<br />
PFS was 59%. 35 Unfortunately, this approach was never<br />
evaluated in a randomized study against other forms <strong>of</strong><br />
therapy. Two randomized studies (Table 3) addressed the<br />
role <strong>of</strong> radioimmunotherapy as consolidation therapy in<br />
patient responders to chemotherapy. In the FIT study, 36,37<br />
administration <strong>of</strong> a single course <strong>of</strong> ibritumomab tiutexan<br />
resulted in high rates <strong>of</strong> complete response (87.4% with<br />
either complete or unconfirmed complete response) and<br />
significant improvements in PFS (median 15 months in the<br />
observation arm compared with 49 months in the experimental<br />
arm). OS was not different between the two groups<br />
but six myeloid malignancies were observed in patients that<br />
had received radioimmunotherapy, compared with one in<br />
the control arm (p � 0.0001). 37 The major pitfall <strong>of</strong> this<br />
study was the fact that only 15% <strong>of</strong> patients had received<br />
rituximab with induction chemotherapy. The results <strong>of</strong> the<br />
SWOG0016 Intergroup study comparing R-CHOP (267 patients)<br />
with CHOP followed by tositumomab (265 patients)<br />
have been recently presented. 38 With a median follow-up <strong>of</strong><br />
4.9 years, 2-year PFS and OS were not significantly different<br />
between the groups (80% and 93%, respectively, in the<br />
radioimmunotherapy arm, 76% and 97%, respectively, in the<br />
R-CHOP arm). A slight but nonsignificant excess <strong>of</strong> secondary<br />
myeloid malignancies was observed in patients in the<br />
radioimmunotherapy arm. 38<br />
Based on these results, it is therefore difficult to establish<br />
the role <strong>of</strong> radioimmunotherapy in the first-line treatment <strong>of</strong><br />
patients with follicular lymphoma. Studies assessing the<br />
role <strong>of</strong> consolidation therapy with radiolabeled antibodies<br />
against rituximab maintenance in the context <strong>of</strong> induction<br />
chemotherapy with rituximab should be conducted.<br />
Is There a Role for High-Dose Therapy Supported<br />
with Stem Cell Transplantation?<br />
Although this strategy was investigated before the rituximab<br />
era with some indication <strong>of</strong> prolonged PFS in at least<br />
two studies, 39 autologous stem cell transplantation does not<br />
seem to improve OS. 40 This strategy therefore does not<br />
represent a standard option as consolidation therapy for<br />
patients with responsive disease in first-line settings.<br />
Future Prospects for the Management <strong>of</strong> Indolent<br />
Lymphoma in First-Line Therapy<br />
Although the results achieved with the combination <strong>of</strong><br />
chemotherapy and maintenance or consolidation therapies<br />
are favorable, with 5-year OS estimates exceeding 90%,<br />
several questions remain. The first question regards the<br />
stratification <strong>of</strong> patients, which is currently performed with<br />
clinical criteria. It is our hope that new biologic tools will<br />
translate recent advances in the understanding <strong>of</strong> the biology<br />
<strong>of</strong> follicular lymphoma into clinical practice and provide<br />
useful predictors <strong>of</strong> patient outcome that will help tailor<br />
therapy. 41 A second question lies in the evaluation <strong>of</strong> treatment<br />
efficacy and the benefit <strong>of</strong> achieving complete response<br />
to first-line therapy. 42 To improve the accuracy <strong>of</strong> this<br />
evaluation, minimal residual disease has not gained a wide<br />
acceptance outside clinical trials given its technical limitations.<br />
But recent studies underline the value <strong>of</strong> 18 Ffluorodeoxy-D-glucose<br />
positron emission tomography in<br />
deciphering the quality <strong>of</strong> response in patients that achieved<br />
Fig 1. Outcome <strong>of</strong> patients randomly assigned to observation or rituximab maintenance therapy in the PRIMA trial according to the Follicular<br />
Lymphoma Prognostic Index (FLIPI).<br />
Abbreviations: PFS, progression free survival; NA, not applicable.<br />
491
partial or incomplete response. 43,44 Whether modifying<br />
treatment for patients with disease that does not achieved<br />
complete metabolic response will improve outcomes has yet<br />
to be determined. Finally, the emergence <strong>of</strong> new biologic<br />
Authors’ Disclosures <strong>of</strong> Potential Conflicts <strong>of</strong> Interest<br />
Author<br />
Employment or<br />
Leadership<br />
Positions<br />
Consultant or<br />
Advisory Role<br />
Gilles Salles Calistoga<br />
Pharmaceuticals;<br />
Celgene;<br />
Janssen-Cilag;<br />
Roche<br />
Hervé Ghesquières*<br />
Emmanuel Bachy*<br />
*No relevant relationships to disclose.<br />
1. Pulte D, Gondos A, Brenner H. Expected long-term survival <strong>of</strong> older<br />
patients diagnosed with non-Hodgkin lymphoma in 2008-<strong>2012</strong>. Cancer Epidemiol.<br />
<strong>2012</strong>;36:e19-25.<br />
2. Fisher RI, LeBlanc M, Press OW, et al. New treatment options have<br />
changed the survival <strong>of</strong> patients with follicular lymphoma. J Clin Oncol.<br />
2005;23:8447-8452.<br />
3. Sebban C, Brice P, Delarue R, et al. Impact <strong>of</strong> rituximab and/or<br />
high-dose therapy with autotransplant at time <strong>of</strong> relapse in patients with<br />
follicular lymphoma: A GELA study. J Clin Oncol. 2008;26:3614-3620.<br />
4. Friedberg JW, Taylor MD, Cerhan JR, et al. Follicular lymphoma in the<br />
United States: first report <strong>of</strong> the national LymphoCare study. J Clin Oncol.<br />
2009;27:1202-1208.<br />
5. Bertoni F, Coiffier B, Salles G, et al. MALT lymphomas: pathogenesis<br />
can drive treatment. <strong>Oncology</strong>. 2011;25:1134-1142, 1147.<br />
6. Peinert S, Seymour JF. Indolent lymphomas other than follicular and<br />
marginal zone lymphomas. Hematol Oncol Clin North Am. 2008;22:903-940.<br />
7. Vaughan Hudson B, Vaughan Hudson G, MacLennan KA, et al. <strong>Clinical</strong><br />
stage 1 non-Hodgkin’s lymphoma: long-term follow-up <strong>of</strong> patients treated by<br />
the British National Lymphoma Investigation with radiotherapy alone as<br />
initial therapy. Br J Cancer. 1994;69:1088-1093.<br />
8. Mac Manus MP, Hoppe RT. Is radiotherapy curative for stage I and II<br />
low-grade follicular lymphoma? Results <strong>of</strong> a long-term follow-up study <strong>of</strong><br />
patients treated at Stanford University. J Clin Oncol. 1996;14:1282-1290.<br />
9. Advani R, Rosenberg SA, Horning SJ. Stage I and II follicular non-<br />
Hodgkin’s lymphoma: long-term follow-up <strong>of</strong> no initial therapy. J Clin Oncol.<br />
2004;22:1454-1459.<br />
10. Pugh TJ, Ballon<strong>of</strong>f A, Newman F, et al. Improved survival in patients<br />
with early stage low-grade follicular lymphoma treated with radiation: a<br />
Surveillance, Epidemiology, and End Results database analysis. Cancer.<br />
2010;116:3843-3851.<br />
11. Friedberg J, Byrtek M, Link B, et al. Should radiation therapy (Xrt) be<br />
the standard therapeutic approach for stage I follicular lymphoma (FL)? A<br />
comparative effectiveness analysis <strong>of</strong> the National Lymphocare Study (Nlcs).<br />
In: Internationational Conference on Malignat Lymphoma; 2011; Lugano;<br />
2011;p. iv91.<br />
12. Dreyling M, Ghielmini M, Marcus R, et al. Newly diagnosed and<br />
relapsed follicular lymphoma: ESMO <strong>Clinical</strong> Practice Guidelines for diagnosis,<br />
treatment and follow-up. Ann Oncol. 2011;22:vi59-63 (suppl 6).<br />
13. Zelenetz AD, Abramson JS, Advani RH, et al. NCCN <strong>Clinical</strong> Practice<br />
Guidelines in <strong>Oncology</strong>: Non-Hodgkin’s lymphomas. JNCCN. 2010;8:288-334.<br />
14. Lowry L, Smith P, Qian W, et al. Reduced dose radiotherapy for local<br />
control in non-Hodgkin lymphoma: a randomised phase III trial. Radiother<br />
Oncol. 2011;100:86-92.<br />
15. Young RC, Longo DL, Glatstein E, et al. The treatment <strong>of</strong> indolent<br />
lymphomas: watchful waiting v aggressive combined modality treatment.<br />
Semin Hematol. 1988;25:11-16 (suppl 2).<br />
16. Brice P, Bastion Y, Lepage E, et al. Comparison in low-tumor-burden<br />
follicular lymphomas between an initial no-treatment policy, prednimustine,<br />
or interferon alfa: a randomized study from the Groupe d’Etude des Lymphomes<br />
Folliculaires. Groupe d’Etude des Lymphomes de l’Adulte. J Clin<br />
Oncol. 1997;15:1110-1117.<br />
17. Ardeshna KM, Smith P, Norton A, al. E. Long-term effect <strong>of</strong> a watch<br />
and wait policy versus immediate systemic treatment for asymptomatic<br />
492<br />
therapies, such engineered monoclonal antibodies, 45 immunomodulatory<br />
agents, 46 or kinase inhibitors, 47 represent<br />
innovative options that might change the current treatment<br />
landscape.<br />
Stock<br />
Ownership Honoraria<br />
REFERENCES<br />
Celgene;<br />
Janssen-Cilag;<br />
Pfizer; Roche<br />
Research<br />
Funding<br />
SALLES, GHESQUIÈRES, AND BACHY<br />
Expert<br />
Testimony<br />
Other<br />
Remuneration<br />
advanced-stage non-Hodgkin lymphoma: a randomised controlled trial. Lancet.<br />
2003;362:516-522.<br />
18. Hainsworth JD, Burris HA, 3rd, Morrissey LH, et al. Rituximab<br />
monoclonal antibody as initial systemic therapy for patients with low-grade<br />
non-Hodgkin lymphoma. Blood. 2000;95:3052-3056.<br />
19. Colombat P, Salles G, Brousse N, et al. Rituximab (anti-CD20 monoclonal<br />
antibody) as single first-line therapy for patients with follicular<br />
lymphoma with a low tumor burden: clinical and molecular evaluation. Blood.<br />
2001;97:101-106.<br />
20. Ghielmini M, Schmitz SF, Cogliatti S, et al. Effect <strong>of</strong> single-agent<br />
rituximab given at the standard schedule or as prolonged treatment in<br />
patients with mantle cell lymphoma: a study <strong>of</strong> the Swiss Group for <strong>Clinical</strong><br />
Cancer Res (SAKK). J Clin Oncol. 2005;23:705-711.<br />
21. Martinelli G, Schmitz SF, Utiger U, et al. Long-term follow-up <strong>of</strong><br />
patients with follicular lymphoma receiving single-agent rituximab at two<br />
different schedules in trial SAKK 35/98. J Clin Oncol. 2010;28:4480-4484.<br />
22. Ardeshna KM, Smith P, Qian W, et al. An intergroup randomised trial<br />
<strong>of</strong> rituximab versus a watch and wait strategy in patients with stage II, III,<br />
IV, asymptomatic, non-bulky follicular lymphoma (grades 1, 2 and 3a). A<br />
preliminary analysis. ASH Annual Meeting Abstracts. 2011;116 (abstr 6).<br />
23. Ardeshna KM, Qian W, Stephens R, et al. Preliminary Results Of<br />
Quality Of Life (Qol) Analyses From The Intergroup Phase III Randomised<br />
Trial Of Rituximab Vs A Watch And Wait Approach In Patients With<br />
Advanced Stage, Asymptomatic, Non-Bulky Follicular Lymphoma (FL). Ann<br />
Oncol. 2011;22:iv88 (suppl 4).<br />
24. Kahl BS, Hong F, Williams ME, et al. Results <strong>of</strong> Eastern Cooperative<br />
<strong>Oncology</strong> Group Protocol E4402 (RESORT): a randomized phase III study<br />
comparing two different rituximab dosing strategies for low tumor burden<br />
follicular lymphoma. ASH Annual Meeting Abstracts. 2011:118 (suppl LBA-<br />
6).<br />
25. Leonard JP, Coleman M, Kostakoglu L, et al. Abbreviated chemotherapy<br />
with fludarabine followed by tositumomab and iodine I 131 tositumomab<br />
for untreated follicular lymphoma. J Clin Oncol. 2005;23:5696-5704.<br />
26. Zinzani PL, Tani M, Pulsoni A, et al. Fludarabine and mitoxantrone<br />
followed by yttrium-90 ibritumomab tiuxetan in previously untreated patients<br />
with follicular non-Hodgkin lymphoma trial: a phase II nonrandomised<br />
trial (FLUMIZ). Lancet Oncol. 2008;9:352-358.<br />
27. Link BK, Martin P, Kaminski MS, et al. Cyclophosphamide, vincristine,<br />
and prednisone followed by tositumomab and iodine-131-tositumomab in<br />
patients with untreated low-grade follicular lymphoma: eight-year follow-up<br />
<strong>of</strong> a multicenter phase II study. J Clin Oncol. 2010;28:3035-3041.<br />
28. Marcus R, Imrie K, Solal-Celigny P, et al. Phase III study <strong>of</strong> R-CVP<br />
compared with cyclophosphamide, vincristine, and prednisone alone in patients<br />
with previously untreated advanced follicular lymphoma. J Clin Oncol.<br />
2008;26:4579-4586.<br />
29. Salles G, Seymour JF, Offner F, et al. Rituximab maintenance for 2<br />
years in patients with high tumour burden follicular lymphoma responding to<br />
rituximab plus chemotherapy (PRIMA): a phase 3, randomised controlled<br />
trial. Lancet. 2011;377:42-51.<br />
30. Federico M, Luminari S, Dondi A, et al. R-CVP vs R-CHOP vs R-FM for<br />
the initial treatment <strong>of</strong> patients with advanced stage follicular lymphoma.<br />
Preliminary results <strong>of</strong> Foll05 Iil Trial. Ann Oncol. 2011;22:iv128 (suppl 4).<br />
31. Nastoupil L, Sinha R, Byrtek M, et al. A comparison <strong>of</strong> the effectiveness
TREATMENT OPTIONS IN INDOLENT LYMPHOMA<br />
<strong>of</strong> first-line chemoimmunotherapy regimens for follicular lymphoma (FL)<br />
used in the United States. ASH Annual Meeting Abstracts. 2011;118 (abstr<br />
97).<br />
32. Morschhauser F, Seymour JF, Feugier P, et al. Impact <strong>of</strong> induction<br />
chemotherapy regimen on response, safety and outcome in the Prima study.<br />
Ann Oncol. 2011;22:iv88 (suppl 4).<br />
33. Rummel MJ, Niederle N, Maschmeyer G, et al. Bendamustine plus<br />
rituximab is superior in respect <strong>of</strong> progression free survival and CR rate when<br />
compared to CHOP plus rituximab as first-line treatment <strong>of</strong> patients with<br />
advanced follicular, indolent, and mantle cell lymphomas: final results <strong>of</strong> a<br />
randomized phase III study <strong>of</strong> the StiL (Study Group Indolent Lymphomas,<br />
Germany). ASH Annual Meeting Abstracts. 2009;114 (abstr 405).<br />
34. Vidal L, Gafter-Gvili A, Salles G, et al. Rituximab maintenance for the<br />
treatment <strong>of</strong> patients with follicular lymphoma: an updated systematic<br />
review and meta-analysis <strong>of</strong> randomized trials. J Natl Cancer Inst. 2011;103:<br />
1799-1806.<br />
35. Kaminski MS, Tuck M, Estes J, et al. 131I-tositumomab therapy as<br />
initial treatment for follicular lymphoma. N Engl J Med. 2005;352:441-449.<br />
36. Morschhauser F, Radford J, Van Ho<strong>of</strong> A, et al. Phase III trial <strong>of</strong><br />
consolidation therapy with yttrium-90-ibritumomab tiuxetan compared with<br />
no additional therapy after first remission in advanced follicular lymphoma.<br />
J Clin Oncol. 2008;26:5156-5164.<br />
37. Hagenbeek A, Radford J, Van Ho<strong>of</strong> A, et al. 90Y-Ibritumomab tiuxetan<br />
(Zevalin(R)) consolidation <strong>of</strong> first remission in advanced-stage follicular<br />
non-Hodgkin’s lymphoma: updated results after a median follow-up <strong>of</strong> 66.2<br />
months from the international, randomized, phase III First-Line Indolent<br />
Trial (FIT) in 414 Patients. ASH Annual Meeting Abstracts. 2010;116 (abstr<br />
594).<br />
38. Press OW, Unger JM, Rimsza LM, et al. A phase III randomized<br />
intergroup trial (SWOG S0016) <strong>of</strong> CHOP chemotherapy plus rituximab vs.<br />
CHOP chemotherapy plus iodine-131-tositumomab for the treatment <strong>of</strong> newly<br />
diagnosed follicular non-Hodgkin’s lymphoma. ASH Annual Meeting Abstracts.<br />
2011;118 (abstr 98).<br />
39. Bachy E, Salles G. Marrow-ablative treatment and autologous stem cell<br />
transplantation in follicular NHL. Best Pract Res Clin Haematol. 2011;24:<br />
257-270.<br />
40. Al Khabori M, de Almeida JR, Guyatt GH, et al. Autologous stem cell<br />
transplantation in follicular lymphoma: a systematic review and metaanalysis.<br />
J Natl Cancer Inst. <strong>2012</strong>;104:18-28.<br />
41. Relander T, Johnson NA, Farinha P, et al Prognostic factors in<br />
follicular lymphoma. J Clin Oncol. 2010;28:2902-2913.<br />
42. Bachy E, Brice P, Fournier M, et al. Long term follow-up <strong>of</strong> the GELF86<br />
French and Belgian trials: complete remission after first line treatment with<br />
conventional chemotherapy in newly diagnosed follicular lymphoma patients<br />
correlates with prolonged overall survival compared with partial remission.<br />
ASH Annual Meeting Abstracts. 2008;112 (abstr 2590).<br />
43. Trotman J, Fournier M, Lamy T, et al. Positron emission tomographycomputed<br />
tomography (PET-CT) after induction therapy is highly predictive<br />
<strong>of</strong> patient outcome in follicular lymphoma: analysis <strong>of</strong> PET-CT in a subset <strong>of</strong><br />
PRIMA trial participants. J Clin Oncol. 2011;29:3194-200.<br />
44. Dupuis J, Meignan M, Julian A, et al. Significant prognostic impact <strong>of</strong><br />
[18F] fluorodeoxyglucose-PET scan performed during and at the end <strong>of</strong><br />
treatment with R-CHOP in high-tumor mass follicular lymphoma patients: a<br />
GELA-GOELAMS study. ASH Annual Meeting Abstracts. 2011;118 (abstr<br />
877).<br />
45. Salles GA, Morschhauser F, Thieblemont C, et al. Efficacy and safety <strong>of</strong><br />
obinutuzumab (GA101) monotherapy in relapsed/refractory indolent non-<br />
Hodgkin’s lymphoma: results from a phase I/II study (BO20999). ASH Annual<br />
Meeting Abstracts. 2011;118 (abstr 268).<br />
46. Samaniego F, Hagemeister F, Mclaughlin P, et al. High response rates<br />
with lenalidomide plus rituximab for untreated indolent B-cell non-Hodgkin<br />
lymphoma, including those meeting GELF criteria. J Clin Oncol. 2011;29<br />
(suppl; abstr 8030).<br />
47. Furman RR, Byrd JC, Flinn IW, et al. Interim results from a phase I<br />
study <strong>of</strong> CAL-101, a selective oral inhibitor <strong>of</strong> phosphatidylinositol 3-kinase<br />
p110d is<strong>of</strong>orm, in patients with relapsed or refractory hematologic malignancies.<br />
J Clin Oncol.2010; 28 (suppl; abstr 3032).<br />
493
What Is the Role <strong>of</strong> Transplantation for<br />
Indolent Lymphoma?<br />
By Ryan D. Cassaday, MD, and Ajay K. Gopal, MD<br />
Overview: Despite advances in chemoimmunotherapy, indolent<br />
B-cell non-Hodgkin lymphomas (B-NHLs) are generally<br />
not considered curable with this approach. Much attention has<br />
been paid to the prospect <strong>of</strong> hematopoietic cell transplantation<br />
(HCT) as a way to improve long-term outcomes for this<br />
group <strong>of</strong> diseases. Autologous (auto) HCT provides intensive<br />
conditioning therapy followed by rescue <strong>of</strong> hematopoiesis,<br />
and this has been shown in randomized studies to prolong<br />
survival compared with more standard chemotherapy, albeit<br />
with increased short-term toxicity and the potential for higher<br />
rates <strong>of</strong> secondary malignancies. Allogeneic (allo) HCT can<br />
provide anticancer effects beyond the conditioning therapy<br />
through the immune-mediated graft-versus-lymphoma (GVL)<br />
B-NHLs are <strong>of</strong>ten classified into indolent and aggressive<br />
subtypes. Indolent lymphomas typically include several<br />
different specific histologies: follicular lymphoma (FL),<br />
lymphoplasmacytic lymphoma (LPL), small lymphocytic<br />
lymphoma (SLL), and marginal zone lymphoma (MZL). 1 The<br />
vast majority <strong>of</strong> patients present with disseminated disease<br />
requiring systemic therapy, which can improve overall survival<br />
(OS) but is not thought to be curative. Over time,<br />
indolent B-NHL becomes increasingly refractory to chemotherapy,<br />
with shallower, shorter remissions the norm.<br />
HCT is a strategy that aims to overcome the limitations <strong>of</strong><br />
standard chemoimmunotherapy in two general forms. Auto<br />
HCT relies on the steep dose-response curve observed in<br />
hematopoietic malignancies by using MA doses <strong>of</strong> chemoradiotherapy<br />
followed by auto hematopoietic stem cell rescue.<br />
In contrast, allo HCT also employs an immune-mediated<br />
GVL effect to eradicate disease following preparative regimens<br />
ranging in intensity from high-dose MA combinations<br />
to very low-dose (2 Gy) radiation. The less intensive conditioning<br />
strategies are designed to suppress the patient’s<br />
immune system sufficiently to allow engraftment <strong>of</strong> the<br />
donor hematopoietic cells and relies almost exclusively on<br />
the GVL effect. 2,3 However, this same immunologic phenomenon<br />
can affect normal healthy tissues in the patient,<br />
inducing GVHD—the major source <strong>of</strong> morbidity and mortality<br />
<strong>of</strong> this approach. 4 Nevertheless, allo HCT remains a<br />
viable option for select patients as it can <strong>of</strong>fer durable<br />
remissions and the potential for cure in patients with<br />
relapsed or refractory indolent lymphomas.<br />
Indications for Auto HCT<br />
Initial Consolidation<br />
Auto HCT is generally not recommended as initial consolidation<br />
<strong>of</strong> response for any indolent B-NHL outside the<br />
context <strong>of</strong> a clinical trial. This conclusion is supported by at<br />
least four prospective randomized controlled trials (RCTs)<br />
that investigated this approach as part <strong>of</strong> first-line management<br />
(Table 1). 5-8 Although most <strong>of</strong> these studies show<br />
improved progression-free survival (PFS) or event-free survival<br />
(EFS) when compared with standard chemotherapy,<br />
overall survival (OS) has not been shown to be significantly<br />
improved (p � 0.5 in the 3 studies where this comparison<br />
was reported). Moreover, there appeared to be an increased<br />
494<br />
effect. It can be administered following myeloablative (MA)<br />
conditioning or reduced-intensity (RI) regimens aimed at sufficiently<br />
suppressing the patient’s immune system to allow<br />
engraftment <strong>of</strong> donor hematopoiesis. However, this same<br />
potentially curative alloreactivity <strong>of</strong> the engrafted immune<br />
system can lead to graft-versus-host disease (GVHD), a significant<br />
cause <strong>of</strong> morbidity and mortality following allo HCT.<br />
This article will discuss the current role <strong>of</strong> both auto HCT and<br />
allo HCT in the management <strong>of</strong> indolent lymphoma as well as<br />
the relative risks and benefits <strong>of</strong> each approach such that the<br />
reader can place this in context <strong>of</strong> the multitude <strong>of</strong> options<br />
available for patients with indolent B-NHL.<br />
risk <strong>of</strong> secondary malignancies (both myeloid and solid<br />
tumors) associated with auto HCT, which is becoming increasingly<br />
relevant in indolent lymphoma as survival has<br />
continued to improve with modern chemoimmunotherapy<br />
regimens. Though select retrospective analyses have implicated<br />
etoposide, alkylating agents, and total body irradiation<br />
as increasing this risk, the study by Sebban and others<br />
used these treatments as part <strong>of</strong> the transplant arm and saw<br />
no increased incidence <strong>of</strong> secondary cancers compared to<br />
the non-transplant arm. 8 In the trial that noted the highest<br />
rates <strong>of</strong> secondary malignancies, Gyan and colleagues postulated<br />
that the practice <strong>of</strong> in vitro purging <strong>of</strong> B lymphocytes<br />
may have been responsible for this finding, perhaps by<br />
affecting post-transplant immunosurveillance. 5 This process<br />
is used rarely today with the application <strong>of</strong> in vivo B-cell<br />
purging by rituximab.<br />
Relapsed Disease<br />
The major role <strong>of</strong> auto HCT remains its use in patients<br />
with relapsed but chemotherapy-sensitive indolent B-NHL.<br />
Unfortunately, most <strong>of</strong> the data supporting this approach<br />
come from single-arm studies suggesting that a subset <strong>of</strong><br />
patients can again achieve long-term remissions with<br />
this treatment. For example, in the previously mentioned<br />
study by Ladetto et al, patients initially treated in the<br />
chemotherapy-only arm were allowed to cross over to auto<br />
HCT at the time <strong>of</strong> relapse. 6 Among the 28 patients salvaged<br />
with this approach, the 3-year projected EFS and OS were<br />
68% and 81%, respectively. Likewise, a retrospective study<br />
from the United Kingdom in which patients received auto<br />
HCT for FL in second or later remission showed a plateau in<br />
the freedom from progression curve at 48% at 12 years<br />
(median follow-up <strong>of</strong> 13.5 years), <strong>of</strong>fering further evidence<br />
that relapsed FL can be salvaged with auto HCT. 9 To date,<br />
there has been a single small, but important, prospective<br />
From the <strong>Clinical</strong> Research Division, Fred Hutchinson Cancer Research Center, Seattle,<br />
WA; Division <strong>of</strong> Medical <strong>Oncology</strong>, Department <strong>of</strong> Medicine, University <strong>of</strong> Washington,<br />
Seattle, WA.<br />
Authors’ disclosures <strong>of</strong> potential conflicts <strong>of</strong> interest are found at the end <strong>of</strong> this article.<br />
Address reprint requests to Ajay Gopal, MD, 1100 Fairview Ave. N., Mailstop G3200,<br />
Seattle, WA 98109; email: agopal@uw.edu.<br />
© <strong>2012</strong> by <strong>American</strong> <strong>Society</strong> <strong>of</strong> <strong>Clinical</strong> <strong>Oncology</strong>.<br />
1092-9118/10/1-10
ROLE OF HCT FOR INDOLENT LYMPHOMA<br />
Table 1. Summary <strong>of</strong> Results from Selected Randomized Controlled Trials <strong>of</strong> Autologous Hematopoietic Cell Transplantation for<br />
Follicular Lymphoma<br />
Study<br />
(Sample Size) Induction Therapy ^<br />
Lenz et al 20047 (n � 307)<br />
Sebban et al 8 2006<br />
(n � 401)<br />
Ladetto et al 6 2008<br />
(n � 136)<br />
Gyan et al 5 2009<br />
(n � 166)<br />
Schouten et al 10 2003<br />
(n � 89)<br />
RCT that evaluated auto HCT for relapsed, chemotherapysensitive<br />
FL showing a significant benefit over standard<br />
chemotherapy in both PFS (p � 0.001) and OS (p � 0.026)<br />
with nearly 6 years <strong>of</strong> follow-up (Table 1). 10<br />
Therefore, these data suggest that for most patients with<br />
FL, auto HCT should be considered part <strong>of</strong> the treatment <strong>of</strong><br />
relapse. Outside <strong>of</strong> the previously mentioned studies focusing<br />
on FL, there are little data describing the utility <strong>of</strong> auto<br />
KEY POINTS<br />
● Indolent lymphomas are not considered curable with<br />
standard chemotherapy approaches, but hematopoietic<br />
cell transplantation (HCT) can be considered for<br />
selected patients with relapsed or refractory indolent<br />
lymphoma.<br />
● In randomized studies compared with standard nontransplant<br />
approaches, autologous HCT as initial<br />
consolidation has shown improved progression-free<br />
(PFS) survival but not significantly better overall<br />
survival (OS).<br />
● Autologous transplantation for relapsed chemotherapy-sensitive<br />
indolent lymphoma has been<br />
shown to improve both PFS and OS in a single phase<br />
III trial.<br />
● Allogeneic HCT can <strong>of</strong>fer durable remissions even for<br />
patients who have relapsed after autologous HCT.<br />
● The use <strong>of</strong> reduced-intensity conditioning before allogeneic<br />
HCT may mitigate some <strong>of</strong> the treatmentrelated<br />
toxicity, but graft-versus-host disease<br />
remains a significant cause <strong>of</strong> morbidity and mortality<br />
with this approach.<br />
Consolidation Therapy ^<br />
HCT in other indolent B-NHL. Approximately 11% <strong>of</strong> the<br />
patients enrolled in the RCT from the German Low Grade<br />
Lymphoma Study Group had LPL: this subgroup had a<br />
5-year PFS <strong>of</strong> 65% after auto HCT compared with 73% with<br />
maintenance interferon-alpha (p � 0.98). 7 The National<br />
Comprehensive Cancer Network (NCCN) recommends considering<br />
the use <strong>of</strong> auto HCT to consolidate a second or later<br />
remission for FL and MZL and for salvage treatment <strong>of</strong><br />
LPL. 11 Beyond the NCCN guidelines, the Fourth International<br />
Workshop on Waldenstrom’s Macroglobulinemia reviewed<br />
the available data and generated treatment<br />
recommendations for LPL. 12 These guidelines go into<br />
slightly more detail regarding the potential role for HCT,<br />
but this group’s conclusions were similar to that <strong>of</strong> NCCN,<br />
stating that prospective trials are needed. Open questions<br />
in the field <strong>of</strong> auto HCT for indolent B-NHL include the<br />
optimal conditioning regimen (chemotherapy alone compared<br />
with radiation plus chemotherapy) and the role <strong>of</strong><br />
high-dose therapy in patients with rituximab-refractory<br />
disease. This latter point is particularly interesting, given<br />
that this agent was not widely available at the time that<br />
most <strong>of</strong> the above-noted prospective studies began. The<br />
study by Ladetto and others showed no improvement in OS<br />
with initial consolidative auto HCT when both groups received<br />
rituximab, but did not address the issue <strong>of</strong> postrituximab<br />
relapse. 6 One could hypothesize that intensive<br />
chemoradiotherapy may have a greater relative benefit in<br />
rituximab-refractory, yet chemosensitive indolent lymphoma,<br />
though prospective randomized trials evaluating<br />
this specific scenario are needed to confirm this premise.<br />
Indications for Allo HCT<br />
EFS/PFS OS<br />
Secondary<br />
Malignancies<br />
CHOP or MCP �4–6<br />
Disease Setting: Initial Consolidation<br />
Dexa-BEAM, then Cy/TBI-Auto HCT 65% NR NR 4.2 yrs<br />
CHOP or MCP �2, then IFN 33% NR NR<br />
p � 0.001 NR NR<br />
CHOP �4 CE, then CE/TBI-Auto HCT 38% 76% 6% 7.7 yrs<br />
CHVP � IFN �6 CHVP � IFN �6 28% 71% 7%<br />
p � 0.11 p � 0.53 NR<br />
APO �2 � DHAP �2* CE � R, then Mito/Mel-Auto HCT 61% 81% 8% 4.2 yrs<br />
CHOP �6, then R �4 None 28% 80% 3%<br />
p � 0.001 p � 0.96 NR<br />
VCAP �2–3 � IMV �1 or DHAP �2–3* TBI/Cy-Auto HCT 56% 76% 14% 9 yrs<br />
CHVP �6 CHVP �6, IFN � 18 mo 39% 80% 1%<br />
Disease Setting: Relapsed Disease<br />
p � 0.03 p � 0.55 NR<br />
CHOP or MIME �3 Cy/TBI-Auto HCT 57% 70% NR 5.8 yrs<br />
CHOP or MIME �3 17% 42% NR<br />
p � 0.001 p � 0.026 NR<br />
Abbreviations: EFS, event-free survival; PFS, progression-free survival; OS, overall survival; F/U, follow-up; CHOP, cyclophosphamide, doxorubicin, vincristine, and<br />
prednisone; MCP, mitoxantrone, chlorambucil, and prednisone; Dexa-BEAM, dexamethasone, carmustine, etoposide, cytarabine, and melphalan; Cy, cyclophosphamide;<br />
TBI, total body irradiation; auto, autologous; HCT, hematopoietic cell transplantation; IFN, interferon-alpha; NR, not reported; yrs, years; CE, cyclophosphamide<br />
and etoposide; CHVP, cyclophosphamide, doxorubicin, teniposide, and prednisolone; APO, doxorubicin, prednisone, and vincristine; DHAP, dexamethasone,<br />
cytarabine, and cisplatin; R, rituximab; mito, mitoxantrone; mel, melphalan; VCAP, vindesine, cyclophosphamide, doxorubicin, and prednisone; IMV, ifosfamide,<br />
methotrexate, and etoposide; MIME, mesna, ifosfamide, mitoxantrone, and etoposide.<br />
^ The numbers following the chemotherapy regimen abbreviations denote the number <strong>of</strong> cycles administered.<br />
* In these studies, DHAP was given only to subjects who did not achieve a sufficient response to the initial chemotherapy (in the study by Gyan et al 5 , DHAP was given<br />
instead <strong>of</strong> IMV).<br />
Median<br />
F/U<br />
Although allo HCT has an established role in the management<br />
<strong>of</strong> myeloid neoplasms, its exact place in the treatment<br />
<strong>of</strong> lymphoid malignancy is less clear. As stated above, the<br />
495
potential for GVL-mediated disease eradication does make it<br />
an attractive option for patients with increasingly shorter<br />
remission durations or young patients with relapsed disease<br />
where the likelihood <strong>of</strong> decades <strong>of</strong> disease control with<br />
currently available standard therapies is low. The associated<br />
toxicities (particularly GVHD) and the limitations<br />
imposed by finding human leukocyte antigen (HLA)matched<br />
donors reduce the number <strong>of</strong> potential candidates<br />
for this treatment. For this reason, the NCCN guidelines<br />
reserve allo HCT for highly selected patients beyond second<br />
relapse. 11 This section will discuss the literature available<br />
that compares auto HCT with allo HCT, the use <strong>of</strong> MA or RI<br />
conditioning, and options for treating relapse after allo HCT.<br />
Auto HCT Compared with Allo HCT<br />
There have been no sufficiently powered RCTs addressing<br />
the relative benefits <strong>of</strong> auto compared with allo HCT for<br />
indolent lymphoma. There have been several analyses comparing<br />
outcomes <strong>of</strong> these two approaches, but they are<br />
obviously limited by their nonrandomized and typically<br />
retrospective design. For example, allo HCT is reserved for<br />
later in the treatment course (<strong>of</strong>ten having relapsed after<br />
auto HCT), so this generally selects for more advanced or<br />
chemorefractory disease. Alternatively, because it is understood<br />
that allo HCT is a higher-risk treatment, it may not be<br />
<strong>of</strong>fered to patients felt to be more frail. It is important to<br />
appreciate such biases when assessing studies <strong>of</strong> this type,<br />
but several potentially useful themes do emerge when reviewing<br />
the results.<br />
The largest auto HCT/MA allo HCT comparison included<br />
904 patients with FL from the International Bone Marrow<br />
Transplant Registry/Autologous Bone Marrow Transplant<br />
Registry (IBMTR/ABMTR). 13 These were patients who received<br />
transplantation during the 1990s, before the widespread<br />
use <strong>of</strong> rituximab and RI allo HCT. The patients<br />
receiving allo HCT were more likely to have worse performance<br />
status, abnormal lactate dehydrogenase (LDH) levels,<br />
advanced disease, and chemotherapy-resistant disease.<br />
As mentioned above, the relative risk <strong>of</strong> TRM with allo HCT<br />
was significantly higher compared with auto HCT (p �<br />
0.001), while the relative risk <strong>of</strong> relapse was significantly<br />
lower with allo HCT (p � 0.03). This yielded similar OS<br />
between the groups, though there were more long-term<br />
disease-free survivors with few late relapses in the allo HCT<br />
cohort. Similar findings were reported in a smaller retrospective<br />
analysis from investigators in the United Kingdom.<br />
14 Interestingly, both studies suggested lower rates <strong>of</strong><br />
secondary malignancies in the allo HCT group potentially<br />
because <strong>of</strong> the allogeneic graft replacing damaged host<br />
hematopoiesis; though confounding factors—such as the use<br />
total-body irradiation conditioning regimens—could have<br />
contributed to these differences. 13,14<br />
A prospective cohort study <strong>of</strong> auto HCT compared with<br />
allo HCT in 216 patients with FL from the NCCN Outcomes<br />
Database Project was recently presented in abstract form. 15<br />
Importantly, these patients were all treated in the postrituximab<br />
era. Additionally, the patients receiving allo HCT<br />
received a mix <strong>of</strong> MA and RI conditioning, though the groups<br />
were reported in aggregate. These investigators found that<br />
allo HCT recipients were generally younger and more heavily<br />
pretreated than those in the auto HCT cohort. With a<br />
median follow-up <strong>of</strong> 2.9 years, they noted an overall nonrelapse<br />
mortality (NRM) rate <strong>of</strong> 10% with auto HCT compared<br />
496<br />
CASSADAY AND GOPAL<br />
with 33% with allo HCT (p � 0.0001). The 3-year estimate <strong>of</strong><br />
failure-free survival (with “failure” defined as relapse, transformation,<br />
disease progression, or death) was not different<br />
between the two groups (p � 0.3), with a rate <strong>of</strong> 55% for auto<br />
HCT compared with 56% for allo HCT. However, the 3-year<br />
estimate <strong>of</strong> OS was significantly different (p � 0.001), with<br />
a rate <strong>of</strong> 85% for auto HCT compared with 64% for allo HCT.<br />
Using a multivariate analysis to adjust for age, number <strong>of</strong><br />
prior therapies, and disease status at the time <strong>of</strong> HCT, allo<br />
HCT was still associated with an increased risk <strong>of</strong> death<br />
(hazard ratio [HR] � 2.2, p � 0.01). Taken together, these<br />
studies demonstrate the potential to control relapsed or<br />
refractory disease with allo HCT, but the relative toxicities<br />
compared with that <strong>of</strong> auto HCT make it a less attractive<br />
option earlier in the disease course.<br />
Lastly, the Bone Marrow Transplant <strong>Clinical</strong> Trials Network<br />
performed a prospective study comparing auto HCT<br />
with RI allo HCT in patients with relapsed but<br />
chemotherapy-sensitive FL, but it closed early because <strong>of</strong><br />
slow accrual. 16 This study used a biologic treatment assignment,<br />
where patients with an HLA-identical sibling received<br />
RI allo HCT, and those without received auto HCT. Because<br />
it closed early, statistical comparisons between the groups<br />
could not be performed. Nevertheless, they did find that the<br />
3-year OS and PFS for auto HCT was 73% and 63% (respectively;<br />
20 patients), compared with 100% and 86% (respectively;<br />
7 patients) for RI allo HCT. Although firm conclusions<br />
cannot be drawn from this study, these data do suggest at<br />
least comparable short-term efficacy between auto HCT and<br />
RI allo HCT for chemotherapy-sensitive relapsed FL.<br />
MA Allo HCT Compared with RI Allo HCT<br />
Given the relative risks and benefits <strong>of</strong> allo HCT, it stands<br />
to reason that this treatment approach would <strong>of</strong>fer more<br />
benefit if the toxicity could somehow be mitigated. With the<br />
advent <strong>of</strong> RI conditioning regimens, this has become an<br />
appealing modality for management <strong>of</strong> relapsed or refractory<br />
indolent lymphoma. 2,3 Biologically, the growth kinetics<br />
<strong>of</strong> this group <strong>of</strong> diseases would seem to make them amenable<br />
to an approach in which the bulk <strong>of</strong> antimalignancy effects<br />
are mediated by GVL. In particular, diseases that progress<br />
more slowly may be less likely to outpace the engraftment<br />
<strong>of</strong> the donor immune system, which is required to see GVL,<br />
and less likely to require high-dose conditioning to induce<br />
early disease control. Although limited prospective data<br />
exist to support this rationale, again there are retrospective<br />
analyses that can <strong>of</strong>fer some evidence that this hypothesis is<br />
correct.<br />
The largest <strong>of</strong> these studies for indolent lymphoma comparing<br />
RI with MA conditioning was a review <strong>of</strong> the IBMTR/<br />
ABMTR registry. 17 They compiled 120 MA and 88 RI cases,<br />
all <strong>of</strong> whom had FL. Patients in the RI cohort were noted to<br />
be significantly older and later in the disease course, while<br />
the MA cohort more frequently included patients with primary<br />
refractory disease and bone marrow involvement.<br />
Even though the type <strong>of</strong> conditioning did not correlate with<br />
risk <strong>of</strong> TRM, PFS, or OS, recipients <strong>of</strong> RI allo HCT did have<br />
a significantly higher risk <strong>of</strong> progression on multivariate<br />
analysis (p � 0.044). These results are in contrast to a cohort<br />
<strong>of</strong> 220 patients (84 <strong>of</strong> whom had indolent disease including<br />
chronic lymphocytic leukemia [CLL] and T-cell lymphomas)<br />
treated with either MA or RI allo HCT at Fred Hutchinson<br />
Cancer Research Center (FHCRC). 18 Among patients with
ROLE OF HCT FOR INDOLENT LYMPHOMA<br />
Table 2. Summary <strong>of</strong> Results from Selected Prospective Trials <strong>of</strong> Reduced-Intensity Allogeneic Hematopoietic Cell Transplantation<br />
for Indolent Lymphoma<br />
Study<br />
(Sample Size)<br />
Morris et al23 2004<br />
(n � 41 low-grade^ ; 88 total)*<br />
Rezvani et al22 2008<br />
(n � 62; 16 with HT)<br />
Khouri et al21 2008<br />
(n � 47, all FL)<br />
Piñana et al24 2010<br />
(n � 37, all FL)<br />
Thomson et al26 2010<br />
(n � 82, all FL)*<br />
Shea et al25 2011<br />
(n � 16 FL; 47 total)<br />
Conditioning<br />
Regimen<br />
GVHD<br />
Prophylaxis<br />
indolent histologies, there was a significantly higher NRM<br />
(p � 0.02), a trend toward higher overall mortality (p �<br />
0.07), and a nonsignificantly lower risk <strong>of</strong> relapse after MA<br />
(p � 0.33) compared with RI allo HCT. These authors also<br />
used the Hematopoietic Cell Transplant-Comorbidity Index<br />
(HCT-CI) to estimate risk based on pretransplant comorbidities,<br />
but this analysis was performed for the entire study<br />
cohort, irrespective <strong>of</strong> histology. 19 In patients with a lowrisk<br />
HCT-CI score, there was no difference in NRM, relapse<br />
rate, or OS at 3 years between patients receiving RI or MA<br />
allo HCT. However, in patients with an intermediate- or<br />
high-risk HCT-CI score, the RI patients had significantly<br />
lower NRM (p � 0.001) and significantly higher OS (p �<br />
0.007) compared with MA patients, but relapse rates were<br />
not statistically different (p � 0.26). Another relatively large<br />
registry study from the European Group for Blood and<br />
Marrow Transplantation came to comparable conclusions,<br />
though this study was unique in that it restricted the<br />
analysis to patients with HLA-matched unrelated donors. 20<br />
In 131 patients with FL, MA allo HCT was associated with<br />
significantly worse outcomes for NRM, PFS, and OS by<br />
multivariate analysis (p � 0.01 for all three endpoints).<br />
Although the data are not conclusive, it would appear that<br />
RI allo HCT may <strong>of</strong>fer long-term disease control with less<br />
risk <strong>of</strong> serious toxicity compared with MA allo, perhaps most<br />
relevant in patients with significant comorbidities.<br />
Focus on RI Allo HCT<br />
There have been several prospective studies focusing on<br />
the outcome <strong>of</strong> indolent lymphomas following RI allo HCT,<br />
but all have been single-arm trials (Table 2). 21-26 One<br />
important finding seen in each <strong>of</strong> these studies is that<br />
relapses <strong>of</strong> indolent lymphoma following RI allo HCT were<br />
rarely seen past 3 years. The study by Khouri et al was<br />
recently updated and presented in abstract form. They have<br />
seen only one additional recurrence <strong>of</strong> FL with an added<br />
3 years <strong>of</strong> follow-up, yielding 10-year estimates <strong>of</strong> OS <strong>of</strong><br />
78% and PFS <strong>of</strong> 72%. 27 In the study by Thomson et al, the<br />
Prior Tx: Median<br />
(Range)<br />
Prior<br />
Auto HCT aGVHD, cGVHD NRM/TRM EFS/PFS OS<br />
Flu/Mel Alem � CSP 3 (1–6)# 37% 15%, 5%# 11% 65% 73% 3 yrs<br />
TBI (18%),<br />
Flu/TBI (82%)#<br />
CSP � MMF 6 (1–19)# 44%# 63%, 47%# 42%# 43% 52% 3 yrs<br />
FCR TAC � MTX 3 (2–7) 19% 11%<br />
36%<br />
15% 83% 85% 5 yrs<br />
Flu/Mel CSP � MTX 3 (NR) 46% 51%, 53% 37% 55% 57% 4.3 yrs<br />
Flu/Mel Alem � CSP 4 (1–8) 26% 13%, 18% 15% 76% 76% 3.6 yrs<br />
FC TAC 2 (1–3) 0% 29%, 18%# 14%# 75% 81% 4.6 yrs<br />
Abbreviations: GVHD, graft-versus-host disease (a � acute �Grades 2–4�, c � chronic �extensive�); Tx, treatment; auto HCT, autologous hematopoietic cell<br />
transplantation; aGVHD, acute GVHD (grades 2–4); cGVHD, chronic GVHD (extensive); NRM, nonrelapse mortality; TRM, treatment-related mortality; EFS, event-free<br />
survival; PFS, progression-free survival; OS, overall survival; F/U, follow-up; Flu, fludarabine; Mel, melphalan; Alem, alemtuzumab; CSP, cyclosporine; yrs, years; HT,<br />
histologic transformation; TBI, total body irradiation; MMF, mycophenolate m<strong>of</strong>etil; FL, follicular lymphoma; FCR, fludarabine, cyclophosphamide, and rituximab; TAC,<br />
tacrolimus; MTX, methotrexate; NR, not reported; FC, fludarabine and cyclophosphamide.<br />
^<br />
Specific histologies comprised within this group are as follows: 29 follicular lymphomas, three lymphoplasmacytic lymphomas, and 9 chronic lymphocytic or<br />
prolymphocytic leukemias.<br />
* Nineteen <strong>of</strong> the patients treated in the study by Thomson et al from 2010 were also described in the report by Morris et al from 2004 but with more than 5 years<br />
<strong>of</strong> additional follow-up. Since this comprised a minority <strong>of</strong> the patients in the Morris study, it was included as its own reference.<br />
# The data reported were not segregated by histology and therefore represent the result for the entire study cohort, not exclusively for patients with low-grade<br />
lymphoma<br />
Median<br />
F/U<br />
latest relapse seen was at 43 months, with no other relapses<br />
seen past 3 years. 26 A few other salient features from these<br />
studies are worth mentioning.<br />
First, patients included in these studies were generally<br />
heavily pretreated, with a substantial portion in some <strong>of</strong> the<br />
studies having received prior auto HCT. This is in keeping<br />
with the NCCN guidelines that allo HCT should generally<br />
be reserved for patients with relapsed for refractory disease.<br />
11 There were varying degrees <strong>of</strong> chemotherapy sensitivity<br />
across the trials. The studies by Khouri et al and Shea<br />
et al exclusively enrolled patients who were in either complete<br />
remission (CR) or partial remission (PR), which may in<br />
part explain why these studies had some <strong>of</strong> the best survival<br />
outcomes. 21,25 The report by Piñana et al noted a nonsignificant<br />
trend in 4-year OS according to disease status (71% for<br />
those in CR, 48% for those in PR, and 29% for refractory or<br />
progressive disease [PD]; p � 0.09). 24 In the largest <strong>of</strong> these<br />
studies, Thomson et al noted that the absence <strong>of</strong> prior auto<br />
HCT predicted for improved OS and PFS in a univariate<br />
analysis, but this did not persist following multivariate<br />
analysis. 26 The studies by Morris et al and Rezvani et al<br />
supported the concept that chemotherapy sensitivity and<br />
disease status at the time <strong>of</strong> transplantation are predictive<br />
<strong>of</strong> outcome. 22,23 However, even in these seemingly higherrisk<br />
patients, salvage and long-term remission is possible<br />
with this approach.<br />
Another issue in these studies worth addressing is the<br />
effect <strong>of</strong> GVHD on outcome. Classically, GVHD is an indication<br />
<strong>of</strong> an immunologically active allograft, which should<br />
also be capable <strong>of</strong> mediating GVL effects. However, GVHD<br />
remains the major cause <strong>of</strong> morbidity and NRM in allo HCT.<br />
Although none <strong>of</strong> these studies were designed to determine<br />
the best method <strong>of</strong> GVHD prophylaxis, two <strong>of</strong> these studies<br />
from the same group used a strategy <strong>of</strong> partial in vivo T-cell<br />
depletion using the anti-CD52 antibody alemtuzumab. 23,26<br />
In these studies, the authors noted a relatively low incidence<br />
<strong>of</strong> GVHD. However, relapse rates with this approach were<br />
higher than that seen in the other prospective studies<br />
497
Fig. 1. Proposed algorithm to incorporate the use <strong>of</strong> hematopoietic cell transplantation in patients with indolent B-cell non-Hodgkin<br />
lymphoma. ^ As the best treatment for relapsed indolent lymphoma remains unclear, these patients should be treated in the context <strong>of</strong> a clinical<br />
trial if feasible. * If a patient has not responded to second-line chemotherapy or has significant disease burden, enrollment in a clinical trial<br />
should be strongly considered before use <strong>of</strong> HCT. Either auto or allo HCT could then be considered to consolidate a response to this intervention.<br />
Abbreviations: HCT, hematopoietic cell transplantation; CR1, first complete remission; HCT-CI, hematopoietic cell transplant-comorbidity<br />
index; auto, autologous; allo, allogeneic; RI, reduced-intensity; MA, myeloablative; BSC, best supportive care; WIS, withdrawal <strong>of</strong> immunosuppression;<br />
DLI, donor lymphocyte infusion.<br />
reviewed in this article, though these relapses were <strong>of</strong>ten<br />
salvageable with further immunomanipulation (see additional<br />
details below). It is noteworthy that none <strong>of</strong> these<br />
prospective studies showed that the presence <strong>of</strong> either acute<br />
or chronic GVHD was associated with decreased relapse or<br />
improved disease-free survival.<br />
With these data suggesting that relapse rates are higher<br />
in patients not in CR, investigators hypothesized that radioimmunotherapy<br />
(RIT) as part <strong>of</strong> the preparative regimen<br />
could deliver targeted radiation to tumor sites with minimal<br />
nonhematologic toxicity to improve early post-transplant<br />
disease control and potentially lead to improved PFS.<br />
Bethge et al performed a phase II study <strong>of</strong> yttrium-90<br />
conjugated to the anti-CD20 antibody ibritumomab tiuxetan<br />
(90-YIT) plus RI allo HCT. 28 A total <strong>of</strong> 40 patients were<br />
accrued (17 with FL, 1 with MZL, 1 with LPL, 13 with CLL,<br />
and 8 with mantle cell lymphoma). All had received at least<br />
two prior chemotherapy regimens, and 82% had active<br />
disease at the time <strong>of</strong> HCT. This treatment approach yielded<br />
2-year estimated EFS <strong>of</strong> 43% and OS <strong>of</strong> 51% for the entire<br />
cohort. NRM at 2 years was relatively high compared with<br />
other studies at 45%, which the authors attributed to a<br />
relatively high-risk patient population and high frequency<br />
(68%) receiving grafts from unrelated donors. Additionally, a<br />
cohort <strong>of</strong> 40 patients with high-risk B-cell lymphomas was<br />
treated at FHCRC with 90-YIT along with RI allo HCT, 45%<br />
498<br />
CASSADAY AND GOPAL<br />
<strong>of</strong> whom had indolent histologies. 29 For the indolent subset,<br />
only 17% (3 patients) had chemotherapy-sensitive disease at<br />
the time <strong>of</strong> HCT, but the 3-month response rate was still<br />
83%, with over 75% alive and approximately 50% alive and<br />
progression-free at 2.5 years. The cumulative incidence<br />
estimate <strong>of</strong> NRM at 2.5 years was 16%, more in line with<br />
other studies reported here. These studies suggest the<br />
feasibility and potential efficacy <strong>of</strong> the use <strong>of</strong> RIT added to RI<br />
preparative regimens for patients with indolent B-NHL not<br />
in CR.<br />
Management <strong>of</strong> Relapse after Allo HCT<br />
Relapse <strong>of</strong> hematologic malignancies after allo HCT poses<br />
a particularly daunting challenge, and indolent lymphomas<br />
are no exception. Apart from additional chemoimmunotherapy<br />
or radiation, manipulation <strong>of</strong> the allografted immune<br />
system is an intervention unique to this particular<br />
disease state. This can be attempted using withdrawal <strong>of</strong><br />
immunosuppression (WIS) or donor lymphocyte infusion<br />
(DLI). Many <strong>of</strong> the studies mentioned above include a small<br />
population <strong>of</strong> patients where such interventions were performed<br />
with or without additional anticancer therapy, occasionally<br />
with dramatic and durable responses. 21,26 DLI in<br />
particular is advocated in patients who received alemtuzumab<br />
as part <strong>of</strong> their transplant conditioning. A recent<br />
publication from FHCRC specifically analyzed their experi-
ROLE OF HCT FOR INDOLENT LYMPHOMA<br />
ence with managing lymphoma relapsing after allo HCT,<br />
revealing that those with indolent histologies had significantly<br />
less mortality compared to aggressive NHL in a<br />
multivariate model (p � 0.008). 30 This analysis also demonstrated<br />
potential benefit <strong>of</strong> WIS and DLI. In the absence <strong>of</strong><br />
significant GVHD, WIS and/or DLI should be considered in<br />
patients with indolent lymphoma who have relapsed after<br />
allo HCT and desire further treatment <strong>of</strong> their disease.<br />
Special Circumstance: Histologic Transformation<br />
One <strong>of</strong> the most challenging aspects <strong>of</strong> managing indolent<br />
lymphoma is the occurrence <strong>of</strong> histologic transformation<br />
(HT) to an aggressive large-cell lymphoma. The general<br />
treatment approach is similar to that for diffuse large B-cell<br />
lymphoma (DLBCL), with one notable exception. Since HT<br />
is thought to have a worse prognosis than de novo DLBCL,<br />
the NCCN guidelines recommend considering consolidation<br />
with either auto or allo HCT in first remission (either<br />
complete or partial). 11 This may be particularly true for<br />
patients who are unable to receive anthracycline-based<br />
therapy. Few <strong>of</strong> the studies mentioned above included patients<br />
with HT, as this manifestation takes on the characteristics<br />
<strong>of</strong> an intermediate- or high-grade lymphoma. A<br />
recent prospective study in Norway evaluated 47 patients<br />
with indolent B-NHL that had relapsed with HT following<br />
chemotherapy. 31 All patients treated in this trial received<br />
salvage chemotherapy with a plan to consolidate responders<br />
with auto HCT; however, only 30 met this criterion and<br />
received sufficient treatment to be included in the analysis.<br />
Among these transplanted patients, the 5-year PFS and OS<br />
rates were 32% and 47%, respectively, with a plateau in the<br />
PFS curve at 30% beyond 3.5 years. It is noteworthy that, in<br />
the 17 patients from the initial cohort who did not undergo<br />
auto HCT, there were three patients (18%)—referred to by<br />
the doctors as “long-term survivors”—who received salvage<br />
chemotherapy and/or radiation alone. The adverse prognosis<br />
<strong>of</strong> HT before RI allo HCT was evaluated by Rezvani et al,<br />
noting that the risk <strong>of</strong> relapse was nearly 5 times higher for<br />
HT than for those with nontransformed relapsed/refractory<br />
indolent lymphoma (p � 0.001). 22 Moreover, OS at 3 years<br />
Authors’ Disclosures <strong>of</strong> Potential Conflicts <strong>of</strong> Interest<br />
Employment or<br />
Leadership<br />
Positions<br />
Consultant or<br />
Advisory Role<br />
for relapsed/refractory indolent disease was 52% compared<br />
with 18% for HT (p � 0.02). Although these data underscore<br />
the poor prognosis following HT, long-term remissions are<br />
possible in a subset <strong>of</strong> patients if aggressive therapy can be<br />
tolerated.<br />
Conclusion<br />
Although the long-term prognosis <strong>of</strong> indolent lymphomas<br />
has improved with standard chemoimmunotherapy approaches<br />
alone, HCT does <strong>of</strong>fer the potential for durable<br />
remissions for a subset <strong>of</strong> patients who are eligible for such<br />
treatment. Figure 1 shows a proposed algorithm for the<br />
application <strong>of</strong> HCT in the management <strong>of</strong> these diseases,<br />
conceding that rigorous prospective trials are not readily<br />
available for all decision points. It is important to remember<br />
that both auto and allo HCT require a large amount <strong>of</strong><br />
preparation (both for the provider and for the patient), so<br />
their application should be considered early in the course <strong>of</strong><br />
treating patients with indolent lymphoma.<br />
In summary, auto HCT for chemotherapy-sensitive relapse<br />
can provide improved outcomes, though concerns regarding<br />
secondary malignancies remain. The development<br />
<strong>of</strong> RI allo HCT has allowed the benefits <strong>of</strong> GVL and a<br />
potential cure to be <strong>of</strong>fered to patients that would have been<br />
deemed too unfit for traditional MA conditioning. However,<br />
despite this advance, GVHD remains a significant cause <strong>of</strong><br />
morbidity and mortality. Challenges also remain in improving<br />
disease control for those with chemotherapy-resistant or<br />
transformed disease as well placing HCT in the context <strong>of</strong><br />
ever-expanding nontransplant options for patients with indolent<br />
B-NHL. Despite these hurdles, both auto and allo<br />
HCT remain viable, effective options for many patients with<br />
indolent lymphoma.<br />
Acknowledgment<br />
Dr. Cassaday is supported by the National Institutes <strong>of</strong><br />
Health T32 training grant number T32CA009515–27. Dr. Gopal<br />
is supported by the following: P01 CA044991, Leukemia &<br />
Lymphoma <strong>Society</strong> SCOR grant 7040, and a gift from Frank and<br />
Betty Vandermeer. Dr. Gopal is a Scholar in <strong>Clinical</strong> Research<br />
<strong>of</strong> the Leukemia & Lymphoma <strong>Society</strong>.<br />
Stock<br />
Ownership Honoraria<br />
Author<br />
Ryan D. Cassaday*<br />
Ajay K. Gopal Seattle Genetics Millennium;<br />
Seattle Genetics<br />
*No relevant relationships to disclose.<br />
1. Swerdlow SH, Campo E, Harris NL, et al (eds). WHO Classification <strong>of</strong><br />
Tumours <strong>of</strong> Haematopoietic and Lymphoid Tissues. Lyon, France; International<br />
Agency for Research on Cancer: 2008.<br />
2. Khouri IF, Keating M, Korbling M, et al. Transplant-lite: Induction <strong>of</strong><br />
graft-versus-malignancy using fludarabine-based nonablative chemotherapy<br />
and allogeneic blood progenitor-cell transplantation as treatment for lymphoid<br />
malignancies. J Clin Oncol. 1998;16:2817-2824.<br />
REFERENCES<br />
Research<br />
Funding<br />
Abbott<br />
Laboratories;<br />
Biomarin;<br />
Cephalon;<br />
GlaxoSmithKline;<br />
Merck; Pfizer;<br />
Piramal; Seattle<br />
Genetics;<br />
Spectrum<br />
Pharmaceuticals<br />
Expert<br />
Testimony<br />
Other<br />
Remuneration<br />
3. McSweeney PA, Niederwieser D, Shizuru JA, et al. Hematopoietic cell<br />
transplantation in older patients with hematologic malignancies: Replacing<br />
high-dose cytotoxic therapy with graft-versus-tumor effects. Blood. 2001;97:<br />
3390-3400.<br />
4. Sun Y, Tawara I, Toubai T, et al. Pathophysiology <strong>of</strong> acute graft-versushost<br />
disease: Recent advances. Transl Res. 2007;150:197-214.<br />
5. Gyan E, Foussard C, Bertrand P, et al. High-dose therapy followed by<br />
499
autologous purged stem cell transplantation and doxorubicin-based chemotherapy<br />
in patients with advanced follicular lymphoma: A randomized multicenter<br />
study by the GOELAMS with final results after a median follow-up <strong>of</strong><br />
9 years. Blood. 2009;113:995-1001.<br />
6. Ladetto M, De Marco F, Benedetti F, et al. Prospective, multicenter<br />
randomized GITMO/IIL trial comparing intensive (R-HDS) versus conventional<br />
(CHOP-R) chemoimmunotherapy in high-risk follicular lymphoma at<br />
diagnosis: The superior disease control <strong>of</strong> R-HDS does not translate into an<br />
overall survival advantage. Blood. 2008;111:4004-4013.<br />
7. Lenz G, Dreyling M, Schiegnitz E, et al. Myeloablative radiochemotherapy<br />
followed by autologous stem cell transplantation in first remission<br />
prolongs progression-free survival in follicular lymphoma: Results <strong>of</strong> a prospective,<br />
randomized trial <strong>of</strong> the German Low-Grade Lymphoma Study<br />
Group. Blood. 2004;104:2667-2674.<br />
8. Sebban C, Mounier N, Brousse N, et al. Standard chemotherapy with<br />
interferon compared with CHOP followed by high-dose therapy with autologous<br />
stem cell transplantation in untreated patients with advanced follicular<br />
lymphoma: The GELF-94 randomized study from the Groupe d’Etude des<br />
Lymphomes de l’Adulte (GELA). Blood. 2006;108:2540-2544.<br />
9. Rohatiner AZ, Nadler L, Davies AJ, et al. Myeloablative therapy with<br />
autologous bone marrow transplantation for follicular lymphoma at the time<br />
<strong>of</strong> second or subsequent remission: Long-term follow-up. J Clin Oncol.<br />
2007;25:2554-2559.<br />
10. Schouten HC, Qian W, Kvaloy S, et al. High-dose therapy improves<br />
progression-free survival and survival in relapsed follicular non-Hodgkin’s<br />
lymphoma: Results from the randomized European CUP trial. J Clin Oncol.<br />
2003;21:3918-3927.<br />
11. Zelentz AD, Abramson JA, Advani RA, et al. NCCN <strong>Clinical</strong> Practice<br />
Guidelines in <strong>Oncology</strong>: Non-Hodgkin’s Lymphomas. http://www.nccn.org/<br />
pr<strong>of</strong>essionals/physician_gls/f_guidelines.asp. Accessed January 11, <strong>2012</strong>.<br />
12. Dimopoulos MA, Gertz MA, Kastritis E, et al. Update on treatment<br />
recommendations from the Fourth International Workshop on Waldenstrom’s<br />
Macroglobulinemia. J Clin Oncol. 2009;27:120-126.<br />
13. van Besien K, Loberiza FR, Jr., Bajorunaite R, et al. Comparison <strong>of</strong><br />
autologous and allogeneic hematopoietic stem cell transplantation for follicular<br />
lymphoma. Blood. 2003;102:3521-3529.<br />
14. Ingram W, Devereux S, Das-Gupta EP, et al. Outcome <strong>of</strong> BEAMautologous<br />
and BEAM-alemtuzumab allogeneic transplantation in relapsed<br />
advanced stage follicular lymphoma. Br J Haematol. 2008;141:235-243.<br />
15. Evens AM, Vanderplas A, LaCasce AS, et al. Outcomes with autologous<br />
(auto) and allogeneic (allo) stem cell transplantation (SCT) for relapsed/<br />
refractory follicular lymphoma (FL) in the post-rituximab era: A comparative<br />
analysis from the National Comprehensive Cancer Network (NCCN) Non-<br />
Hodgkin’s Lymphoma (NHL) Outcomes Database Project. Blood. 2011;118<br />
(abstr 4112).<br />
16. Tomblyn MR, Ewell M, Bredeson C, et al. Autologous versus reducedintensity<br />
allogeneic hematopoietic cell transplantation for patients with<br />
chemosensitive follicular non-Hodgkin lymphoma beyond first complete response<br />
or first partial response. Biol Blood Marrow Transplant. 2011;17:<br />
1051-1057.<br />
17. Hari P, Carreras J, Zhang MJ, et al. Allogeneic transplants in follicular<br />
lymphoma: Higher risk <strong>of</strong> disease progression after reduced-intensity compared<br />
to myeloablative conditioning. Biol Blood Marrow Transplant. 2008;<br />
14:236-245.<br />
18. Sorror ML, Storer BE, Maloney DG, et al. Outcomes after allogeneic<br />
500<br />
CASSADAY AND GOPAL<br />
hematopoietic cell transplantation with nonmyeloablative or myeloablative<br />
conditioning regimens for treatment <strong>of</strong> lymphoma and chronic lymphocytic<br />
leukemia. Blood. 2008;111:446-452.<br />
19. Sorror ML, Maris MB, Storb R, et al. Hematopoietic cell transplantation<br />
(HCT)-specific comorbidity index: A new tool for risk assessment before<br />
allogeneic HCT. Blood. 2005;106:2912-2919.<br />
20. Avivi I, Montoto S, Canals C, et al. Matched unrelated donor stem cell<br />
transplant in 131 patients with follicular lymphoma: An analysis from the<br />
Lymphoma Working Party <strong>of</strong> the European Group for Blood and Marrow<br />
Transplantation. Br J Haematol. 2009;147:719-728.<br />
21. Khouri IF, McLaughlin P, Saliba RM, et al. Eight-year experience with<br />
allogeneic stem cell transplantation for relapsed follicular lymphoma after<br />
nonmyeloablative conditioning with fludarabine, cyclophosphamide, and<br />
rituximab. Blood. 2008;111:5530-5536.<br />
22. Rezvani AR, Storer B, Maris M, et al. Nonmyeloablative allogeneic<br />
hematopoietic cell transplantation in relapsed, refractory, and transformed<br />
indolent non-Hodgkin’s lymphoma. J Clin Oncol. 2008;26:211-217.<br />
23. Morris E, Thomson K, Craddock C, et al. Outcomes after alemtuzumabcontaining<br />
reduced-intensity allogeneic transplantation regimen for relapsed<br />
and refractory non-Hodgkin lymphoma. Blood. 2004;104:3865-3871.<br />
24. Pinana JL, Martino R, Gayoso J, et al. Reduced intensity conditioning<br />
HLA identical sibling donor allogeneic stem cell transplantation for patients<br />
with follicular lymphoma: Long-term follow-up from two prospective multicenter<br />
trials. Haematologica. 2010;95:1176-1182.<br />
25. Shea T, Johnson J, Westervelt P, et al. Reduced-intensity allogeneic<br />
transplantation provides high event-free and overall survival in patients with<br />
advanced indolent B cell malignancies: CALGB 109901. Biol Blood Marrow<br />
Transplant. 2011;17:1395-1403.<br />
26. Thomson KJ, Morris EC, Milligan D, et al. T-cell-depleted reducedintensity<br />
transplantation followed by donor leukocyte infusions to promote<br />
graft-versus-lymphoma activity results in excellent long-term survival in<br />
patients with multiply relapsed follicular lymphoma. J Clin Oncol. 2010;28:<br />
3695-3700.<br />
27. Khouri IF, Saliba RM, Valverde R, et al. Nonmyeloablative allogeneic<br />
stem cell transplantation with or without 90-yttrium ibritumomab tiuxetan<br />
(90YIT) is curative for relapsed follicular lymphoma: Median 9-year follow-up<br />
results. Blood. 2011;118 (abstr 662).<br />
28. Bethge WA, Lange T, Meisner C, et al. Radioimmunotherapy with<br />
yttrium-90-ibritumomab tiuxetan as part <strong>of</strong> a reduced-intensity conditioning<br />
regimen for allogeneic hematopoietic cell transplantation in patients with<br />
advanced non-Hodgkin lymphoma: Results <strong>of</strong> a phase 2 study. Blood. 2010;<br />
116:1795-1802.<br />
29. Gopal AK, Guthrie KA, Rajendran J, et al. Y-Ibritumomab tiuxetan,<br />
fludarabine, and TBI-based nonmyeloablative allogeneic transplantation<br />
conditioning for patients with persistent high-risk B-cell lymphoma. Blood.<br />
2011;118:1132-1139.<br />
30. Ram R, Gooley TA, Maloney DG, et al. Histology and time to progression<br />
predict survival for lymphoma recurring after reduced-intensity conditioning<br />
and allogeneic hematopoietic cell transplantation. Biol Blood Marrow<br />
Transplant. 2011;17:1537-1545.<br />
31. Eide MB, Lauritzsen GF, Kvalheim G, et al. High-dose chemotherapy<br />
with autologous stem cell support for patients with histologically transformed<br />
B-cell non-Hodgkin lymphomas. A Norwegian multi centre phase II study.<br />
Br J Haematol. 2011;152:600-610.
CONTROVERSIES IN MYELOMA: INDUCTION,<br />
TRANSPLANT, AND MAINTENANCE<br />
CHAIR<br />
Amrita Krishnan, MD<br />
City <strong>of</strong> Hope National Medical Center<br />
Duarte, CA<br />
SPEAKERS<br />
S. Vincent Rajkumar, MD<br />
Mayo Clinic<br />
Rochester, MN<br />
Michel Attal, MD<br />
CHU Purpan<br />
Toulouse, France
Stem Cell Transplantation for Multiple<br />
Myeloma: Who, When, and What Type?<br />
Overview: Early randomized trials <strong>of</strong> high-dose chemotherapy<br />
with autologous stem cell rescue showed improved<br />
progression-free survival (PFS) over conventional chemotherapy.<br />
However, in the era <strong>of</strong> novel agents for myeloma in<br />
conjunction with the evolution <strong>of</strong> hematopoietic stem cell<br />
transplantation, many new questions arise. First, how can<br />
novel agents be incorporated into the transplant paradigm?<br />
Given the efficacy <strong>of</strong> new induction regimens, should transplant<br />
be delayed until relapse? Also, in the era <strong>of</strong> individualized<br />
medicine, chronologic age alone should not drive<br />
decisions regarding transplantation. Therefore, the feasibility<br />
and role <strong>of</strong> transplantation in older patients with myeloma is<br />
THE CONCEPT <strong>of</strong> dose-intensive chemotherapy to treat<br />
a malignant disease is a paradigm that has been<br />
explored for over a quarter century. A group from The Royal<br />
Marsden Hospital 1 demonstrated that increased doses <strong>of</strong><br />
melphalan could induce responses in nine patients with<br />
myeloma; however, this was at the cost <strong>of</strong> significant hematologic<br />
toxicity with a median <strong>of</strong> 46 days <strong>of</strong> neutropenia. In<br />
later trials with the use <strong>of</strong> autologous hematopoietic peripheral<br />
blood stem cell support, this approach became safer.<br />
Ultimately, a randomized trial by the Intergroupe Francophone<br />
du Myelome (IFM) demonstrated the superiority <strong>of</strong><br />
high-dose chemotherapy in terms <strong>of</strong> both OS and diseasefree<br />
survival over conventional chemotherapy. 2<br />
The IFM trial is now more than a decade old, and in this<br />
era <strong>of</strong> novel agents, many new questions remain unanswered.<br />
For example, the original IFM trial included patients<br />
65 years and younger. In this modern era, should<br />
there be an age limit for myeloma transplant? Also, what is<br />
the optimal induction regimen, and how many cycles <strong>of</strong><br />
therapy should be administered (i.e., when do you transplant<br />
a patient)? Lastly, what is the best way to incorporate<br />
novel agents into the post-transplant setting?<br />
When to Transplant<br />
It is common knowledge that the use <strong>of</strong> melphalancontaining<br />
induction regimens should be avoided in patients<br />
who could potentially undergo autologous transplantation. 3<br />
In contrast, there is little consensus on the optimal nonmelphalan-containing<br />
induction regimen. The summary by<br />
S. Vincent Rajkumar, MD, will discuss in depth the choice <strong>of</strong><br />
an optimal induction regimen. In regard to the optimal<br />
number <strong>of</strong> cycles before transplant, there remains no consensus.<br />
There are numerous phase I and II clinical trials <strong>of</strong><br />
two-, three-, and even four-drug induction regimens. All<br />
incorporate at least one novel agent and traditionally still<br />
include a steroid backbone, although with an intent to be<br />
more steroid-sparing by using lower doses <strong>of</strong> steroids. The<br />
lenalidomide, bortezomib, and dexamethasone (RVD) regimen<br />
was initially used in a phase I trial in the relapsed<br />
setting and demonstrated good tolerability and high response<br />
rates. 3 In the upfront setting, response rates are<br />
extremely high (100%), as is depth <strong>of</strong> response (complete<br />
remission [CR]/near CR [nCR] 40%). 4 In that initial trial,<br />
patients had the option to proceed to transplant after four<br />
502<br />
By Amrita Krishnan, MD<br />
being studied. The controversy <strong>of</strong> transplant type (i.e., autologous<br />
compared with reduced intensity allogeneic transplant)<br />
remains unresolved. Several large international trials have<br />
demonstrated conflicting results in regard to an overall survival<br />
(OS) benefit with the allogeneic approach. The role <strong>of</strong><br />
allogeneic transplant remains under study especially in the<br />
high-risk population, which has high relapse rates with traditional<br />
autologous approaches. Future directions to reduce<br />
relapse include post-transplantation consolidation and maintenance<br />
therapy with either approved agents or new agents<br />
and immunotherapy, either vaccine based or natural killer (NK)<br />
and T-cell based.<br />
cycles, and the response rate was reported after four cycles.<br />
However, the investigators report 75% <strong>of</strong> patients had a<br />
further upgrade <strong>of</strong> responses at six or eight cycles. It is<br />
unknown whether this would ultimately affect posttransplant<br />
PFS.<br />
A new three-drug regimen that may soon gain traction is<br />
the combination <strong>of</strong> the new proteasome inhibitor carfilzomib,<br />
lenalidomide, and dexamethasone (CRD). 5 Preliminary<br />
results showed very high response rates (100% �<br />
partial response [PR]) in phase I and II trials, as well as the<br />
highest depth <strong>of</strong> response seen outside a transplant setting<br />
(79% CR/nCR). Twenty-four <strong>of</strong> 49 patients in this trial had<br />
stem cells collected. However, too few patients have gone on<br />
to transplant to be able to assess the effect <strong>of</strong> this induction<br />
on post-transplant survival either with early or delayed<br />
transplant.<br />
Indirectly, one can extrapolate from phase III trials that<br />
improved responses before transplant can mean further<br />
improvements post-transplant, at least in terms <strong>of</strong> PFS.<br />
There have been several large phase III trials in Europe<br />
comparing traditional regimens to newer three-drug regimens<br />
incorporating novel agents. For example, Cavo et al<br />
compared bortezomib, thalidomide, and dexamethasone<br />
(VTD) with thalidomide and dexamethasone (TD). 6 Both<br />
arms received three courses <strong>of</strong> induction before transplant,<br />
followed by two cycles <strong>of</strong> consolidation with VTD or TD. Not<br />
surprisingly, the CR/nCR rate pretransplant was higher in<br />
the VTD arm (31% vs. 11%; p � 0.0001 for TD). Posttransplant<br />
this translated into an improved PFS at three<br />
years (69% vs. 37%) but not an OS benefit. Though it<br />
remains unknown whether this was due to induction or<br />
consolidation or both.<br />
In the MRCIX trial <strong>of</strong> CTD versus CVAD, the post-SCT<br />
CR rate was 50% versus 37% in the CVAD arm. PFS was<br />
greater in patients who received a CR post-transplant.<br />
Further statistical modeling suggested that with longer<br />
follow up this translated into a small PFS benefit.<br />
From the City <strong>of</strong> Hope Cancer Center, Duarte, CA.<br />
Author’s disclosure <strong>of</strong> potential conflicts <strong>of</strong> interest are found at the end <strong>of</strong> this article.<br />
Address reprint requests Amrita Krishnan, MD, 1500 E. Duarte Road, Duarte, CA,<br />
91010; email: akrishnan@coh.org.<br />
© <strong>2012</strong> by <strong>American</strong> <strong>Society</strong> <strong>of</strong> <strong>Clinical</strong> <strong>Oncology</strong>.<br />
1092-9118/10/1-10
TRANSPLANTATION FOR MYELOMA<br />
Alternately for less aggressive disease, many experts in<br />
the field advocate the use <strong>of</strong> a two-drug regimen <strong>of</strong> either<br />
bortezomib and dexamethasone or lenalidomide and dexamethasone<br />
depending on the patients’ comorbidities. The<br />
lenalidomide plus low-dose dexamethasone approach has<br />
good tolerability and 2-year OS data in both young and<br />
elderly patients. 7 Although the point at which patients<br />
should receive transplant on this induction regimen remains<br />
controversial, it is clear that prolonged exposure to lenalidomide<br />
may impair stem cell yields. Therefore, if a transplant<br />
is planned, transplant physicians recommend that stem cell<br />
collection should occur before the six months <strong>of</strong> lenalidomide<br />
exposure. 8<br />
Who Should Undergo a Transplant?<br />
Age Limits for Transplantation?<br />
Myeloma is a disease <strong>of</strong> elderly patients with a median <strong>of</strong><br />
72 years at presentation. Therefore, if arbitrary age cut<strong>of</strong>fs<br />
are used, a large portion <strong>of</strong> patients with myeloma would be<br />
ineligible for a transplant. In earlier eras, autologous transplantation<br />
was generally reserved for patients younger than<br />
60 or 65 years. The original IFM trial—which served as the<br />
platform for high-dose therapy in multiple myeloma—enrolled<br />
patients 65 years and younger. However, since that<br />
time, there have been several trials suggesting that age is<br />
not a prognostic variable to transplant outcome in myeloma,<br />
but rather it is disease-related indices. A group from M. D.<br />
Anderson Cancer Center published their experience with 84<br />
patients with the median age 72. All patients met the<br />
standard organ function criteria for transplant and received<br />
melphalan doses between 140 mg/m 2 and 200 mg/m 2 .<br />
Nonrelapse-related mortality was 3%, and 5-year OS was<br />
67%. 9 A larger registry-based study comparing patients<br />
older than 60 years and younger than 60 years showed no<br />
difference in transplant-related mortality or PFS between<br />
older and younger patients. Though the median age in the<br />
older than age 60 group was still on the younger side at 63<br />
years. 10<br />
Specific issues about elderly patients to consider begin<br />
with the choice <strong>of</strong> induction therapy for a patient that one<br />
KEY POINTS<br />
● The use <strong>of</strong> novel agents as induction therapy for<br />
myeloma can improve post-transplant progressionfree<br />
survival.<br />
● Studies are ongoing to determine the optimal timing<br />
<strong>of</strong> transplantation after an effective induction strategy<br />
(i.e., early compared with delayed transplant).<br />
● Patients with high-risk myeloma—especially those<br />
with 17p deletion—remain a challenge and may be a<br />
subgroup for whom allogeneic transplantation should<br />
be considered.<br />
● Post-transplant strategies to reduce relapse, such as<br />
consolidation or immunotherapy, are being studied.<br />
● Age should not be considered a barrier to transplantation,<br />
but appropriate assessment <strong>of</strong> the older patient<br />
should include geriatric assessments and<br />
traditional functional testing.<br />
would consider “transplant eligible.” As previously mentioned,<br />
melphalan-based regimens are to be avoided because<br />
<strong>of</strong> potential stem cell toxicity. High-dose dexamethasone is<br />
also particularly difficult for the older patient because <strong>of</strong><br />
toxicity. In addition, increasing age has been correlated<br />
inversely with CD34 counts. 11 Therefore, stem cell collection<br />
should be considered early in the disease course, and the use<br />
<strong>of</strong> agents such as plerixafor may be necessary.<br />
Before transplant, Karn<strong>of</strong>sky performance status is the<br />
most traditional evaluation in addition to organ function<br />
testing. However, in the older patient, performance status as<br />
the only marker <strong>of</strong> functional status may be inadequate. The<br />
concept <strong>of</strong> frailty is important to consider. For example, the<br />
very fit older person who exercises and is fully independent<br />
has different needs than the mildly frail individual who may<br />
need help for household tasks. Frailty in and <strong>of</strong> itself is a<br />
predictor <strong>of</strong> outcome in elderly patients. 12 The geriatric<br />
assessment tool may be helpful as part <strong>of</strong> pretransplant<br />
assessment. Although it has not yet been validated in the<br />
transplant setting, it has been a predictor <strong>of</strong> chemotherapy<br />
toxicity with conventional chemotherapy. This tool, in addition<br />
to comorbidities, encompasses cognition, psychologic<br />
status, social functioning and support, and nutritional status.<br />
13 Optimal assessment <strong>of</strong> all these factors could help<br />
predict the older patient who may pass all the functional<br />
testing but may have the potential for significant debility<br />
with the transplant. In addition, it may help with needs<br />
assessment for post-transplant care.<br />
Patients at High Risk<br />
Risk stratification is an important part <strong>of</strong> the initial<br />
assessment <strong>of</strong> a myeloma patient and obviously influences<br />
the choices <strong>of</strong> induction therapy, but what role should it play<br />
in the transplant paradigm? Generally, risk stratification<br />
encompasses International Staging System (ISS) staging,<br />
cytogenetics, and fluorescence in situ hybridization (FISH)<br />
analysis and possibly gene expression pr<strong>of</strong>iling (GEP). The<br />
early transplant studies did not differentiate between patients<br />
at standard risk and those at high risk. Also, although<br />
later trials did attempt to stratify patients and as the<br />
definition <strong>of</strong> “high-risk” disease evolves, the applicability <strong>of</strong><br />
those older trials is limited. Recent trials tend to define<br />
“high risk” as t(4,14), t(14,16), deletion 13, or deletion 17p. A<br />
retrospective study reviewed approximately 500 patients<br />
who received bortezomib plus dexamethasone before highdose<br />
melphalan compared with a group treated with vincristine,<br />
doxorubicin, dexamethasone (VAD) induction before<br />
high-dose melphalan. 14 The use <strong>of</strong> bortezomib could abrogate<br />
some <strong>of</strong> the high-risk prognosis conferred by t(4,14).<br />
Event-free survival (EFS) was 28 months for patients<br />
treated with bortezomib compared with 16 months for the<br />
VAD group (p � 0.001). However, this still remained a poor<br />
prognostic factor compared with patients without t(4,14). In<br />
contrast, no improvement in either EFS or OS was seen for<br />
patients with 17p deletion treated with the bortezomibbased<br />
induction. However, in the HOVON-65 GMMG-HD4<br />
trial, bortezomib-based treatment before and after stem cell<br />
transplant markedly improved DFS and OS in patients with<br />
17p deletion compared with those treated with VAD induction<br />
and thalidomide maintenance. 15<br />
GEP has also been studied as a prognostic indicator in the<br />
era <strong>of</strong> novel agents plus high-dose therapy as part <strong>of</strong> the<br />
Arkansas Total Therapy 3 protocol. 16 For patients with<br />
503
Fig. 1. BMT CTN 012 trial: tandem autologous transplant compared with autologous plus reduced intensity allogeneic transplant.<br />
Abbreviations: PFS, progression-free survival; OS, overall survival.<br />
low-risk gene expression pr<strong>of</strong>ile myeloma, the addition <strong>of</strong><br />
bortezomib to the Total Therapy regimen improved clinical<br />
outcomes. However, for GEP-defined high-risk disease, EFS<br />
was uniformly poor at 27% for patients with ISS stage III<br />
disease and 36% for ISS stage I. Therefore, different treatment<br />
approaches are still needed for this group <strong>of</strong> patients.<br />
Whether the optimal approach is Total Therapy 5 (increased<br />
dose density) or allogeneic approaches, which are discussed<br />
below, remains to be seen.<br />
Plasma cell leukemia is also generally considered a poor<br />
prognostic indicator. The role <strong>of</strong> transplantation for this<br />
group <strong>of</strong> patients is controversial. However, novel agents for<br />
induction may improve the prognosis and make transplant a<br />
viable option. A Center for International Bone Marrow<br />
Transplant Research retrospective review <strong>of</strong> 160 patients<br />
with plasma cell leukemia who received either autologous or<br />
allogeneic transplantation, demonstrated surprisingly good<br />
survival in the autologous arm: 64% at a median follow-up <strong>of</strong><br />
38 months. 17<br />
In summary, novel agents as part <strong>of</strong> induction and possibly<br />
consolidation and maintenance may improve autologous<br />
transplant outcomes <strong>of</strong> certain subgroups <strong>of</strong> patients with<br />
high-risk disease. However, for the group with 17p deletion,<br />
improvement may be demonstrated, with bortezomib alternate<br />
strategies are still needed.<br />
Allogeneic Transplantation<br />
Should allogeneic transplantation be considered for patients<br />
with poor-risk disease, including those with 17p<br />
deletion? Traditional allogeneic transplant with myeloablative<br />
conditioning was associated with high transplantrelated<br />
mortality but did demonstrate a higher rate <strong>of</strong><br />
molecular remission and lower rates <strong>of</strong> relapse than autologous<br />
transplantation. 18 Reduced intensity allogeneic transplant<br />
may harness the immunologic benefits <strong>of</strong> the<br />
allogeneic approach but with reduced transplant-related<br />
mortality. It remains controversial whether this ultimately<br />
will lead to improved OS, especially for patients with highrisk<br />
disease.<br />
504<br />
Three large trials have shown conflicting results. The<br />
French Group (IFM) ran two parallel phase II trials in<br />
patients with high-risk myeloma, defined by deletion <strong>of</strong><br />
chromosome 13 or beta-2 microglobulin greater than 3 mg/L.<br />
IFM99–03 analyzed 65 patients with available human leukocyte<br />
antigen (HLA)–matched sibling donors who received<br />
an allogeneic hematopoietic cell transplant (HCT) with reduced<br />
intensity conditioning after an autologous HCT. Outcomes<br />
<strong>of</strong> these patients were compared with 219 patients<br />
enrolled in IFM99–04, an auto-auto trial. EFS and OS were<br />
not significantly different for patients receiving auto-allo<br />
and auto-auto in the IFM trial. 19<br />
In contrast, Bruno et al in an Italian biologic assignment<br />
trial demonstrated superior EFS and OS in 58 and 46<br />
recipients <strong>of</strong> auto-allo transplant compared with tandem<br />
autologous transplant. However, they did not stratify patients<br />
as high risk. 20 The United States’ BMT CTN 0102<br />
trial did stratify patients as high risk and standard risk. The<br />
definition <strong>of</strong> high risk was the same as during the era when<br />
the trial started and, therefore, was defined as deletion 13 by<br />
classic karyotyping and beta-2 microglobulin greater than<br />
3.0. The trial’s primary endpoint was 3-year PFS in patients<br />
with standard-risk disease, and for this group there was no<br />
PFS or OS benefit to auto-allo transplant (Fig. 1). Subgroup<br />
analysis <strong>of</strong> patients with high-risk, protocol-defined disease<br />
also did not show any benefit for the allogeneic route.<br />
However, since deletion 17p was not protocol defined, it<br />
remains unclear whether this group <strong>of</strong> patients could benefit<br />
from the allogeneic approach. 21 It also remains unclear<br />
whether with longer follow-up the benefit <strong>of</strong> an allogeneic<br />
approach may become apparent as more autologous patients<br />
continue to relapse. Indeed, in the trial reported by Bjorkstrand<br />
et al, 5-year follow-up was an OS benefit <strong>of</strong> 65% in the<br />
auto-allo arm compared with 58% in auto-auto arm. 22<br />
Early Compared with Delayed Transplant<br />
AMRITA KRISHNAN<br />
This remains the most difficult and, as yet, not wholly<br />
answered question. One randomized trial demonstrated<br />
comparable efficacy in terms <strong>of</strong> OS for early transplant
TRANSPLANTATION FOR MYELOMA<br />
compared with transplant delayed until first relapse, although<br />
early transplant improved quality <strong>of</strong> life scores. 23<br />
However, this was in the era predating novel agents. The<br />
novel agents clearly have improved tolerability and efficacy<br />
over the historic conventional agents. On the converse side,<br />
it is also unknown whether we can effectively use high-dose<br />
melphalan to salvage patients who relapse after initial<br />
induction with the novel agents. Therefore, are patients<br />
missing the optimal window to benefit from high-dose therapy<br />
if they delay transplant?<br />
The Mayo Clinic studied 290 patients with newly diagnosed<br />
myeloma treated with an IMiD-based induction regimen.<br />
Patients who underwent stem cell harvest and<br />
transplant within 12 months <strong>of</strong> diagnosis were considered<br />
early transplant, and those who underwent these procedures<br />
more than 12 months after diagnosis were considered<br />
delayed transplant. The overall response to transplant was<br />
the same in the early and delayed transplant groups. The<br />
time to progression after transplant also was not significantly<br />
different between the early and delayed transplant<br />
groups (20 months vs. 16 months). However, the retrospective<br />
nature <strong>of</strong> the trial precludes accurate assessment <strong>of</strong> why<br />
some patients opted for early rather than delayed transplant.<br />
In addition, the patients in the delayed transplant<br />
group primarily had transplant at first relapse, so it is<br />
unknown whether comparable results for deferred transplant<br />
would be seen after multiple relapses.<br />
The current US-French Intergroup trial would be the most<br />
relevant trial in the era <strong>of</strong> novel agents to answer the early<br />
or delayed transplant question. However, it too cannot<br />
address the question <strong>of</strong> the comparability <strong>of</strong> transplant after<br />
multiple relapses. In this trial, patients with newly diagnosed<br />
myeloma will receive one cycle <strong>of</strong> RVD and then be<br />
randomly selected to receive either additional RVD followed<br />
by stem cell apheresis and autologous transplant or further<br />
RVD and stem cell collection and no transplant until relapse<br />
(Fig. 2). This large trial will also look at important<br />
surrogate and prognostic features, such as cytogenetics and<br />
GEP.<br />
What Is the Optimal Post-transplant Therapy?<br />
Despite the increased response rates seen with autologous<br />
transplant after induction therapy using novel agents, most<br />
Fig. 2. Abbreviations: ASCT, autologous stem cell transplant; MEL,<br />
melphalan; RVD, lenalidomide, bortezomib, and dexamethasone.<br />
Fig. 3. Abbreviations: MM, multiple myeloma; SCT, stem cell transplantation;<br />
RVD, lenalidomide, bortezomib, and dexamethasone;<br />
ASCT, autologous stem cell transplant.<br />
patients will ultimately relapse. Options to reduce or delay<br />
post-transplant relapse include consolidation therapy and/or<br />
maintenance therapy. This article will focus on consolidation<br />
therapy.<br />
Consolidation Therapy<br />
This approach is akin to the leukemia therapy (i.e.,<br />
further intensive chemotherapy after initial induction therapy).<br />
However, in contrast to leukemia consolidation, there<br />
is no standard in myeloma in regard to either the agents or<br />
the number <strong>of</strong> cycles administered. Before the approval <strong>of</strong><br />
novel agents, one “consolidative approach” was in fact a<br />
second transplant as part <strong>of</strong> a tandem transplant in patients<br />
who failed to achieve a very good partial remission (VGPR)<br />
or better. The Arkansas Total Therapy 3 approach uses<br />
tandem transplant but with novel agents as consolidation:<br />
VTD PACE; bortezomib, thalidomide, dexamethasone, cisplatin,<br />
adriamycin, cyclophosphamide, and etoposide as induction<br />
and post-tandem transplant. 24<br />
Novel agents used as consolidation may serve to both<br />
improve depth <strong>of</strong> response and prolong responses. A pro<strong>of</strong> <strong>of</strong><br />
principle was the work by Ladetto et al, who treated 39<br />
patients in a VGPR postsingle autologous transplant. These<br />
selected patients had a detectable quantifiable molecular<br />
marker based on immunoglobulin heavy chain rearrangement<br />
that could be followed. They were treated with four<br />
courses <strong>of</strong> VTD, and the molecular markers were followed by<br />
polymerase chain reaction (PCR) using tumor clone specific<br />
primers. CR increased from 15% post-transplant to 49%<br />
post-VTD, and even more striking was the increase in<br />
molecular remissions from 3% to 18%. 25 In addition, patients<br />
who had a quantitative depletion in tumor burden<br />
above the median as measured by PCR had an improved<br />
PFS.<br />
The Nordic study group used a different post-transplant<br />
consolidative strategy in a randomized phase III trial. In one<br />
arm, patients received bortezomib in the traditional schedule<br />
for three cycles followed by weekly dosing for four cycles,<br />
and the observation arm had no therapy. The treatment arm<br />
had a higher response rate and PFS but no benefit in OS. 26<br />
The ongoing BMT CTN 0702 STAMINA trial may be illuminating<br />
regarding the benefit <strong>of</strong> consolidation. In this threearmed<br />
trial, all patients will receive a single autologous<br />
transplant, one arm will receive a subsequent second transplant,<br />
one arm will receive RVD for four cycles, and one arm<br />
505
will go straight to maintenance. All three arms will receive<br />
lenalidomide maintenance for 3 years or until disease progression<br />
(Fig. 3).<br />
Immunotherapy<br />
The post-transplant period is the ideal time point for<br />
immunotherapy, as theoretically the disease burden is low.<br />
Immune function remains depressed post-high-dose therapy<br />
for many months. Ex vivo expansion and subsequent transfer<br />
<strong>of</strong> autologous stimulated T cells may enhance host<br />
antitumor immunity and may also allow for enhancement<br />
<strong>of</strong> a post-transplant vaccination strategy against tumordirected<br />
antigens. A phase I/II trial <strong>of</strong> this strategy in 54<br />
patients has been conducted. 27 At day 2, patients who were<br />
post-transplant received an infusion <strong>of</strong> ex vivo stimulated<br />
antiCD3/antiCD28 T cells. Patients positive for HLA-A2<br />
antigen received pneumococcal vaccination and a multipeptide<br />
tumor antigen vaccine. Patients who were negative<br />
received the pneumococcal vaccine only. Significant T-cell<br />
recovery was seen at day 14. A subset <strong>of</strong> patients developed<br />
immune responses to tumor antigens, but this did not<br />
Author’s Disclosure <strong>of</strong> Potential Conflicts <strong>of</strong> Interest<br />
Author<br />
Employment or<br />
Leadership<br />
Positions<br />
Consultant or<br />
Advisory Role<br />
translate into improved EFS. However, the potential <strong>of</strong> this<br />
approach is suggested by the demonstration <strong>of</strong> the rapid<br />
recovery <strong>of</strong> cellular and humoral immunity as well as<br />
immune responses to the cancer vaccine. Future trials will<br />
focus on vaccines for other possibly more potent tumor<br />
antigens and the use <strong>of</strong> adjuvant therapy, such as immunomodulatory<br />
agents, in conjunction with T-cell infusions.<br />
Conclusion<br />
Where does this leave the treating physician in <strong>2012</strong> as he<br />
or she eyes the future? This future will likely include a more<br />
risk-adapted approach, with risk stratification determined<br />
by clinical features, cytogenetics, and GEP. The goal <strong>of</strong><br />
therapy would be to minimize toxicity in the patient with<br />
low-risk disease and minimize risk <strong>of</strong> relapse in the patient<br />
with poor-risk disease. Post-transplant disease assessment<br />
by detection <strong>of</strong> minimal residual disease may also help guide<br />
therapy. Options would include post-transplant consolidation<br />
with either current or new agents, reduced intensity<br />
allogeneic transplant, NK- or T-cell-directed immunotherapy,<br />
and long-term maintenance with new agents.<br />
Stock<br />
Ownership Honoraria<br />
Amrita Krishnan Merck Celgene Celgene;<br />
Genentech;<br />
Millennium<br />
1. McElwain TJ, Powles RL. High-dose intravenous melphalan for plasma<br />
cell leukemia and myeloma. The Lancet. 1983;322:822-823.<br />
2. Attal M, Harousseau JL, Stooa AM, et al. A prospective randomized trial<br />
<strong>of</strong> autologous bone marrow transplantation chemotherapy in multiple myeloma<br />
Intergroupe Francasi du Myelome. N Engl J Med. 1996;335:91-97.<br />
3. Rajkumar SV. Multiple Myeloma <strong>2012</strong> update on diagnosis, risk stratification<br />
and management. Am J Hematology. <strong>2012</strong>;87:78-88.<br />
4. Richardson PG, Weller E, Lonial S, et al. Lenalidomide, bortezomib, and<br />
dexamethasone combination therapy in patients with newly diagnosed myeloma.<br />
Blood. 2010;16:679-686.<br />
5. Jakubowiak A, Dytfield D, Jagannath S, et al. Final results <strong>of</strong> a frontline<br />
phase 1/2 study <strong>of</strong> carfilzomib, lenalidomide, and low-dose dexamethasone<br />
(CRD) in multiple myeloma. Blood. 2011;118 (abstr 631).<br />
6. Cavo M, Tacchetti P, Patriarca F, et al. Bortezomib with thalidomide<br />
plus dexamethasone compared with thalidomide plus dexamethasone as<br />
induction therapy before and consolidation therapy after double autologous<br />
stem cell transplantation in newly diagnosed multiple myeloma. A randomized<br />
phase 3 study. Lancet. 2010;376:2075-2085.<br />
7. Rajkumar SV, Jacobus S, Callander NS, et al. Lenalidomide plus<br />
high-dose dexamethasone versus lenalidomide plus low-dose dexamethasone<br />
as initial therapy for newly diagnosed multiple myeloma; An open-label<br />
randomized controlled trial. Lancet Oncol. 2010;11:29-37.<br />
8. Giralt S, Stadtmauer EA, Harousseau JL, et al. International Myeloma<br />
Working Group (IMWG) consensus statement and guidelines regarding the<br />
current status <strong>of</strong> stem cell collection and high-dose therapy for multiple<br />
myeloma and the role <strong>of</strong> plerixafor (AMD3100). Leukemia. 2009;23:1904-<br />
1912.<br />
9. Bashir Q, Shah N, Parmar S, et al. Feasibility <strong>of</strong> autologous stem cell<br />
transplantation in patients � 70 years with multiple myeloma. Leuk Lymphoma.<br />
<strong>2012</strong>;53:118-122.<br />
10. Reece DE, Bredeson C, Perez WS, et al. Autologous stem cell transplantation<br />
in multiple myeloma patients � 60 vs. � 60 years <strong>of</strong> age. Bone<br />
Marrow Transplant. 2003;32:1135-1143.<br />
11. Morris CL, Siegel E, Barlogie B, et al. Mobilization <strong>of</strong> CD34� cells in<br />
elderly patients with multiple myeloma: influence <strong>of</strong> age, prior therapy,<br />
platelet count and mobilization regimen. Br J Haematol. 2003;120:413-423.<br />
506<br />
REFERENCES<br />
Research<br />
Funding<br />
Expert<br />
Testimony<br />
AMRITA KRISHNAN<br />
Other<br />
Remuneration<br />
12. Fried LP, Tangen CM, Watson J, et al. Fraility in older adults: evidence<br />
for a phenotype. J Gerontol Biol Sci Med Sci. 2001;56:M146-M156.<br />
13. Hurria A, Togawa Kayo, Supriya G, et al. Predicting chemotherapy<br />
toxicity in older adults with cancer: a prospective multicenter study. J Clin<br />
Oncol. 2011;29:3457-3465.<br />
14. Avet-Loiseau H, Leleu X, Roussel M, et al. Bortezomib plus dexamethasone<br />
induction improves outcome <strong>of</strong> patients with t(4,14) myeloma but not<br />
outcome <strong>of</strong> patients with del(17p). J Clin Oncol. 2010;28:4630-4634.<br />
15. Neben K, Lokhorst H, Jauch A. et al. Administration <strong>of</strong> bortezomib<br />
before and after autologous stem cell transplantation improves outcome in<br />
multiple myeloma patients with 17p deletion. Blood. 2011;119:3-19.<br />
16. Waheed S, Shaugnessy JD, van Rhee F, et al. International staging<br />
system and metaphase cytogenetic abnormalities in the era <strong>of</strong> gene expression<br />
pr<strong>of</strong>iling data in multiple myeloma treated with total therapy 2 and 3<br />
protocols. Cancer. 2011;117:1001-1009.<br />
17. Mahindra A, Vesole D, Kalaycio M, et al. Autologous hematopoietic<br />
stem cell transplantation is a safe and effective treatment for primary plasma<br />
cell leukemia. Blood. 2009;114 (abstr 532).<br />
18. Bensinger WI, Buckner CD, Anasetti C, et al. Allogeneic marrow<br />
transplantation for multiple myeloma: An analysis <strong>of</strong> risk factors on outcome.<br />
Blood. 1996;88:2787-2793.<br />
19. Garban F, Attal M, Michallet M, et al. Prospective comparison <strong>of</strong><br />
autologous stem cell transplantation followed by dose-reduced allograft<br />
(IFM99-03 trial) with tandem autologous stem cell transplantation<br />
(IFM99-04 trial) in high-risk de novo multiple myeloma. Blood. 2006;107:<br />
3474-3480.<br />
20. Bruno B, Storer B, Patriarca F, et al. Long-term follow up <strong>of</strong> a<br />
comparison <strong>of</strong> non-myeloablative allografting with autografting for newly<br />
diagnosed myeloma. Blood. 2010;116 (abstr 525).<br />
21. Krishnan A, Pasquini M, Logan B, et al. Autologous hematopoietic stem<br />
cell transplantation followed by allogeneic or autologous hematopoietic stem<br />
cell transplantation in patients with multiple myeloma (BMTCTN0102): a<br />
phase 3 biological assignment trial. Lancet Oncol. 2011;12:1195-1203.<br />
22. Bjorkstrand B, Iacobelli S, Hegenbart U, et al. Tandem autologous/<br />
reduced-intensity conditioning allogeneic stem-cell transplantation versus
TRANSPLANTATION FOR MYELOMA<br />
autologous transplantation in myeloma: long-term follow-up. J Clin Oncol.<br />
2011;29:3016-3022.<br />
23. Fermand JP, Ravaud P, Chevret S, et al. High-dose therapy and<br />
autologous peripheral blood stem cell transplantation in multiple myeloma;<br />
up-front or rescue treatment? Results <strong>of</strong> a multicenter sequential randomized<br />
clinical trial. Blood. 1998;92:3131-3136.<br />
24. Van Rhee F, Szymonika J, Anaissie E, et al. Total Therapy Three for<br />
multiple myeloma: prognostic implications <strong>of</strong> cumulative dosing and premature<br />
discontinuation <strong>of</strong> VTD maintenance components bortezomib, thalidomide,<br />
and dexamethasone relevant to all phases <strong>of</strong> therapy. Blood. 2010;115:<br />
1220-1227.<br />
25. Ladetto M, Pagliano G, Ferrero S, et al. Major tumor shrinking and<br />
persistent molecular remissions after consolidation with bortezomib, thalidomide,<br />
and dexamethasone in patients with autograft for myeloma. J Clin<br />
Oncol. 2010;28:2077-2084.<br />
26. Mellqvist UH, Ginsing P, Hjertner O, et al. Improved progression<br />
free survival with bortezomib consolidation after high-dose melphalan; results<br />
<strong>of</strong> a randomized phase III trial. Haematologica. 2011;96:S31 (suppl;<br />
abstr 11).<br />
27. Rapoport A, Aqui N, Stadtmauer E, et al. Combination immunotherapy<br />
using adoptive T cell transfer and tumor antigen vaccination on the basis <strong>of</strong><br />
HTERT and surviving after ASCT for myeloma. Blood. 2011;117:788-797.<br />
507
Upfront Therapy for Myeloma: Tailoring<br />
Therapy across the Disease Spectrum<br />
Overview: The treatment <strong>of</strong> multiple myeloma is evolving<br />
rapidly. Despite the number <strong>of</strong> regimens and combinations<br />
available, there is lack <strong>of</strong> data from phase III trials demonstrating<br />
superiority <strong>of</strong> one regimen over the other in terms <strong>of</strong><br />
overall survival and/or quality <strong>of</strong> life. The only clear survival<br />
signals have come from studies that compared newer regimens<br />
with historic ones such as melphalan-prednisone (MP)<br />
or vincristine-doxorubicin hydrochloride-thalidomide (VAD).<br />
Thus, the choice <strong>of</strong> therapy at present is <strong>of</strong>ten made based on<br />
physician discretion, bias, and limited data from phase II<br />
studies. Further, the regimens available have considerably<br />
MULTIPLE MYELOMA is a malignant plasma-cell<br />
proliferative disorder that accounts for approximately<br />
10% <strong>of</strong> all hematologic malignancies. 1,2 Over 20,000<br />
new cases are diagnosed annually in the United States, and<br />
most patients continue to die <strong>of</strong> the disease. The risk <strong>of</strong><br />
multiple myeloma is two-fold higher in blacks compared<br />
with whites. Multiple myeloma evolves from an asymptomatic<br />
premalignant stage termed monoclonal gammopathy <strong>of</strong><br />
undetermined significance (MGUS). In the spectrum <strong>of</strong><br />
plasma cell disorders, there is an additional intermediate<br />
arbitrary clinical stage–termed smoldering multiple myeloma<br />
(SMM) that comprises <strong>of</strong> a mixture <strong>of</strong> patients with<br />
premalignancy (i.e., MGUS) and malignancy (i.e., early<br />
stage multiple myeloma). The SMM category is clinically<br />
important because it is not possible in most instances to<br />
determine which patient with SMM has premalignancy<br />
compared with malignancy; therefore, these patients are<br />
currently observed without therapy until overt signs or<br />
malignancy develop. They are also candidates for clinical<br />
trials, testing the value <strong>of</strong> early intervention. The diagnostic<br />
criteria for multiple myeloma, MGUS, and SMM are given in<br />
Table 1. 3<br />
Prognosis and Risk-Stratification<br />
Prognosis in myeloma depends on four major factors: host<br />
characteristics (age, performance status, comorbidities),<br />
stage, disease aggressiveness, and response to therapy. 4<br />
Each <strong>of</strong> these four factors result in poor prognosis through<br />
different mechanisms, and therefore the strategy to overcome<br />
each <strong>of</strong> these factors will need to be quite different. For<br />
example, it is not possible to overcome the poor prognosis<br />
associated with host factors such as advanced age or multiple<br />
comorbidities by increasing the aggressiveness <strong>of</strong> the<br />
therapy used. Similarly, it is not possible to overcome the<br />
poor prognostic effect conferred by acute renal failure without<br />
taking into account the specific drugs than can be used<br />
safely, and the specific dose modifications that may be<br />
required. In other words, one needs to know precisely the<br />
mechanism by which a given factor confers its adverse<br />
prognostic value before we can develop strategies to overcome<br />
such an effect.<br />
Staging <strong>of</strong> myeloma by using the Durie-Salmon Staging<br />
(DSS) or the International Staging System (ISS) both provide<br />
prognostic information, but are typically not helpful in<br />
making therapeutic choices because the risk groups are<br />
508<br />
By S. Vincent Rajkumar, MD<br />
different pr<strong>of</strong>iles in terms <strong>of</strong> safety, convenience, and cost.<br />
Given the dramatic variations in expected outcome depending<br />
on the various known prognostic factors, a risk-adapted<br />
strategy is required to provide the best available therapy to<br />
each patient based on host factors as well as prognostic<br />
markers <strong>of</strong> disease aggressiveness. This article reviews the<br />
current status <strong>of</strong> myeloma therapy and risk stratification.<br />
Results from major phase III trials are reviewed, and a<br />
risk-adapted individualized approach to therapy is presented<br />
and discussed.<br />
heterogeneous and defy any rational uniform approach to<br />
risk-adapted therapy. For instance, stage III ISS (defined on<br />
the basis <strong>of</strong> a high beta-2 microglobulin level) consists <strong>of</strong> a<br />
heterogeneous mix <strong>of</strong> patients, some <strong>of</strong> whom derive the<br />
stage III label because <strong>of</strong> renal failure, whereas some meet<br />
criteria for stage III because <strong>of</strong> a high tumor burden; both<br />
renal failure and high tumor burden have the same effect on<br />
beta-2 microglobulin. Thus staging, whereas useful for prognosis,<br />
is not particularly useful in deciding choice <strong>of</strong> therapy<br />
in multiple myeloma; the one exception to this rule is stage<br />
I Durie-Salmon disease, which is a heterogeneous group <strong>of</strong><br />
patients with either SMM (no therapy needed) or solitary<br />
plasmacytoma (typically treated with radiation alone).<br />
Response to therapy is also not useful in determining<br />
choice <strong>of</strong> therapy since this information is only available<br />
post-hoc, and there are no good reliable tests that allow us<br />
to predict response to specific drugs ahead <strong>of</strong> time. Thus<br />
risk-adapted, individualized therapy for myeloma is currently<br />
done primarily on the basis <strong>of</strong> two factors: host factors<br />
and markers <strong>of</strong> disease aggressiveness.<br />
Host factors are currently used in multiple myeloma<br />
primarily to determine eligibility for autologous stem cell<br />
transplantation (ASCT). In general, patients need to meet<br />
minimum requirements for age, organ function, and performance<br />
status to be considered eligible for ASCT. The initial<br />
therapy varies according to eligibility for transplantation.<br />
Disease aggressiveness has a major effect on prognosis and<br />
can also be used to individualize therapy. Table 2 provides a<br />
risk stratification model based on markers <strong>of</strong> disease aggressiveness<br />
(mSMART). 5 This requires fluorescent in situ hybridization<br />
(FISH) or similar studies be done on the bone<br />
marrow at the time <strong>of</strong> initial diagnosis to detect t(11;14),<br />
t(4;14), t(14;16), t(6;14), t(14;20), and deletion <strong>of</strong> 17p. 5 Gene<br />
expression pr<strong>of</strong>iling (GEP), if available, can provide additional<br />
prognostic value. Patients with standard-risk myeloma<br />
(75% <strong>of</strong> myeloma) have a median overall survival (OS)<br />
<strong>of</strong> an excess <strong>of</strong> 7 years whereas those with high-risk disease<br />
From the Division <strong>of</strong> Hematology, Mayo Clinic, Rochester, MN.<br />
Author’s disclosure <strong>of</strong> potential conflicts <strong>of</strong> interest are found at the end <strong>of</strong> this article.<br />
Address reprint requests to S. Vincent Rajkumar, MD, Division <strong>of</strong> Hematology, Mayo<br />
Clinic, 200 First Street SW, Rochester, MN 55905; email: rajkumar.vincent@mayo.edu.<br />
© <strong>2012</strong> by <strong>American</strong> <strong>Society</strong> <strong>of</strong> <strong>Clinical</strong> <strong>Oncology</strong>.<br />
1092-9118/10/1-10
CURRENT STATUS OF MYELOMA THERAPY<br />
KEY POINTS<br />
Table 1. Diagnostic Criteria for Multiple Myeloma and Related Disorders<br />
Disorder Disease Definition<br />
Monoclonal gammopathy <strong>of</strong> undetermined<br />
significance (MGUS)<br />
● The diagnosis <strong>of</strong> multiple myeloma requires 10% or<br />
more clonal plasma cells on bone marrow examination<br />
or a biopsy-proven plasmacytoma plus evidence<br />
<strong>of</strong> end-organ damage felt to be related to the underlying<br />
plasma cell disorder.<br />
● Patients with 17p deletion, t(14;16), t(14;20), or highrisk<br />
gene expression pr<strong>of</strong>iling signature have highrisk<br />
myeloma; patients with t(4;14) translocation<br />
have intermediate-risk disease; all others are considered<br />
to have standard-risk myeloma.<br />
● A risk-adapted strategy to multiple myeloma is essential<br />
because prognosis varies dramatically across<br />
risk groups, the various modern regimens in use have<br />
different safety and cost pr<strong>of</strong>iles, and there are no<br />
clear survival or quality-<strong>of</strong>-life data from randomized<br />
trials showing a clear superiority <strong>of</strong> one strategy over<br />
the other.<br />
● Risk-stratified therapy relies on both host factors as<br />
well as prognostic markers <strong>of</strong> disease aggressiveness.<br />
● Bortezomib-based regimens appear to be able to overcome<br />
the poor prognostic effect <strong>of</strong> t4;14 translocation.<br />
All three criteria must be met:<br />
• Serum monoclonal protein �3gm/dL<br />
• Clonal bone marrow plasma cells �10%, and<br />
• Absence <strong>of</strong> end-organ damage such as hypercalcemia, renal insufficiency, anemia, and bone lesions (CRAB) that can be<br />
attributed to the plasma cell proliferative disorder; or in the case <strong>of</strong> IgM MGUS, no evidence <strong>of</strong> anemia, constitutional<br />
symptoms, hyperviscosity, lymphadenopathy, or hepatosplenomegaly that can be attributed to the underlying<br />
lymphoproliferative disorder.<br />
Light chain MGUS All criteria must be met:<br />
• Abnormal FLC ratio (�0.26 or �1.65)<br />
• Increased level <strong>of</strong> the appropriate involved light chain (increased kappa FLC in patients with ratio �1.65 and increased<br />
lambda FLC in patients with ratio �0.26)<br />
• Absence <strong>of</strong> end-organ damage such as hypercalcemia, renal insufficiency, anemia, and bone lesions (CRAB) that can be<br />
attributed to the plasma cell proliferative disorder<br />
• No immunoglobulin heavy chain expression on immun<strong>of</strong>ixation<br />
Smoldering multiple myeloma (also referred<br />
to as asymptomatic multiple myeloma)<br />
Both criteria must be met:<br />
• Serum monoclonal protein (IgG or IgA) �3gm/dL and/or clonal bone marrow plasma cells �10%, and<br />
• Absence <strong>of</strong> end-organ damage such as lytic bone lesions, anemia, hypercalcemia, or renal failure that can be attributed<br />
to a plasma cell proliferative disorder<br />
Solitary Plasmacytoma All four criteria must be met<br />
• Biopsy-proven solitary lesion <strong>of</strong> bone or s<strong>of</strong>t tissue with evidence <strong>of</strong> clonal plasma cells<br />
• Normal bone marrow with no evidence <strong>of</strong> clonal plasma cells<br />
• Normal skeletal survey and MRI <strong>of</strong> spine and pelvis (except for the primary solitary lesion)<br />
• Absence <strong>of</strong> end-organ damage such as hypercalcemia, renal insufficiency, anemia, or other bone lesions (CRAB) that<br />
can be attributed to a lympho-plasma cell proliferative disorder<br />
Multiple myeloma All three criteria must be met except as noted:<br />
• Clonal bone marrow plasma cells �10% and/or biopsy proven plasmacytoma<br />
• Presence <strong>of</strong> serum and/or urinary monoclonal protein (except in patients with true non-secretory multiple myeloma), and<br />
• Evidence <strong>of</strong> end organ damage that can be attributed to the underlying plasma cell proliferative disorder, specifically<br />
o Hypercalcemia: Serum calcium �11.5 mg/dL or<br />
o Renal insufficiency: Serum creatinine �1.73 mmol/L (or �2 mg/dL) or estimated creatinine clearance less than 40<br />
mL/min<br />
o Anemia: Normochromic, normocytic with a hemoglobin value <strong>of</strong> �2 g/dL below the lower limit <strong>of</strong> normal or a<br />
hemoglobin value �10 g/dL<br />
o Bone lesions: Lytic lesions, severe osteopenia or pathologic fractures<br />
Modified from Kyle and Rajkumar. 1<br />
(15% <strong>of</strong> myeloma) have a median OS <strong>of</strong> 2 to 3 years, despite<br />
initial therapy containing bortezomib, tandem ASCT, and<br />
routine use <strong>of</strong> maintenance therapy. 1<br />
Risk-Adapted Therapy<br />
With the number <strong>of</strong> regimens and combinations in myeloma<br />
available, and the lack <strong>of</strong> phase III trials demonstrating<br />
superiority in the only two endpoints that matter<br />
(overall survival and/or quality <strong>of</strong> life), the choice <strong>of</strong> therapy<br />
is <strong>of</strong>ten made based on physician discretion, bias, and<br />
limited data from phase II studies. Further, the regimens<br />
available have considerably different pr<strong>of</strong>iles in terms <strong>of</strong><br />
safety, convenience, and cost, thereby dictating the need for<br />
Modified from Rajkumar. 2<br />
Table 2. Risk-Stratification <strong>of</strong> Myeloma<br />
A. Standard Risk<br />
• Hyperdiploidy<br />
• t (11;14)<br />
• t (6;14)<br />
B. Intermediate Risk<br />
• t (4;14)<br />
C. High Risk<br />
• 17p deletion<br />
• t (14;16)<br />
• t (14;20)<br />
• High-risk gene expression pr<strong>of</strong>iling signature<br />
509
Table 3. Rationale for Individualized, Risk-Adapted Therapy in Multiple Myeloma<br />
Factor Individualized Approach to Initial Therapy Rationale<br />
Age, performance status, co-morbidities Decision on Transplant Eligible versus Transplant Ineligible • The value <strong>of</strong> ASCT has been demonstrated in<br />
• ASCT <strong>of</strong>fered to patients age 65 (or up to age 75 in<br />
USA) who have adequate performance status and<br />
acceptable organ function<br />
• Nature and duration <strong>of</strong> initial therapy varies<br />
according to eligibility for ASCT<br />
a judicious approach to the use <strong>of</strong> the available options.<br />
Thus, given the dramatic variations in prognosis depending<br />
on the various known prognostic factors, a “one-size-fits-all”<br />
approach is both unreasonable and obsolete in multiple<br />
myeloma. Even skeptics <strong>of</strong> risk-adapted or individualized<br />
therapy do in reality practice such an approach, albeit<br />
subconsciously. Table 3 shows how therapy can be tailored<br />
by using host factors and markers <strong>of</strong> disease aggressiveness<br />
and the rationale for such a risk-adapted strategy.<br />
A risk-adapted strategy does not and should not mean<br />
giving suboptimal (weak) therapy to patients with standard<br />
risk and reserving the best therapy for patients at high risk.<br />
To the contrary, it means not subjecting (or recommending<br />
to) patients with standard risk regimens that have not been<br />
proven to be useful in randomized trials. For example, VRD<br />
has shown promise in newly diagnosed myeloma in a phase<br />
II trial. That is sufficient to design phase III trials with this<br />
regimen, but insufficient to recommend as an option outside<br />
a clinical trial setting to patients with newly diagnosed<br />
myeloma, especially to patients with standard-risk disease<br />
in whom the value <strong>of</strong> this regimen, given its cost and<br />
toxicity, requires data from randomized trials. One could<br />
randomized trials only in this patient population<br />
• Patient safety<br />
Advanced age (�80 yr) Regimens with the least impact on quality <strong>of</strong> life • Although phase III data are not available, oral<br />
Acute renal failure due to light chain<br />
cast-nephropathy<br />
•Rd<br />
•MP<br />
Rapidly acting regimens using drugs that are safe to use<br />
in renal failure<br />
regimens such as Rd or MP are well tolerated and<br />
effective in patients with advanced age<br />
• Rapid reversal <strong>of</strong> renal failure is possible with this<br />
approach<br />
• VTD • Persistent renal failure carries an adverse prognosis<br />
• VCD • Patient safety dictates that in the presence <strong>of</strong> acute<br />
Possible plasmapheresis<br />
Possible hemodialysis<br />
Goal <strong>of</strong> therapy is rapid reduction <strong>of</strong> involved serum-free<br />
light chains to less than 50 mg/dL<br />
Standard-risk myeloma Wide choice <strong>of</strong> induction regimens based on patient<br />
preference, cost, toxicity, availability, etc.<br />
•Rd<br />
•VD<br />
• VTD<br />
• VCD<br />
• Other<br />
renal failure, drugs that are renally excreted, such as<br />
lenalidomide, are best avoided particularly when<br />
reversal <strong>of</strong> renal failure is a stated goal<br />
• Little or no phase III data showing overall survival is<br />
affected based on which <strong>of</strong> these induction regimens<br />
are used as initial therapy<br />
• Cost, toxicity, and availability concerns do matter and<br />
should be taken into consideration in the absence <strong>of</strong><br />
compelling overall survival data dictating the use <strong>of</strong><br />
one specific regimen over the other, given the<br />
estimated median survival <strong>of</strong> �7–10 yr<br />
Intermediate-risk myeloma Bortezomib-based induction required Almost complete reversal <strong>of</strong> the poor prognostic impact<br />
•VD<br />
• VTD<br />
• VCD<br />
ASCT<br />
Bortezomib-based maintenance therapy<br />
<strong>of</strong> t4;14 has been noted with this strategy<br />
High-risk myeloma Experimental approaches The median OS with approaches used in either<br />
Plasma cell leukemia or multiple sites <strong>of</strong><br />
extramedullary disease at<br />
presentation<br />
Regimens with promising phase II data<br />
• VRD<br />
Initial therapy with multi-agent chemotherapy such as<br />
VDT-PACE<br />
standard- or intermediate-risk disease produce a<br />
median survival <strong>of</strong> only 2–3 yr<br />
The median OS is poor with approaches used in either<br />
standard- or intermediaterisk disease<br />
ASCT VDT-PACE can control disease rapidly and predictably<br />
Maintenance therapy<br />
Abbreviations: ASCT, autologous stem cell transplantation; MP, melphalan plus prednisone; Rd, lenalidomide plus low dose dexamethasone; VCD, bortezomib,<br />
cyclophosphamide, dexamethasone; VD, bortezomib plus dexamethasone; VDT-PACE, bortezomib, dexamethasone, thalidomide, cisplatin, doxorubicin, cyclophosphamide,<br />
etoposide; VRD, bortezomib, lenalidomide, dexamethasone; VTD, bortezomib, thalidomide, dexamethasone.<br />
510<br />
S. VINCENT RAJKUMAR<br />
argue, however, that the outcome <strong>of</strong> patients with high-risk<br />
myeloma is so poor with current available (phase III-vetted)<br />
regimens that a new regimen with high activity such as VRD<br />
is reasonable despite the added cost and toxicity. In other<br />
words, the principle <strong>of</strong> risk-adapted therapy is to individualize<br />
the treatment options for each patient in a manner<br />
that carefully balances the risks and benefits depending on<br />
prognosis and patient expectations. Patients with high risk<br />
may be willing to take the added risks <strong>of</strong> unproven therapy<br />
more <strong>of</strong>ten than patients with standard risk. Patients with<br />
standard risk should be treated by using regimens that have<br />
been tested and found effective in phase III trials.<br />
More importantly, there is a critical ongoing “cure compared<br />
with control” debate in the myeloma community on<br />
whether we should treat the disease with an aggressive<br />
multidrug strategy targeting complete response (CR) or a<br />
sequential disease control approach that emphasizes quality<br />
<strong>of</strong> life as well as OS. 6 A risk-adapted strategy allows patients<br />
with standard-risk myeloma to have the choice<br />
between pursuing either approach since there are no randomized<br />
data demonstrating the clear superiority <strong>of</strong> one<br />
approach over the other. It also highlights the need for such
CURRENT STATUS OF MYELOMA THERAPY<br />
Fig 1. Approach to the treatment <strong>of</strong> newly diagnosed myeloma (A)<br />
and in newly diagnosed patients with myeloma with special circumstances<br />
(B).<br />
Abbreviations: ASCT, autologous stem cell transplantation; CR,<br />
complete response; Dex, dexamethasone; Rd, lenalidomide plus<br />
low-dose dexamethasone; VCD, bortezomib, cyclophosphamide,<br />
dexamethasone; VDT-PACE, bortezomib, dexamethasone, thalidomide,<br />
cisplatin, doxorubicin, cyclophosphamide, etoposide; VGPR,<br />
very good partial response; VRD, bortezomib, lenalidomide, dexamethasone;<br />
VTD, bortezomib, thalidomide, dexamethasone.<br />
* For patients who choose delayed ASCT, continue Rd.<br />
† When continuing Rd, dexamethasone usually discontinued after<br />
12 months, and continued long-term lenalidomide is an option for<br />
patients who are tolerating treatment well.<br />
trials. On the other hand, patients with high-risk require a<br />
complete response for long-term survival and hence clearly<br />
need an aggressive strategy designed to eradicate all plasma<br />
cells, if possible.<br />
Approach to Initial Therapy<br />
The approach to treatment <strong>of</strong> symptomatic newly diagnosed<br />
multiple myeloma is outlined in Fig. 1 and is dictated<br />
by host factors that govern eligibility for ASCT and markers<br />
<strong>of</strong> disease aggressiveness that determine risk-stratification.<br />
1 The major regimens used for therapy and the data to<br />
support their use are listed in Tables 4 and 5. In general,<br />
all patients are probably best served by participation in<br />
clinical trials, and every effort should be made to encourage<br />
participation in such studies. The algorithms outlined in<br />
this review are to be considered primarily when a suitable<br />
clinical trial is unavailable or impractical.<br />
Initial Treatment in Patients Eligible for ASCT<br />
Typically, patients eligible for ASCT are treated with<br />
approximately two to four cycles <strong>of</strong> induction therapy before<br />
stem cell harvest. In general, regardless <strong>of</strong> the regimen<br />
used, once weekly subcutaneous bortezomib and dexamethasone<br />
at a 40-mg once-a-week schedule (low-dose dexamethasone)<br />
is preferred in most patients for initial therapy,<br />
unless there is an urgent need for rapid disease control.<br />
Several other regimens besides the ones discussed in this<br />
review have been tested in newly diagnosed multiple myeloma,<br />
but there are no clear data from randomized controlled<br />
trials that they have an effect on long-term<br />
endpoints.<br />
Standard risk. Patients with standard risk who are eligible<br />
for ASCT can be treated with a variety <strong>of</strong> active regimens.<br />
The main options for initial therapy are lenalidomide<br />
plus low-dose dexamethasone (Rd) or one <strong>of</strong> several<br />
bortezomib-based regimens. Lenalidomide plus dexamethasone<br />
is active in newly diagnosed myeloma. Rd, which uses<br />
low-dose dexamethasone, has less toxicity and better overall<br />
survival than lenalidomide plus high-dose dexamethasone. 7<br />
One caveat is that Rd may impair collection <strong>of</strong> peripheral<br />
blood stem cells for transplant in some patients when<br />
mobilized with granulocyte–colony stimulating factor (G-<br />
CSF) alone. Thus, patients over the age <strong>of</strong> 65 and those who<br />
have received more than four cycles <strong>of</strong> Rd) stem cells must be<br />
mobilized with either cyclophosphamide plus G-CSF or with<br />
plerixafor. All patients require antithrombosis prophylaxis<br />
with aspirin; low molecular–weight heparin or coumadin is<br />
needed in patients at high risk <strong>of</strong> DVT. 8<br />
Several bortezomib-containing regimens can also be used<br />
as initial therapy instead <strong>of</strong> Rd. Harousseau and colleagues<br />
compared bortezomib plus dexamethasone (VD) with vincristine,<br />
doxorubicin hydrochloride, dexamethasone (VAD)<br />
as pre-transplant–induction therapy. 9 Post-induction verygood-partial-response<br />
(VGPR) was superior with VD compared<br />
with VAD, 38% compared with 15%, respectively.<br />
However, progression-free survival (PFS) improvement was<br />
modest, 36 months compared with 30 months, respectively,<br />
and did not reach statistical significance. No overall survival<br />
benefit was noted. Three-drug regimens such as bortezomibcyclophosphamide-dexamethasone<br />
(VCD), and bortezomibthalidomide-dexamethasone<br />
(VTD), and bortezomiblenalidomide-dexamethasone<br />
(VRD) are in various stages<br />
<strong>of</strong> development. 10 In randomized trials, VTD has shown<br />
better response rates and PFS compared with TD, 11 as well<br />
as VD. 12 VCD has impressive activity in newly diagnosed<br />
multiple myeloma and is less expensive than either VTD or<br />
VRD. Preliminary studies indicate that VCD is well tolerated<br />
and has similar activity compared with VRD, making<br />
it an excellent choice when considering a bortezomibcontaining<br />
regimen for frontline use. 13<br />
There are no data so far comparing Rd with any <strong>of</strong> the<br />
bortezomib-containing regimens. Results from a Southwest<br />
<strong>Oncology</strong> Group (SWOG) randomized trial that compared<br />
VRD to Rd are awaited. One drawback <strong>of</strong> bortezomibcontaining<br />
regimens is the risk <strong>of</strong> neurotoxicity early in the<br />
disease course. However, recent studies show that the neurotoxicity<br />
<strong>of</strong> bortezomib can be greatly diminished by administering<br />
bortezomib using a once-weekly schedule, 14,15 and<br />
by administering the drug subcutaneously. 16 Unlike lenalidomide,<br />
bortezomib does not appear to have any adverse<br />
511
effect on stem cell mobilization and does not cause DVT.<br />
Since there are no data from randomized trials, the choice <strong>of</strong><br />
therapy is driven primarily by expert opinion, feasibility,<br />
cost, and patient preference.<br />
Based on cost and toxicity considerations, Rd or VCD are<br />
probably the best options for this patient population at<br />
present. After four cycles <strong>of</strong> therapy with Rd or VCD, stem<br />
cells should be harvested, and patients can either undergo<br />
frontline ASCT (preferred) or resume induction therapy,<br />
delaying ASCT until first relapse.<br />
Intermediate risk. In patients with standard risk, there<br />
are no data from randomized trials that a bortezomibcontaining<br />
regimen provides superior OS compared with Rd,<br />
or vice versa. By contrast, so far only bortezomib-containing<br />
regimens appear to be able overcome the poor prognosis<br />
associated with the t4;14 translocation. 11 There are also<br />
data that the best results in the t4;14 population come from<br />
studies that used post-transplant bortezomib-based maintenance<br />
therapy, and double ASCT.<br />
High risk. Even regimens as aggressive as total therapy 3<br />
(bortezomib-based induction therapy, tandem autologous<br />
transplantation, and bortezomib-based maintenance therapy)<br />
have not made any meaningful improvement in the<br />
outcome <strong>of</strong> patients with high-risk disease. Some <strong>of</strong> these<br />
Table 4. Major Treatment Regimens in Multiple Myeloma<br />
Regimen Usual Dosing Schedule*<br />
Melphalan-prednisone (7-d schedule) Melphalan 8–10 mg oral days 1–7<br />
Prednisone 60 mg/d oral days 1–7<br />
Repeated every 6 wk<br />
Thalidomide-dexamethasone** Thalidomide 200 mg oral days 1–28<br />
Dexamethasone 40 mg oral days 1, 8, 15, 22<br />
Repeated every 4 wk<br />
Lenalidomide-dexamethasone Lenalidomide 25 mg oral days 1–21 every 28 d<br />
Dexamethasone 40 mg oral days 1, 8, 15, 22 every 28 d<br />
Repeated every 4 wk<br />
Bortezomib-dexamethasone** Bortezomib 1.3 mg/m 2 subcutaneous or intravenous days 1, 8, 15, 22<br />
Dexamethasone 20 mg on day <strong>of</strong> and day after bortezomib (or 40 mg days 1, 8, 15, 22)<br />
Repeated every 4 wk<br />
Melphalan-prednisone-thalidomide Melphalan 0.25 mg/kg oral days 1–4 (use 0.20 mg/kg/d oral days 1–4 in patients over the age <strong>of</strong> 75)<br />
Prednisone 2 mg/kg oral days 1–4<br />
Thalidomide 100–200 mg oral days 1–28 (use 100 mg dose in patients �75)<br />
Repeated every 6 wk<br />
Bortezomib-melphalan-prednisone** Bortezomib 1.3 mg/m 2 subcutaneous or intravenous days 1, 8, 15, 22<br />
Melphalan 9 mg/m 2 oral days 1–4<br />
Prednisone 60 mg/m 2 oral days 1 to 4<br />
Repeated every 35 d<br />
Bortezomib-thalidomide-dexamethasone** Bortezomib 1.3 mg/m 2 subcutaneous or intravenous days 1, 8, 15, 22<br />
Thalidomide 100–200 mg oral days 1–21<br />
Dexamethasone 20 mg on day <strong>of</strong> and day after bortezomib (or 40 mg days 1, 8, 15, 22)<br />
Repeated every 4 wk � 4 cycles as pre-transplant induction therapy<br />
Bortezomib-cyclophosphamide-dexamethasone** (VCD) Cyclophosphamide 300 mg/m 2 orally on days 1, 8, 15 and 22<br />
Bortezomib 1.3 mg/m 2 subcutaneous or intravenously on days 1, 8, 15, 22<br />
Dexamethasone 40 mg orally on days on days 1, 8, 15, 22<br />
Repeated every 4 wk†<br />
Bortezomib-lenalidomide-dexamethasone** Bortezomib 1.3 mg/m 2 subcutaneous or intravenous days 1, 8, 15<br />
Lenalidomide 25 mg oral days 1–14<br />
Dexamethasone 20 mg on day <strong>of</strong> and day after bortezomib (or 40 mg days 1, 8, 15, 22)<br />
Repeated every 3 wk‡<br />
Reproduced from Rajkumar. 2<br />
* All doses need to be adjusted for performance status, renal function, blood counts, and other toxicities.<br />
** Doses <strong>of</strong> dexamethasone and/or bortezomib reduced based on subsequent data showing lower toxicity and similar efficacy with reduced doses.<br />
† Omit day 22 dose if counts are low or when the regimen is used as maintenance therapy; when used as maintenance therapy for patients at high risk, delays can<br />
be instituted between cycles.<br />
‡ Omit day 15 dose if counts are low or when the regimen is used as maintenance therapy; when used as maintenance therapy for patients at high risk , lenalidomide<br />
dose may be decreased to 10–15 mg per day, and delays can be instituted between cycles as done in total therapy protocols. 17,28<br />
512<br />
S. VINCENT RAJKUMAR<br />
patients may be content with the best quality <strong>of</strong> life possible,<br />
given the grave prognosis, and choose a gentle approach<br />
such as Rd. However, it would be reasonable to consider<br />
expensive, albeit unproven, approaches such as VRD in this<br />
patient population. It is not clear whether a full myeloablative<br />
allogeneic transplantation may be <strong>of</strong> value, but some<br />
patients are willing to take on the high risk <strong>of</strong> transplantrelated<br />
mortality in exchange for a chance <strong>of</strong> long-term<br />
survival.<br />
Plasma cell leukemia or multiple extramedullary plasmacytomas.<br />
Multiagent combination chemotherapy should be<br />
considered for these patients as initial therapy since rapid<br />
and reliable disease control is essential. VDT-PACE (bortezomib,<br />
dexamethasone, thalidomide, cisplatin, doxorubicin,<br />
cyclophosphamide, and etoposide) has been tested extensively<br />
as part <strong>of</strong> the total therapy 3 trials and is a valuable<br />
option. 17 Following two cycles <strong>of</strong> therapy, these patients<br />
should be considered candidates for ASCT and subsequently<br />
for maintenance therapy with a bortezomib-based regimen.<br />
Options for Initial Treatment in Patients Not Eligible<br />
for ASCT<br />
In patients with newly diagnosed multiple myeloma who<br />
are considered ineligible for ASCT because <strong>of</strong> age or other
CURRENT STATUS OF MYELOMA THERAPY<br />
Trial Regimen<br />
Table 5. Results <strong>of</strong> Recent Randomized Studies in Newly Diagnosed Myeloma<br />
No. <strong>of</strong><br />
Patients<br />
Overall<br />
Response<br />
Rate (%)<br />
CR plus<br />
VGPR (%)<br />
comorbidities, the major options at present are either<br />
alkylator- or bortezomib-based combination therapies given<br />
for 9 to 18 months, or Rd. 1 With Rd, the optimum duration<br />
<strong>of</strong> therapy is unclear; most patients continue therapy until<br />
progression, provided the treatment is well tolerated. In<br />
such patients, dexamethasone is usually lowered to a minimal<br />
dose or discontinued after the first year. Results <strong>of</strong> a<br />
randomized trial comparing Rd for 18 months compared<br />
with Rd until progression are awaited. Thalidomide plus<br />
dexamethasone is inferior to melphalan plus prednisone<br />
(MP) in terms <strong>of</strong> overall survival and is not recommended.<br />
The addition <strong>of</strong> lenalidomide to MP (MPR) does not improve<br />
PFS compared with MP alone, and is also not recommended.<br />
Standard risk. Six randomized studies have found that<br />
melphalan, prednisone, thalidomide (MPT) improves response<br />
rates with MP. 18-23 Four <strong>of</strong> these trials have shown a<br />
prolongation <strong>of</strong> PFS with MPT, 18-20,22 and an OS advantage<br />
has been observed in the two Intergroupe Francophone<br />
Myelome (IFM) trials and in the trial by Wijermans and<br />
colleagues. 18-22 Two meta-analyses <strong>of</strong> these randomized trials<br />
have been conducted and they show a clear superiority<br />
<strong>of</strong> MPT over MP. 24,25 Grade 3–4 adverse events occur in<br />
approximately 55% <strong>of</strong> patients treated with MPT, compared<br />
to 22% with MP. 20<br />
Bortezomib-containing combinations serve as an alternative<br />
to MPT as initial therapy in this patient population. In<br />
a large phase III trial, bortezomib, melphalan, prednisone<br />
(VMP) led to improved OS compared with MP. 26 Neuropathy<br />
is an important risk with VMP therapy; grade 3 neuropathy<br />
occurred in 13% <strong>of</strong> patients compared to 0% with MP. 26<br />
Thus, as with transplant-eligible patients, the once-weekly,<br />
subcutaneous schedule is preferred. There is no significant<br />
Progression-free<br />
Survival<br />
(Median in Months)<br />
P for Progression<br />
Free Survival<br />
3 yr Overall<br />
Survival<br />
Rate (%)*<br />
Overall Survival<br />
(Median in Months)<br />
Rajkumar et al 7 RD 223 81 50 19.1 75 NR<br />
Rd 222 70 40 25.3 0.026 74 NR 0.47<br />
Harousseau et al 9 VAD 242 63 15 30 77 NR<br />
VD 240 79 38 36 0.06 81 NR 0.46<br />
Cavo et al 11 TD 238 79 28 40 84 NR<br />
VTD 236 93 62 NR 0.006 86 NR 0.3<br />
Moreau et al 12 VD 99 81 35 N/A N/A N/A<br />
VTD 100 90 51 N/A N/A N/A<br />
Facon et al 18 MP 196 35 7 17.8 48 33.2<br />
Mel 100 126 65 43 19.4 52 38.3<br />
MPT 125 76 47 27.5 �0.001 66 51.6 �0.001<br />
Hulin et al 19 MP � Placebo 116 31 7 18.5 40 29.1<br />
MPT 113 62 21 24.1 0.001 55 44 0.028<br />
Wijermans et al 22 MP 168 45 10 9 43 31<br />
MPT 165 66 27 13 �0.001 55 40 0.05<br />
Palumbo et al 29 MP 164 48 11 14.5 65 47.6<br />
MPT 167 69 29 21.8 0.004 65 45 0.79<br />
Waage et al 21 MP � Placebo 175 33 7 14 43 32<br />
MPT 182 34 23 15 NS 43 29 0.16<br />
San Miguel et al** 26,30 MP 331 35 8 16.6 54 43<br />
VMP 337 71 41 24 �0.001 69 NR �0.001<br />
Reproduced from Rajkumar. 2<br />
Abbreviations: CR, complete response; MP, melphalan plus prednisone; MPT, melphalan plus prednisone plus thalidomide; N/A, not available; NS, not significant; Rd,<br />
lenalidomide plus dexamethasone; TD, thalidomide plus dexamethasone; VGPR, very good partial response; VMP, bortezomib plus melphalan plus prednisone; VTD,<br />
bortezomib, thalidomide, dexamethasone.<br />
* Estimated from survival curves when not reported.<br />
** Progression-free survival not reported; numbers indicate time to progression.<br />
P for<br />
Overall<br />
Survival<br />
advantage <strong>of</strong> either bortezomib, thalidomide, prednisone<br />
(VTP) or VTD over VMP. VCD is a minor variation <strong>of</strong> the<br />
VMP combination that can be used in place <strong>of</strong> VMP to<br />
minimize side effects.<br />
Rd is also an attractive option for the treatment <strong>of</strong> elderly<br />
patients with newly diagnosed myeloma because <strong>of</strong> its<br />
excellent tolerability, convenience, and efficacy. The threeyear<br />
OS rate with Rd in patients age 70 and older who did<br />
not receive ASCT is 70%, and is comparable to results with<br />
MPT and VMP. An ongoing phase III trial is currently<br />
comparing MPT with Rd for 18 months compared with Rd<br />
until progression.<br />
No results from randomized trials are available yet comparing<br />
Rd, MPT, VMP or VCD. Although MPT and VMP<br />
have the strongest data, they are also less well-tolerated<br />
than Rd and VCD, respectively. Based on cost and toxicity<br />
considerations, Rd or VCD are probably the best options for<br />
this patient population at present.<br />
Intermediate risk. Bortezomib-containing regimens can<br />
overcome to some extent the poor prognosis associated with<br />
the t4;14 translocation. 11,17,27 As a result, in patients with<br />
intermediate risk, bortezomib-containing initial therapy<br />
such as VCD for approximately 24 months is preferred.<br />
High risk, plasma cell leukemia, multiple extramedullary<br />
plasmacytomas. Patients ineligible for ASCT with high-risk<br />
disease, plasma cell leukemia or multiple extramedullary<br />
plasmacytomas have poor prognosis. In these patients, depending<br />
on tolerability, it would be reasonable to consider<br />
VCD or VRD as initial therapy. The duration <strong>of</strong> therapy<br />
will be dependent on the tolerability <strong>of</strong> these regimens in a<br />
given patient. If possible, some form <strong>of</strong> continuous therapy is<br />
probably needed.<br />
513
Author’s Disclosures <strong>of</strong> Potential Conflicts <strong>of</strong> Interest<br />
Author<br />
S. Vincent Rajkumar*<br />
*No relevant relationships to disclose.<br />
Employment or<br />
Leadership<br />
Positions<br />
Consultant or<br />
Advisory Role<br />
1. Rajkumar SV. Treatment <strong>of</strong> Multiple Myeloma. Nature Rev Clin Oncol.<br />
2011;8:479-491.<br />
2. Rajkumar SV. Multiple myeloma: <strong>2012</strong> update on diagnosis, riskstratification,<br />
and management. <strong>American</strong> Journal <strong>of</strong> Hematology. <strong>2012</strong>;87:<br />
78-88.<br />
3. Kyle RA, Rajkumar SV. Criteria for diagnosis, staging, risk stratification<br />
and response assessment <strong>of</strong> multiple myeloma. Leukemia. 2009;23:3-9.<br />
4. Russell SJ, Rajkumar SV. Multiple myeloma and the road to personalised<br />
medicine. Lancet Oncol. 2011;12:617-619.<br />
5. Kumar SK, Mikhael JR, Buadi FK, et al. Management <strong>of</strong> newly diagnosed<br />
symptomatic multiple myeloma: updated Mayo stratification <strong>of</strong><br />
Myeloma and Risk-Adapted Therapy (mSMART) consensus guidelines. Mayo<br />
Clinic Proceedings. 2009;84:1095-110.<br />
6. Rajkumar SV, Gahrton G, Bergsagel PL. Approach to the treatment <strong>of</strong><br />
multiple myeloma: a clash <strong>of</strong> philosophies. Blood. 2011;118:3205-3211.<br />
7. Rajkumar SV, Jacobus S, Callander NS, et al. Lenalidomide plus<br />
high-dose dexamethasone versus lenalidomide plus low-dose dexamethasone<br />
as initial therapy for newly diagnosed multiple myeloma: an open-label<br />
randomised controlled trial. Lancet Oncol. 2010;11:29-37.<br />
8. Palumbo A, Rajkumar SV, Dimopoulos MA, et al. Prevention <strong>of</strong><br />
thalidomide- and lenalidomide-associated thrombosis in myeloma. Leukemia.<br />
2008;22:414-423.<br />
9. Harousseau JL, Attal M, Avet-Loiseau H, et al. Bortezomib pPlus<br />
dexamethasone is superior to vincristine plus doxorubicin plus dexamethasone<br />
as induction treatment prior to autologous stem-cell transplantation in<br />
newly diagnosed multiple myeloma: results <strong>of</strong> the IFM 2005-01 phase III<br />
trial. J Clin Oncol. 2010;28:4621-4629.<br />
10. Richardson PG, Weller E, Lonial S, et al. Lenalidomide, bortezomib,<br />
and dexamethasone combination therapy in patients with newly diagnosed<br />
multiple myeloma. Blood. 2010;116:679-686.<br />
11. Cavo M, Tacchetti P, Patriarca F, et al. Bortezomib with thalidomide<br />
plus dexamethasone compared with thalidomide plus dexamethasone as<br />
induction therapy before, and consolidation therapy after, double autologous<br />
stem-cell transplantation in newly diagnosed multiple myeloma: a randomised<br />
phase 3 study. Lancet. 2010;376:2075-2085.<br />
12. Moreau P, Facon T, Attal M, et al. Comparison <strong>of</strong> reduced-dose<br />
bortezomib plus thalidomide plus dexamethasone (vTD) to bortezomib plus<br />
dexamethasone (VD) as induction treatment prior to ASCT in de novo<br />
multiple myeloma (MM): Results <strong>of</strong> IFM2007-02 study. J Clin Oncol. 2010;28<br />
(suppl; abstr 8014).<br />
13. Kumar S, Flinn IW, Richardson PG, et al. Novel Three- and Four-Drug<br />
Combination Regimens <strong>of</strong> Bortezomib, Dexamethasone, Cyclophosphamide,<br />
and Lenalidomide, for Previously Untreated Multiple Myeloma: Results From<br />
the Multi-Center, Randomized, Phase 2 EVOLUTION Study. Abstract presented<br />
at ASH Annual Meeting; December 2010; Orlando, FL.<br />
14. Mateos MV, Oriol A, Martínez-López J, et al. Bortezomib, melphalan,<br />
and prednisone versus bortezomib, thalidomide, and prednisone as induction<br />
therapy followed by maintenance treatment with bortezomib and thalidomide<br />
versus bortezomib and prednisone in elderly treated patients with untreated<br />
multiple myeloma: a randomized trial. Lancet Oncol. 2010;11:934-941.<br />
15. Palumbo A, Bringhen S, Rossi D, et al. Bortezomib-melphalanprednisone-thalidomide<br />
followed by maintenance with bortezomibthalidomide<br />
compared with bortezomib-melphalan-prednisone for initial<br />
treatment <strong>of</strong> multiple myeloma: a randomized controlled trial. J Clin Oncol.<br />
2010;28:5101-5109.<br />
16. Moreau P, Pylypenko H, Grosicki S, et al. Subcutaneous versus intra-<br />
514<br />
Stock<br />
Ownership Honoraria<br />
REFERENCES<br />
Research<br />
Funding<br />
Expert<br />
Testimony<br />
S. VINCENT RAJKUMAR<br />
Other<br />
Remuneration<br />
venous administration <strong>of</strong> bortezomib in patients with relapsed multiple<br />
myeloma: a randomised, phase 3, non-inferiority study. Lancet Oncol. 2011;<br />
12:431-440.<br />
17. Barlogie B, Anaissie E, van Rhee F, et al. Incorporating bortezomib<br />
into upfront treatment for multiple myeloma: early results <strong>of</strong> total therapy 3.<br />
British Journal <strong>of</strong> Haematology. 2007;138:176-185.<br />
18. Facon T, Mary JY, Hulin C, et al. Melphalan and prednisone plus<br />
thalidomide versus melphalan and prednisone alone or reduced-intensity<br />
autologous stem cell transplantation in elderly patients with multiple myeloma<br />
(IFM 99-06): a randomised trial. Lancet. 2007;370:1209-1218.<br />
19. Hulin C, Facon T, Rodon P, et al. Efficacy <strong>of</strong> melphalan and prednisone<br />
plus thalidomide in patients older than 75 years with newly diagnosed<br />
multiple myeloma: IFM 01/01 Trial. J Clin Oncol. 2009;27:3664-3670.<br />
20. Palumbo A, Bringhen S, Caravita T, et al. Oral melphalan and<br />
prednisone chemotherapy plus thalidomide compared with melphalan and<br />
prednisone alone in elderly patients with multiple myeloma: randomised<br />
controlled trial. Lancet. 2006;367:825-831.<br />
21. Waage A, Gimsing P, Fayers P, et al. Melphalan and prednisone plus<br />
thalidomide or placebo in elderly patients with multiple myeloma. Blood.<br />
2010;116:1405-1412.<br />
22. Wijermans P, Schaafsma M, Termorshuizen F, et al. Phase III study<br />
<strong>of</strong> the value <strong>of</strong> thalidomide added to melphalan plus prednisone in elderly<br />
patients with newly diagnosed multiple myeloma: the HOVON 49 Study.<br />
J Clin Oncol. 2010;28:3160-3166.<br />
23. Beksac M, Haznedar R, Firatli-Tuglular T, et al. Addition <strong>of</strong> thalidomide<br />
to oral melphalan/prednisone in patients with multiple myeloma not<br />
eligible for transplantation: results <strong>of</strong> a randomized trial from the Turkish<br />
Myeloma Study Group. Eur J Haematol. 2011;86:16-22.<br />
24. Kapoor P, Rajkumar SV, Dispenzieri A, et al. Melphalan and prednisone<br />
versus melphalan, prednisone and thalidomide for elderly and/or transplant<br />
ineligible patients with multiple myeloma: A meta-analysis. Leukemia.<br />
2011;25:689-696.<br />
25. Fayers PM, Palumbo A, Hulin C, et al. Thalidomide for previously<br />
untreated elderly patients with multiple myeloma: meta-analysis <strong>of</strong> 1685<br />
individual patient data from 6 randomized clinical trials. Blood. 2011;118:<br />
1239-1247.<br />
26. San Miguel JF, Schlag R, Khuageva NK, et al. Bortezomib plus<br />
melphalan and prednisone for initial treatment <strong>of</strong> multiple myeloma. N Engl<br />
J Med. 2008;359:906-917.<br />
27. Nair B, van Rhee F, Shaughnessy JD Jr., et al. Superior results <strong>of</strong> Total<br />
Therapy 3 (2003-33) in gene expression pr<strong>of</strong>iling-defined low-risk multiple<br />
myeloma confirmed in subsequent trial 2006-66 with VRD maintenance.<br />
Blood. 2010;115:4168-4173.<br />
28. van Rhee F, Szymonifka J, Anaissie E, et al. Total therapy 3 for<br />
multiple myeloma: Prognostic implications <strong>of</strong> cumulative dosing and premature<br />
discontinuation <strong>of</strong> VTD maintenance components, bortezomib, thalidomide<br />
and dexamethasone, relevant to all phases <strong>of</strong> therapy. Blood. 2010;116:<br />
1220-1227.<br />
29. Palumbo A, Bringhen S, Liberati AM, et al. Oral melphalan, prednisone,<br />
and thalidomide in elderly patients with multiple myeloma: updated<br />
results <strong>of</strong> a randomized controlled trial. Blood. 2008;112:3107-3114.<br />
30. Mateos MV, Richardson PG, Schlag R, et al. Bortezomib plus melphalan<br />
and prednisone compared with melphalan and prednisone in previously<br />
untreated multiple myeloma: updated follow-up and impact <strong>of</strong><br />
subsequent therapy in the phase III VISTA trial. J Clin Oncol. 2010;28:2259-<br />
2266.
Maintenance Therapy for Myeloma: How<br />
Much, How Long, and at What Cost?<br />
By Michel Attal, MD, and Murielle Roussel, MD<br />
Overview: Maintenance therapy in multiple myeloma has<br />
been under investigation for more than 3 decades and has<br />
been without evidence <strong>of</strong> clear advantage in terms <strong>of</strong><br />
progression-free survival (PFS) until the mid-2000s. Neither<br />
conventional chemotherapy, prednisone, nor interferon-based<br />
maintenance regimens <strong>of</strong>fered any benefit after conventional<br />
or high-dose therapy. Thalidomide was the first drug, mainly<br />
given as maintenance after high dose therapy, to demonstrate<br />
clinical benefits in terms <strong>of</strong> PFS and, in some studies, <strong>of</strong><br />
overall survival (OS). The role <strong>of</strong> other novel agents such as<br />
lenalidomide and bortezomib as maintenance therapy is<br />
emerging. Lenalidomide has been shown to reduce the risk <strong>of</strong><br />
relapse with longer follow-up needed to see if this will<br />
THERAPEUTIC ADVANCES in the treatment <strong>of</strong> multiple<br />
myeloma have improved remission duration and<br />
OS. 1 While agents such as thalidomide, bortezomib, and<br />
lenalidomide have had a major effect on response rates, most<br />
patients with multiple myeloma will ultimately relapse.<br />
Further strategies are needed to both improve response<br />
depth and prolong response duration in order to induce a<br />
clonal extinction and, at last, cure the disease.<br />
Maintenance therapy can be defined as any treatment<br />
given after completion <strong>of</strong> induction treatment, for a prolonged<br />
period <strong>of</strong> time, with the goal <strong>of</strong> extending the duration<br />
<strong>of</strong> response, prolonging PFS and OS, while maintaining<br />
a good quality <strong>of</strong> life. 2,3 In addition, an optimal maintenance<br />
treatment should have an acceptable toxicity pr<strong>of</strong>ile and not<br />
compromise treatment at time <strong>of</strong> relapse. Unfortunately,<br />
this strategy remains elusive in multiple myeloma.<br />
Early attempts at maintenance therapy with conventional<br />
chemotherapy has always failed to prolong PFS and OS 4 and<br />
the role <strong>of</strong> long-term steroids has also been controversial. 5<br />
The effect <strong>of</strong> alpha-interferon was shown to be modest in<br />
terms <strong>of</strong> prolonging PFS and OS. 6,7 Most investigators<br />
concluded that the benefit was small and needed to be<br />
balanced against the cost and potential toxicity <strong>of</strong> prolonged<br />
treatment with alpha-interferon.<br />
The advent <strong>of</strong> “novel” agents (thalidomide, bortezomib,<br />
and lenalidomide) renewed the concept <strong>of</strong> maintenance or<br />
long-term treatment, even for elderly patients or those not<br />
eligible for high dose therapy (HDT). Several phase II and<br />
III trials on maintenance therapy are ongoing.<br />
Maintenance Therapy with Immunomodulatory<br />
Drugs (IMiDs)<br />
The IMiDs thalidomide and lenalidomide are the most<br />
frequently studied maintenance drugs, given their ease <strong>of</strong><br />
oral administration and established antimyeloma efficacy<br />
(Table 1).<br />
Thalidomide and Lenalidomide Prolong Duration <strong>of</strong> Response<br />
after HDT<br />
Thalidomide. There are five randomized studies with<br />
thalidomide maintenance after HDT that have been published<br />
to date; one additional trial was reported during the<br />
<strong>American</strong> <strong>Society</strong> <strong>of</strong> Hematology meeting in 2010. 8-13 De-<br />
translate into a survival benefit. At present, a number <strong>of</strong> key<br />
questions remain unanswered. What are the optimal dose and<br />
duration <strong>of</strong> those treatments? Is the risk <strong>of</strong> toxicity and<br />
second primary malignancies acceptable? Will the disease be<br />
more aggressive at time <strong>of</strong> relapse? Is the clinical benefit<br />
predicted by initial prognostic factors and response to previous<br />
therapy? Does maintenance therapy work by further<br />
eradication <strong>of</strong> minimal residual disease or by immunological<br />
control <strong>of</strong> the malignant clone? Ongoing randomized trials are<br />
evaluating lenalidomide and bortezomib, both in the transplant<br />
and nontransplant settings, to better define the role <strong>of</strong><br />
these drugs as maintenance in multiple myeloma.<br />
spite differences, such as duration <strong>of</strong> treatment and/or<br />
concomitant use <strong>of</strong> prednisone/prednisolone, thalidomide<br />
maintenance was consistently associated with a longer PFS.<br />
The benefit with respect to OS, however, was variable.<br />
Along with the Arkansas group, the Intergroupe Francophone<br />
du Myélome (IFM) was one <strong>of</strong> the first to show that<br />
thalidomide as maintenance after tandem autologous stem<br />
cell transplantation (ASCT) was superior to no maintenance<br />
or pamidronate alone (IFM 9902 trial 9 ). Continuous thalidomide<br />
increased the complete response (CR) and very good<br />
partial response (VGPR) rate (67% vs. 55% and 57%, respectively;<br />
p � 0.03), the 3-year PFS (52% vs. 36% and 37%,<br />
respectively; p � 0.009), and the 4-year OS (87% vs. 77% and<br />
74%, respectively; p � 0.04). The Australian group 10 obtained<br />
similar results when comparing thalidomide (for 12<br />
months) plus prednisolone (until progression) with prednisolone<br />
alone. The thalidomide-containing arm achieved<br />
higher rates <strong>of</strong> VGPR with increased PFS and OS. Within<br />
the total therapy 2 program, 8 thalidomide was given as<br />
maintenance until disease progression or intolerance. In the<br />
initial report, CR rate and 5-year PFS were notably better in<br />
the thalidomide arm (62% vs. 43% and 56% vs. 44%, respectively)<br />
but there was no improvement in OS. After relapse,<br />
survival was shorter in patients pre-exposed to thalidomide.<br />
However, in an updated analysis (with a median follow-up <strong>of</strong><br />
72 months), prolonged OS was eventually confirmed even in<br />
a subgroup <strong>of</strong> patients with poor-risk cytogenetics. 14<br />
Lastly, a recently presented meta-analysis <strong>of</strong> five transplant<br />
studies revealed a significant improvement both in<br />
PFS (p � 0.001) and OS (p � 0.002) with thalidomide<br />
maintenance therapy. 15<br />
Lenalidomide. Given the more favorable efficacy pr<strong>of</strong>ile,<br />
lenalidomide was considered an attractive agent for maintenance.<br />
This has prompted several ongoing trials designed<br />
to compare continuous treatment until relapse with no<br />
maintenance after ASCT. Two large randomized phase III<br />
From the Hematology Department, CHU Purpan, Toulouse, France.<br />
Authors’ disclosures <strong>of</strong> potential conflicts <strong>of</strong> interest are found at the end <strong>of</strong> this article.<br />
Address reprint requests to M. Attal, MD, Hématologie Clinique, CHU Purpan, Place du<br />
Dr Baylac, TSA 40031, 31059 Toulouse Cedex 9, France; email: attal.m@chu-toulouse.fr.<br />
© <strong>2012</strong> by <strong>American</strong> <strong>Society</strong> <strong>of</strong> <strong>Clinical</strong> <strong>Oncology</strong>.<br />
1092-9118/10/1-10<br />
515
trials, one conducted by the IFM 16 and the second by the<br />
Cancer and Leukemia Group B (CALGB), 17 were updated<br />
during the last International Myeloma Workshop in Paris,<br />
France.<br />
The IFM 2005–02 phase III trial 16 included 614 patients<br />
under age 65 who were randomly selected after HD melphalan<br />
200 mg/m 2 followed by ASCT and consolidation with<br />
lenalidomide (25 mg/day, 21 days/month for 2 months) to<br />
receive either maintenance therapy with lenalidomide (10 to<br />
15 mg/day, 28 days/month) or placebo until relapse. The first<br />
interim analysis showed a statistically significant (p � 10 �7 )<br />
benefit favoring maintenance therapy with lenalidomide<br />
and the study was unblinded in July 2010. There was no<br />
planned cross-over and all patients stayed in their allocated<br />
arm until January 2011. After a median follow-up <strong>of</strong> 36<br />
months from randomization, median PFS was 41 months for<br />
patients in the lenalidomide arm versus 24 months in the<br />
placebo arm (p � 0.0001). Thus, 3 years after randomization,<br />
61% <strong>of</strong> patients in the lenalidomide maintenance group did<br />
not relapse versus 34% in the placebo group. The benefit <strong>of</strong><br />
lenalidomide maintenance in terms <strong>of</strong> PFS, the primary<br />
objective <strong>of</strong> this study, was notable, but there was no OS<br />
benefit recorded (4-year OS estimates in the lenalidomide<br />
arm were 79% vs. 73% in the placebo arm). The benefit <strong>of</strong><br />
lenalidomide maintenance was confirmed by the CALGB<br />
100104 study. 17 The protocol was similar to the IFM study<br />
with the exception <strong>of</strong> consolidation component. This trial<br />
was also closed prematurely because <strong>of</strong> a significant advantage<br />
in the lenalidomide arm in terms <strong>of</strong> time to progression<br />
(TTP) calculated from the day 0 <strong>of</strong> intensification. The study<br />
was unblinded in December 2009 and 86 patients in the<br />
placebo arm went on to receive lenalidomide. Based on an<br />
intent-to-treat analysis, after a median follow-up <strong>of</strong> 28<br />
KEY POINTS<br />
● Maintenance therapy with thalidomide can improve<br />
the depth <strong>of</strong> response and reduce the risk <strong>of</strong> progression<br />
or death in both intensive and nonintensive<br />
programs.<br />
● On average, patients tolerate thalidomide for no more<br />
than 1 year; the optimal dose <strong>of</strong> thalidomide should<br />
be the minimal effective dose that is associated with<br />
superior tolerance and less toxicity (<strong>of</strong>ten around 50<br />
mg/day).<br />
● Maintenance therapy with lenalidomide is clearly<br />
effective in prolonging the duration <strong>of</strong> response and<br />
diminishes the risk <strong>of</strong> relapse by 35% after frontline<br />
therapy, irrespective <strong>of</strong> the patient’s age; however,<br />
patients treated with lenalidomide have the ongoing<br />
risk <strong>of</strong> new cancer and no clear OS benefit has yet<br />
been shown.<br />
● Bortezomib maintenance therapy can also increase<br />
response rate and prolong PFS after initial therapy<br />
but questions regarding scheduling, duration <strong>of</strong> therapy,<br />
and combination with other drugs, remain.<br />
● At present, none <strong>of</strong> the drugs evaluated are approved<br />
for maintenance therapy.<br />
516<br />
months, the TTP was 48 months in the lenalidomide arm<br />
versus 31 months in the placebo arm (p � 0.0001).<br />
One can thus conclude that lenalidomide maintenance can<br />
reduce the risk <strong>of</strong> relapse in the transplant setting.<br />
What Group <strong>of</strong> Patients Will Benefit from IMiD Maintenance?<br />
Thalidomide. A number <strong>of</strong> concerns exist regarding the<br />
use <strong>of</strong> thalidomide maintenance after ASCT, which include<br />
the presence <strong>of</strong> adverse cytogenetic abnormalities or the<br />
quality <strong>of</strong> response after HDT.<br />
In the IFM 99–02 trial, 9 only patients who failed to<br />
achieve at least VGPR and who lacked the del 13q cytogenetic<br />
abnormality had a longer PFS after receiving the<br />
thalidomide. Within the intensive pathway <strong>of</strong> the British<br />
MRC IX trial, 13 thalidomide maintenance had no effect on<br />
PFS (9 months vs. 12 months, p � 0.48) in patients with<br />
adverse iFISH [t (4; 14), t (14; 16), del17p, gain (1q21)] and<br />
was associated with worse OS (p � 0.009). Conversely in the<br />
Australian trial, 10 the PFS advantage with thalidomide<br />
maintenance was irrespective <strong>of</strong> response to HDT or cytogenetic<br />
abnormalities.<br />
On balance however, the evidence would support avoiding<br />
thalidomide maintenance in patients with adverse cytogenetics.<br />
Lenalidomide. Interestingly, with lenalidomide maintenance,<br />
subgroup analysis has shown that the benefit <strong>of</strong><br />
maintenance therapy was seen irrespective <strong>of</strong> response to<br />
HDT and/or consolidation and initial prognostic factors. In<br />
the IFM 2005–02 trial, 16 3-year PFS estimates were higher<br />
in all stratified subgroups <strong>of</strong> patients who received lenalidomide<br />
maintenance compared with those who received placebo.<br />
This included patients who had achieved VGPR at time<br />
<strong>of</strong> randomization (64% vs. 49%; p � 0.004) or not (51% vs.<br />
18%; p � 0.0001), and patients with 13q deletion (53% vs.<br />
24%; p � 0.0001) or not (67% vs. 44%; p � 0.0001).<br />
Can We Ultimately Prolong the OS with<br />
IMiD Maintenance?<br />
ATTAL AND ROUSSEL<br />
The shorter postrelapse survival observed in several studies<br />
may be caused by a number <strong>of</strong> factors, such as the<br />
duration <strong>of</strong> maintenance treatment, selection <strong>of</strong> more resistant<br />
clones (especially in high-risk patients or patients who<br />
received IMiDs as part <strong>of</strong> the induction), the age <strong>of</strong> patients<br />
at time <strong>of</strong> relapse, toxicities from previous treatments, and<br />
the availability <strong>of</strong> salvage treatments. The Australian trial 10<br />
showed that maintenance thalidomide for only 1 year did not<br />
adversely affect the outcome after relapse. However, three<br />
additional studies suggested that the long-term use <strong>of</strong> thalidomide<br />
may induce more resistant disease at relapse. 8,11,13<br />
Interestingly, within the meta-analysis described previously,<br />
outcomes did not differ between trials that used<br />
thalidomide during the maintenance phase only and those<br />
that used thalidomide both for induction and maintenance<br />
treatment. 15 For OS, a major effect variability between<br />
trials was noted (test for heterogeneity, p � 0.03) and the<br />
positive result for overall effect must be interpreted with<br />
caution.<br />
Conversely, based on our own experience, it seems that<br />
continuous therapy with lenalidomide does not have a negative<br />
influence on survival after relapse. It will be important<br />
to determine whether patients given lenalidomide maintenance<br />
will have same median OS at first progression and
MAINTENANCE THERAPY FOR MULTIPLE MYELOMA<br />
Trial<br />
Author and Year<br />
Table 1. IMiDs Thalidomide and Lenalidomide Maintenance Studies after ASCT and Conventional Chemotherapy<br />
n<br />
Age Maintenance Dose Comparator<br />
Duration <strong>of</strong><br />
Maintenance PFS/EFS OS Survival after Relapse<br />
Thalidomide Trials<br />
Post ASCT Thal� Thal� Thal� Thal� Thal� Thal�<br />
IFM 95–02 597 Thalidomide Pamidronate or Until progression 52%*<br />
37%* 87%*<br />
75%* 75% 75%<br />
200–400 mg/d Observation<br />
(3-yr from<br />
(4-yr from<br />
(1-yr OS)<br />
Attal et al 2006 � 65 y � Pamidronate<br />
randomization)<br />
enrollment)<br />
Total therapy 2 668 Thalidomide IFN Until progression 56%* 44% Not reached 7 yr 27% 23%<br />
100 mg/d<br />
1st year<br />
then 50 mg<br />
alternate days<br />
(5-yr)<br />
(median, yr)<br />
(5-yr OS)<br />
Barlogie et al<br />
2006–2008<br />
� 75 y � IFN<br />
ALLG MM6 243 Thalidomide<br />
100–200 mg/d<br />
Spencer et al 2009 � 70 y � Prednisolone<br />
NCIC CTG MY.10 332 Thalidomide<br />
200 mg/d<br />
Stewart et al 2010 � 65 y �prednisone<br />
Prednisolone 12 mo<br />
(prednisolone<br />
until progression)<br />
Observation 48 mo or until<br />
progression<br />
42%*<br />
(3-yr PFS)<br />
28 mo*<br />
(median, mo)<br />
23%* 86%*<br />
(3-yr OS)<br />
17 mo NR<br />
(median, yr)<br />
75%* 79%<br />
(1-yr OS)<br />
5 y N/A<br />
HOVON-50 556 Thalidomide IFN Until progression 34 mo* 25 mo* 73 mo 60 mo 20 mo* 31 mo*<br />
Lokhorst et al 2010 � 65 y 50 mg/d (median, mo) (median, mo) (median, mo)<br />
MRC myeloma IX<br />
Morgan et al<br />
<strong>2012</strong><br />
493 Thalidomide Observation Until progression 30 mo* 27 mo* 75% 80% 20 mo* 36 mo*<br />
� 65 y 50–100 mg/d (median, mo) (3-yr OS) (median, mo)<br />
Post Chemotherapy Thal� Thal� Thal� Thal� Thal� Thal�<br />
GIMEMA 331 MPT� Thalidomide<br />
100 mg/d<br />
Palumbo et al<br />
2008<br />
� 65 y<br />
MP �<br />
observation<br />
Until progression 22 mo*<br />
(median, mo)<br />
14.5 mo* 45 mo<br />
(median, mo)<br />
77%<br />
48 mo 11.5 mo* 24 mo*<br />
HOVON 49 333 MPT� MP �<br />
Until progression 34%* 14%* 40 mo* 31 mo* N/A<br />
Wijermans et al � 65 y Thalidomide observation<br />
(2-yr PFS)<br />
(median, mo)<br />
2010<br />
50 mg/d<br />
NMSG 357 MPT � MP � Until progression 15 mo 14 mo 29 mo 32 mo N/A<br />
Waage et al 2010 � 65 y Thalidomide<br />
200 mg/j<br />
placebo (median, mo) (median, mo)<br />
CEMSG 128 Thalidomide<br />
200 mg/d<br />
Ludwig et al 2010 � 65 y � IFN<br />
MRC myeloma IX<br />
Morgan et al<br />
<strong>2012</strong><br />
IFN Until progression 28 mo*<br />
(median, mo)<br />
13 mo* 53 mo<br />
(median, mo)<br />
51 mo 8 mo<br />
(median, mo)<br />
25 mo<br />
327 Thalidomide Observation Until progression 11 mo* 9 mo* 38 mo 39 mo 21 mo 26 mo<br />
� 65 y 50–100 mg/d (median, mo) (median, mo) (median, mo)<br />
Post ASCT<br />
Lenalidomide Trials<br />
Len� Len� Len� Len� Len� Len�<br />
IFM 2005–02 614 Lenalidomide placebo Until progression 41 mo* 24 mo* 79% 73% 12 mo 12 mo<br />
Attal et al 2011 �65y 10–15 mg/d<br />
(median, mo) (5-yr OS) (median, mo)<br />
CALBG 100104 568 Lenalidomide placebo Until progression 43 mo* 31 mo* N/A N/A<br />
McCarthy et al �65y 10–15 mg/d<br />
(median, mo) 23 versus<br />
2011<br />
39 deaths*<br />
Post chemotherapy Len� Len� Len� Len� Len� Len�<br />
MM 015 499 MPR� MPR or MP � Until progression 31 mo* 14/13 mo* 73% 65% N/A<br />
Palumbo et al �65 y Lenalidomide observation (median, mo) (3-yr OS)<br />
2011<br />
10md/d 21–28d<br />
Abbreviations: PFS, progression-free survival; EFS, event-free survival; OS, overall survival; y, years; mo, months; N/A, not available.<br />
* Statistically significant.<br />
517
Trial<br />
Author and Year<br />
Table 2. Bortezomib Maintenance Studies after ASCT and Conventional Chemotherapy<br />
n<br />
age Maintenance Dose Comparator<br />
will respond to increased doses <strong>of</strong> lenalidomide (25 mg) and<br />
the addition <strong>of</strong> dexamethasone at the time <strong>of</strong> myeloma<br />
progression. Longer follow-up is currently needed to assess<br />
that issue.<br />
Can Maintenance Therapy with IMiDs Also Prolong<br />
Duration <strong>of</strong> Response in Elderly Patients or Those<br />
Ineligible for HDT?<br />
Fewer data are available concerning maintenance therapy<br />
with IMiDS in elderly patients and in the nontransplant<br />
setting.<br />
Thalidomide. Three melphalan, prednisone, and thalidomide<br />
(MPT) trials were reported that included thalidomide<br />
maintenance (MPT�T). 18-20 The MPT�T schedule was compared<br />
to melphalan and prednisone (MP) with placebo and<br />
thus it may be difficult to determine the value <strong>of</strong> maintenance<br />
therapy per se as the studies were designed to<br />
evaluate the entire treatment program. Median PFS was<br />
improved in MPT�T arm in 2 <strong>of</strong> the 3 trials but there was<br />
no OS advantage with long-term thalidomide use. The MRC<br />
Myeloma IX maintenance study, was properly designed to<br />
evaluate the influence <strong>of</strong> thalidomide maintenance in the<br />
nonintensive pathway. 13 Results indicated that thalidomide<br />
had only a marginal effect in improving PFS in nonintensively<br />
treated patients (median PFS 11 months vs. 9<br />
months, p � 0.014). Furthermore, it suggests that it may be<br />
related to thalidomide exposure at some point during the<br />
treatment course rather than directly attributable to its use<br />
in maintenance.<br />
Lenalidomide. Prolonged treatment with lenalidomide<br />
has shown a clear benefit in elderly patients, making it a<br />
good choice for long-term maintenance. Several trials are<br />
examining this role with the most mature data emerging<br />
from the International study MM015. 21<br />
In this large, phase III trial, a total <strong>of</strong> 459 patients older<br />
Duration <strong>of</strong><br />
Maintenance PFS/EFS OS<br />
Bortezomib Trials<br />
Post ASCT Bor� Bor� Bor� Bor�<br />
HOVON 65/HD4 800 Bortezomib Thal 50 2 yr 48%* 42%* 78%* 71%*<br />
Sonneveld et al 2010 � 65y 1.3 mg/m2/2 wk (3-yr PFS) (3-yr OS)<br />
PETHEMA/GEM 266 Bortezomib Thalidomide 100 mg/d 3 yr 78%* 63/49%* N/A<br />
Rosinol et al 2011 � 65y 1.3 mg/m2/3 mo Or (2-yr PFS) No difference<br />
Post Chemotherapy<br />
�<br />
Thalidomide 100 mg/d<br />
IFN<br />
GIMEMA 511 VMPT VMP Until<br />
NR* 27 mo* 89% 87%<br />
Palumbo et al 2010 � 65 y �VT<br />
Bortezomib 1.3 mg/m2,<br />
d1,15/4 wk<br />
Thalidomide 50 mg/d<br />
�observation progression<br />
(3-yr OS)<br />
GEM2005MAS65 260 VMP or VTP VMP or VTP 3 yr VT 39 mo* VT NR<br />
Mateos et al 2011 � 65 y �VT �VP VP 32 mo* VP 60 mo<br />
Bortezomib 1.3 mg/m2<br />
d 1, 4, 8, 11/3 mo<br />
Bortezomib 1.3 mg/m2<br />
Thalidomide 50 mg/d d 1, 4, 8, 11/3 mo<br />
Prednisone 50 mg<br />
alternate day<br />
(median, mo) (median, mo)<br />
Abbreviations: PFS, progression-free survival; EFS, event-free survival; OS, overall survival; y, years; mo, months; N/A, not available.<br />
* Statistically significant.<br />
518<br />
ATTAL AND ROUSSEL<br />
than or equal to age 65 with newly diagnosed multiple<br />
myeloma (MM) were enrolled. Induction consisted <strong>of</strong> nine<br />
28-day cycles <strong>of</strong> melphalan 0.18 mg/kg (d1–4), prednisone<br />
2 mg/kg (d1–4), and lenalidomide 10 mg (d1–21) (MPR) or<br />
MP. After induction, patients receiving MPR-R received<br />
lenalidomide 10 mg (d1–21) maintenance until progression;<br />
patients receiving MPR and MP received placebo. The MPR<br />
regimen resulted in significantly higher response rates<br />
(MPR-R: 77%, MPR: 68%, MP: 50%, p � 0.001). Sixty percent<br />
<strong>of</strong> responses were achieved within 3 months following<br />
induction treatment initiation. Further improvements in the<br />
quality <strong>of</strong> response occurred with continued treatment, particularly<br />
during the first year, with a few patients achieving<br />
further tumor reduction thereafter. A landmark analysis<br />
showed that the addition <strong>of</strong> lenalidomide maintenance to<br />
MPR decreased the risk <strong>of</strong> progression by 68% (p � 0.001)<br />
and significantly prolonged the median PFS (MPR-R 31<br />
months vs. MPR 15 months (p � 0.001) and MP 12 months<br />
(p � 0.001), respectively). In addition, a subgroup landmark<br />
analysis showed that the benefit <strong>of</strong> MPR-R over MPR was<br />
maintained regardless <strong>of</strong> Internatinal Staging System (ISS)<br />
stage (ISS I and II vs. III), response (� VGPR vs. PR) and<br />
age (65–75 vs. �75 years old). With a median follow-up <strong>of</strong><br />
41 months, the estimated 4-year OS was similar between the<br />
3 groups (58% to 59%).<br />
Maintenance Therapy with Proteasome Inhibitors<br />
Data concerning maintenance with bortezomib are less<br />
mature. Several randomized studies by European and<br />
<strong>American</strong> study groups are ongoing. The IV formulation<br />
makes bortezomib a less attractive option for long-term<br />
treatment. However, SC administration and the future<br />
availability <strong>of</strong> oral forms <strong>of</strong> proteasome inhibitors will certainly<br />
boost potential maintenance trials with these agents<br />
(Table 2).
MAINTENANCE THERAPY FOR MULTIPLE MYELOMA<br />
Only two phase III trials <strong>of</strong> bortezomib maintenance<br />
post-ASCT presenting PFS and OS data have been reported<br />
to date. 22,23 In the HOVON 65 MM/GMMG-HD4 trial, 22<br />
bortezomib 1.3 mg/m 2 was given every 2 weeks for 2 years<br />
after bortezomib-adriamycin-dexamethasone (PAD) induction<br />
and one or two transplantations. All outcomes were<br />
significantly better in the bortezomib-containing arm (3year<br />
PFS: 48% vs. 42%, p � 0.047 and OS: 78% vs. 71%, p �<br />
0.048) compared with the control arm using vincristineadriamycin-dexamethasone<br />
(VAD) induction, ASCT, and<br />
low-dose thalidomide maintenance (50 mg daily). Patients<br />
with high-risk MM derived a particular benefit from bortezomib:<br />
the 3-year PFS for patients with t(4;14) was 32%<br />
compared with 22% in the controls. Unfortunately, the<br />
design <strong>of</strong> the study does not allow a clear dissection <strong>of</strong> the<br />
role <strong>of</strong> bortezomib maintenance therapy.<br />
The Spanish group PETHEMA/GEM 23 recently presented<br />
updated results <strong>of</strong> their large phase III randomized trial<br />
comparing induction with thalidomide/dexamethasone (TD)<br />
versus bortezomib/thalidomide/dexamethasone (VTD) versus<br />
alternate polychemotherapy regimen with vincristine/<br />
melphalan, cyclophosphamide/prednisone, vincristine/<br />
carmustine/doxorubicin/dexamethasone, and bortezomib<br />
(VBMCP/VBAD/B) in patients less than age 65 with newly<br />
diagnosed symptomatic MM. The maintenance program<br />
after HDT consisted <strong>of</strong> bortezomib and thalidomide (VT; 1<br />
cycle <strong>of</strong> bortezomib-1.3 mg/m 2 on days 1, 4, 8, and 11 every<br />
3 months plus thalidomide 100 mg daily) compared with<br />
thalidomide (T; 100 mg daily) compared with alfa2b-IFN<br />
(subcutaneous, 3 MU � 3 per week). The planned maintenance<br />
duration was 3 years or until disease progression or<br />
toxicity.<br />
The CR rate with maintenance was improved by 23% with<br />
VT, 11% with T and 19% with alfa2-IFN (p � not significant).<br />
After a median follow-up <strong>of</strong> 24 months, the PFS was<br />
significantly longer with VT compared with T and alfa2-IFN<br />
(PFS at 2 years: 78% vs. 63% vs. 49%, respectively, p � 0.01).<br />
However, OS was not significantly different among the 3<br />
arms.<br />
For elderly or nontransplant eligible patients, three large<br />
phase III trials studied the role <strong>of</strong> bortezomib maintenance<br />
alone or in combination with either T or prednisone. 24<br />
UPFRONT 24,25 is a randomized, open-label, multicenter,<br />
phase IIIb study conducted in the U.S. community practice<br />
setting, designed to evaluate the use <strong>of</strong> single-agent bortezomib<br />
as a maintenance therapy following bortezomib/<br />
dexamethasone (VD), VTD, or bortezomib/melphalan/<br />
prednisone (VMP) induction in newly diagnosed MM<br />
patients older than age 65 or ineligible for HDT. Bortezomib<br />
1.6 mg/m 2 was given once a week for five 35-day cycles. After<br />
a median follow-up <strong>of</strong> 26 months, the 1-year PFS estimates<br />
for VD, VTD, and VMP were 57.4%, 63.8%, and 67.3%,<br />
respectively. Additonally, the 2-year OS estimates for VD,<br />
VTD, and VMP were 73.7%, 73.6%, and 77.6%, respectively.<br />
In the GEM2005MAS65, the Spanish myeloma group<br />
investigated, a 3-year maintenance program with bortezomib/thalidomide<br />
(VT) or bortezomib/prednisone (VP)<br />
(bortezomib: 1,3 mg/m 2 /d 1,4,8,11/3 months; thalidomide:<br />
50 mg/d; prednisone: 50 mg alternating days). Updated data<br />
were recently presented. 25 This maintenance regimen increased<br />
the CR from 24% to 42%. After a median follow-up<br />
<strong>of</strong> 46 months, there was no difference between VT and VP<br />
maintenance with respect to PFS (39 months vs. 32 months)<br />
or OS (not reached vs. 60 months). Neither maintenance<br />
regimens overcame the poor prognosis <strong>of</strong> high-risk cytogenetics.<br />
Maintenance with VT was also tested in an Italian study<br />
comparing VMPT/VT with VMP. Median PFS was not<br />
reached in the VT continuous therapy versus 27 months in<br />
the no maintenance arm. 26<br />
Does Maintenance Therapy Enhance the Depth <strong>of</strong><br />
Response and/or Control a Subclinical Relapse?<br />
All published trials with thalidomide and bortezomib<br />
maintenance have shown that response rates can be increased<br />
but it is unclear whether this is the result <strong>of</strong> a late<br />
consolidation effect within the first months <strong>of</strong> maintenance<br />
(as proposed in the thalidomide trials) and/or the result <strong>of</strong> a<br />
gradual clearance <strong>of</strong> residual disease over time. Further<br />
trials examining sequential minimal residual disease evaluation<br />
are needed to address this issue.<br />
With lenalidomide maintenance, less data concerning response<br />
rates are available. However, it appears that lenalidomide<br />
may also further enhance the depth <strong>of</strong> response<br />
and not only control the residual tumor cells through active<br />
immune surveillance. Within the IFM 2005–02 trial, the<br />
best response during maintenance therapy was slightly, but<br />
not significantly, higher (CR: 25% vs. 22%, p � 0.4; VGPR:<br />
77% vs. 70%, p � 0.08). 16 Nevertheless, the IFM recently<br />
reported the results <strong>of</strong> a phase II study suggesting that<br />
lenalidomide maintenance can increase response rate in<br />
nearly 30% <strong>of</strong> patients, even up to stringent CR based on<br />
flow assessment. 27<br />
Can Those Agents Be Administered Safely for a Long<br />
Period <strong>of</strong> Time? The Benefit/Risk Ratio<br />
There is a urgent need for future studies aimed at establishing<br />
the appropriate dose and optimal duration <strong>of</strong> maintenance<br />
therapy in both intensive and nonintensive<br />
programs (Table 3).<br />
Despite the longer PFS, the major caveat that precludes<br />
widespread use <strong>of</strong> thalidomide maintenance is the toxicity<br />
related to long-term administration. Peripheral neuropathy,<br />
sedation, and constipation are common and <strong>of</strong>ten lead to a<br />
reduction in quality-<strong>of</strong>-life parameters and premature discontinuation<br />
<strong>of</strong> therapy even when low doses are used. In<br />
the French trial, patients received 200 mg/day <strong>of</strong> thalidomide,<br />
for a median <strong>of</strong> 15 months (range: 0.1–50 months). In<br />
the other trials, the median duration <strong>of</strong> treatment varied<br />
between 9 months and 2 years. The optimal dose <strong>of</strong> thalidomide<br />
should be the minimal effective dose that is associated<br />
with superior tolerance and least toxicity; 50 mg/day for less<br />
than 1 year appears to be the appropriate dose and duration.<br />
Subsequently, lenalidomide became a logical and attractive<br />
alternative to thalidomide because <strong>of</strong> its lack <strong>of</strong> neurologic<br />
toxicity and a favorable prerequisite for long-term use.<br />
Within the phase III reported trials, toxicity was acceptable<br />
and only 5% <strong>of</strong> patients in the MM 015 trial, 21 12% in the<br />
CALBG trial, 17 and 21% <strong>of</strong> patients in the IFM trial 16<br />
discontinued lenalidomide maintenance before myeloma<br />
progression. Neutropenia was the most common side effect<br />
noted, but febrile neutropenia was rare (2%–6%). Thromboembolic<br />
disease has been reported and long-term use <strong>of</strong><br />
prophylactic antiplatelets agent or low molecular weight<br />
heparin should be considered.<br />
519
Trial<br />
Author and Year<br />
An unexpected finding from the three lenalidomide maintenance<br />
trials was an increase in the incidence <strong>of</strong> secondary<br />
cancers (SPMs), including hematological malignancies and<br />
solid tumors; nearly 8% in both post-ASCT studies compared<br />
with 2% in the placebo groups. The difference compared with<br />
Table 3. Toxical Patterns <strong>of</strong> Maintenance Trials<br />
Adverse Events* (%)<br />
*All Grade Unless Specified<br />
Discontinuation <strong>of</strong><br />
Treatment due to Drug-Related<br />
Adverse Events (%)<br />
Median Duration <strong>of</strong><br />
Treatment and Median<br />
Received Dose<br />
Thalidomide Maintenance<br />
Post ASCT<br />
IFM 95–02 Neuropathy (68%), fatigue (34%), constipation (20%), neutropenia<br />
(7%)<br />
39% 15 mo (0.1–50)<br />
Attal et al 2006 NB: grade 3/4 neuropathy (7%), infection (6%) (mainly related to neuropathy) 200 mg/d<br />
Total therapy 2 �grade2 neuropathy (15%), thrombosis (6%) N/A 80% stopped thalidomide<br />
Barlogie et al 2006–2008<br />
within 2 yr<br />
ALLG MM6 Neuropathy (52%), infection 23%, fatigue (14%), constipation (18%) 30% 58% still on therapy at 12 mo<br />
Spencer et al 2009 NB: grade 3/4 neuropathy (10%) (mainly related to neuropathy) 100 mg/d<br />
HOVON-50 Neuropathy (64%) 33% 2 yr<br />
Lokhorst et al 2010 (mainly related to neuropathy)<br />
NCIC CTG MY.10<br />
Stewart et al 2010<br />
Post ASCT or Chemotherapy<br />
Grade 3/4 neuropathy (10%), thrombosis (7%), fatigue (7%) 16 mo<br />
MRC myeloma IX N/A 52% 7 mo (0–50)<br />
Morgan et al <strong>2012</strong><br />
Post Chemotherapy<br />
(mainly related to neuropathy) 50 mg/d<br />
GIMEMA 55% AE grade 3/4 N/A 8 mo (0.03–39.40)<br />
Palumbo et al 2008 within the entire program<br />
HOVON 49<br />
Wijermans et al 2010<br />
�grade 2 neuropathy (54%) N/A 8.4 mo (1.4–35.9)<br />
NMSG Grade 3–4 non-hematological AE (40%, neuropathy (6%), thrombosis 56% at 1 yr (mainly related to 7,5 mo<br />
Waage et al 2010 (8%)<br />
neuropathy)<br />
CEMSG Leucopenia (59%), fatigue (78%), neurological (69%), infection (28%),<br />
23% 13 mo<br />
Ludwig et al 2010<br />
Post ASCT<br />
constipation (44%), skin (33%)<br />
Lenalidomide Maintenance<br />
75 mg/d<br />
IFM 2005–02 Neutropenia (68%), febrile neutropenia (2%), thrombocytopenia<br />
(24%), diarrhea (40%), fatigue (47%), upper respiratory infection<br />
(70%), neuropathy (23%), skin rash (20%), cramps (39%)<br />
25% at time <strong>of</strong> unblinding N/A<br />
Attal et al 2011 SPMs n � 26<br />
CALBG 100104 Grade 3/4 Neutropenia (43%), febrile neutropenia (6%),<br />
thrombocytopenia (13%), diarrhea (4%), fatigue (5%), infection<br />
(33%), skin rash (4%)<br />
12% N/A<br />
McCarthy et al 2011<br />
Post Chemotherapy<br />
SPMs n � 15<br />
MM 015 Grade 4 hematologic AEs: thrombocytopenia (5%), neutropenia (4%),<br />
5% N/A<br />
Palumbo et al 2011<br />
Post ASCT<br />
anemia (3%); Grade 3/4 nonhematologic AEs: bone pain (5%),<br />
diarrhea (5%) SPMs 8%<br />
Bortezomib Maintenance<br />
HOVON 65/HD4 Grade 2/4 neuropathy (23%), GI symptoms (23%), infections (53%) 9% 49% during 2 yr<br />
Sonneveld et al 2010 (mainly related to neuropathy)<br />
PETHEMA/GEM Grade 1/3 neuropathy (12.2%) 15.6% N/A<br />
Rosinol et al 2011<br />
Post Chemotherapy<br />
(mainly related to neuropathy)<br />
GIMEMA Grade 3/4 neutropenia (38%), thrombocytopenia (22%), N/A N/A<br />
Palumbo et al 2010 Infections (13%), neuropathy (8%), fatigue (6%), DVT (5%, Rash (4%)<br />
GEM2005MAS65 VT: grade 3/4 neuropathy (9%), GI symptoms (4%), neutropenia (4%) VT 13% 20 mo (1–36)<br />
Mateos et al 2011 VP 9% (mainly related to neuropathy)<br />
Abbreviation: N/A, not available.<br />
520<br />
ATTAL AND ROUSSEL<br />
placebo was statistically significant (p � 0.001), mainly after<br />
24 months <strong>of</strong> treatment.<br />
Intense investigation regarding the risk <strong>of</strong> SPMs in recipients<br />
<strong>of</strong> lenalidomide has ensued. In addition to a suspected<br />
intrinsic risk <strong>of</strong> second cancers in this disease, antimyeloma
MAINTENANCE THERAPY FOR MULTIPLE MYELOMA<br />
drugs other than lenalidomide may predispose to SPMs. In<br />
the IFM 2005–02 trial, the cumulative incidence <strong>of</strong> SPMs<br />
was 3.1 per 100 patient-years. Risk factors identified in<br />
multivariate analysis were treatment with lenalidomide,<br />
being over age 55, male gender, ISS III, and induction DCEP<br />
(IFM 2005–01 protocol). In the MM 015 study, the cumulative<br />
incidence was 2.5 per 100 patient-years.<br />
Palumbo A and colleagues 28 recently presented the pooled<br />
data <strong>of</strong> 1,798 patients from nine European protocols. The<br />
risk/benefit <strong>of</strong> lenalidomide maintenance therapy was<br />
largely in favor <strong>of</strong> the IMiDs, either regarding the risk <strong>of</strong><br />
progression or the risk <strong>of</strong> death as a result <strong>of</strong> myeloma.<br />
Therefore, the strong efficacy <strong>of</strong> continuous therapy with<br />
lenalidomide outweighs the potential risk <strong>of</strong> SPMs. Longer<br />
follow-up is needed to definitively assess the risk <strong>of</strong> SPMs<br />
in patients receiving lenalidomide, especially in context <strong>of</strong><br />
alkylating agents.<br />
With respect to bortezomib maintenance, only sporadic<br />
SPMs were reported with an incidence rate <strong>of</strong> 0.3 per 100<br />
person-years. 29 This advantage proved partially abrogated<br />
by its specific toxicity pr<strong>of</strong>ile, in particular the neurotoxicity.<br />
After ASCT, in the HOVON joint trial, 22 57% <strong>of</strong> patients<br />
started maintenance, <strong>of</strong> which 27% required dose reductions<br />
and 9% discontinued bortezomib because <strong>of</strong> toxicity, mainly<br />
peripheral neuropathy. Less than half <strong>of</strong> the patients could<br />
receive the planned 2 years <strong>of</strong> maintenance.<br />
In elderly patients, similar limitations were observed. In<br />
the UPFRONT study, 24 50% <strong>of</strong> patients effectively received<br />
the maintenance with 13% to 25% requiring dose reductions.<br />
Surprisingly, maintenance with single-agent bortezomib was<br />
well tolerated, with limited additional toxicity compared<br />
with induction. In the Spanish trial GEM2005MAS65, 25 up<br />
to 10% <strong>of</strong> patients experienced grade 3/4 peripheral neuropathy<br />
and nearly 60% <strong>of</strong> patients discontinued bortezomib<br />
Authors’ Disclosures <strong>of</strong> Potential Conflicts <strong>of</strong> Interest<br />
Author<br />
Employment or<br />
Leadership<br />
Positions<br />
Consultant or<br />
Advisory Role<br />
because <strong>of</strong> adverse events. Of note, 5% <strong>of</strong> deaths were<br />
related to drug toxicity. The median duration <strong>of</strong> bortezomib<br />
maintenance was 20 months (range: 1–36 months).<br />
Overall, the neurotoxicity related to bortezomib will likely<br />
limits its use as a maintenance agent and dose adjustments<br />
as well as a limited administration are therefore suggested.<br />
Conclusion<br />
Currently, maintenance therapy with novel agents is not<br />
approved in most countries. Thalidomide and bortezomib<br />
effectively increase response rates and prolong PFS, but that<br />
benefit is hampered by their neurologic toxicity likely making<br />
them more useful as consolidation agents rather than<br />
long-term maintenance.<br />
Maintenance therapy with lenalidomide is clearly effective<br />
in prolonging the duration <strong>of</strong> response and diminishing<br />
the risk <strong>of</strong> relapse in frontline multiple myeloma patients<br />
irrespective <strong>of</strong> their age. However, patients treated with<br />
lenalidomide have the ongoing risk <strong>of</strong> new cancer and a clear<br />
OS benefit has yet to be shown. In the case <strong>of</strong> elderly<br />
patients, where one hopes to maximize the response to<br />
upfront treatment, the current robust data supports the use<br />
<strong>of</strong> lenalidomide maintenance in these patients, this balancing<br />
the potential risk <strong>of</strong> SPMs.<br />
In younger patients, whether lenalidomide maintenance<br />
therapy should be routinely <strong>of</strong>fered to patients is controversial.<br />
The increased incidence <strong>of</strong> SPMs is an important risk<br />
and we await for more mature survival data before making<br />
firm recommendations in this regard.<br />
Acknowledgment<br />
We are indebted to Christopher P. Venner, MD, for his critical<br />
reading <strong>of</strong> the manuscript.<br />
Stock<br />
Ownership Honoraria<br />
Michel Attal Celgene;<br />
Janssen-Cilag<br />
Murielle Roussel Celgene;<br />
Janssen<br />
1. Kumar SK, Rajkumar SV, Dispenzieri A, et al. Improved survival in<br />
multiple myeloma and the impact <strong>of</strong> novel therapies. Blood. 2008;111:2516-<br />
2520.<br />
2. Cavo M, Rajkumar SV, Palumbo A, et al. International Myeloma<br />
Working Group consensus approach to the treatment <strong>of</strong> multiple myeloma<br />
patients who are candidates for autologous stem cell transplantation. Blood.<br />
2011;117:6063-6073.<br />
3. Mihelic R, Kaufman JL, Lonial S. Maintenance therapy in multiple<br />
myeloma. Leukemia. 2007;21:1150-1157.<br />
4. Belch A, Shelley W, Bergsagel D, et al. A randomized trial <strong>of</strong> maintenance<br />
versus no maintenance melphalan and prednisone in responding<br />
multiple myeloma patients. Br J Cancer. 1988;57:94-99.<br />
5. Berenson JR, Crowley JJ, Grogan TM, et al. Maintenance therapy with<br />
alternate-day prednisone improves survival in multiple myeloma patients.<br />
Blood. 2002;99:3163-168.<br />
6. Fritz E, Ludwig H. Interferon-alpha treatment in multiple myeloma:<br />
meta-analysis <strong>of</strong> 30 randomised trials among 3948 patients. Ann Oncol.<br />
2000;11:1427-1436.<br />
7. Group MTC. Interferon as therapy for multiple myeloma: an individual<br />
patient data overview <strong>of</strong> 24 randomized trials and 4012 patients. Br J<br />
Haematol. 2001;113:1020-1034.<br />
REFERENCES<br />
Research<br />
Funding<br />
Celgene;<br />
Janssen-Cilag<br />
Expert<br />
Testimony<br />
Other<br />
Remuneration<br />
8. Barlogie B, Tricot G, Anaissie E, et al. Thalidomide and hematopoieticcell<br />
transplantation for multiple myeloma. N Engl J Med. 2006;354:1021-<br />
1030.<br />
9. Attal M, Harousseau JL, Leyvraz S, et al. Maintenance therapy with<br />
thalidomide improves survival in patients with multiple myeloma. Blood.<br />
2006;108:3289-3294.<br />
10. Spencer A, Prince HM, Roberts AW, et al. Consolidation therapy with<br />
low-dose thalidomide and prednisolone prolongs the survival <strong>of</strong> multiple<br />
myeloma patients undergoing a single autologous stem-cell transplantation<br />
procedure. J Clin Oncol. 2009;27:1788-1793.<br />
11. Lokhorst HM, van der Holt B, Zweegman S, et al. A randomized phase<br />
3 study on the effect <strong>of</strong> thalidomide combined with adriamycin, dexamethasone,<br />
and high-dose melphalan, followed by thalidomide maintenance in<br />
patients with multiple myeloma. Blood. 2010;115:1113-1120.<br />
12. Stewart AK, Trudel S, Bahlis NJ, et al. A randomized phase III trial <strong>of</strong><br />
thalidomide and prednisone as maintenance therapy following autologous<br />
stem cell transplantation (ASCT) in patients with multiple myeloma (MM):<br />
The NCIC CTG MY. 10 Trial. ASH Annual Meeting Abstracts. 2010;116:39.<br />
13. Morgan GJ, Gregory WM, Davies FE, et al. The role <strong>of</strong> maintenance<br />
thalidomide therapy in multiple myeloma: MRC Myeloma IX results and<br />
meta-analysis. Blood. <strong>2012</strong>;119:7-15.<br />
521
14. Zangari M, van Rhee F, Anaissie E, et al. Eight-year median survival in<br />
multiple myeloma after total therapy 2: roles <strong>of</strong> thalidomide and consolidation<br />
chemotherapy in the context <strong>of</strong> total therapy 1. Br J Haematol. 2008;141:433-<br />
444.<br />
15. Hahn-Ast C, Lilienfeld-Toal M, Heteren P, et al. Improved progressionfree<br />
and overall survival with thalidomide maintenance therapy after autologous<br />
stem cell transplantation in multiple myeloma: a metaanalysis <strong>of</strong> five<br />
randomized trials. Haematologica. 2011;96:1-678, 884.<br />
16. Attal M, Olivier P, Cances Lauwers V, et al. Maintenance treatment<br />
with lenalidomide after transplantation for myeloma: analysis <strong>of</strong> secondary<br />
malignancies within the IFM 2005-02 trial and updated data. Haematologica.<br />
2011;96:S23.<br />
17. McCarthy PL, Owzar K, Anderson KC, et al. Phase III intergroup study<br />
<strong>of</strong> lenalidomide versus placebo maintenance therapy following single autologous<br />
stem cell transplant (ASCT) for multiple myeloma (MM): CALGB ECOG<br />
BMT-CTN 100104. Haematologica. 2011;96:S23.<br />
18. Palumbo A, Bringhen S, Liberati AM, et al. Oral melphalan, prednisone,<br />
and thalidomide in elderly patients with multiple myeloma: updated<br />
results <strong>of</strong> a randomized controlled trial. Blood. 2008;112:3107-114.<br />
19. Wijermans P, Schaafsma M, Termorshuizen F, et al. Phase III study<br />
<strong>of</strong> the value <strong>of</strong> thalidomide added to melphalan plus prednisone in elderly<br />
patients with newly diagnosed multiple myeloma: the HOVON 49 Study.<br />
J Clin Oncol. 2010;28:3160-3166.<br />
20. Waage A, Gimsing P, Fayers P, et al. Melphalan and prednisone plus<br />
thalidomide or placebo in elderly patients with multiple myeloma. Blood.<br />
2010;116:1405-1412.<br />
21. Palumbo A, Adam Z, Kropff M, et al. A phase 3 study evaluating the<br />
efficacy and safety <strong>of</strong> lenalidomide (Len) combined with melphalan and<br />
prednisone followed by continuous lenalidomide maintenance (MPR-R) in<br />
patients (Pts) � 65 years (yrs) with newly diagnosed multiple myeloma<br />
(NDMM): updated results for pts aged 65-75 yrs enrolled in MM-015. ASH<br />
Annual Meeting Abstracts. 2011;118:475.<br />
22. Sonneveld P, Schmidt-Wolf I, van der Holt B, et al. HOVON-65/GMMG-<br />
HD4 randomized phase III trial comparing bortezomib, doxorubicin, dexa-<br />
522<br />
ATTAL AND ROUSSEL<br />
methasone (PAD) Vs VAD followed by high-dose melphalan (HDM) and<br />
maintenance with bortezomib or thalidomide in patients with newly diagnosed<br />
multiple myeloma (MM). ASH Annual Meeting Abstracts. 2010;116:40.<br />
23. Rosinol L, Cibeira MT, Mateos MV, et al. A phase III PETHEMA/GEM<br />
randomized trial <strong>of</strong> postransplant (ASCT) maintenance in multiple myeloma:<br />
superiority <strong>of</strong> bortezomib/thalidomide compared with thalidomide and alfa-2b<br />
interferon. ASH Annual Meeting Abstracts. 2011;118:3962.<br />
24. Niesvizky R, Flinn IW, Rifkin R, et al. Efficacy and safety <strong>of</strong> three<br />
bortezomib-based combinations in elderly, newly diagnosed multiple myeloma<br />
patients: results from all randomized patients in the community-based,<br />
phase 3b UPFRONT study. ASH Annual Meeting Abstracts. 2011;118:478.<br />
25. Mateos M-V, Oriol A, Teruel A-I, et al: Maintenance Therapy with<br />
Bortezomib Plus Thalidomide (VT) or Bortezomib Plus Prednisone (VP) In<br />
Elderly Myeloma Patients Included In the GEM2005MAS65 Spanish Randomized<br />
Trial. ASH Annual Meeting Abstracts. 2011;118:477.<br />
26. Palumbo A, Bringhen S, Rossi D, et al. Bortezomib-melphalanprednisone-thalidomide<br />
followed by maintenance with bortezomibthalidomide<br />
compared with bortezomib-melphalan-prednisone for initial<br />
treatment <strong>of</strong> multiple myeloma: a randomized controlled trial. J Clin Oncol.<br />
2010;28:5101-5109.<br />
27. Roussel M, Robillard N, Moreau P, et al: Bortezomib, lenalidomide, and<br />
dexamethasone (VRD) consolidation and lenalidomide maintenance in frontline<br />
multiple myeloma patients: updated results <strong>of</strong> the IFM 2008 phase II<br />
VRD intensive program. ASH Annual Meeting Abstracts. 2011;118:1872.<br />
28. Palumbo A, Larocca A, Zweegman S, et al. Second primary malignancies<br />
in newly diagnosed multiple myeloma patients treated with lenalidomide:<br />
analysis <strong>of</strong> pooled data in 2459 patients. ASH Annual Meeting Abstracts.<br />
2011;118:996.<br />
29. San Miguel JF, Richardson PG, Orlowski RZ, et al. Risk <strong>of</strong> second<br />
primary malignancies (SPMs) following bortezomib (Btz)-based therapy:<br />
analysis <strong>of</strong> four phase 3 randomized controlled trials in previously untreated<br />
or relapsed multiple myeloma (MM). ASH Annual Meeting Abstracts. 2011;<br />
118:2933.
NEW OPTIONS, NEW QUESTIONS: HOW TO<br />
SELECT AND SEQUENCE THERAPIES FOR<br />
METASTATIC MELANOMA<br />
CHAIR<br />
Michael B. Atkins, MD<br />
Beth Israel Deaconess Medical Center<br />
Boston, MA<br />
SPEAKERS<br />
Keith T. Flaherty, MD<br />
Massachusetts General Hospital<br />
Boston, MA<br />
Jeff A. Sosman, MD<br />
Vanderbilt Medical Center<br />
Nashville, TN
New Options and New Questions: How to<br />
Select and Sequence Therapies for Patients<br />
with Metastatic Melanoma<br />
By Keith T. Flaherty, MD, Jeff A. Sosman, MD, and Michael B. Atkins, MD<br />
Overview: Recent advances in the understanding <strong>of</strong> the melanoma<br />
biology and tumor immunology have yielded new treatment<br />
strategies for patients with advanced melanoma. Within<br />
the past year, the selective BRAF inhibitor vemurafenib and<br />
immune checkpoint inhibitor ipilimumab have been added to<br />
the treatment armamentarium. In addition, other molecularly<br />
targeted agents and immunotherapies are showing considerable<br />
promise. The availability <strong>of</strong> multiple, effective treatment<br />
MELANOMA IS currently the fifth and seventh most<br />
common cancer in <strong>American</strong> men and women, respectively.<br />
1 Although early-stage patients can be treated<br />
successfully with surgical resection, the prognosis for metastatic<br />
melanoma is dismal, with an overall 5-year mortality<br />
rate <strong>of</strong> 90%. An estimated 8000 <strong>American</strong>s will die <strong>of</strong><br />
melanoma in <strong>2012</strong>. Many <strong>of</strong> these patients are young (median<br />
around 56 years) and otherwise healthy; therefore,<br />
their loss represents a societal burden disproportionate to<br />
other cancers. Historically, common treatment approaches<br />
have included cytotoxic chemotherapy, interleukin-2-based<br />
immunotherapy, and combinations <strong>of</strong> chemotherapy and<br />
immunotherapy—so called biochemotherapy. Although some<br />
treatment approaches, most notably, high-dose interleukin<br />
(IL)-2, have produced extremely durable tumor responses in<br />
a small percentage <strong>of</strong> patients; unfortunately, no treatment<br />
approach was able to reproducibly show a survival advantage<br />
in phase III trials. 2 Hence, there has been an urgent<br />
need for new treatment approaches for patients with advanced<br />
melanoma.<br />
Molecular Biology <strong>of</strong> Melanoma<br />
Elucidation <strong>of</strong> the somatic genetic alterations responsible<br />
for the formation <strong>of</strong> melanoma has provided a matrix for<br />
developing molecularly targeted therapies. Therapies aimed<br />
at the mutated gene products themselves have been developed.<br />
In cases where oncogenes <strong>of</strong> interest cannot be feasibly<br />
targeted directly or when pathways are activated by loss <strong>of</strong><br />
tumor suppressor–gene function, critical nodes <strong>of</strong> signaling<br />
are being sought for development <strong>of</strong> indirect targeting strategies.<br />
Three well-defined driver oncogenes have been identified<br />
in melanoma: BRAF, NRAS, and CKIT. Each is capable <strong>of</strong><br />
activating the MAP kinase pathway, whereas NRAS and<br />
CKIT activate the PI3 kinase pathways among others.<br />
BRAF is the most commonly activated oncogene in melanoma,<br />
with approximately 50% <strong>of</strong> advanced melanomas<br />
harboring such mutations. 3 Eighty to 90% <strong>of</strong> the BRAF<br />
mutations found in melanoma result in a substitution <strong>of</strong><br />
glutamate for valine at the 600 position <strong>of</strong> the kinase domain<br />
within the BRAF amino–acid sequence. The second most<br />
common mutation results in a lysine substitution in the<br />
same position. The two substitutions account for 95% <strong>of</strong> all<br />
BRAF mutations in melanoma, and both create a kinase<br />
that is constitutively active, phosphorylating MEKindependent<br />
<strong>of</strong> upstream activation by receptor tyrosine<br />
kinases or RAS. Being a serine-threonine kinase that con-<br />
524<br />
options for patients with melanoma, although long sought, has<br />
complicated treatment decisions. This article will review the<br />
advances in our understanding <strong>of</strong> melanoma biology and<br />
tumor immunology, the current status <strong>of</strong> immunotherapy, the<br />
advances in molecularly targeted therapy for patients with<br />
BRAF mutant melanomas, the possible approaches to patients<br />
with BRAF wild-type (WT) tumors, and the current considerations<br />
for treatment selection <strong>of</strong> individual patients.<br />
sumes ATP as an energy source, BRAF is amenable to<br />
targeting with ATP competitive, small molecular inhibitors.<br />
4 The two most clinically advanced inhibitors, vemurafenib<br />
and dabrafenib, will be discussed in detail below.<br />
NRAS, which activates RAF kinases, as well several other<br />
downstream pathways, is activated by mutations that<br />
disable the GTPase activity within the molecule. 5 Such<br />
mutations are found in 20% <strong>of</strong> advanced melanomas. 6 Pharmacologic<br />
agents that can restore the function <strong>of</strong> the GTPase<br />
have been technically difficult to generate. Thus, direct<br />
targeting strategies for NRAS remain elusive. Indirect therapeutic<br />
strategies will be discussed in detail later in this<br />
manuscript. Lastly, CKIT mutations, many <strong>of</strong> which overlap<br />
with the ones found in gastrointestional tumors, are found<br />
in approximately 1% <strong>of</strong> all melanomas. They are found<br />
nearly exclusively in melanomas that arise on the nail beds,<br />
palms, soles, and mucosal surfaces. In the advanced melanoma<br />
population, melanomas arising from these sites account<br />
for only 5% to 10% <strong>of</strong> all cases. 7 Within these subsets,<br />
CKIT mutations are found in approximately 10%. Mutations<br />
in exons 11 and 13 are common, and many have previously<br />
been described as responses to imatinib and other small<br />
molecule KIT inhibitors. With roughly 70% <strong>of</strong> all melanomas<br />
harboring BRAF, NRAS, orCKIT mutations, the remaining<br />
tumors have yet to be associated with a described driver<br />
oncogene, and thus lack a primary target for therapy.<br />
Important secondary or complementary mutations have<br />
been identified in two additional pathways, PI3 kinase and<br />
p16/Rb. Both <strong>of</strong> these pathways have been implicated in<br />
melanoma formation, and preliminary evidence suggests<br />
that they serve as important modifiers <strong>of</strong> resistance to<br />
single-agent targeted therapies aimed at driver oncogenes,<br />
although direct targeting <strong>of</strong> these pathways in preclinical<br />
models is not associated with significant antitumor effects.<br />
Activating mutations <strong>of</strong> PI3 kinase itself have not been<br />
described in melanoma, and activating AKT mutations have<br />
been observed only rarely. 8 Amplification <strong>of</strong> one AKT is<strong>of</strong>orm,<br />
AKT3, has been identified in a subset <strong>of</strong> melanoma cell<br />
From the Massachusetts General Hospital Cancer Center, Boston, MA; Vanderbilt<br />
University Medical Center, Nashville, TN; Beth Israel Deaconess Medical Center, Boston,<br />
MA.<br />
Authors’ disclosures <strong>of</strong> potential conflicts <strong>of</strong> interest are found at the end <strong>of</strong> this article.<br />
Address reprint requests to Keith T. Flaherty, MD, Massachusetts General Hospital<br />
Cancer Center, 55 Fruit St., Yawkey 9E, Boston MA 02114; email: kflaherty@partners.org.<br />
© <strong>2012</strong> by <strong>American</strong> <strong>Society</strong> <strong>of</strong> <strong>Clinical</strong> <strong>Oncology</strong>.<br />
1092-9118/10/1-10
THERAPIES FOR PATIENTS WITH METASTATIC MELANOMA<br />
lines, but has not yet been validated in a large population <strong>of</strong><br />
human tumor samples. 9 PTEN deletion and inactivating<br />
mutation are the common causes <strong>of</strong> PI3 kinase pathway<br />
activation, occurring in approximately 20% <strong>of</strong> all melanomas.<br />
6 These aberrations commonly coexist with BRAF mutation,<br />
and on occasion with NRAS mutation, and are<br />
critical c<strong>of</strong>actors in melanoma formation in experimental<br />
models. 10,11 Mutations <strong>of</strong> p16 are well described in familial<br />
melanoma, and mutations or deletion <strong>of</strong> p16 are commonly<br />
found in sporadic melanoma as well (approximately 40% <strong>of</strong><br />
advanced tumors). 6 Loss <strong>of</strong> p16 function co-occurs with<br />
BRAF mutation as well as BRAF WT tumors in many cases.<br />
In cooperation with BRAF mutation, p16 loss can facilitate<br />
formation <strong>of</strong> invasive melanomas in mouse models. 12 CDK4,<br />
a key regulator <strong>of</strong> the cell cycle, is negatively regulated by<br />
p16. Thus, loss <strong>of</strong> p16 function permits CDK4 to drive<br />
progression through the cell cycle, independent <strong>of</strong> its usual<br />
upstream regulation. CDK4 itself is amplified or can harbor<br />
activating mutations in some melanomas (approximately<br />
5%) and cyclin D, the binding partner <strong>of</strong> CDK4, is amplified<br />
in a distinct subset (20%), resulting in the same cell cycle<br />
dysregulation as p16 loss. 6 To intercept the consequences <strong>of</strong><br />
p16 loss, cyclin D amplification or CDK4 mutation, or<br />
selective CDK4 inhibitor are thought be relevant.<br />
Based on currently available evidence, targeting BRAF,<br />
CKIT, and the pathways downstream <strong>of</strong> NRAS is a promising<br />
starting point for melanoma-targeted therapies. At least<br />
for BRAF, cotargeting <strong>of</strong> constituents <strong>of</strong> the PI3 kinase<br />
pathway or CDK4 appears to be a plausible strategy for<br />
overcoming resistance to BRAF inhibitors and antagonizing<br />
multiple oncogenic pathways to achieve greater degrees <strong>of</strong><br />
tumor control.<br />
Melanoma Tumor Immunology<br />
Studies <strong>of</strong> immune regulation and the mechanisms <strong>of</strong><br />
tumor-induced immune suppression have identified specific<br />
obstacles to effective immunotherapy <strong>of</strong> melanoma. These<br />
include the physiologic down-modulation <strong>of</strong> the immune<br />
response through the upregulation <strong>of</strong> the expression <strong>of</strong><br />
KEY POINTS<br />
● Delineation <strong>of</strong> genetic alterations in melanoma has<br />
provided the opportunity to develop targeted therapy<br />
in the treatment <strong>of</strong> refractory disease.<br />
● For the approximately 50% <strong>of</strong> melanoma patients<br />
whose tumors harbor mutated BRAF, BRAF inhibitors<br />
produce unprecedented response rates and a<br />
survival advantage compared with chemotherapy.<br />
● Advances in the understanding <strong>of</strong> immune-system<br />
function and tolerance have led to the development <strong>of</strong><br />
a new generation <strong>of</strong> targeted immunotherapies.<br />
● Ipilimumab represents the first immunotherapy to<br />
improve survival in phase III trials, with durable<br />
responses observed in a subset <strong>of</strong> patients.<br />
● Selection <strong>of</strong> optimal therapy for individual patients<br />
requires consideration <strong>of</strong> patients and disease characteristics<br />
to devise best first-line and sequential<br />
therapy approaches.<br />
molecules, such as CTLA4 on the surface <strong>of</strong> activated T cells.<br />
Mechanisms identified for tumor-induced immune suppression<br />
have included stimulation <strong>of</strong> CD4�, CD25� T–regulatory<br />
cell (Treg) production, which serves to limit immune<br />
activation, inhibition <strong>of</strong> T-cell receptor signaling, and melanoma<br />
cell expression <strong>of</strong> PDL1, blocking the cytolytic function<br />
<strong>of</strong> tumor-infiltrating T lymphocytes (Lisee; Greenwald, Freeman).<br />
These biologic discoveries have been made clinically<br />
relevant through the development <strong>of</strong> therapeutic agents that<br />
directly or indirectly target these immunomodulatory pathways,<br />
potentially enabling the restoration <strong>of</strong> effective,<br />
tumor-specific immune destruction. Ipilimumab, an antibody<br />
directed against CTLA4, has shown a significant survival<br />
benefit in two phase III studies <strong>of</strong> patients with<br />
advanced melanoma, leading to its U.S. Food and Drug<br />
Administration (FDA) approval in 2011. In addition, antibodies<br />
that block the effects <strong>of</strong> tumor PDL1 expression<br />
(anti-PD1 antibodies) have shown encouraging response<br />
rates and toxicity pr<strong>of</strong>iles in phase I trials, prompting<br />
extensive additional investigation. Other efforts have attempted<br />
to selectively deplete Treg cells through lymphodepletion<br />
with nonmyeloablative chemotherapy coupled<br />
with tumor specific adoptive T-cell immunotherapy. Another<br />
strategy has focused on identifying the patients most likely<br />
to respond to immunotherapy, based on melanoma genomic<br />
pr<strong>of</strong>ile or gene-expression pattern; immune-cell infiltration;<br />
or, checkpoint inhibitor or plasma cytokine expression.<br />
Treatment is restricted to those most likely to benefit or<br />
focuses on identifying ways <strong>of</strong> manipulating the immune<br />
system to convert tumors to a more immune-responsive<br />
phenotype.<br />
Immunotherapy for Melanoma<br />
IL 2–Based Therapy<br />
High-dose bolus IL2 (HD IL-2) received FDA approval in<br />
1998 for the treatment <strong>of</strong> patients with metastatic melanoma,<br />
largely based on its ability to produce durable complete<br />
responses in 5% to 10% <strong>of</strong> patients. In a retrospective<br />
review <strong>of</strong> 270 patients treated on multiple phase II studies,<br />
the objective response rate was 16%, with a median duration<br />
<strong>of</strong> 9 months (range from 4 months to over 106 months).<br />
Despite the low objective response rate, 59% <strong>of</strong> complete<br />
responders remained progression-free at 7 years, and no<br />
patient responding for longer than 30 months had progressed,<br />
suggesting that some patients were “cured.” 2 Treatment,<br />
however, was associated with significant toxicity<br />
limiting its application to a select group <strong>of</strong> patients treated<br />
in specialized centers.<br />
Efforts to improve on the efficacy <strong>of</strong> IL-2 in patients with<br />
melanoma have included combinations with chemotherapy,<br />
i.e., biochemotherapy, vaccines, and adoptive-cell therapy.<br />
Although several phase II trials, a small phase III trial, and<br />
two meta-analyses suggested that combinations <strong>of</strong> IL-2<br />
and cisplatin-based biochemotherapy <strong>of</strong>fered benefit relative<br />
to either chemotherapy or IL-2 alone, several multiinstitutional<br />
phase III trials have failed to confirm this<br />
benefit.<br />
Another approach to improving the activity <strong>of</strong> HD IL-2<br />
involved the addition <strong>of</strong> a gp100 peptide vaccine. A recently<br />
reported phase III trial randomly assigned 185 patients with<br />
metastatic melanoma to HD IL-2 given alone every three<br />
weeks or in combination with a gp100 peptide vaccine. 13<br />
525
Because <strong>of</strong> the specificity <strong>of</strong> the peptide vaccine for certain<br />
histocompatibility receptor subtypes, enrollment was limited<br />
to HLA-A*0201 patients. The study reported an objective<br />
response rate <strong>of</strong> 16% for the combination compared with<br />
6% for HD IL-2 alone. There were eight complete responses<br />
(9%) in the combination arm, but only one (1%) among those<br />
treated with IL-2 alone. There was a trend toward increased<br />
overall survival (median 17.8 months vs. 11.1 months; p �<br />
0.06), although the trial was not adequately powered to<br />
assess this endpoint. The clinical significance <strong>of</strong> this finding<br />
is uncertain, considering the relatively poor response rate<br />
in patients treated with HD IL-2 alone, the current lack <strong>of</strong><br />
availability <strong>of</strong> the specific formulation <strong>of</strong> vaccine adjuvant<br />
used in this trial, and the observations that this same<br />
vaccine did not improve the efficacy <strong>of</strong> ipilimumab in a phase<br />
III trial (see below).<br />
Others have explored the efficacy <strong>of</strong> HD IL-2 in combination<br />
with adoptive transfer <strong>of</strong> tumor derived–tumor reactive<br />
T cells. These approaches have included preparative regimens<br />
involving myeloablative chemotherapy with or without<br />
total-body irradiation in order to delete host immune<br />
cells and promote engraftment <strong>of</strong> adoptively transferred<br />
tumor-reactive T cells. 14 Autologous hematopoietic progenitorcell<br />
support was used in patients who received TBI. The<br />
Surgery Branch <strong>of</strong> the National Cancer Institute recently<br />
reported the combined results from three separate trials.<br />
There were 52 objective responses in 93 patients (56%<br />
response rate), including 20 (22%) complete responses. Complete<br />
responses were ongoing at 37 months to 82 months in<br />
19 <strong>of</strong> the 20 responders, and the 3- and 5-year actuarial<br />
survival rates for patients achieving a complete response<br />
were 100% and 93%, respectively. Efforts to confirm these<br />
results at other centers, including in a proposed multicenter<br />
trial, as well as to develop a more practical treatment<br />
regimen are currently underway.<br />
Ipilimumab<br />
The CTLA-4 receptor on T lymphocytes is a negative<br />
regulator <strong>of</strong> T-cell activation that blocks positive stimulatory<br />
effects to these cells, mediated through their costimulatory<br />
and antigen specific–T-cell receptors. The monoclonal antibodies<br />
ipilimumab and tremelimumab bind to CTLA-4 and<br />
thus prevent this feedback inhibition. Both have been studied<br />
in patients with melanoma, with the most extensive data<br />
and promising results being observed with ipilimumab.<br />
Ipilimumab was studied in a placebo-controlled phase III<br />
trial in which 676 patients with previously treated advanced<br />
melanoma were randomly assigned in a 3:1:1 ratio to ipilimumab<br />
plus gp100 peptide vaccine, ipilimumab alone or<br />
gp100 vaccine alone. 15 Ipilimumab (3 mg/kg) and/or vaccine<br />
were given every three weeks for four doses. Patients with<br />
confirmed partial or complete response or stable disease for<br />
three months or more after completion <strong>of</strong> the 12-week<br />
induction period were allowed to receive reinduction with<br />
their original treatment if they subsequently had disease<br />
progression.<br />
In this study, overall survival was significantly increased<br />
in the two groups who received ipilimumab (median 10.0<br />
months and 10.1 months vs. 6.4 months, in the ipilimumab<br />
plus gp100, ipilimumab alone, and gp100 groups, hazard<br />
ratios for death 0.68 and 0.66 compared with gp100 alone,<br />
respectively). Treatment benefits appeared to be independent<br />
<strong>of</strong> sex, age (� age 65 or � age 65), stage at presentation<br />
526<br />
FLAHERTY, SOSMAN, AND ATKINS<br />
(M0, M1a, and M1b compared with M1c), baseline lactate<br />
dehydrogenase (LDH), or prior use <strong>of</strong> IL-2. Tumor response<br />
rate was also significantly improved in both groups <strong>of</strong><br />
patients treated with ipilimumab compared with gp100<br />
alone (5.7% and 10.9% vs. 1.5%, respectively). Further,<br />
objective partial or complete responses were maintained for<br />
at least 2 years in four <strong>of</strong> 23 (17%) patients treated with<br />
ipilimumab plus gp100 and nine <strong>of</strong> 15 (60%) patients with<br />
ipilimumab alone. Among 31 patients who initially received<br />
ipilimumab either alone or with gp100 and then underwent<br />
reinduction therapy with ipilimumab, six (21%) had an<br />
objective response to retreatment, and 15 (48%) had stable<br />
disease. Although this phase III trial limited enrollment to<br />
patients who were HLA-A*0201 positive, a retrospective<br />
analysis <strong>of</strong> four phase II trials involving ipilimumab alone<br />
showed similar activity regardless <strong>of</strong> HLA type. Although<br />
patients in this trial did not have tumor pr<strong>of</strong>iling for BRAF<br />
mutations, limited unpublished data from Bristol-Myers<br />
Squibb (BMS) suggests that the activity <strong>of</strong> ipilimumab is<br />
independent <strong>of</strong> BRAF mutational status. As a consequence<br />
<strong>of</strong> this study, ipilimumab was approved for the treatment <strong>of</strong><br />
all patients with advanced melanoma.<br />
The presumed mechanism <strong>of</strong> action <strong>of</strong> ipilimumab is to<br />
break down tolerance to tumor-associated antigens in the<br />
melanoma. At the same time, this break down <strong>of</strong> tolerance<br />
may result in autoimmune reactions against self-antigens. A<br />
wide range <strong>of</strong> immune-mediated adverse events have been<br />
observed. The most common, serious manifestations include<br />
enterocolitis, hepatitis, dermatitis, and endocrinopathies. In<br />
this trial using a 3 mg/kg dose <strong>of</strong> ipilimumab, immunerelated<br />
adverse events occurred in approximately 60% <strong>of</strong><br />
patients treated with ipilimumab. Grade 3 or 4 toxicity was<br />
seen in 10% to 15% <strong>of</strong> ipilimumab-treated patients, compared<br />
to 3% <strong>of</strong> those receiving only gp100. These side effects<br />
were typically not seen until 6 or more weeks into therapy.<br />
A somewhat higher incidence <strong>of</strong> side effects was observed<br />
with a dose <strong>of</strong> 10 mg/kg every 3 weeks in a randomized phase<br />
II trial that assessed the effects <strong>of</strong> dose on activity and<br />
toxicity. 16 Several investigators have suggested that the<br />
development <strong>of</strong> immune-related toxicities correlated with<br />
benefit from therapy.<br />
Although patients with untreated brain metastases were<br />
excluded from the phase III trial, other studies have observed<br />
antitumor activity with ipilimumab for patients with<br />
brain metastases. 17 Finally, data from phase II trials suggested<br />
that a number <strong>of</strong> patients (up to 10% <strong>of</strong> those treated)<br />
exhibited apparent disease progression after 12 weeks <strong>of</strong><br />
ipilimumab (with either larger lesions or new lesions),<br />
followed by subsequent disease regression. The overall survival<br />
outcome <strong>of</strong> these patients was similar to those exhibiting<br />
a tumor response. This led to the establishment <strong>of</strong><br />
Immune-related Response Criteria that endeavored to capture<br />
these patients in the subset <strong>of</strong> patients achieving<br />
treatment benefit. 18<br />
A second phase III trial involved previously untreated<br />
patients who were randomly assigned to dacarbazine plus<br />
either ipilimumab or placebo. 19 In this study, overall survival<br />
was significantly increased in patients assigned to the<br />
dacarbazine plus ipilimumab arm (median 11.2 months vs.<br />
9.1 months). The overall incidence <strong>of</strong> grade 3 or 4 toxicity<br />
was significantly higher with dacarbazine plus ipilimumab<br />
(56% vs. 28%). In particular, hepatic toxicity was significantly<br />
more common with the combination than with dacar-
THERAPIES FOR PATIENTS WITH METASTATIC MELANOMA<br />
bazine alone or than that observed with ipilimumab alone.<br />
The increase in hepatic toxicity may be due to its combination<br />
with dacarbazine, which is also known to be hepatotoxic.<br />
On other hand, the incidence <strong>of</strong> other immune-related<br />
toxicities (colitis, rash, hypophysitis) was less than that seen<br />
in prior studies with ipilimumab alone, perhaps suggesting<br />
that dacarbazine may have blunted the immune-toxicity<br />
pr<strong>of</strong>ile,and/or the higher incidence <strong>of</strong> hepatotoxicity may<br />
have pre-empted or altered the immune toxicity pr<strong>of</strong>ile.<br />
Whether this blunting <strong>of</strong> immune toxicity by dacarbazine<br />
might have also blunted the antitumor effect <strong>of</strong> ipilimumab<br />
is a matter <strong>of</strong> speculation. However, the overall pattern <strong>of</strong><br />
toxicity and efficacy <strong>of</strong> this trial do not support the addition<br />
<strong>of</strong> dacarbazine to ipilimumab. The relative value <strong>of</strong> the use<br />
<strong>of</strong> ipilimumab at the 10 mg/kg dose used in this study and in<br />
multiple phase II studies versus the already approved 3<br />
mg/kg dose awaits the completion <strong>of</strong> an ongoing phase III<br />
trial directly comparing the two doses.<br />
Other Immune Regulatory Checkpoints<br />
Monoclonal antibodies targeted against a number <strong>of</strong> other<br />
regulatory checkpoints are being evaluated in patients with<br />
advanced melanoma based on our current understanding <strong>of</strong><br />
the development <strong>of</strong> cellular immunity. The PD-1 receptor<br />
acts as an inhibitory receptor <strong>of</strong> T cells, in a manner<br />
analogous to CTLA-4S. 20 In a preliminary report <strong>of</strong> a phase<br />
I study, a monoclonal antibody directed against the PD-1<br />
receptor (MDX-1106) caused significant regression <strong>of</strong> metastatic<br />
melanoma lesions in three cases and appears to be<br />
associated with less autoimmunity than reported with ipilimumab.<br />
21 Ongoing studies are exploring various doses <strong>of</strong><br />
this antibody as well as a variety <strong>of</strong> different antibodies<br />
targeting either PD1 or PDL1 patients with melanoma. A<br />
member <strong>of</strong> the tumor necrosis factor (TNF) family, 4–1BB<br />
(CD137), acts as a costimulatory molecule that causes T-cell<br />
proliferation. A humanized Monoclonal Ab (MAb), BMS-<br />
663513, targeted at CD137 acts as an agonist and can cause<br />
costimulation <strong>of</strong> CD8� and CD4� cells. In a preliminary<br />
report <strong>of</strong> the initial phase I study with this agent, three<br />
partial responses were observed among 54 patients with<br />
melanoma. 22 OX-40 is another TNF receptor, which also<br />
acts as a costimulatory factor for T cells. A MAb that targets<br />
this receptor has begun a phase I study. 23<br />
Treatment Selection Options<br />
Considerable effort has focused on identifying patients<br />
who respond to immunotherapy in the hope <strong>of</strong> restricting<br />
such treatment to those most likely to benefit. IL-2 responsiveness<br />
has been shown to be more likely in patients with<br />
normal serum LDH, or low plasma vascular endothelial<br />
growth (VEGF) and fibronectin levels. 24 In addition, response<br />
appears to be more frequent in patients whose<br />
tumors contain mutations in BRAF or NRAS, or possess an<br />
inflammatory gene-expression signature. 25 More recent<br />
studies have suggested that response to IL-2 is associated<br />
with enhancement <strong>of</strong> a pre-existing gene expression pattern<br />
within the tumor associated with immune-mediated tissuespecific<br />
destruction under the control <strong>of</strong> IFNgamma. 26<br />
Benefit from vaccination has also been linked to tumors<br />
expressing an IFN driven chemokine signature. 27 Preliminary<br />
results suggest that both PD1 antibody responsiveness<br />
and IL-2 responsive in patients with renal cell carcinoma<br />
(RCC) may be correlated with tumor cell–surface expression<br />
<strong>of</strong> PDL1. Furthermore, research suggests that tumor PDL1<br />
expression is not constitutive, but is related to the secretion<br />
<strong>of</strong> IFNg by <strong>of</strong> tumor reactive CD8 T cells in the microenvironment.<br />
Thus, effective immunotherapy may require<br />
pre-existence <strong>of</strong> tumor-specific immunity within the microenvironment<br />
and the use <strong>of</strong> agents that can either drive<br />
T-cell function (HD IL-2 or vaccines) or block inherent<br />
immunoregulatory signals (ipilimumab, or anti-PD1). Several<br />
current studies are underway to validate these predictive<br />
biomarkers for specific immunotherapies as well as<br />
to determine if combinations <strong>of</strong> immunotherapy with either<br />
other immunotherapies or molecularly targeted agents<br />
could convert nonimmune responsive tumors into those<br />
capable <strong>of</strong> responding.<br />
BRAF Inhibitor Therapy for BRAFV600<br />
Mutant Melanoma<br />
As previously described, approximately 50% <strong>of</strong> patients<br />
with cutaneous melanoma, have tumors that express the<br />
mutated BRAF oncogene. Those patients with melanoma<br />
carrying a V600E/K mutation are responsive to selective<br />
BRAF inhibitors <strong>of</strong> which vemurafenib is the first to obtain<br />
FDA-approval. 28,29 Other BRAF inhibitors are in the late<br />
stages <strong>of</strong> development and look equally promising. 30<br />
Through a series <strong>of</strong> clinical trials performed in rapid succession<br />
from phase I to phase II and III, the BRAF inhibitor,<br />
vemurafenib, was shown to effectively inhibit the MAPK<br />
pathway constitutively activated by the BRAF mutation;<br />
affect clinical tumor regression in most patients with objective<br />
clinical responses in over 50% <strong>of</strong> patients; and, finally,<br />
improve overall survival compared to dacarbazine chemotherapy.<br />
28,29 The phase I trial established that vemurafenib<br />
inhibited its target and demonstrated clinical efficacy in<br />
most <strong>of</strong> the 32 patients with a BRAFV600E/K mutation. 28 A<br />
much larger phase II trial enrolled patients with 132<br />
BRAFV600E/K mutated-melanoma who had failed either<br />
standard immunotherapy or chemotherapy and documented<br />
an objective response rate <strong>of</strong> 53%, <strong>of</strong> which 6% were complete<br />
responses. 31 The median progression-free survival for<br />
the entire population was 6.8 months. The median overall<br />
survival was a remarkable 15.9 months (95% CI 11.8–18.3).<br />
Finally, a randomized phase III trial allocated 675 patients<br />
to open label vemurafenib at 960 mg twice daily orally or<br />
dacarbazine at 1,000 mg/m 2 every 3 weeks with no crossover<br />
allowed. At the first interim analysis, the HR for death<br />
was 0.37 (95% CI 0.26–0.55; p � 0.0001). But the median<br />
follow-up was short, only 3.75 months. Vemurafenib was<br />
approved in August <strong>of</strong> 2011.<br />
Although vemurafenib is generally well tolerated, it can<br />
cause frequent skin toxicities dominated by hyperproliferative<br />
skin lesions, including keratoacanthomas, a low-grade<br />
cutaneous squamous cell carcinoma. 32 Approximately 25%<br />
<strong>of</strong> patients develop such lesions in a median time <strong>of</strong> only 8<br />
weeks, but they can be easily treated by excision without<br />
interrupting treatment. However, there may be a more<br />
significant issue in the adjuvant setting that could be<br />
representative <strong>of</strong> the agent’s potential to enhance the<br />
growth <strong>of</strong> other subclinical malignancies.<br />
With a median progression-free survival <strong>of</strong> 6 to 8 months,<br />
many responses to BRAF inhibitor therapy are short-lived.<br />
A minority are maintained for more than 12 months, and<br />
these patients are typically those with baseline normal LDH<br />
and nonbulky disease. Resistance to these BRAF inhibitors<br />
527
has been intensively studied. These studies have proposed<br />
and validated a number <strong>of</strong> resistance mechanisms in small<br />
numbers <strong>of</strong> patient tumor samples including: (1) Mutations<br />
<strong>of</strong> NRAS, the protein upstream <strong>of</strong> BRAF in the MAPK<br />
pathway. 33 (2) Activation <strong>of</strong> downstream signaling proteins<br />
including MEK1 through mutation or activation by overexpression<br />
<strong>of</strong> MAP kinases 34,35 (3) The expression <strong>of</strong> alternate<br />
shortened forms <strong>of</strong> the BRAFV600 mutated molecule that<br />
are capable <strong>of</strong> binding together and reactivating the pathway<br />
even in the presence <strong>of</strong> vemurafenib. 36 All three <strong>of</strong> these<br />
mechanisms appear to reactivate the MAPK pathway, providing<br />
a direction to pursue in order to effectively delay or<br />
overcome these forms <strong>of</strong> resistance. (4) Alternatively, activation<br />
or overexpression <strong>of</strong> alternate pathways through<br />
activation <strong>of</strong> receptor tyrosine kinases, such as IGFR1 or<br />
PDGFRb. Both <strong>of</strong> these RTK appear to signal at least in part<br />
through the PI3K/Akt/mTOR pathway. 33,37<br />
Ongoing trials are combining BRAF and MEK inhibitors<br />
with the goal <strong>of</strong> delaying or overcoming resistance as a<br />
result <strong>of</strong> reactivation <strong>of</strong> the MAPK pathway. On the other<br />
hand, combinations <strong>of</strong> BRAF-inhibitor therapy with either<br />
PI3 kinase, PI3 kinase/mTOR, Akt, mTORC1- and mTORC2dual<br />
inhibitors could be effective inhibiting mechanisms <strong>of</strong><br />
resistance involving alternate pathways and different receptor<br />
tyrosine kinases.<br />
Although BRAF inhibitors induce rapid disease regression<br />
and symptom benefit in patients with bulky disease, markedly<br />
delay progression, and improve overall survival for the<br />
overall metastatic melanoma population, there is clearly<br />
need for further improvement <strong>of</strong> these outcomes.<br />
For patients with BRAFV600E/K mutant melanoma,<br />
there is a choice <strong>of</strong> therapy between new and older immunotherapy<br />
approaches, including high dose IL-2, ipilimumab,<br />
or a clinical trial investigating anti-PD1 compared<br />
with a BRAF inhibitor. It is apparent that patients with<br />
symptomatic, bulky, rapidly growing, and high-serum LDH–<br />
associated melanoma are much more likely to have rapid<br />
clinical improvement when treated with vemurafenib. The<br />
improvement may be relatively short term but can last up to<br />
12 months in some patients. These patients are unlikely to<br />
benefit from immunotherapy, in part, because immunebased<br />
treatments can take a period <strong>of</strong> time and some<br />
experience progression initially before later improvement.<br />
Furthermore, the percent <strong>of</strong> patients having an objective<br />
clinical response is much lower with IL-2 (15% to 20%) or<br />
ipilimumab (5% to 10%) than with vemurafenib (� 50%).<br />
The greater dilemma surrounds the patient with BRAFmutant<br />
melanoma who has nonbulky, asymptomatic disease,<br />
and with a normal LDH. These patients have a choice.<br />
If young enough, with excellent organ function and at a<br />
treatment center with extensive experience, IL-2 may be the<br />
first choice. Or, in others who do not fulfill IL-2 criteria,<br />
ipilimumab. However, because <strong>of</strong> the delay in beneficial<br />
effect, it may require waiting up to 4 to 6 months before one<br />
can be certain that the melanoma is not responsive. Also, in<br />
this subset <strong>of</strong> patients, the response rate to vemurafenib is<br />
above 60% in the phase II trial, and overall survival is quite<br />
long with a median <strong>of</strong> 15.9 months. 31 This compares to 10<br />
months on the previously treated trial and 11.2 months on<br />
the treatment-naive trial with ipilimumab. The definite<br />
answer to the question <strong>of</strong> which drug to use first will only be<br />
answered through trials including combinations <strong>of</strong> the two<br />
agents.<br />
528<br />
Treatment Options for BRAF WT Melanoma<br />
BRAF WT melanoma is a heterogeneous disease that<br />
includes patients with NRAS mutated melanoma (about<br />
one-third <strong>of</strong> these BRAF WT patients) as well as very small<br />
subset <strong>of</strong> patients with CKIT mutant melanoma from mucosal,<br />
acral, and chronic sun-damaged sites. Characterizing<br />
these mutations can be justified currently for the purposes<br />
<strong>of</strong> directing patients to relevant clinical trials. There is<br />
supporting literature showing that a fraction <strong>of</strong> patients<br />
with L597 or K642 mutations in CKIT are responsive to<br />
imatinib, some <strong>of</strong> which the results are quite durable (� 12<br />
months). 38,39 NRAS approaches will be discussed below.<br />
Immunotherapy remains a strong consideration in BRAF<br />
WT patients. Those who are eligible for IL-2 should consider<br />
this treatment at an experienced center. Since it has the<br />
longest follow-up, we are confident <strong>of</strong> both its response rate<br />
and the frequency <strong>of</strong> extremely durable disease responses.<br />
Those patients are frequently assessed at 7 to 8 weeks and<br />
again at 11 to 12 weeks and in the case <strong>of</strong> progression or<br />
even stable disease, this treatment is rarely continued.<br />
Toxicities are acute and rarely, if ever, delayed. For these<br />
reasons, we believe that consideration <strong>of</strong> ipilimumab should<br />
come after these patients have received IL-2 if the treatment<br />
is appropriate. As stated previously, new trials with anti-<br />
PD1 antibodies are ongoing and demonstrate exciting, early<br />
results with response rates over 20% <strong>of</strong> which most are<br />
durable lasting over 12 months.<br />
It is critical to continue efforts to define other driver<br />
kinases within the BRAF WT population that could be<br />
targeted with drug therapy. For now, efforts are underway<br />
to investigate approaches to NRAS mutant melanoma,<br />
These include MEK inhibitors (GSK 1120212, TAK733, or<br />
others) alone or in combination with inhibitors <strong>of</strong> the PI3K/<br />
AKT, mTOR pathway. These trials are still in dose-finding<br />
phases, but soon phase II efforts will be undertaken in<br />
NRAS mutant melanoma.<br />
Finally, chemotherapy still represents an option for patients<br />
without either good standard options, as listed above,<br />
or clinical phase I and II trial options. Because <strong>of</strong> its limited<br />
long-term benefit, chemotherapy is best considered for patients<br />
who are symptomatic or with rapidly growing disease<br />
with a goal <strong>of</strong> improving quality <strong>of</strong> life. Agents, including<br />
dacarbazine, temozolomide, and taxane-based regimens,<br />
have all demonstrated similar antitumor activity with below<br />
20% response rates. Rarely, long-lasting responses are observed.<br />
Conclusion<br />
FLAHERTY, SOSMAN, AND ATKINS<br />
Given the current availability <strong>of</strong> multiple treatment options,<br />
patients and physicians now <strong>of</strong>ten have choices regarding<br />
initial treatment and the sequence <strong>of</strong> various<br />
treatments. For patients with BFAFV600E mutant melanoma,<br />
options could include HD IL-2, ipilimumab or vemurafenib.<br />
At the moment, there appears to be only limited<br />
information to guide this choice. Although vemurafenib<br />
appears to be equally active in patients whose disease has<br />
progressed following IL 2-based immunotherapy, there is no<br />
data on its activity following resistance to ipilimumab.<br />
Similarly, there is no data on the activity <strong>of</strong> ipilimumab (or<br />
even the feasibility <strong>of</strong> stopping vemurafenib and administering<br />
ipilimumab) following disease progression on vemurafenib.<br />
It is conceivable that patients treated initially with
THERAPIES FOR PATIENTS WITH METASTATIC MELANOMA<br />
ipilimumab may get the benefit <strong>of</strong> a durable response (up to<br />
25%) without compromising the benefit <strong>of</strong> subsequent vemurafenib<br />
for those patients who exhibit disease progression,<br />
whereas those started on vemurafenib may not be able to<br />
benefit from subsequent ipilimumab at time <strong>of</strong> progression.<br />
It is also possible that the high response rate and prolonged<br />
progression free survival associated with vemurafenib therapy<br />
relative to ipilimumab may overwhelm any benefit that<br />
might be achieved with initial ipilimumab therapy. Further,<br />
it is possible that subsets <strong>of</strong> patients, defined by tumor<br />
and/or clinical characteristics (for example, PTEN loss, immune<br />
infiltration, high serum LDH), might do better with<br />
one initial therapy or sequence than with the other. A<br />
randomized phase III trial <strong>of</strong> ipilimumab (followed by vemurafenib)<br />
versus vemurafenib (followed by ipilimumab) is<br />
being planned within the Cooperative Group mechanism in<br />
an effort to obtain real data to address these important<br />
therapeutic questions and guide future treatment choices for<br />
this patient population. In addition, this population repre-<br />
Authors’ Disclosures <strong>of</strong> Potential Conflicts <strong>of</strong> Interest<br />
Author<br />
Employment or<br />
Leadership<br />
Positions<br />
Consultant or<br />
Advisory Role<br />
Keith T. Flaherty Genentech;<br />
GlaxoSmithKline;<br />
Metamark<br />
Genetics<br />
Jeff A. Sosman GlaxoSmithKline;<br />
Roche<br />
Michael B. Atkins Bristol-Myers<br />
Squibb; Celgene;<br />
Curetech;<br />
Genentech;<br />
Merck; Novartis;<br />
Prometheus<br />
1. Siegel R, Ward E, Brawley O, et al. Cancer Statistics, 2011: the impact<br />
<strong>of</strong> eliminating socioeconomic and racial disparities on premature cancer<br />
deaths. CA Cancer J Clin. 2011;61:212-236.<br />
2. Atkins MB, Lotze MT, Dutcher JP, et al. High-dose recombinant interleukin<br />
2 therapy for patients with metastatic melanoma: analysis <strong>of</strong> 270<br />
patients treated between 1985 and 1993. J Clin Oncol. 1999;17:2105-2116.<br />
3. Davies H, Bignell GR, Cox C, et al. Mutations <strong>of</strong> the BRAF gene in<br />
human cancer. Nature. 2002;417:949-954.<br />
4. Tsai J, Lee JT, Wang W, et al. Discovery <strong>of</strong> a selective inhibitor <strong>of</strong><br />
oncogenic B-Raf kinase with potent antimelanoma activity. Proc Natl Acad<br />
SciUSA.2008;105:3041-3046.<br />
5. Neal SE, Eccleston JF, Webb MR. Hydrolysis <strong>of</strong> GTP by p21NRAS, the<br />
NRAS protooncogene product, is accompanied by a conformational change in<br />
the wild-type protein: use <strong>of</strong> a single fluorescent probe at the catalytic site.<br />
Proc Natl Acad Sci USA.1990;87:3562-3565.<br />
6. Curtin JA, Fridlyand J, Kageshita T, et al. Distinct sets <strong>of</strong> genetic<br />
alterations in melanoma. N Engl J Med. 2005;353:2135-2147.<br />
7. Curtin JA, Busam K, Pinkel D, et al. Somatic activation <strong>of</strong> KIT in<br />
distinct subtypes <strong>of</strong> melanoma. J Clin Oncol. 2006;24:4340-4346.<br />
8. Davies MA, Stemke-Hale K, Tellez C, et al. A novel AKT3 mutation in<br />
melanoma tumours and cell lines. Br J Cancer. 2008;99:1265-1268.<br />
9. Stahl JM, Sharma A, Cheung M, et al. Deregulated Akt3 activity<br />
promotes development <strong>of</strong> malignant melanoma. Cancer Res. 2004;64:7002-<br />
7010.<br />
10. Tsao H, Goel V, Wu H, et al. Genetic interaction between NRAS and<br />
BRAF mutations and PTEN/MMAC1 inactivation in melanoma. J Invest<br />
Dermatol. 2004;122:337-341.<br />
11. Dankort D, Curley DP, Cartlidge RA, et al. Braf(V600E) cooperates<br />
with Pten loss to induce metastatic melanoma. Nat Genet. 2009;41:544-552.<br />
12. Dhomen N, Reis-Filho JS, da Rocha Dias S, et al. Oncogenic Braf<br />
sents an ideal group for combination treatment approaches.<br />
Studies involving selective BRAF inhibitors with immunotherapy<br />
(ipilimumab, HD IL-2, PD1 pathway blockers),<br />
VEGF pathway inhibitors, or other molecularly targeted<br />
agents (MEK or PI3Kinase inhibitors or apoptosis inducers)<br />
are currently underway and represent opportunities to further<br />
advance treatment results.<br />
Options are more limited for patients with BRAF WT<br />
melanoma. For the few patients with mutations in C-Kit,<br />
imatinib or participation in a C-kit-inhibitor clinical trial<br />
seems to represent the best choice. For those with NRAS<br />
mutations, HD IL-2 might be a reasonable option, although<br />
data supporting this selection requires validation. Additional<br />
research is needed to determine the activity <strong>of</strong> ipilimumab<br />
or other immunotherapies in this population.<br />
Efforts are also underway to identify potential therapeutic<br />
targets downstream <strong>of</strong> NRAS or for BRAF, NRAS, or CKIT<br />
triple WT tumors that may allow for development <strong>of</strong> personalized<br />
treatment options for this patient population.<br />
Stock<br />
Ownership Honoraria<br />
Genentech;<br />
GlaxoSmithKline<br />
REFERENCES<br />
Roche<br />
Research<br />
Funding<br />
Expert<br />
Testimony<br />
Other<br />
Remuneration<br />
induces melanocyte senescence and melanoma in mice. Cancer Cell. 2009;15:<br />
294-303.<br />
13. Schwartzentruber DJ, Lawson DH, Richards JM, et al. Gp100 peptide<br />
vaccine and interleukin-2 in patients with advanced melanoma. N Engl<br />
J Med. 2011;364:2119-2127.<br />
14. Rosenberg SA, Yang JC, Sherry RM, et al. Durable complete responses<br />
in heavily pretreated patients with metastatic melanoma using T-cell transfer<br />
immunotherapy. Clin Cancer Res. 2011;17:4550-4557.<br />
15. Hodi FS, O’Day SJ, McDermott DF, et al. Improved survival with<br />
ipilimumab in patients with metastatic melanoma. N Engl J Med. 2010;363:<br />
711-723.<br />
16. Wolchok JD, Neyns B, Linette G, et al. Ipilimumab monotherapy in<br />
patients with pretreated advanced melanoma: A randomised, double-blind,<br />
multicentre, phase 2, dose-ranging study. Lancet Oncol. 2010;11:155-64.<br />
17. Lawrence DPH, McDermott DF. Phase II trial <strong>of</strong> ipilimumab monotherapy<br />
in melanoma patients with brain metastasis. J Clin Oncol. 2010;28<br />
(suppl; abstr 8523).<br />
18. Wolchok JD, Hoos A, O’Day S, et al. Guidelines for the evaluation <strong>of</strong><br />
immune therapy activity in solid tumors: immune-related response criteria.<br />
Clin Cancer Res. 2009;15:7412-7420.<br />
19. Robert C, Thomas L, Bondarenko I, et al. Ipilimumab plus dacarbazine<br />
for previously untreated metastatic melanoma. N Engl J Med. 364:2517-2526.<br />
20. Sharpe AHW, EJ, Ahmed R, Freeman GJ. The function <strong>of</strong> programmed<br />
cell death 1 and its ligands in regulating autoimmunity and infection. Nat<br />
Immun. 2007;8:239-245.<br />
21. Brahmer JRD, Wollner I, Powderly JD, et al. Phase I study <strong>of</strong> singleagent<br />
anti-programmed death-1 (MDX-1106) in refractory solid tumors:<br />
safety, clinical activity, pharmacodynamics, and immunologic correlates.<br />
J Clin Oncol. 2010;28:3167-3175.<br />
22. Sznol MH, Margolin K, McDermott D, et al. Phase I study <strong>of</strong> BMS-<br />
529
663513, a fully human anti-CD137 agonist monoclonal antibody in patients<br />
with advanced cancer. J Clin Oncol. 2008;26;15s (suppl; abstr 3007).<br />
23. Curti BW, Morris N. A phase I trial <strong>of</strong> monoclonal antibody to OX40 in<br />
patients with advanced cancer (abstract). International <strong>Society</strong> for Biological<br />
Therapy <strong>of</strong> Cancer Annual Meeting. 2007.<br />
24. Sabatino MK, Panelli MC. Serum vascular endothelial growth factor<br />
and fibronectin predict clinical response to high-dose interleukin-2 therapy.<br />
J Clin Oncol. 2009;27:2645-2652.<br />
25. Joseph RW, Sullivan RJ, Harrell R, et al. Correlation <strong>of</strong> NRAS mutations<br />
with clinical response to high-dose IL-2 in patients with advanced<br />
melanoma. J Immunother. <strong>2012</strong>;35:66-72.<br />
26. Sullivan RJ, Hoshida Y, Brunet J, et al. A single center experience with<br />
high-dose IL-2 treatment for patients with advanced melanoma and pilot<br />
investigation <strong>of</strong> a novel gene expression signature as a predictor <strong>of</strong> response.<br />
J Clin Oncol. 2009;27:15s (suppl; abstr 9003).<br />
27. Gajewski TF, Louahed J, Brichard VG. Gene signature in melanoma<br />
associated with clinical activity: A potential clue to unlock cancer immunotherapy.<br />
Cancer J. 2010;16:399-403.<br />
28. Flaherty KT, Puzanov I, Kim KB, et al. Inhibition <strong>of</strong> mutated, activated<br />
BRAF in metastatic melanoma. N Engl J Med. 2010;363:809-819.<br />
29. Chapman PB, Hauschild A, Robert C, et al. Improved survival with<br />
vemurafenib in melanoma with BRAF V600E mutation. N Engl J Med.<br />
2011;364:2507-2516.<br />
30. Kefford RA, Brown MP, Millward M, et al. Phase I/II study <strong>of</strong><br />
GSK2118436, a selective inhibitor <strong>of</strong> oncogenic mutant BRAF kinase, in<br />
patients with metastatic melanoma and other solid tumors. J Clin Oncol.<br />
2010;28;15s (suppl; abstr 8503).<br />
530<br />
FLAHERTY, SOSMAN, AND ATKINS<br />
31. Ribas AK, Schuchter L, Gonzalez R, et al. BRIM 2: An open-label,<br />
multicenter phase II study <strong>of</strong> vemurafenib (PLX4032, RG7204) in previously<br />
treated patients with BRAFV600E mutation-positive metastatic melanoma.<br />
J Clin Oncol. 2011;29:15s (abstr; suppl 8509).<br />
32. Su F, Viros A, Milagre C, et al. RAS mutations in cutaneous squamouscell<br />
carcinomas in patients treated with BRAF inhibitors. N Engl J Med.<br />
<strong>2012</strong>;366:207-215.<br />
33. Nazarian R, Shi H, Wang Q, et al. Melanomas acquire resistance to<br />
B-RAF(V600E) inhibition by RTK or N-RAS upregulation. Nature. 2010;468:<br />
973-977.<br />
34. Wagle N, Emery C, Berger MF, et al. Dissecting therapeutic resistance<br />
to RAF inhibition in melanoma by tumor genomic pr<strong>of</strong>iling. J Clin Oncol.<br />
2011;29:3085-3096.<br />
35. Johannessen CM, Boehm JS, Kim SY, et al. COT drives resistance to<br />
RAF inhibition through MAP kinase pathway reactivation. Nature. 2010;468:<br />
968-972.<br />
36. Poulikakos PI, Persaud Y, Janakiraman M, et al. RAF inhibitor<br />
resistance is mediated by dimerization <strong>of</strong> aberrantly spliced BRAF(V600E).<br />
Nature. 2011;480:387-390.<br />
37. Villanueva J, Vultur A, Lee JT, et al. Acquired resistance to BRAF<br />
inhibitors mediated by a RAF kinase switch in melanoma can be overcome by<br />
cotargeting MEK and IGF-1R/PI3K. Cancer Cell. 2010;18:683-695.<br />
38. Guo J, Si L, Kong Y, et al. Phase II, open-label, single-arm trial <strong>of</strong><br />
imatinib mesylate in patients with metastatic melanoma harboring c-Kit<br />
mutation or amplification. J Clin Oncol. 2011;29:2904-2909.<br />
39. Carvajal RD, Antonescu CR, Wolchok JD, et al. KIT as a therapeutic<br />
target in metastatic melanoma. JAMA. 2011;305:2327-2334.
ANTIEMETICS: CURRENT STANDARDS, EMERGING<br />
APPROACHES, AND PERSISTENT GAPS<br />
CHAIR<br />
Ethan Basch, MD, MSc<br />
Memorial Sloan-Kettering Cancer Center<br />
New York, NY<br />
SPEAKERS<br />
Steven M. Grunberg, MD<br />
Fletcher Allen Health Care<br />
Burlington, VT<br />
Rebecca Anne Clark-Snow, RN<br />
University <strong>of</strong> Kansas Cancer Center<br />
Kansas City, KS<br />
Matti S. Aapro, MD<br />
Clinique de Genolier<br />
Genolier, Switzerland
Antiemetic Use in <strong>Oncology</strong>: Updated<br />
Guideline Recommendations from ASCO<br />
By Ethan Basch, MD, Ann Alexis Prestrud, MPH, Paul J. Hesketh, MD, Mark G. Kris, MD,<br />
Mark R. Somerfield, PhD, and Gary H. Lyman, MD<br />
Overview: In 2011, ASCO updated its guideline for the use <strong>of</strong><br />
antiemetics in oncology, informed by a systematic review <strong>of</strong><br />
the medical literature. This is an abbreviated version <strong>of</strong> that<br />
guideline, which is available in full at www.asco.org/guide<br />
lines/antiemetics. Key changes from the prior update in 2006<br />
include the following: Combined anthracycline and cyclophosphamide<br />
regimens were reclassified as highly emetic. Patients<br />
who receive this combination or any highly emetic agents<br />
should receive a 5-HT3 receptor antagonist, dexamethasone,<br />
and an NK1 receptor antagonist. A large trial validated the<br />
equivalency <strong>of</strong> fosaprepitant, a single-day intravenous formulation,<br />
with aprepitant; either therapy is appropriate. Preferential<br />
use <strong>of</strong> palonosetron is recommended for moderate<br />
LIKE ALL ASCO guidelines, these recommendations are<br />
based on a systematic review <strong>of</strong> the literature, with<br />
recommendations developed by an expert panel, followed by<br />
review by ASCO’s <strong>Clinical</strong> Practice Guideline Committee<br />
(CPGC), blinded external review, ASCO Board <strong>of</strong> Directors<br />
review, and blinded peer review by the Journal <strong>of</strong> <strong>Clinical</strong><br />
<strong>Oncology</strong>.<br />
Literature Review and Analysis Methods<br />
The initial search for this guideline’s systematic review<br />
was based on an Agency for Healthcare Research and<br />
Quality (AHRQ)-funded Evidence-Based Practice Center report<br />
completed at Oregon Health and Science University<br />
(OHSU). 3 The dates <strong>of</strong> the OHSU literature search <strong>of</strong> Medline<br />
were 1966 through October 2008. That evidence review<br />
was limited to trials including the newer antiemetics:<br />
aprepitant (the NK 1 receptor antagonist) and the 5-HT 3<br />
receptor antagonists. Initially, two literature searches were<br />
completed by ASCO staff in Medline. The first included all<br />
relevant search terms, overlapping minimally with the<br />
OHSU search, from September 2008 through December<br />
2009. A second search, excluding the OHSU intervention<br />
search terms, overlapped briefly with the search for the 2006<br />
ASCO update, ranging from February 2004 to February<br />
2010. This second search was designed to identify new<br />
adjunctive therapy. The Cochrane Collaboration Library<br />
electronic database was also searched through 2011, using<br />
the terms emesis, vomiting, and nausea. Data presented at<br />
the ASCO and the Multinational Association <strong>of</strong> Supportive<br />
Care in Cancer (MASCC) annual meetings was also searched<br />
systematically using the terms “vomiting,” “emesis,” and<br />
“nausea,” but only presentations or posters were included.<br />
Data presented only in abstract form was excluded.<br />
Eligible reports were identified in multiple rounds <strong>of</strong><br />
review by ASCO staff. Full-text copies were obtained for<br />
assessment <strong>of</strong> inclusion/exclusion criteria. Articles that provisionally<br />
met inclusion criteria underwent data extraction<br />
by ASCO staff for patient characteristics, study design and<br />
quality, interventions, outcomes, and adverse events. Evidence<br />
summary tables (Data Supplement, available online<br />
at www.asco.org/guidelines/antiemetics) were reviewed for<br />
accuracy and completeness by an ASCO staff member who<br />
was not involved in their original preparation.<br />
532<br />
emetic risk regimens, combined with dexamethasone. For<br />
low-risk agents, patients can be <strong>of</strong>fered dexamethasone before<br />
the first dose <strong>of</strong> chemotherapy. Patients undergoing high<br />
emetic risk radiation therapy should receive a 5-HT 3 receptor<br />
antagonist before each fraction and for 24 hours following<br />
treatment and may receive a 5-day course <strong>of</strong> dexamethasone<br />
during fractions 1 to 5. Continued symptom monitoring<br />
throughout therapy is recommended. Clinicians <strong>of</strong>ten underestimate<br />
the incidence <strong>of</strong> nausea, which is not as well controlled<br />
as vomiting. Detailed information about the development <strong>of</strong> the<br />
guideline as well as practice tools are available at www.asco.<br />
org/guidelines/antiemetics.<br />
Literature Review Results<br />
The literature search yielded a total <strong>of</strong> 271 unique citations<br />
from Medline and 48 from the MASCC and ASCO<br />
meetings. Additional materials evaluated were from the<br />
personal libraries <strong>of</strong> Update Committee members. Of those,<br />
36 reports met inclusion criteria (previously described) and<br />
were selected for full-text review. Eleven (30.6%) <strong>of</strong> those<br />
included were either posters or presentations from meetings.<br />
Nine studies evaluated antiemetic regimens according<br />
to emetic risk, six <strong>of</strong> which applied to highly emetic<br />
chemotherapy 4-8<br />
and the remainder to moderately<br />
emetic. 9-11 Five trials evaluated the comparative efficacy <strong>of</strong><br />
5-HT 3 receptor antagonists including a systematic review<br />
from the Cochrane Collaboration 12-16 ; five described findings<br />
from dosing studies specifically for palonosetron. 17-21 Two<br />
trials described new delivery methods <strong>of</strong> two previously<br />
approved therapies. 22,23 A number <strong>of</strong> studies assessed special<br />
populations. Three trials detailed results in patients<br />
undergoing myeloablative therapy before transplant, 24-26<br />
two described efforts in patients receiving multiday chemotherapy<br />
regimens, 27,28 and three trials evaluated antiemetic<br />
therapies in pediatric patients undergoing cancer<br />
therapy. 29-31<br />
Three studies examined complementary therapies in patients<br />
receiving cancer treatment. 32-34 Two studies that<br />
specifically considered therapy for delayed nausea and vomiting<br />
were identified. 35,36 Among the studies reviewed, only<br />
one trial evaluating therapy for patients undergoing radiation<br />
was identified. 37<br />
Update Panel<br />
A multidisciplinary expert panel was assembled in accordance<br />
with ASCO’s Conflict <strong>of</strong> Interest Management Proce-<br />
From the Memorial-Sloan Kettering Cancer Center, New York, NY; <strong>American</strong> <strong>Society</strong> <strong>of</strong><br />
<strong>Clinical</strong> <strong>Oncology</strong>, Alexandria, VA; Lahey Clinic Medical Center; Burlington, MA; Duke<br />
University, Durham, NC.<br />
Authors’ disclosures <strong>of</strong> potential conflicts <strong>of</strong> interest are found at the end <strong>of</strong> this article.<br />
Address reprint requests to Ethan Basch, MD, MSc, Genitourinary <strong>Oncology</strong> Service,<br />
Department <strong>of</strong> Medicine, Memorial Sloan-Kettering Cancer Center, 1275 York Avenue, New<br />
York, NY 10065; email: basche@mskcc.org.<br />
© <strong>2012</strong> by <strong>American</strong> <strong>Society</strong> <strong>of</strong> <strong>Clinical</strong> <strong>Oncology</strong>.<br />
1092-9118/10/1-10
ANTIEMETICS: ASCO GUIDELINE UPDATE<br />
dures for <strong>Clinical</strong> Practice Guidelines (“Procedures,”<br />
summarized at www.asco.org/guidelinescoi). The panel developed<br />
the recommendations based on the systematic review,<br />
in keeping with ASCO standard practice (www.<br />
asco.org/guidelines).<br />
Guideline Recommendations<br />
Highly and moderately emetogenic antineoplastic agents<br />
have the potential to induce both acute (� 24 hours) and<br />
delayed (� 24 hours) nausea and vomiting following chemotherapy.<br />
The guideline recommendations include prophylaxis<br />
for both types <strong>of</strong> nausea and vomiting where<br />
appropriate.<br />
This guideline update includes the most recent recommendations<br />
developed by the Update Committee (Table 1).<br />
Detailed discussions <strong>of</strong> the evidence are available in the full<br />
guideline, available at www.asco.org/guidelines/antiemetics<br />
or in the published guideline in the Journal <strong>of</strong> <strong>Clinical</strong><br />
<strong>Oncology</strong>.<br />
A table with intravenous agents organized by emetic<br />
risk (Table 2) is included. The intravenous risk stratification<br />
schema was originally published in 1997 38 and was<br />
updated at the MASCC/European <strong>Society</strong> for Medical <strong>Oncology</strong><br />
(ESMO) 2009 consensus conference. 39 The modified<br />
stratification from MASCC was adopted by ASCO for this<br />
guideline update. 40 Dosing schedules are also detailed<br />
herein (Table 3).<br />
Chemotherapy-Induced Nausea and Vomiting (CINV)<br />
<strong>Clinical</strong> Question 1. What is the optimal treatment to<br />
prevent nausea and vomiting from highly emetogenic antineoplastic<br />
agents?<br />
KEY POINTS<br />
● A list <strong>of</strong> drugs and radiation therapies associated<br />
with high, moderate, and low emetic risk, as well as a<br />
list <strong>of</strong> recommended antiemetic regimens based on<br />
emetic risk, is provided in the ASCO Antiemetic<br />
Guideline (www.asco.org/guidelines/antiemetics) and<br />
in this abbreviated version <strong>of</strong> the guideline.<br />
● Continued symptom monitoring throughout therapy<br />
is recommended. Clinicians <strong>of</strong>ten underestimate the<br />
incidence <strong>of</strong> nausea.<br />
● Patients receiving highly emetic agents should receive<br />
a 5-HT 3 receptor antagonist, dexamethasone,<br />
and an NK 1 receptor antagonist. Combined anthracycline<br />
and cyclophosphamide regimens are now classified<br />
as highly emetic.<br />
● Preferential use <strong>of</strong> palonosetron is now recommended<br />
for moderate emetic risk regimens, combined with<br />
dexamethasone. For low-risk agents, patients can be<br />
<strong>of</strong>fered dexamethasone before the first dose <strong>of</strong> chemotherapy.<br />
● Patients undergoing high emetic risk radiation therapy<br />
should receive a 5-HT 3 receptor antagonist before<br />
each fraction and for 24 hours following treatment<br />
and may receive a 5-day course <strong>of</strong> dexamethasone<br />
during fractions 1 to 5.<br />
Recommendation 1. The three-drug combination <strong>of</strong> an<br />
NK 1 receptor antagonist (day 1 through 3 for aprepitant; day<br />
1 only for fosaprepitant), a 5-HT 3 receptor antagonist (day 1<br />
only), and dexamethasone (day 1 through 3 or 1–4) is<br />
recommended for patients receiving highly emetogenic chemotherapy.<br />
This recommendation is unchanged since the<br />
2006 update, but reworded for clarification. The Update<br />
Committee also recommended reclassification <strong>of</strong> the combined<br />
anthracycline and cyclophosphamide (AC) regimen as<br />
highly emetogenic.<br />
<strong>Clinical</strong> Question 2. What is the optimal treatment to<br />
prevent nausea and vomiting from moderately emetogenic<br />
antineoplastic agents?<br />
Recommendation 2. The two-drug combination <strong>of</strong> palonosetron<br />
(day 1 only) and dexamethasone (day 1 through 3) is<br />
recommended for patients receiving moderately emetogenic<br />
chemotherapy. If palonosetron is not available, clinicians<br />
may substitute a first-generation 5-HT 3 receptor antagonist,<br />
preferably granisetron or ondansetron. Limited evidence<br />
also supports adding aprepitant to the combination. Should<br />
clinicians opt to add aprepitant in patients receiving<br />
moderate-risk chemotherapy, any one <strong>of</strong> the 5-HT 3 receptor<br />
antagonists is appropriate.<br />
<strong>Clinical</strong> Question 3. What is the optimal treatment to<br />
prevent nausea and vomiting from low emetogenic antineoplastic<br />
agents?<br />
Recommendation 3. A single 8 mg dose <strong>of</strong> dexamethasone<br />
before chemotherapy is suggested. No change since 2006.<br />
<strong>Clinical</strong> Question 4. What is the optimal treatment to<br />
prevent nausea and vomiting from minimally emetogenic<br />
antineoplastic agents?<br />
Recommendation 4. No antiemetic should be administered<br />
routinely before or after chemotherapy. No change<br />
from the original guideline.<br />
<strong>Clinical</strong> Question 5. What is the optimal treatment to<br />
prevent nausea and vomiting from combination chemotherapy?<br />
Recommendation 5. Patients should be administered<br />
antiemetics appropriate for the component chemotherapeutic<br />
(antineoplastic) agent <strong>of</strong> greatest emetic risk. No<br />
change from the original guideline. Anthracycline-cyclophosphamide<br />
combinations are now classified as highly<br />
emetogenic.<br />
<strong>Clinical</strong> Question 6. What is the role <strong>of</strong> adjunctive drugs<br />
for nausea and vomiting induced by cancer treatments?<br />
Recommendation 6. Lorazepam or diphenhydramine are<br />
useful adjuncts to antiemetic drugs, but are not recommended<br />
as single-agent antiemetics. No change since 2006.<br />
<strong>Clinical</strong> Question 7. What is the role <strong>of</strong> complementary<br />
and alternative medicine therapies to prevent or control<br />
nausea and vomiting induced by chemotherapy?<br />
Recommendation 7. No published randomized controlled<br />
trial data which met inclusion criteria are currently available<br />
to support a recommendation about such therapies.<br />
Special Populations<br />
<strong>Clinical</strong> Question 8. What is the optimal treatment to<br />
prevent nausea and vomiting associated with cancer therapy<br />
for pediatric patients?<br />
Recommendation 8. The combination <strong>of</strong> a 5-HT 3 receptor<br />
antagonist plus a corticosteroid is suggested before chemotherapy<br />
in children receiving chemotherapy <strong>of</strong> high or mod-<br />
533
Table 1. Summary <strong>of</strong> Recommendations<br />
Chemotherapy-Induced Nausea and Vomiting<br />
Highly Emetogenic Agents The three-drug combination <strong>of</strong> a 5-HT 3 receptor antagonist,<br />
dexamethasone, and aprepitant is recommended before<br />
chemotherapy. In all patients receiving cisplatin and all other<br />
agents <strong>of</strong> high emetic risk, the two-drug combination <strong>of</strong><br />
dexamethasone and aprepitant is recommended. The Update<br />
Committee no longer recommends the combination <strong>of</strong> a 5-<br />
HT3 serotonin receptor antagonist and dexamethasone on<br />
days 2 and 3.<br />
Moderately Emetogenic Agents The three-drug combination <strong>of</strong> a 5-HT3 receptor antagonist,<br />
dexamethasone, and aprepitant is recommended for patients<br />
receiving AC.<br />
For patients receiving chemotherapy <strong>of</strong> moderate emetic risk<br />
other than AC, we recommend the two-drug combination <strong>of</strong> a<br />
5-HT3 receptor antagonist and dexamethasone. In patients<br />
receiving AC, aprepitant as a single agent is recommended<br />
on days 2 and 3. For all other chemotherapies <strong>of</strong> moderate<br />
emetic risk, single-agent dexamethasone or a 5-HT3 receptor<br />
antagonist is suggested for the prevention <strong>of</strong> emesis on days 2<br />
and 3.<br />
Low Emetogenic Agents Dexamethasone 8 mg is suggested. No routine preventive use <strong>of</strong><br />
antiemetics for delayed emesis is suggested.<br />
Minimally Emetogenic Agents No change from the original guideline. No antiemetic should be<br />
administered routinely before or after chemotherapy.<br />
Combination Chemotherapy No change from the original guideline. Patients should be<br />
administered antiemetics appropriate for the chemotherapeutic<br />
agent <strong>of</strong> greatest emetic risk.<br />
Adjunctive Drugs Lorazepam and diphenhydramine are useful adjuncts to<br />
antiemetic drugs, but are not recommended as single agents.<br />
2006 2011 Update<br />
The three-drug combination <strong>of</strong> an NK 1 receptor antagonist (days<br />
1–3 for aprepitant; day 1 only for fosaprepitant), a 5-HT 3<br />
receptor antagonist (day 1 only), and dexamethasone (days<br />
1–3 or 1–4) is recommended for patients receiving highly<br />
emetogenic chemotherapy. This recommendation is<br />
unchanged since the 2006 update, but reworded for<br />
clarification.<br />
The Update Committee also recommended reclassification <strong>of</strong> the<br />
combined anthracycline and cyclophosphamide regimen as<br />
highly emetogenic.<br />
The two-drug combination <strong>of</strong> palonosetron (day 1 only) and<br />
dexamethasone (days 1–3) is recommended for patients<br />
receiving moderately emetogenic chemotherapy. If<br />
palonosetron is not available, clinicians may substitute a first<br />
generation 5-HT3 receptor antagonist, preferably granisetron<br />
or ondansetron.<br />
Limited evidence also supports adding aprepitant to the<br />
combination. Should clinicians opt to add aprepitant in<br />
patients receiving moderate-risk chemotherapy, any one <strong>of</strong> the<br />
5-HT3 receptor antagonists is appropriate.<br />
A single 8 mg dose <strong>of</strong> dexamethasone before chemotherapy is<br />
suggested. No change since 2006.<br />
No antiemetic should be administered routinely before or after<br />
chemotherapy. No change from the original guideline.<br />
Patients should be administered antiemetics appropriate for the<br />
component chemotherapeutic (antineoplastic) agent <strong>of</strong> greatest<br />
emetic risk. No change from the original guideline.<br />
Anthracycline-cyclophosphamide combinations are now<br />
classified as highly emetogenic.<br />
Lorazepam or diphenhydramine are useful adjuncts to<br />
antiemetic drugs, but are not recommended as single agent<br />
antiemetics. No change since 2006.<br />
Complementary Therapy New question for 2011 update. No published randomized controlled trial data which met<br />
inclusion criteria are currently available to support a<br />
recommendation about such therapies.<br />
Pediatric Patients The combination <strong>of</strong> a 5-HT 3 antagonist plus a corticosteroid is<br />
suggested before chemotherapy in children receiving<br />
chemotherapy <strong>of</strong> high or moderate emetic risk. Due to<br />
variation <strong>of</strong> pharmacokinetic parameters in children, higher<br />
weight-based doses <strong>of</strong> 5-HT3 antagonists than those used in<br />
adults may be required for antiemetic protection.<br />
High-Dose Chemotherapy with<br />
Stem Cell or Bone Marrow<br />
Transplant<br />
No change from original guideline. A 5-HT 3 receptor antagonist<br />
antiemetic combined with dexamethasone is suggested.<br />
Aprepitant should be considered although evidence to support<br />
its use specifically in these patients is lacking.<br />
Multiday Chemotherapy No change from the original guideline. It is suggested that<br />
antiemetics appropriate for the risk class <strong>of</strong> the chemotherapy,<br />
as outlined above, be administered for each day <strong>of</strong> the<br />
chemotherapy and for 2 d after, if appropriate.<br />
534<br />
BASCH ET AL<br />
The combination <strong>of</strong> a 5-HT3 receptor antagonist plus a<br />
corticosteroid is suggested before chemotherapy in children<br />
receiving chemotherapy <strong>of</strong> high or moderate emetic risk. Due<br />
to variation <strong>of</strong> pharmacokinetic parameters in children, higher<br />
weight-based doses <strong>of</strong> 5-HT3 receptor antagonists than those<br />
used in adults may be required for antiemetic protection. No<br />
change since 2006.<br />
A 5-HT3 receptor antagonist combined with dexamethasone is<br />
suggested. Aprepitant should be considered, although<br />
evidence to support its use is limited.<br />
It is suggested that antiemetics appropriate for the emetogenic<br />
risk class <strong>of</strong> the chemotherapy be administered for each day<br />
<strong>of</strong> the chemotherapy and for two days after, if appropriate.<br />
No change from the original guideline.<br />
The Update Committee suggests, based on limited data, that<br />
patients receiving five-day cisplatin regimens be treated with a<br />
5-HT 3 receptor antagonist in combination with<br />
dexamethasone and aprepitant.
ANTIEMETICS: ASCO GUIDELINE UPDATE<br />
Chemotherapy-Induced Nausea and Vomiting (cont’d)<br />
Emesis or Nausea Despite<br />
Optimal Prophylaxis<br />
Anticipatory Nausea and<br />
Vomiting<br />
erate emetic risk. Because <strong>of</strong> variation <strong>of</strong> pharmacokinetic<br />
parameters in children, higher weight-based doses <strong>of</strong> 5-HT 3<br />
receptor antagonists than those used in adults may be<br />
required for antiemetic protection. No change since 2006.<br />
<strong>Clinical</strong> Question 9. What is the optimal treatment to<br />
prevent nausea and vomiting in patients who are undergoing<br />
high-dose chemotherapy with stem cell or bone marrow<br />
transplant?<br />
Recommendation 9. A 5-HT 3 receptor antagonist combined<br />
with dexamethasone is recommended. Aprepitant<br />
should be considered, although evidence to support its use is<br />
limited.<br />
Question 10. What is the optimal treatment to prevent<br />
nausea and vomiting for patients receiving multiday chemotherapy?<br />
Recommendation 10. It is suggested that antiemetics appropriate<br />
for the emetogenic risk class <strong>of</strong> the chemotherapy<br />
be administered for each day <strong>of</strong> the chemotherapy and for<br />
2 days after, if appropriate. No change from the original<br />
Table 1. Summary <strong>of</strong> Recommendations (Cont’d)<br />
2006 2011<br />
No change from original guideline. The Update Committee<br />
suggests that clinicians (1) conduct a careful re-evaluation <strong>of</strong><br />
emetic risk, disease status, concurrent illnesses, and<br />
medications; (2) ascertain that the best regimen is being<br />
administered for the emetic risk; (3) consider adding an<br />
lorazepam or alprazolam to the regimen; and (4) consider<br />
substituting a high-dose intravenous metoclopramide for the<br />
5-HT 3 antagonist or adding a dopamine antagonist to the<br />
regimen.<br />
No change since the original guideline. Use <strong>of</strong> the most active<br />
antiemetic regimens appropriate for the chemotherapy being<br />
administered to prevent acute or delayed emesis is suggested.<br />
Such regimens may be used with the initial chemotherapy,<br />
rather than assessing the patient’s emetic response with less<br />
effective treatment. If anticipatory emesis occurs, behavioral<br />
therapy with systematic desensitization is effective and<br />
suggested.<br />
Radiation Induced Nausea and Vomiting<br />
High Risk No change from original guideline. The Update Committee<br />
suggests administration a 5-HT 3 antagonist with or without a<br />
corticosteroid before each fraction and for at least 24 h after.<br />
There is no change from the original guideline.<br />
Moderate Risk The Update Committee recommends a 5-HT 3 antagonist before<br />
each fraction.<br />
Low Risk No change from original guideline. The Update Committee<br />
recommends a 5-HT 3 antagonist before each fraction.<br />
Minimal Risk No change from original guideline. The Update Committee<br />
suggests that treatment be administered on an as-needed<br />
basis only. Dopamine or serotonin receptor antagonists are<br />
advised. Antiemetics should be continued prophylactically for<br />
each remaining radiation treatment day.<br />
Combined Chemotherapy and<br />
Radiation Therapy<br />
No change. Patients should receive rescue therapy with a<br />
dopamine- receptor antagonists or a 5-HT 3 receptor<br />
antagonist. Antiemetics should be continued prophylactically<br />
for each remaining radiation treatment day.<br />
Clinicians should:<br />
(1) Re-evaluate emetic risk, disease status, concurrent<br />
illnesses, and medications;<br />
(2) Ascertain that the best regimen is being administered for<br />
the emetic risk;<br />
(3) Consider adding lorazepam or alprazolam to the<br />
regimen; and<br />
(4) Consider adding olanzapine to the regimen or substituting<br />
high-dose intravenous metoclopramide for the 5-HT3 receptor antagonist or adding a dopamine antagonist to<br />
the regimen.<br />
Use <strong>of</strong> the most active antiemetic regimens appropriate for the<br />
chemotherapy being administered to prevent acute or delayed<br />
emesis is suggested. Such regimens should be used with initial<br />
chemotherapy, rather than assessing the patient’s emetic<br />
response with less effective treatment. If anticipatory emesis<br />
occurs, behavioral therapy with systematic desensitization is<br />
effective and suggested. No change since the original<br />
guideline.<br />
Based on extrapolation <strong>of</strong> evidence, the Update Committee<br />
recommends that all patients should receive a 5-HT3 receptor<br />
antagonist before each fraction and for at least 24 h after<br />
completion <strong>of</strong> radiotherapy. Patients should also receive a<br />
five-day course <strong>of</strong> dexamethasone before fractions 1–5.<br />
The Update Committee recommends that patients receive a 5-<br />
HT3 receptor antagonist before each fraction for the entire<br />
course <strong>of</strong> radiotherapy. Patients may be <strong>of</strong>fered a short<br />
course (fractions 1–5) <strong>of</strong> dexamethasone before treatment.<br />
The Update Committee recommends a 5-HT3 receptor antagonist<br />
alone as either prophylaxis or rescue. For patients who<br />
experience RINV while receiving rescue therapy only,<br />
prophylactic treatment should continue until radiotherapy is<br />
complete.<br />
Patients should receive rescue therapy with either a dopamine<br />
receptor antagonist or a 5-HT3 receptor antagonist.<br />
Prophylactic antiemetics should continue throughout radiation<br />
treatment if a patient experiences RINV while receiving rescue<br />
therapy.<br />
Patients should receive antiemetic prophylaxis according to the<br />
emetogenicity <strong>of</strong> chemotherapy, unless the emetic risk with the<br />
planned radiotherapy is higher. No change from the original<br />
guideline.<br />
guideline. The Update Committee suggests, based on limited<br />
data, that patients receiving 5-day cisplatin regimens be<br />
treated with a 5-HT 3 receptor antagonist in combination<br />
with dexamethasone and aprepitant.<br />
<strong>Clinical</strong> Question 11. What is the optimal antiemetic<br />
regimen for patients who experience nausea and vomiting<br />
secondary to cancer therapy despite optimal prophylaxis?<br />
Recommendation 11. Language from the 2006 guideline<br />
was reformatted for clarity. Clinicians should:<br />
(1) Re-evaluate emetic risk, disease status, concurrent<br />
illnesses, and medications<br />
(2) Ascertain that the best regimen Is being administered<br />
for the emetic risk<br />
(3) Consider adding lorazepam or alprazolam to the regimen<br />
(4) Consider adding olanzapine to the regimen or substituting<br />
high-dose intravenous metoclopramide for the 5-HT 3<br />
receptor antagonist or adding a dopamine antagonist to the<br />
regimen.<br />
535
Table 2. Emetic Risk <strong>of</strong> Intravenous Antineoplastic Agents*<br />
High � 90% Carmustine Dactinomycin<br />
Cisplatin Mechlorethamine<br />
Cyclophosphamide �1500 mg/m 2 Streptozotocin<br />
Dacarbazine<br />
Moderate Azacitidine Daunorubicin**<br />
Alemtuzumab Doxorubicin**<br />
Bendamustine Epirubicin**<br />
Carboplatin Idarubicin**<br />
Cl<strong>of</strong>arabine Ifosfamide<br />
Cyclophosphamide �1500 mg/m 2 Irinotecan<br />
Cytarabine �1000 mg/m 2 Oxaliplatin<br />
Low 5-Fluorouracil Methotrexate<br />
Bortezomib Mitomycin<br />
Carbezetaxel Mitoxantrone<br />
Catumaxumab Paclitaxel<br />
Cytarabine �1000 mg/m 2 Panitumumab<br />
Docetaxel Pemetrexed<br />
Doxorubicin HCL liposome injection Temsirolimus<br />
Etoposide Topotecan<br />
Gemcitabine Trastuzumab<br />
Ixabepilone<br />
Minimal 2-Chlorodeoxyadenosine Pralatrexate<br />
Bevacizumab Rituximab<br />
Bleomycin Vinblastine<br />
Busulfan Vincristine<br />
Cetuximab Vinorelbine<br />
Fludarabine<br />
* List is not exhaustive.<br />
** These anthracyclines, when combined with cyclophosphamide, are now<br />
designated as high emetic risk.<br />
<strong>Clinical</strong> Question 12. What treatment options are available<br />
for patients who experience anticipatory nausea and<br />
vomiting?<br />
Recommendation 12. Use <strong>of</strong> the most active antiemetic<br />
regimens appropriate for the chemotherapy being administered<br />
to prevent acute or delayed emesis is suggested. Such<br />
regimens should be used with initial chemotherapy, rather<br />
than assessing the patient’s emetic response with less effective<br />
treatment. If anticipatory emesis occurs, behavioral<br />
therapy with systematic desensitization is effective and<br />
suggested. No change since the original guideline.<br />
Radiation-Induced Nausea and Vomiting (RINV)<br />
This guideline update includes an updated risk stratification<br />
table according to site <strong>of</strong> radiation treatment. MASCC<br />
updated the radiation therapy emetic risk table at the<br />
MASCC/ESMO 2009 consensus conference (Table 4) and it<br />
was adopted by ASCO for this guideline update. 40 Dosing<br />
schedules, according to risk level, are detailed in Table 5.<br />
<strong>Clinical</strong> Question 13. What is the optimal prophylaxis for<br />
nausea and vomiting caused by high emetic risk radiation<br />
therapy?<br />
Recommendation 13. Based on extrapolation <strong>of</strong> indirect<br />
evidence, the Update Committee recommends that all patients<br />
should receive a 5-HT 3 receptor antagonist before<br />
each fraction and for at least 24 hours after completion <strong>of</strong><br />
radiotherapy. Patients should also receive a 5-day course <strong>of</strong><br />
dexamethasone during fractions 1 to 5.<br />
<strong>Clinical</strong> Question 14. What is the optimal prophylaxis for<br />
nausea and vomiting caused by moderate emetic risk radiation<br />
therapy?<br />
Recommendation 14. The Update Committee recommends<br />
that patients receive a 5-HT 3 receptor antagonist<br />
536<br />
Table 3. Antiemetic Dosing by Chemotherapy Risk Category*<br />
Day <strong>of</strong> Chemotherapy Subsequent Days<br />
High Emetic Risk (HEC) a<br />
NK1 Antagonist<br />
Aprepitant 125 mg oral 80 mg oral; days<br />
2 and 3<br />
Fosaprepitant 150 mg IV<br />
5-HT3 Receptor Antagonist<br />
Granisetron 2mgoralOR<br />
1 mg or 0.01 mg/kg IV<br />
Ondansetron 8 mg oral twice daily OR<br />
8 mg or 0.15 mg/Kg IV<br />
Palonosetron 0.50 mg oral OR<br />
0.25 mg IV<br />
Dolasetron 100 mg oral only<br />
Tropisetron 5mgoralOR5mgIV<br />
Ramosetron 0.3mgIV<br />
Corticosteroid b<br />
Dexamethasone<br />
Moderate Emetic Risk (MEC)<br />
12 mg oral or IV 8 mg oral or IV; days<br />
2–3 or days 2–4<br />
c<br />
5-HT3 Receptor Antagonist<br />
Palonosetron<br />
Corticosteroid<br />
0.25 mg IV OR<br />
0.50 mg oral<br />
Dexamethasone<br />
Low Emetic Risk<br />
Corticosteroid<br />
8mgoralorIV 8 mg; days 2 and 3<br />
Dexamethasone 8mgoralorIV<br />
* For patients receiving multiday chemotherapy, clinicians must first determine<br />
the emetic risk <strong>of</strong> the agent(s) included in the regimen. Patients should receive<br />
the agent <strong>of</strong> the highest therapeutic index daily during chemotherapy and for two<br />
days thereafter. Patients can also be <strong>of</strong>fered the granisetron transdermal patch<br />
(Sancuso) that delivers therapy over multiple days rather than taking a serotonin<br />
antagonist daily.<br />
a Includes AC (combination <strong>of</strong> anthracycline and cyclophosphamide).<br />
b The dexamethasone dose is for patients who are receiving the recommended<br />
three-drug regimen for highly emetic chemotherapy. If patients do not receive<br />
aprepitant, the dexamethasone dose should be adjusted to 20 mg on day 1 and<br />
16 mg on days 2–4.<br />
c Clinicians who choose to utilize an NK1 antagonist should follow HEC dosing.<br />
Importantly, corticosteroid is only given on day one; dexamethasone dose is<br />
12 mg.<br />
before each fraction for the entire course <strong>of</strong> radiotherapy.<br />
Patients may be <strong>of</strong>fered a short course <strong>of</strong> dexamethasone<br />
during fractions 1 to 5.<br />
<strong>Clinical</strong> Question 15. What is the optimal treatment to<br />
manage nausea and vomiting associated with low emetic<br />
risk radiation therapy?<br />
Recommendation 15. The Update Committee recommends<br />
a 5-HT 3 receptor antagonist alone as either prophylaxis<br />
or rescue. For patients who experience RINV while<br />
Table 4. Emetic Risk by Site <strong>of</strong> Radiation Therapy 40<br />
High Total Body Irradiation<br />
Total Nodal Irradiation<br />
Moderate Upper Abdomen<br />
Upper Body Irradiation<br />
Half Body Irradiation<br />
Low Cranium<br />
Craniospinal<br />
Head and Neck<br />
Lower Thorax Region<br />
Pelvis<br />
Minimal Extremities<br />
Breast<br />
BASCH ET AL
ANTIEMETICS: ASCO GUIDELINE UPDATE<br />
High Emetic Risk<br />
5-HT3 Receptor Antagonist<br />
Granisetrona 2 mg oral<br />
1 mg or 0.01 mg/kg IV<br />
receiving rescue therapy only, prophylactic treatment<br />
should continue until radiotherapy is complete.<br />
<strong>Clinical</strong> Question 16. What is the optimal treatment to<br />
manage nausea and vomiting associated with minimal<br />
emetic risk radiation therapy?<br />
Recommendation 16. Patients should receive rescue therapy<br />
with either a dopamine receptor antagonist or a 5-HT 3<br />
receptor antagonist. Prophylactic antiemetics should continue<br />
throughout radiation treatment if a patient experiences<br />
RINV while receiving rescue therapy.<br />
<strong>Clinical</strong> Question 17. What is the optimal treatment to<br />
manage nausea and vomiting during concurrent radiation<br />
and chemotherapy?<br />
Recommendation 17. Patients should receive antiemetic<br />
prophylaxis according to the emetogenicity <strong>of</strong> chemotherapy,<br />
unless the emetic risk with the planned radiotherapy is<br />
higher. No change from the original guideline.<br />
Drug Formulations, Agent Dosing<br />
A study published in 2007 compared an orally disintegrating<br />
tablet (ODT) <strong>of</strong> ondansetron with a standard tablet<br />
(Data Supplement). 22 No differences were reported in emesis<br />
or nausea control between the two agents. Notably, it is<br />
not clear whether this study was designed as a nonequivalence<br />
trial a priori. The ODT formulation is an acceptable<br />
alternative to the standard ondansetron tablet.<br />
Two antiemetic agents received regulatory approval in<br />
alternative formulations since the 2006 update. Granisetron<br />
Table 5. Antiemetic Dosing by Radiation Risk Category<br />
Dose Schedule<br />
5-HT 3 antagonist before each fraction throughout XRT. Continue for at<br />
least 24 h following completion <strong>of</strong> XRT.<br />
Ondansetrona 8 mg oral twice daily<br />
8 mg or 0.15 mg/kg IV<br />
Palonosetronb 0.50 mg oral<br />
0.25 mg IV<br />
Dolasetron 100 mg oral only<br />
Tropisetron<br />
Corticosteroid<br />
5 mg oral or IV<br />
Dexamethasone<br />
Moderate Emetic Risk<br />
5-HT3 Receptor Antagonist<br />
4 mg oral or IV Before fractions 1–5<br />
Any <strong>of</strong> the above listed agents are acceptable, note preferred options b Corticosteroid<br />
5-HT3 antagonist before each fraction throughout XRT<br />
Dexamethasone<br />
Low Emetic Risk<br />
5-HT3 Receptor Antagonist<br />
4 mg IV or oral Before fractions 1–5<br />
Any <strong>of</strong> the above listed agents are acceptable, note preferred options<br />
Minimal Emetic Risk<br />
5-HT3 Receptor Antagonist<br />
5-HT3 either as rescue or prophylaxis. If rescue is utilized, then<br />
prophylactic therapy should be given until the end <strong>of</strong> XRT.<br />
Any <strong>of</strong> the above listed agents are acceptable, note preferred options Patients should be <strong>of</strong>fered either class as rescue therapy. If rescue is<br />
Dopamine Receptor Antagonist<br />
utilized, then prophylactic therapy should be given until the end <strong>of</strong> XRT.<br />
Metoclopramide 20 mg oral<br />
Prochlorperazine 10 oral or IV<br />
Abbreviations: IV, intravenous; XRT, radiation therapy; bid, twice daily; qid, four times daily; q, every; h, hours;<br />
a<br />
Preferred agents.<br />
b<br />
No data are currently available on the appropriate dosing frequency with palonosetron in this setting. The Update Committee suggests dosing every second or third<br />
day may be appropriate for this agent.<br />
is also available as a transdermal patch that delivers therapy<br />
over 7 days. As described earlier, this is an option for<br />
patients receiving multiday, high-risk chemotherapy regimens.<br />
47 The panel also suggests that the granisetron patch<br />
may be useful for patients undergoing high- or moderaterisk<br />
radiation.<br />
Oral palonosetron was approved by the U.S. Food and<br />
Drug Administration (FDA) in 2008. 48 Data detailing antiemetic<br />
similarity <strong>of</strong> the oral and intravenous formulations<br />
and agent safety was presented at the 2007 European<br />
Conference <strong>of</strong> <strong>Clinical</strong> <strong>Oncology</strong> meeting. 18 This trial also<br />
supported the 0.50 mg dose.<br />
Three studies assessed dosing <strong>of</strong> intravenous palonosetron.<br />
A meta-analysis <strong>of</strong> eight studies suggested similar<br />
outcomes (Data Supplement) with respect to complete response<br />
among patients treated with either 0.25 mg or<br />
0.75 mg doses. 20 The other two trials reported findings that<br />
a dose <strong>of</strong> 0.075 mg is clearly inferior to both 0.25 mg and<br />
0.75 mg. 19,21<br />
Patient and Clinician Communication<br />
Clinicians are encouraged to provide patients with a<br />
prescription for a rescue antiemetic therapy before the<br />
patient leaves the treatment facility on the first day <strong>of</strong><br />
treatment. Data suggest that physicians frequently underestimate<br />
rates <strong>of</strong> nausea and vomiting secondary to radiation<br />
therapy and chemotherapy. 49<br />
In order to ensure optimal symptom management, clini-<br />
537
cians should assess symptoms throughout the course <strong>of</strong><br />
therapy. Clinicians and clinical researchers should consider<br />
collecting direct reports <strong>of</strong> symptom presence and severity<br />
by patients with a checklist. Patient response to treatment<br />
may change over time, thus requiring ongoing assessments<br />
and modification to antiemetic strategies. For example, the<br />
National Cancer Institute has developed a Patient-Reported<br />
Outcomes version <strong>of</strong> its Common Terminology Criteria for<br />
Adverse Events (PRO-CTCAE), which includes two items to<br />
assess nausea 50 :<br />
1) In the last 7 days, how OFTEN did you have NAUSEA<br />
(Never/Rarely/Occasionally/Frequently/Almost constantly)<br />
2) In the last 7 days, what was the SEVERITY <strong>of</strong> your<br />
NAUSEA at its WORST<br />
(None/Mild/Moderate/Severe/Very severe).<br />
Clinicians and patients are also encouraged to discuss<br />
costs <strong>of</strong> treatment, particularly to assess if cost is prohibitive,<br />
a hardship to patients, or if it may impact treatment<br />
compliance.<br />
Additional Resources<br />
The data supplement, including evidence tables, and clinical<br />
tools and resources can be found at www.asco.org/<br />
guidelines/antiemetics. Patient information is available<br />
there as well and also at www.cancer.net.<br />
Acknowledgments<br />
Thanks to additional Guideline Panel members Petra C.<br />
Feyer, Maurice Chesney, Rebecca Anne Clark-Snow, Anne Ma-<br />
Authors’ Disclosures <strong>of</strong> Potential Conflicts <strong>of</strong> Interest<br />
Employment or<br />
Leadership<br />
Positions<br />
Consultant or<br />
Advisory Role<br />
rie Flaherty, Barbara Freundlich, Gary Morrow, Kamakshi V.<br />
Rao, and Rowena N Schwartz. The Panel also wishes to thank<br />
Dr. Steven Grunberg, Dr. Kristopher Dennis, Dr. Edward Chow,<br />
Dr. Carlo DeAngelis, Dr. Amy Abernethy, Dr. Michael Danso,<br />
Dr. James L. Abbruzzese, Dr. Robert Langdon, the ASCO<br />
<strong>Clinical</strong> Practice Guidelines Committee, and the ASCO Board <strong>of</strong><br />
Directors for their thoughtful reviews <strong>of</strong> guideline drafts. Special<br />
thanks to Shauniece Morris, Pamela B. Mangu, and Sarah<br />
Temin for their assistance data checking and data extraction.<br />
APPENDIX: Updated Committee Members, 2011<br />
Ethan Basch, MD, Co-Chair Memorial-Sloan Kettering Cancer Center<br />
Gary H. Lyman, MD, Co-Chair Duke University<br />
Paul J. Hesketh, MD,<br />
Steering Committee<br />
Lahey Clinic Medical Center<br />
Mark G. Kris, MD,<br />
Steering Committee<br />
Memorial-Sloan Kettering Cancer Center<br />
Maurice Chesney Patient Representative<br />
Rebecca Anne Clark-Snow, RN Lawrence Memorial Hospital<br />
<strong>Oncology</strong> Center<br />
Petra C. Feyer MD Vivantes Clinic <strong>of</strong> Radiooncology and<br />
Nuclear Medicine<br />
Anne Marie Flaherty, RN Memorial-Sloan Kettering Cancer Center<br />
Barbara Freundlich, BA Patient Representative<br />
Gary Morrow, PhD University <strong>of</strong> Rochester Cancer Center<br />
Kamakshi V. Rao, PharmD, BCOP, CPP University <strong>of</strong> North Carolina Hospital<br />
Rowena N Schwartz, PharmD, BCOP The Johns Hopkins Hospital<br />
Stock<br />
Ownership Honoraria<br />
Research<br />
Funding<br />
Author<br />
Ethan Basch*<br />
Ann Alexis Prestrud*<br />
Paul J. Hesketh Helsinn; Merck;<br />
Tesaro<br />
Mark G. Kris*<br />
Mark R. Somerfield*<br />
Gary H. Lyman Amgen<br />
*No relevant relationships to disclose.<br />
1. Gralla RJ, Osoba D, Kris MG, et al. Recommendations for the use <strong>of</strong><br />
antiemetics: evidence-based, clinical practice guidelines. <strong>American</strong> <strong>Society</strong> <strong>of</strong><br />
<strong>Clinical</strong> <strong>Oncology</strong>. J Clin Oncol. 1999;17:2971-2994.<br />
2. Kris MG, Hesketh PJ, Somerfield MR, et al. <strong>American</strong> <strong>Society</strong> <strong>of</strong> <strong>Clinical</strong><br />
<strong>Oncology</strong> guideline for antiemetics in oncology: Update 2006. J Clin Oncol.<br />
2006;24:2932-2947.<br />
3. Peterson K, McDonagh MS, Carson S, et al. Drug class review: Newer<br />
antiemetics. Update 1. http://www.ohsu.edu/drugeffectiveness/reports/final.<br />
cfm. Accessed September 1, 2011.<br />
4. Grunberg SM, Chua DT, Maru A, et al. Phase III randomized doubleblind<br />
study <strong>of</strong> single-dose fosaprepitant for prevention <strong>of</strong> cisplatin-induced<br />
nausea and vomiting. J Clin Oncol. 2010;28:641s (suppl; abstr 9021).<br />
5. Herrington JD, Jaskiewicz AD, Song J. Randomized, placebo-controlled,<br />
pilot study evaluating aprepitant single dose plus palonosetron and dexamethasone<br />
for the prevention <strong>of</strong> acute and delayed chemotherapy-induced<br />
nausea and vomiting. Cancer. 2008;112:2080-2087.<br />
6. Hoshi E, Takahashi T, Takagi M, et al. Aprepitant prevents<br />
chemotherapy-induced nausea and vomiting in Japanese cancer patients<br />
receiving high-does cisplatin: a multicenter randomized, double-blind, placebocontrolled<br />
study (abstract P-20). Presented at: 20th Anniversary International<br />
MASCC/ISOO Symposium, St Gallen, Switzerland; June 2007;30.<br />
538<br />
REFERENCES<br />
Expert<br />
Testimony<br />
BASCH ET AL<br />
Other<br />
Remuneration<br />
7. Navari R, Gray S, Kerr A. Olanzapine versus aprepitant for the prevention<br />
<strong>of</strong> chemotherapy-induced nausea and vomiting (Cinv): a randomized trial<br />
(abstract 02-010). Presented at: 2010 MASCC/ISOO Symposium, Vancouver,<br />
Canada; June 2010;26.<br />
8. Yeo W, Mo FK, Suen JJ, et al. A randomized study <strong>of</strong> aprepitant,<br />
ondansetron and dexamethasone for chemotherapy-induced nausea and vomiting<br />
in Chinese breast cancer patients receiving moderately emetogenic<br />
chemotherapy. Breast Cancer Res Treat. 2009;113:529-535.<br />
9. Aapro M, Fabi A, Nole F, et al. Double-blind, randomised, controlled<br />
study <strong>of</strong> the efficacy and tolerability <strong>of</strong> palonosetron plus dexamethasone for<br />
1 day with or without dexamethasone on days 2 and 3 in the prevention <strong>of</strong><br />
nausea and vomiting induced by moderately emetogenic chemotherapy. Ann<br />
Oncol. 2010;21:1083-1088.<br />
10. Celio L, Bajetta E, Denaro A, et al. Single-day regimen <strong>of</strong> palonosetron<br />
(PALO) and dexamethasone (DEX) for the prevention <strong>of</strong> emesis associated<br />
with moderately emetogenic chemotherapy (MEC): subgroup analysis from a<br />
randomized phase III trial. J Clin Oncol. 2009;27 (suppl; abstr 9620).<br />
11. Rapoport BL, Jordan K, Boice JA, et al. Aprepitant for the prevention<br />
<strong>of</strong> chemotherapy-induced nausea and vomiting associated with a broad range<br />
<strong>of</strong> moderately emetogenic chemotherapies and tumor types: a randomized,<br />
double-blind study. Support Care Cancer. 2010;18:423-431.
ANTIEMETICS: ASCO GUIDELINE UPDATE<br />
12. Aapro MS, Grunberg SM, Manikhas GM, et al. A phase III, doubleblind,<br />
randomized trial <strong>of</strong> palonosetron compared with ondansetron in preventing<br />
chemotherapy-induced nausea and vomiting following highly<br />
emetogenic chemotherapy. Ann Oncol. 2006;17:1441-1449.<br />
13. Saito M, Aogi K, Sekine I, et al. Palonosetron plus dexamethasone<br />
versus granisetron plus dexamethasone for prevention <strong>of</strong> nausea and vomiting<br />
during chemotherapy: a double-blind, double-dummy, randomised, comparative<br />
phase III trial. Lancet Oncol. 2009;10:115-124.<br />
14. Yu Z, Liu W, Wang L, et al. The efficacy and safety <strong>of</strong> palonosetron<br />
compared with granisetron in preventing highly emetogenic chemotherapyinduced<br />
vomiting in the Chinese cancer patients: a phase II, multicenter,<br />
randomized, double-blind, parallel, comparative clinical trial. Support Care<br />
Cancer. 2009;17:99-102.<br />
15. Ezzo J, Richardson MA, Vickers A, et al. Acupuncture-point stimulation<br />
for chemotherapy-induced nausea or vomiting. Cochrane Database <strong>of</strong><br />
Systematic Reviews, Issue 1, 2010.<br />
16. Ho CL, Su WC, Hsieh RK, et al. A randomized, double-blind, parallel,<br />
comparative study to evaluate the efficacy and safety <strong>of</strong> ramosetron plus<br />
dexamethasone injection for the prevention <strong>of</strong> acute chemotherapy-induced<br />
nausea and vomiting. Jpn J Clin Oncol. 2010;40:294-301.<br />
17. Boccia RV, Gonzalez EF, Pluzanska G, et al. Palonosetron (PALO),<br />
administered orally or intravenously (IV), plus dexamethasone for prevention<br />
<strong>of</strong> chemotherapy-induced nausea and vomiting (CINV). J Clin Oncol. 2008;26<br />
(suppl; abstr 20608).<br />
18. Grunberg S, Voisin D, Zufferli M, et al. Oral palonosetron is as effective<br />
as intravenous palonesetron: A Phase III dose ranging trial in patients<br />
receiving moderately emetogenic chemotherapy. Presented at: 14th European<br />
Conference <strong>of</strong> <strong>Clinical</strong> <strong>Oncology</strong> (ECCO); Barcelona, Spain, September 2007.<br />
19. Maemondo M, Masuda N, Sekine I, et al. A phase II study <strong>of</strong> palonosetron<br />
combined with dexamethasone to prevent nausea and vomiting induced<br />
by highly emetogenic chemotherapy. Ann Oncol. 2009;20:1860-1866.<br />
20. Raftopoulos H, Bria E, Gralla RJ, et al. Is there a preferred dose <strong>of</strong><br />
palonosetron? Results <strong>of</strong> an abstracted data (AD) meta-analysis (MA) <strong>of</strong> all<br />
eight randomized double-blind (RDB) studies and the impact on guideline<br />
considerations. J Clin Oncol. 2009;27 (suppl; abstr 9636).<br />
21. Segawa Y, Aogi K, Inoue K, et al. A phase II dose-ranging study <strong>of</strong><br />
palonosetron in Japanese patients receiving moderately emetogenic chemotherapy,<br />
including anthracycline and cyclophosphamide-based chemotherapy.<br />
Ann Oncol. 2009;20:1874-1880.<br />
22. Pectasides D, Dafni U, Aravantinos G, et al. A randomized trial to<br />
compare the efficacy and safety <strong>of</strong> antiemetic treatment with ondansetron and<br />
ondansetron zydis in patients with breast cancer treated with high-dose<br />
epirubicin. Anticancer Res. 2007;27:4411-4417.<br />
23. Boccia RV, Gordan LN, Clark G, et al. Efficacy and tolerability <strong>of</strong><br />
transdermal granisetron for the control <strong>of</strong> chemotherapy-induced nausea and<br />
vomiting associated with moderately and highly emetogenic multi-day chemotherapy:<br />
a randomized, double-blind, phase III study. Support Care Cancer,<br />
2010;19:1609-1617.<br />
24. Giralt S, Mangan K, Maziarz R, et al. Palonosetron (PALO) for<br />
prevention <strong>of</strong> chemotherapy-induced nausea and vomiting (CINV) in patients<br />
receiving high-dose melphalan prior to stem cell transplant (SCT). J Clin<br />
Oncol. 2008;26 (suppl; abstract 9617).<br />
25. Stiff P, Fox-Geiman M, Kiley K, et al. Aprepitant Vs. Placebo Plus Oral<br />
Ondansetron and Dexamethasone for the Prevention <strong>of</strong> Nausea and Vomiting<br />
Associated with Highly Emetogenic Preparative Regimens Prior to Hematopoietic<br />
Stem Cell Transplantation: A Prospective, Randomized Double-Blind<br />
Phase III Trial. Blood. 2009;114 (suppl; abstr 2267).<br />
26. Walsh T, Morris AK, Holle LM, et al. Granisetron vs ondansetron for<br />
prevention <strong>of</strong> nausea and vomiting in hematopoietic stem cell transplant<br />
patients: results <strong>of</strong> a prospective, double-blind, randomized trial. Bone Marrow<br />
Transplant. 2004;34:963-968.<br />
27. Brames MJ, Johnson EL, Nichols CR, et al. A Double-Blind, Placebo-<br />
Controlled, Crossover Study Evaluating the Oral Neurokinin-1 Receptor<br />
Antagonist Aprepitant in Combination with a 5HT3 Antagonist and Dexamethasone<br />
in Patients with Germ Cell Tumors Undergoing 5-day Cisplatin-<br />
Based Chemotherapy Regimens. A Hoosier <strong>Oncology</strong> Group Study (H.O.G.).<br />
Support Care Cancer. 2010 (suppl 3; abstr 02-004):S67-S220.<br />
28. Herrstedt J, Sigsgaard TC, Nielsen HA, et al. Randomized, doubleblind<br />
trial comparing the antiemetic effect <strong>of</strong> tropisetron plus metopimazine<br />
with tropisetron plus placebo in patients receiving multiple cycles <strong>of</strong> multipleday<br />
cisplatin-based chemotherapy. Support Care Cancer. 2007;15:417-426.<br />
29. Gore L, Chawla S, Petrilli A, et al. Aprepitant in adolescent patients for<br />
prevention <strong>of</strong> chemotherapy-induced nausea and vomiting: A randomized,<br />
double-blind, placebo-controlled study <strong>of</strong> efficacy and tolerability. Pediatr<br />
Blood Cancer. 2009;52:242-247.<br />
30. Sepulveda-Vildosola AC, Betanzos-Cabrera Y, Lastiri GG, et al. Palono-<br />
setron hydrochloride is an effective and safe option to prevent chemotherapyinduced<br />
nausea and vomiting in children. Arch Med Res. 2008;39:601-606.<br />
31. Pillai AK, Sharma KK, Gupta YK, et al. Anti-emetic effect <strong>of</strong> ginger<br />
powder versus placebo as an add-on therapy in children and young adults<br />
receiving high emetogenic chemotherapy. Pediatr Blood Cancer. 2011;56:234-<br />
238.<br />
32. Ryan JL, Heckler C, Dakhil SR, et al. Ginger for chemotherapy-related<br />
nausea in cancer patients: A URCC CCOP randomized, double-blind, placebocontrolled<br />
clinical trial <strong>of</strong> 644 cancer patients. J Clin Oncol. 2009;27 (suppl;<br />
abstr 9511).<br />
33. Tan L, Liu J, Liu X, et al. <strong>Clinical</strong> research <strong>of</strong> olanzapine for prevention<br />
<strong>of</strong> chemotherapy-induced nausea and vomiting. J Exp Clin Cancer Res.<br />
2009;28:131.<br />
34. Zick SM, Ruffin MT, Lee J, et al. Phase II trial <strong>of</strong> encapsulated ginger<br />
as a treatment for chemotherapy-induced nausea and vomiting. Support Care<br />
Cancer. 2009;17:563-572.<br />
35. Fabi A, Ciccarese M, Metro G, et al. Oral ondansetron is highly active<br />
as rescue antiemetic treatment for moderately emetogenic chemotherapy:<br />
results <strong>of</strong> a randomized phase II study. Support Care Cancer. 2008;16:1375-<br />
1380.<br />
36. Lajolo PP, de Camargo B, del Giglio A. Omission <strong>of</strong> day 2 <strong>of</strong> antiemetic<br />
medications is a cost saving strategy for improving chemotherapy-induced<br />
nausea and vomiting control: Results <strong>of</strong> a randomized phase III trial. Am J<br />
Clin Oncol. 2009;32:23-26.<br />
37. Wong RK, Paul N, Ding K, et al. 5-hydroxytryptamine-3 receptor<br />
antagonist with or without short-course dexamethasone in the prophylaxis <strong>of</strong><br />
radiation induced emesis: A placebo-controlled randomized trial <strong>of</strong> the National<br />
Cancer Institute <strong>of</strong> Canada <strong>Clinical</strong> Trials Group (SC19). J Clin Oncol.<br />
2006;24:3458-3464.<br />
38. Hesketh PJ, Kris MG, Grunberg SM, et al. Proposal for classifying the<br />
acute emetogenicity <strong>of</strong> cancer chemotherapy. J Clin Oncol. 1997;15:103-109.<br />
39. Roila F, Herrstedt J, Aapro M, et al. Guideline update for MASCC and<br />
ESMO in the prevention <strong>of</strong> chemotherapy- and radiotherapy-induced nausea<br />
and vomiting: results <strong>of</strong> the Perugia consensus conference. Ann Oncol 21<br />
Suppl. 2010;5:v232-243.<br />
40. Gralla RJ, Roila F, Tonato M, et al. MASCC/ESMO Antiemetic Guideline<br />
2010. http://data.memberclicks.com/site/mascc/MASCC_Guidelines_<br />
English_2010.pdf. Accessed September 1, 2011.<br />
41. Grunberg S, Chua D, Maru A, et al. Single-dose fosaprepitant for the<br />
prevention <strong>of</strong> chemotherapy-induced nausea and vomiting associated with<br />
cisplatin therapy: randomized, double-blind study protocol-EASE. J Clin<br />
Oncol. 2011;29:1495-1501.<br />
42. Billio A, Morello E, Clarke MJ. Serotonin receptor antagonists for<br />
highly emetogenic chemotherapy in adults. Art. No. CD006272. DOI: 10.1002/<br />
14651858.CD006272.pub2. Cochrane Database <strong>of</strong> Systematic Reviews, Issue<br />
1, 2010.<br />
43. Jordan K, Hinke A, Grothey A, et al. A meta-analysis comparing the<br />
efficacy <strong>of</strong> four 5-HT3-receptor antagonists for acute chemotherapy-induced<br />
emesis. Support Care Cancer. 2007;15:1023-1033.<br />
44. Prescribing information for EMEND (fosaprepitant dimeglumine)<br />
for injection. http://www.accessdata.fda.gov/drugsatfda_docs/label/2010/<br />
022023s002s003s005lbl.pdf. Accessed September 1, 2011.<br />
45. Celio L, Frustaci S, Denaro A, et al. Palonosetron in combination with<br />
1-day versus 3-day dexamethasone for prevention <strong>of</strong> nausea and vomiting<br />
following moderately emetogenic chemotherapy: a randomized, multicenter,<br />
phase III trial. Support Care Cancer, 2010;19:1217-1225.<br />
46. Phillips RS, Gopaul S. Antiemetic medication for prevention and<br />
treatment <strong>of</strong> chemotherapy induced nausea and vomiting in childhood. Art.<br />
No. CD007786. DOI: 10.1002/14651858.CD007786.pub2. Cochrane Database<br />
<strong>of</strong> Systematic Reviews, Issue 9, 2010.<br />
47. Granisetron Drug Information. http://www.accessdata.fda.gov/drug<br />
satfda_docs/appletter/2008/022198s000ltr.pdf, 2010. Accessed September 1,<br />
2011.<br />
48. Food and Drug Administration Drug Information on Oral Palonosetron<br />
http://www.accessdata.fda.gov/scripts/cder/drugsatfda/index.cfm?fuseaction�<br />
Search.DrugDetails. Accessed September 1, 2011.<br />
49. Basch E. The missing voice <strong>of</strong> patients in drug-safety reporting. N Engl<br />
J Med. 2010;362:865-869.<br />
50. Patient-Reported Outcomes version <strong>of</strong> the CTCAE (PRO-CTCAE).<br />
https://wiki.nci.nih.gov/display/PROCTCAE/Patient-Reported�Outcomes�<br />
version�<strong>of</strong>�the�CTCAE�%28PRO-CTCAE%29. Accessed September 1,<br />
2011.<br />
51. U.S. Cancer Statistics Working Group. United States Cancer Statistics:<br />
1999-2002 Incidence and Mortality Web-based Report http://apps.nccd.cdc.<br />
gov/uscs/ U.S. Department <strong>of</strong> Health and Human Services, Centers for Disease<br />
Control and Prevention and National Cancer Institute. Atlanta, 2005.<br />
539
52. <strong>American</strong> Cancer <strong>Society</strong>. Cancer facts and figures for African <strong>American</strong>s<br />
2005-2006. Atlanta, 2005.<br />
53. Mead H, Cartwright-Smith L, Jones K, et al. Racial and ethnic<br />
disparities in U.S. health care: A chartbook. New York: The Commonwealth<br />
Fund, 2008.<br />
54. Ries LAG, Eisner MP, Kossary CL, et al. SEER cancer statistics review<br />
1973-1997. National Cancer Institute. Bethesda, MD, 2000.<br />
55. Hickok JT, Roscoe JA, Morrow GR, et al. Nausea and emesis remain<br />
540<br />
BASCH ET AL<br />
significant problems <strong>of</strong> chemotherapy despite prophylaxis with 5-hydroxytryptamine-3<br />
antiemetics: A University <strong>of</strong> Rochester James P. Wilmot Cancer<br />
Center Community <strong>Clinical</strong> <strong>Oncology</strong> Program Study <strong>of</strong> 360 cancer patients<br />
treated in the community. Cancer. 2003;97:2880-2886.<br />
56. Guidance for Industry. Patient-Reported Outcome Measures: Use in<br />
Medical Product Development to Support Labeling Claims. http://www.fda.<br />
gov/downloads/Drugs/GuidanceComplianceRegulatoryInformation/Guidances/<br />
UCM193282.pdf. Accessed September 1, 2011.
Chemotherapy-Induced Nausea and Vomiting<br />
Incidence and Prevalence<br />
Overview: Although chemotherapy-induced nausea and vomiting<br />
is recognized as having been an important problem<br />
during the initial introduction <strong>of</strong> chemotherapy into the antineoplastic<br />
armamentarium, the assumption that this problem<br />
has already been solved can restrict optimal management and<br />
further advances. Underestimation <strong>of</strong> nausea and vomiting<br />
may have many causes. If these toxicities are assumed to be<br />
necessary properties <strong>of</strong> chemotherapy, then their incidence<br />
may be taken for granted. If nausea and vomiting appear after<br />
discharge from the clinic several days after chemotherapy,<br />
these toxicities may not be reported because <strong>of</strong> poor recall or<br />
THE JEWISH Passover service 1 contains the story <strong>of</strong> the<br />
four sons—one wise, one evil, one simple, and one who<br />
does not know how to ask a question. In the explanation <strong>of</strong><br />
the story, we learn that the most dangerous son is not the<br />
evil one, but rather the one who does not know how to ask a<br />
question. If we do not know where a lack <strong>of</strong> knowledge lies,<br />
then we cannot begin to address that ignorance or to answer<br />
that question. This is also true in medical science and has<br />
become pertinent in recent years in several areas <strong>of</strong> oncology.<br />
Approximately 80% <strong>of</strong> patients with lung cancer have a<br />
significant smoking exposure, and we have long assumed<br />
that the remainder had a smoking exposure that had not yet<br />
been identified. Now that we realize that there is a different<br />
unanswered and unaddressed question concerning the role<br />
<strong>of</strong> epidermal growth factor receptor mutations 2 and EML4-<br />
ALK translocations, 3 we can begin to address this problem<br />
with appropriate targeted agents. We have also always<br />
known that smoking and drinking are important risk factors<br />
for head and neck cancer, and we assumed that patients<br />
without those risk factors simply had bad luck. Now that we<br />
appreciate the role <strong>of</strong> human papillomavirus infection in<br />
oropharyngeal cancers, 4 we can begin to address the unique<br />
natural history <strong>of</strong> this illness and the possibility <strong>of</strong> identifying<br />
specific remedies.<br />
Although significant progress has been made in the prevention<br />
and control <strong>of</strong> chemotherapy-induced nausea and<br />
vomiting, underappreciation <strong>of</strong> the incidence and prevalence<br />
<strong>of</strong> this problem still limits our effectiveness in addressing its<br />
true magnitude. This misunderstanding begins with the<br />
systems used to grade intrinsic emetogenicity, which generally<br />
divide agents into those that are highly emetogenic,<br />
moderately emetogenic, low emetogenic, or minimally emetogenic.<br />
5 The wide range <strong>of</strong> frequencies <strong>of</strong> emesis within a<br />
given category, particularly moderately emetogenic chemotherapy<br />
(30% to 90% vomiting), can lead to significant<br />
variability in interpreting the emetogenic risk <strong>of</strong> any chemotherapeutic<br />
agent. It must also be remembered that these<br />
grading systems were specifically designed to describe the<br />
acute emetogenicity <strong>of</strong> chemotherapy during the 24 hours<br />
after administration <strong>of</strong> a single intravenous bolus dose<br />
without concurrent administration <strong>of</strong> other chemotherapies<br />
and without the administration <strong>of</strong> antiemetic agents. Numerous<br />
modifying factors are therefore not considered, including<br />
the presence <strong>of</strong> delayed nausea and vomiting (24 to<br />
120 hours after chemotherapy) or anticipatory nausea and<br />
By Steven M. Grunberg, MD<br />
because <strong>of</strong> efforts by patients to avoid unnecessary complaints.<br />
Physician education may be compromised if physicians<br />
see nausea and vomiting as population problems but not<br />
problems for their own patients. Failure to recognize nausea<br />
and vomiting as two distinct entities that may appear independently<br />
<strong>of</strong> each other can also limit understanding <strong>of</strong> the<br />
prevalence <strong>of</strong> these problems and efforts at effective management.<br />
Continued attention to the impact <strong>of</strong> nausea and vomiting<br />
on the patient experience will be necessary to insure<br />
optimal maintenance <strong>of</strong> quality <strong>of</strong> life.<br />
vomiting (before chemotherapy), the possibility <strong>of</strong> extended<br />
infusion or divided doses <strong>of</strong> chemotherapy, or the route <strong>of</strong><br />
administration. A single dose <strong>of</strong> an oral chemotherapy may<br />
be less emetogenic than a single dose <strong>of</strong> an intravenous<br />
formulation, but oral chemotherapy regimens are <strong>of</strong>ten<br />
multiple day–multiple dose regimens and emetogenicity<br />
may gradually increase as the latter days <strong>of</strong> administration<br />
are reached. 5 The emetogenicity <strong>of</strong> combination regimens is<br />
also only rarely considered in such tables.<br />
Of greater importance is the source <strong>of</strong> emetogenicity data<br />
itself. Phase I studies <strong>of</strong> new agents are designed to identify<br />
dose-limiting toxicities (grade 3 and grade 4 toxicities), and<br />
lesser toxicities (such as grade 1 nausea/vomiting) may go<br />
unreported, thereby underestimating the true incidence <strong>of</strong><br />
the problem. Prophylactic use <strong>of</strong> antiemetics may not be<br />
restricted in a phase I study <strong>of</strong> a new cytotoxic chemotherapy,<br />
even if it is not known whether the agent is emetogenic.<br />
This can lead to overestimation <strong>of</strong> the emetogenicity <strong>of</strong> a new<br />
chemotherapy, where administration <strong>of</strong> a prophylactic antiemetic<br />
never was required, or to underestimation <strong>of</strong> the<br />
emetogenicity <strong>of</strong> a new chemotherapy, where prophylactic<br />
administration <strong>of</strong> antiemetic really was partially effective.<br />
Even when the existence <strong>of</strong> the problem is recognized, the<br />
magnitude <strong>of</strong> the problem may not be appreciated, particularly<br />
if pertinent events occur away from direct observation<br />
by medical personnel. Advances in clinical practice have<br />
moved administration <strong>of</strong> chemotherapy for most solid tumor<br />
patients from the hospital to the clinic setting. Although<br />
patients are directly observed on the day <strong>of</strong> treatment, the<br />
following days (delayed period) are <strong>of</strong>ten spent at home. In<br />
an observational study, 6 we asked 24 physicians and nurses<br />
from 14 oncology practices in the United States and Europe<br />
to estimate the percentage <strong>of</strong> patients in their own practices<br />
who were having chemotherapy-induced nausea or vomiting<br />
after highly or moderately emetogenic chemotherapy. We<br />
then surveyed 298 patients from these practices using 5-day<br />
antiemetic diaries to determine the actual frequency <strong>of</strong><br />
From the Division <strong>of</strong> Hematology/<strong>Oncology</strong>, University <strong>of</strong> Vermont, Burlington, VT.<br />
Author’s disclosure <strong>of</strong> potential conflicts <strong>of</strong> interest are found at the end <strong>of</strong> this article.<br />
Address reprint requests to Steven M. Grunberg, MD, Division <strong>of</strong> Hematology/<strong>Oncology</strong>,<br />
University <strong>of</strong> Vermont, 89 Beaumont Avenue - Given Building E214, Burlington, VT 05405;<br />
email: steven.grunberg@uvm.edu<br />
© <strong>2012</strong> by <strong>American</strong> <strong>Society</strong> <strong>of</strong> <strong>Clinical</strong> <strong>Oncology</strong>.<br />
1092-9118/10/1-10<br />
541
acute and delayed nausea or vomiting after administration<br />
<strong>of</strong> chemotherapy. Comparison <strong>of</strong> physician/nurse perception<br />
to patient reality demonstrated that estimations <strong>of</strong> the<br />
incidence <strong>of</strong> acute nausea and vomiting were quite accurate<br />
but that the incidence <strong>of</strong> delayed nausea or vomiting was<br />
markedly underestimated. In fact, the incidence <strong>of</strong> delayed<br />
nausea or vomiting could be twice that estimated by experienced<br />
health pr<strong>of</strong>essionals. Underestimation <strong>of</strong> nausea<br />
was more extreme than underestimation <strong>of</strong> vomiting, and<br />
the problem was particularly apparent in patients receiving<br />
moderately emetogenic chemotherapy (<strong>of</strong>ten patients with<br />
breast cancer receiving a combination <strong>of</strong> cyclophosphamide<br />
and an anthracycline). Studies using similar methodology<br />
have now been repeated in various parts <strong>of</strong> the world,<br />
including Mexico, 7 Venezuela, 8 and Taiwan. 9 Although absolute<br />
values may change (in Taiwan 9 there was an overestimation<br />
<strong>of</strong> the incidence <strong>of</strong> acute vomiting while in<br />
Venezuela 8 delayed vomiting was a greater problem than<br />
delayed nausea), the overall pattern <strong>of</strong> underestimation <strong>of</strong><br />
nausea and vomiting, particularly in the delayed setting<br />
where patients are not being directly observed, remains. We<br />
therefore urge oncology practices to adopt as a standard<br />
operating procedure for administration <strong>of</strong> the first cycle <strong>of</strong> a<br />
new chemotherapy either a single telephone call to the<br />
patient or a single administration <strong>of</strong> a reporting tool such as<br />
the Multinational Association <strong>of</strong> Supportive Care in Cancer<br />
(MASCC) Antiemetic Tool (MAT) (which has been validated<br />
for 3-day recall) several days after administration <strong>of</strong> chemotherapy<br />
10 to more accurately describe problems that should<br />
be addressed during future chemotherapeutic cycles.<br />
Even if the existence <strong>of</strong> greater amounts <strong>of</strong> nausea and<br />
vomiting than expected is accepted intellectually as a problem<br />
for the overall population <strong>of</strong> patients receiving chemotherapy,<br />
failure to accept the problem in terms <strong>of</strong> one’s own<br />
patients may limit adoption <strong>of</strong> appropriate remedies. In a<br />
very interesting study, Mertens and colleagues 11 evaluated<br />
different strategies to encourage greater compliance with<br />
generally accepted antiemetic guidelines in a single hospital.<br />
Distribution <strong>of</strong> written educational materials concerning<br />
the guidelines led to a transient increase in compliance to<br />
these standards, but prescription habits soon returned to<br />
KEY POINTS<br />
● The current incidence <strong>of</strong> chemotherapy-induced nausea<br />
and vomiting, particularly delayed nausea and<br />
vomiting, is underestimated by physicians and<br />
nurses.<br />
● The incidence <strong>of</strong> clinically significant toxicities such<br />
as nausea and vomiting for new agents should be<br />
better defined during the drug development process.<br />
● Real-time recording <strong>of</strong> patient-reported outcomes<br />
through telephone contact or questionnaire is more<br />
accurate than reports <strong>of</strong> toxicities recalled several<br />
weeks after the event.<br />
● Physicians should be aware that patient efforts to<br />
appear strong and cooperative may lead to underreporting<br />
<strong>of</strong> toxicities.<br />
● Nausea and vomiting are separate phenomena that<br />
may require unique remedies.<br />
542<br />
STEVEN M. GRUNBERG<br />
baseline. A single Grand Rounds presentation concerning<br />
antiemetics by a nationally recognized expert speaker had<br />
no effect on antiemetic guideline compliance within the<br />
hospital. The investigators then used diaries to survey<br />
patients receiving chemotherapy concerning their emetic<br />
experience and shared these results with the patients’ own<br />
physicians. When the physicians were able to see persistent<br />
nausea and vomiting not as abstract problems but as events<br />
documented to be happening to their own patients, then a<br />
durable increase in compliance with antiemetic guidelines<br />
as reflected in antiemetic prescribing habits was achieved.<br />
This is consistent with the observations by Stuebe 12 concerning<br />
the value <strong>of</strong> Level I as compared to Level IV evidence<br />
in clinical practice. Physicians learn intellectually from the<br />
randomized double-blind clinical trials that are the basis <strong>of</strong><br />
Level I evidence. However, we remember and react to our<br />
experiences with our own patients, and it is the Level IV<br />
evidence <strong>of</strong> “adverse anecdotes” that is most likely to lead to<br />
a durable improvement in practice. However, even if we are<br />
willing to change practice based on the experience <strong>of</strong> our<br />
patients, we must be able to understand what those experiences<br />
are, and questions <strong>of</strong> communication therefore become<br />
important. Salsman and colleagues 13 interviewed physicians<br />
and patients concerning barriers to effective communication<br />
regarding treatment <strong>of</strong> chemotherapy-induced<br />
nausea and vomiting and found interesting similarities as<br />
well as interesting differences. Both physicians and patients<br />
indicated a desire to minimize the number <strong>of</strong> medications<br />
that were being taken to avoid drug-drug interactions<br />
and side effects. This can be a positive goal as long as it<br />
does not lead to underutilization <strong>of</strong> effective and necessary<br />
remedies. A significant percentage <strong>of</strong> physicians and patients<br />
suggested that the presence <strong>of</strong> nausea and vomiting<br />
is an indication that chemotherapy is working. This is<br />
disturbing since no relationship between the intensity <strong>of</strong><br />
chemotherapy-induced nausea and vomiting and the effectiveness<br />
<strong>of</strong> chemotherapy has ever been demonstrated. Failure<br />
to use available effective antiemetics due to a mistaken<br />
belief that such a course <strong>of</strong> action was preserving therapeutic<br />
efficacy would be a disservice to everyone involved. Of<br />
particular concern was the finding that patients wanted to<br />
“be strong” and not complain to their physicians, while<br />
physicians believed that patients would report serious adverse<br />
events if they occurred. This disconnect in communication<br />
in and <strong>of</strong> itself could lead to undertreatment <strong>of</strong><br />
chemotherapy-induced nausea and vomiting even if potentially<br />
effective agents were available.<br />
An additional concept that may lead to the development <strong>of</strong><br />
more effective antiemetic agents is the realization that<br />
nausea and vomiting are not manifestations <strong>of</strong> the same<br />
phenomenon but rather two separate phenomena. 14 It has<br />
long been assumed that nausea is simply the prodrome <strong>of</strong><br />
vomiting and that agents that relieve vomiting will therefore<br />
certainly relieve nausea as well. This belief is the<br />
justification for the use <strong>of</strong> antiemetic clinical trial endpoints<br />
such as complete response (no vomiting and no use <strong>of</strong> rescue<br />
medication), which do not even mention the term “nausea”<br />
as sufficient for regulatory approval <strong>of</strong> agents not just to<br />
prevent vomiting but rather to prevent “nausea and vomiting.”<br />
In reality, vomiting is an objective endpoint (expulsion<br />
<strong>of</strong> stomach contents), while nausea is a subjective endpoint<br />
that is more difficult to localize or define. Because nausea is
INCIDENCE AND PREVALENCE OF CHEMOTHERAPY-INDUCED NAUSEA AND VOMITING<br />
subjective, it is also much more difficult to identify animal<br />
models that reliably display nausea and that can be quantified<br />
to a sufficient degree to screen new antinausea agents<br />
before clinical testing. It is certainly true that nausea and<br />
vomiting are correlated and that agents that are effective<br />
against one <strong>of</strong> these phenomena are likely to have some<br />
effect against the other. However, it must also be recognized<br />
that neurotransmitter pathways that control vomiting and<br />
nausea are probably not identical and that these differences<br />
may provide guidance to develop new agents specifically<br />
effective against chemotherapy-induced nausea. It is <strong>of</strong><br />
interest that several agents suggested to have significant<br />
activity against chemotherapy-induced nausea (such as<br />
corticosteroids, 15 megestrol acetate, 16 cannabinoids, 17 and<br />
olanzapine 18 ) are not agents that have been highly effective<br />
Author’s Disclosures <strong>of</strong> Potential Conflicts <strong>of</strong> Interest<br />
Author<br />
Employment or<br />
Leadership<br />
Positions<br />
Consultant or<br />
Advisory Role<br />
Steven M. Grunberg A.P. Pharma;<br />
Archimedes;<br />
Helsinn; Merck;<br />
Tesaro<br />
1. Goldberg N. Passover Haggadah. Hoboken NJ: Ktav Publishing House;<br />
1993.<br />
2. Sequist LV, Bell DW, Lynch TJ, et al. Molecular predictors <strong>of</strong> response<br />
to epidermal growth factor receptor antagonists in non-small-cell lung cancer.<br />
J Clin Oncol. 2007;25:587-595.<br />
3. Soda M, Choi YL, Enomoto M, et al. Identification <strong>of</strong> the transforming<br />
EML4-ALK fusion gene in non-small-cell lung cancer. Nature. 2007;448:561-<br />
566.<br />
5. Grunberg SM, Warr D, Gralla RJ, et al. Evaluation <strong>of</strong> new antiemetic<br />
agents and definition <strong>of</strong> antineoplastic agent emetogenicity—state <strong>of</strong> the art.<br />
Support Care Cancer. 2011;19 (suppl 1):S43-S47.<br />
6. Grunberg SM, Deuson RR, Mavros P, et al. Incidence <strong>of</strong> chemotherapyinduced<br />
nausea and emesis after modern antiemetics. Cancer. 2004;100:2261-<br />
2268.<br />
7. Erazo Valle A, Wisniewski T, Figueroa Vadillo JI, et al. Incidence <strong>of</strong><br />
chemotherapy-induced nausea and vomiting in Mexico: healthcare provider<br />
predictions versus observed. Curr Med Res Opin. 2006;22:2403-2410.<br />
8. De Jongh-Garcia C, Poli S, Ananth C, et al. Health care provider<br />
perception <strong>of</strong> nausea and vomiting and patients’ reported incidence: the<br />
Venezuela Emesis Registry. Support Care Cancer. 2005;13:414-415 (abstr).<br />
9. Liau CT, Chu NM, Liu HE, et al. Incidence <strong>of</strong> chemotherapy-induced<br />
nausea and vomiting in Taiwan: physicians’ and nurses’ estimation vs<br />
patients’ reported outcomes. Support Care Cancer. 2005;13:277-286.<br />
10. Molassiotis A, Coventry PA, Stricker CT, et al. Validation and psychometric<br />
assessment <strong>of</strong> a short clinical scale to measure chemotherapy-induced<br />
against chemotherapy-induced vomiting but rather agents<br />
suggested to have activity against cancer anorexia and<br />
cachexia. Perhaps the common ground between nausea and<br />
anorexia will be more promising in leading to increased<br />
understanding <strong>of</strong> this phenomenon than the common ground<br />
between nausea and vomiting.<br />
Although advances in antiemetic care have been based on<br />
identification <strong>of</strong> relevant neurotransmitter/neurotransmitter<br />
receptor pathways and on carefully conducted clinical<br />
trials, 19 it is still necessary to understand the extent <strong>of</strong> the<br />
problem and the nature <strong>of</strong> the problem before the problem<br />
can be fully addressed. Issues <strong>of</strong> communication and reporting<br />
as well as continued efforts to more accurately define<br />
and address the specific problem <strong>of</strong> nausea will therefore be<br />
critical to further advances in this field.<br />
Stock<br />
Ownership Honoraria<br />
Merck Merck<br />
REFERENCES<br />
Research<br />
Funding<br />
Expert<br />
Testimony<br />
Other<br />
Remuneration<br />
nausea and vomiting: the MASCC antiemesis tool. J Pain Symptom Manage.<br />
2007;34:148-159.<br />
11. Mertens WC, Higby DJ, Brown D, et al. Improving the care <strong>of</strong> patients<br />
with regard to chemotherapy-induced nausea and emesis: the effect <strong>of</strong><br />
feedback to clinicians on adherence to antiemetic prescribing guidelines.<br />
J Clin Oncol. 2003;21:1373-1378.<br />
12. Stuebe AM. Level IV evidence—adverse anecdote and clinical practice.<br />
N Engl J Med. 2011;365:8-9.<br />
13. Salsman JM, Grunberg SM, Beaumont JL, et al. Communicating about<br />
chemotherapy-induced nausea and vomiting: a comparison <strong>of</strong> patient and<br />
provider perspectives. J Natl Compr Canc Netw. <strong>2012</strong>;10:149-157.<br />
14. Molassiotis A, Stricker CT, Eaby B, et al. Understanding the concept <strong>of</strong><br />
chemotherapy-related nausea: The patient experience. Eur J Cancer Care.<br />
2008;17:444-453.<br />
15. Parry H, Martin K. Single-dose i.v. dexamethasone—an effective antiemetic<br />
in cancer chemotherapy. Cancer Chemother Pharmacol. 1991;28:231-<br />
232.<br />
16. Loprinzi C, Jatoi A. Antiemetic properties <strong>of</strong> megestrol acetate. J<br />
Palliat Med. 2006;9:239-240.<br />
17. Storr MA, Sharkey KA. The endocannabinoid system and gut-brain<br />
signalling. Curr Opin Pharmacol. 2007;7:575-582.<br />
18. Navari RM, Brenner MC. Treatment <strong>of</strong> cancer-related anorexia with<br />
olanzapine and megestrol acetate: a randomized trial. Support Care Cancer.<br />
2010;18:951-956.<br />
19. Hesketh PJ. Chemotherapy-induced nausea and vomiting. N Engl<br />
J Med. 2008;358:2482-2494.<br />
543
MUCOSAL INJURY IN PATIENTS WITH CANCER:<br />
TARGETING THE BIOLOGY<br />
CHAIR<br />
Douglas E. Peterson, DMD, PhD<br />
University <strong>of</strong> Connecticut Health Center<br />
Farmington, CT<br />
SPEAKERS<br />
Dorothy M. Keefe, MD, MBBS<br />
Royal Adelaide Hospital<br />
Adelaide, Australia<br />
Stephen T. Sonis, DMD, DMSc<br />
Biomodels<br />
Watertown, MA
New Frontiers in Mucositis<br />
By Douglas E. Peterson, DMD, PhD, Dorothy M. Keefe, MD, and<br />
Stephen T. Sonis, DMD, DMSc<br />
Overview: Mucositis is among the most debilitating side<br />
effects <strong>of</strong> radiotherapy, chemotherapy, and targeted anticancer<br />
therapy. Research continues to escalate regarding key<br />
issues such as etiopathology, incidence and severity across<br />
different mucosae, relationships between mucosal and nonmucosal<br />
toxicities, and risk factors. This approach is being<br />
translated into enhanced management strategies. Recent<br />
technology advances provide an important foundation for this<br />
continuum. For example, evolution <strong>of</strong> applied genomics is<br />
fostering development <strong>of</strong> new algorithms to rapidly screen<br />
genomewide single-nucleotide polymorphisms (SNPs) for<br />
patient-associated risk prediction. This modeling will permit<br />
individual tailoring <strong>of</strong> the most effective, least toxic treatment<br />
in the future. The evolution <strong>of</strong> novel cancer therapeutics is<br />
changing the mucositis toxicity pr<strong>of</strong>ile. These agents can be<br />
associated with unique mechanisms <strong>of</strong> mucosal damage.<br />
Additional research is needed to optimally manage toxicity<br />
UCOSITIS” is an umbrella term covering cancer<br />
“Mtreatment regimen-related damage to mucosal<br />
surfaces. Although mucositis has been clinically observed for<br />
several decades, comprehensive advances in pathobiology,<br />
clinical impact, and molecularly targeted treatment have<br />
chiefly occurred the past 15 years. As a result “mucositis”<br />
was incorporated as a medical subject headings (MeSH)<br />
term by the National Library <strong>of</strong> Medicine in 2006.<br />
Although targeted anticancer therapies can cause mucosal<br />
damage, the actual scope <strong>of</strong> injury is not identical to that<br />
caused by conventional cancer therapies. While mucositis<br />
is most <strong>of</strong>ten considered in relation to the oral cavity and<br />
digestive tract, other mucosae such as respiratory and<br />
genitourinary tract can be affected. The degree <strong>of</strong> damage is<br />
not uniform, with oral mucositis typically most prominent<br />
with systemic cancer treatments. Conjunctivae and vaginal<br />
mucosa appear uniquely resistant to damage for reasons<br />
that are not well delineated.<br />
Incidence and severity <strong>of</strong> mucositis is related to the type <strong>of</strong><br />
treatment administered, the risk factors <strong>of</strong> the host, and the<br />
tumor being treated. 1,2 Chemotherapy can cause mucositis<br />
even in a limited number <strong>of</strong> doses if the agent is sufficiently<br />
mucotoxic, whereas radiation tends to cause local toxicity,<br />
only leading to distant toxicities if a sufficiently high dose is<br />
used. 2 The various targeted agents cause mucositis depending<br />
on cross-reactivity <strong>of</strong> receptors in the mucosa concerned,<br />
or presence <strong>of</strong> their known target in mucosa. 3<br />
Standard-dose chemotherapy causes oral mucositis in up<br />
to 40% <strong>of</strong> patients, whereas high-dose chemotherapy related<br />
to stem cell/bone marrow transplantation causes oral mucositis<br />
in up to 80% <strong>of</strong> patients. 4,5 Radiotherapy for head<br />
and neck cancer causes oral mucositis in virtually 100% <strong>of</strong><br />
patients. 6 Oral mucositis incidence secondary to mTOR<br />
inhibitors is approximately 50%, and incidence <strong>of</strong> diarrhea<br />
from targeted agents is approximately 40%, albeit <strong>of</strong> low<br />
grade. 7,8 Chemotherapy causes diarrhea in up to 40% <strong>of</strong><br />
patients as well; it is also typically low-grade but can be<br />
more frequent and severe in relation to irinotecan. 9<br />
Given the clinical and economic importance <strong>of</strong> the toxicity,<br />
there has been an escalating trajectory <strong>of</strong> basic and clinical<br />
caused by agents such as mammalian target <strong>of</strong> rapamycin<br />
(mTOR) inhibitors and tyrosine kinase inhibitors, without reducing<br />
antitumor effect. There has similarly been heightened<br />
attention across the health pr<strong>of</strong>essions regarding clinical<br />
practice guidelines for mucositis management in the years<br />
following the first published guidelines in 2004. New opportunities<br />
exist to more effectively interface this collective guideline<br />
portfolio by capitalizing upon novel technologies such as<br />
an Internet-based Wiki platform. Substantive progress thus<br />
continues across many domains associated with mucosal<br />
injury in oncology patients. In addition to enhancing oncology<br />
patient care, these advances are being integrated into highimpact<br />
educational and scientific venues including the National<br />
Cancer Institute Physician Data Query (PDQ) portfolio as<br />
well as a new Gordon Research Conference on mucosal health<br />
and disease scheduled for June 2013.<br />
research over the past 15 years. 10 This has in turn fostered<br />
development <strong>of</strong> clinical practice guidelines. These have been<br />
disseminated widely via publications and presentations at<br />
scientific conferences. In addition, clinical management interventions<br />
and their biologic basis have been incorporated<br />
into electronic technologies such as the National Cancer<br />
Institute PDQ website “Oral Complications <strong>of</strong> Chemotherapy<br />
and Head/Neck Radiation.” 11 A comprehensive update<br />
<strong>of</strong> this website, including oral mucositis pathobiology, prevention,<br />
and treatment, was posted in March 2011. This<br />
material was the fourth <strong>of</strong> 35 most frequently viewed Supportive<br />
Care PDQ websites in the following 3 months (April<br />
to June 2011), with continued high activity among health<br />
pr<strong>of</strong>essionals and patients into <strong>2012</strong>.<br />
Effect on Quality <strong>of</strong> Life (QoL)<br />
Mucositis is but one <strong>of</strong> the toxicities typically associated<br />
with cancer treatment. It rarely occurs alone; study <strong>of</strong> links<br />
between toxicities provides information on shared pathobiology<br />
and opportunities for intervention in multiple toxicities.<br />
12,13 The importance <strong>of</strong> these toxicities lies in the effect<br />
not only on patients’ QoL, but also on their ability to<br />
withstand effective doses <strong>of</strong> treatment. 14 Dose reductions or<br />
delays lead to reduction in efficacy and survival. Most<br />
clinical trials <strong>of</strong> cancer treatment do not have toxicity as a<br />
primary endpoint, making evaluation <strong>of</strong> true incidence difficult.<br />
15 Health pr<strong>of</strong>essionals’ estimates <strong>of</strong> incidence and<br />
severity <strong>of</strong> toxicity may be lower than those reported by<br />
patients themselves. 16<br />
From the Department <strong>of</strong> Oral Health and Diagnostic Sciences, School <strong>of</strong> Dental Medicine,<br />
Neag Comprehensive Cancer Center, University <strong>of</strong> Connecticut Health Center, Farmington,<br />
CT; University <strong>of</strong> Adelaide, Adelaide, Australia; Harvard School <strong>of</strong> Dental Medicine,<br />
Brigham and Women’s Hospital and the Dana-Farber Cancer Institute, and Biomodels,<br />
LLC, Boston, MA.<br />
Authors’ disclosures <strong>of</strong> potential conflicts <strong>of</strong> interest are found at the end <strong>of</strong> this article.<br />
Address reprint requests to Douglas E. Peterson, DMD, PhD, Department <strong>of</strong> Oral Health<br />
and Diagnostic Sciences, School <strong>of</strong> Dental Medicine, University <strong>of</strong> Connecticut Health<br />
Center, 263 Farmington Ave., Farmington, CT 06030-1605; email: peterson@nso.uchc.edu<br />
© <strong>2012</strong> by <strong>American</strong> <strong>Society</strong> <strong>of</strong> <strong>Clinical</strong> <strong>Oncology</strong>.<br />
1092-9118/10/1–10<br />
545
Painful mouth ulcers and gastrointestinal toxicity including<br />
diarrhea can significantly compromise QoL (p � 0.01). 17<br />
For example, a prospective, multicenter study <strong>of</strong> the burden<br />
<strong>of</strong> mucosal toxicity on patients receiving chemotherapy and<br />
radiotherapy for various cancers demonstrated that for each<br />
increasing grade <strong>of</strong> oral mucositis or diarrhea, there is a<br />
corresponding significant reduction in QoL. 18 Given the<br />
importance <strong>of</strong> these and related toxicities there is important<br />
opportunity in the era <strong>of</strong> the Internet and tablet computers<br />
to enhance collection and analysis <strong>of</strong> toxicity data. This<br />
strategy would enable clinicians to better understand toxicity<br />
being experienced by patients and to intervene more<br />
quickly. 19<br />
Symptoms and Signs <strong>of</strong> Mucositis<br />
Symptoms and signs <strong>of</strong> mucositis depend on the mucosa<br />
affected. Oral mucositis results in pain, erythema, and<br />
ulceration, and can lead to infection, malnutrition, and<br />
dehydration; it can also lead to death <strong>of</strong> selected patients. 20<br />
Esophagitis has comparable features. 21 Gastritis can be<br />
accompanied by bleeding. 22 Small and large bowel mucositis<br />
is accompanied by abdominal pain, bloating, and diarrhea;<br />
pelvic mucositis can result in bleeding as well as tenesmus.<br />
23<br />
Typhlitis, a specific manifestation <strong>of</strong> mucositis <strong>of</strong> the<br />
caecum, causes pain, abdominal bloating, and altered bowel<br />
habit—and can present as an “acute abdomen” in the setting<br />
<strong>of</strong> neutropenia and fever. 24 It can be diagnosed by abdominal<br />
imaging, but treatment is typically conservative as<br />
patients are usually too ill to undergo surgery. 24<br />
New Frontiers<br />
Exciting new frontiers are ahead in relation to mucositis.<br />
Selected examples include (1) development <strong>of</strong> molecularly<br />
targeted mucositis therapeutics, (2) genomics, (3) new types<br />
<strong>of</strong> cancer therapies, (4) studies involving nonalimentary<br />
mucosa, and (5) new strategies for mucositis guideline<br />
production and utilization.<br />
KEY POINTS<br />
● Mucositis is a major toxicity <strong>of</strong> cancer treatment and<br />
affects quality <strong>of</strong> life, cancer outcome, and cost <strong>of</strong><br />
care.<br />
● Mucositis rarely occurs in isolation; linkage <strong>of</strong> research<br />
relative to mechanisms and treatments <strong>of</strong><br />
other toxicities is thus strategically important.<br />
● Targeted anticancer therapies cause mucositis by<br />
new mechanisms and provide a fertile area for ongoing<br />
research.<br />
● Applied genomics allows genome-wide risk prediction–tool<br />
development with translation to personalized<br />
cancer medicine.<br />
● Ongoing systematic evaluation <strong>of</strong> the mucositis literature,<br />
development and dissemination <strong>of</strong> guidelines,<br />
and impact assessment <strong>of</strong> the interventions are<br />
key to optimizing clinical and health care economic<br />
outcomes.<br />
546<br />
PETERSON, KEEFE, AND SONIS<br />
New Frontier 1: Development <strong>of</strong> Molecularly Targeted<br />
Mucositis Therapeutics<br />
There are multiple agents at varying stages <strong>of</strong> development<br />
for the prevention or treatment <strong>of</strong> mucosal injury<br />
(Table 1). The collective strategic feature involves targeting<br />
<strong>of</strong> drug mechanism in relation to the molecular model for<br />
mucositis. The drugs cover the spectrum from topical rinses<br />
to injectable agents, and are used for radiotherapy, chemotherapy,<br />
chemoradiation, and more recently, cancertargeted<br />
therapy. The majority are in trial for radiationinduced<br />
mucosal injury, in large part due to the clinical<br />
severity as well as feasibility in the clinical setting.<br />
New Frontier 2: Genomics<br />
Concepts that define underlying mucositis pathobiology<br />
are rapidly evolving. A reductionist view had existed into the<br />
mid-1990s; namely, regimen-related epithelial damage was<br />
solely attributable to clonogenic cell death <strong>of</strong> progenitor<br />
stem cells in the basal layer or villi. Now it is clear that<br />
mucotoxicity represents the culmination <strong>of</strong> a series <strong>of</strong> biologically<br />
complex events that are comprehensively dependent<br />
on the tissue and its environment. 6,25 Evidence exists<br />
to support a role for at least 14 different canonical pathways<br />
that, once activated by radiation or chemotherapy, contribute<br />
to mucosal damage (Fig. 1). 26 In addition and unlike<br />
internal organs, every mucosal surface is exposed to a<br />
unique external environment (i.e., gut flora or saliva), which<br />
in turn potentially modifies tissue response.<br />
The biologic complexity <strong>of</strong> mucosal injury has provided<br />
opportunities for application <strong>of</strong> genomics to at least three<br />
key areas: (1) understanding the biologic basis for the<br />
condition, (2) predicting toxicity risk among various populations,<br />
and (3) identifying responder and nonresponder populations<br />
to mucositis drug therapy. 27 Of note is that this<br />
modeling is also consonant with one <strong>of</strong> the key clinical<br />
cancer advances in 2011 highlighted by Dr. J. Von Roenn,<br />
namely, study <strong>of</strong> a genetic biomarker that predicts taxaneinduced<br />
neuropathy. 28<br />
Differential changes in gene expression following radiation<br />
or chemotherapy were first comprehensively described<br />
in animal models. 29 Sequential differences in levels and<br />
actions <strong>of</strong> expressed genes provided insights into mechanistic<br />
changes that formed the biologic basis for etiopathology<br />
<strong>of</strong> mucositis. These early studies utilized RNA extracted<br />
from targeted tissue in animals, since sequential, frequent<br />
biopsies from oncology patients was and is not generally<br />
feasible. To overcome this barrier to human studies, peripheral<br />
blood was studied as an RNA source in patients treated<br />
for solid tumors. 26 This provided opportunity for study <strong>of</strong><br />
changes in expression over time and correlation with clinical<br />
and surrogate (i.e., cytokine) endpoints. Furthermore,<br />
rather than analyzing individual genes, generation <strong>of</strong><br />
Bayesian networks and their comparison with established<br />
canonical pathways and mechanistic groupings yielded insight<br />
into pathogenesis and potential drug targets for intervention.<br />
Despite these advances, questions persist. For example,<br />
why do some patients develop mucositis and others do not?<br />
Why is it that two individuals with comparable demographics,<br />
tumor classifications, and treatment regimens have such<br />
disparate toxicity?<br />
Partial insights to these challenging questions do exist.<br />
Age, gender, body mass, and other patient characteristics
MUCOSAL INJURY CAUSED BY CANCER THERAPIES<br />
Table 1. Selected Agents in Development for the Prevention/Treatment <strong>of</strong> Mucositis a<br />
Stage <strong>of</strong> Trial Versus<br />
Cancer Treatment Chemotherapy Radiotherapy<br />
have emerged as poor predictors <strong>of</strong> risk. In contrast, and<br />
based on the fact that virtually every mechanistically<br />
important pathway in human biology is genetically controlled,<br />
differences in gene expression are key risk determinants.<br />
For example, oxidative stress and proinflammatory<br />
cytokines have been implicated in development <strong>of</strong> mucositis.<br />
30,31 Sharp increases in reactive oxygen species are noted<br />
almost immediately after exposure to drug or radiation with<br />
a resultant cascade <strong>of</strong> events that cause tissue destruction. 32<br />
Based on this observation, studies directed at genes controlling<br />
the metabolism <strong>of</strong> reactive oxygen species have identified<br />
deletion-specific nucleotide polymorphisms that, when<br />
present in a patient, are associated with increased risk <strong>of</strong><br />
mucositis induced by both chemotherapy 33 and radiation. 34<br />
In addition, both systemic and tissue-associated increases<br />
<strong>of</strong> proinflammatory cytokines (i.e., TNF-�) are associated<br />
with regimen-related mucosal injury. 30,31 Genetically controlled<br />
variability in pro-inflammatory cytokine production<br />
and its association with disease status has been well established.<br />
For example, the presence <strong>of</strong> SNPs associated with<br />
Targeted Cancer<br />
Therapy<br />
Preclinical AMP-18/NX002 AMP-18/NX002<br />
PMX-30063 CBLB502<br />
TXA127 OralX<br />
TZP-201 Transcutaneous electrical nerve stimulation<br />
Phase I Elsiglutide Dexlansoprazole<br />
Glucarpidase Glucarpidase<br />
Oral selenium SGX201<br />
Phase II AG013 ALD518 Doxycycline hyclate<br />
Buprenorphine Camomilla recutita mouthwash<br />
GelClair Clonidine lauriad<br />
Lactobacillus CD 2 lozenges Dexpanthenol mouthwash<br />
LED therapy GelClair<br />
Omegaven Honey mouthwash<br />
Oral impact IZN-6N4 (botanical extracts)<br />
rhEGF LED therapy<br />
rhGM-CSF L-lysine<br />
Sargramostim (GM-CSF) rhGM-CSF<br />
SCV-07 Sargramostim (GM-CSF)<br />
Selenomethionine SCV-07<br />
Selenomethionine<br />
Phase III Amifostine trihydrate Amifostine trihydrate Caphosol<br />
Caphosol Caphosol<br />
Celecoxib Celecoxib<br />
Fosaprepitant Doxepin hydrochloride rinse<br />
Glutamine Fosaprepitant<br />
Low-level laser light therapy Humidification<br />
Palifermin Hyperbaric oxygen<br />
Hyperimmune colostrum<br />
Iseganan hydrochloride<br />
Palifermin<br />
Pilocarpine<br />
rhEGF<br />
Postmarketing Glutamine popsicles Impact enteral nutrition<br />
Impact enteral nutrition Palifermin<br />
Lenograstim<br />
Morphine<br />
MuGard<br />
Palifermin<br />
Epigallocatechin gallate<br />
Abbreviation: LED, light-emitting diode.<br />
a Source: <strong>Clinical</strong>Trials.gov and industry websites.<br />
TNF-� has a significant effect on severe toxicity risk (p �<br />
0.005), including mucositis, in an allogeneic stem cell transplant<br />
population. 35<br />
Despite the importance <strong>of</strong> these lines <strong>of</strong> research, however,<br />
the strategy <strong>of</strong> identifying “the” gene or SNP that<br />
predicts toxicity is limited by the presumption that a single<br />
genetic entity is controlling overall toxicity expression. Further,<br />
although genetically controlled deficits in metabolic<br />
enzymes have been consistently associated with toxicity<br />
risk, the percentage <strong>of</strong> patients manifesting these genetic<br />
alterations is minute compared with the numbers <strong>of</strong> patients<br />
who develop mucositis.<br />
Perhaps a more comprehensive assessment <strong>of</strong> genetically<br />
based risk could thus be biologically valid. An alternative<br />
approach has thus been recently validated. 36 The methodology<br />
is based on a concept that recognizes that a phenotype is<br />
at least as likely to be the consequence <strong>of</strong> a team <strong>of</strong> genes<br />
working cooperatively as it is due to the expression <strong>of</strong> a<br />
single master gene. The analysis employs Bayesian networks<br />
derived from unsupervised and learned networks <strong>of</strong><br />
547
Fig 1. Bayesian network demonstrating overexpressed genes defined by signaling functions. Major roles for TNF-� have been suggested in<br />
pathoetiology <strong>of</strong> radiation-induced toxicities, including mucositis and fatigue. In addition to its potential primary role as a mediator <strong>of</strong> tissue<br />
injury, the importance <strong>of</strong> TNF-� as a major signaling conduit is supported by its central position in this pathway. Its prominence in more than one<br />
pathway further reinforces its potential importance in radiation-induced toxicities. 26<br />
genes or SNPs. How does this differ from the conventional<br />
approach? First, it is not hypothesis driven. Although this<br />
may sound contrary to conventional wisdom, it actually<br />
provides substantial new potential modeling as we are not<br />
limited by historic knowledge. The unsupervised approach<br />
assumes that we do not know what we do not know.<br />
Consequently and second, there are no threshold expression<br />
values required for a gene or SNP to be included in the<br />
analysis. Third, using aggressive computer programs, algorithms<br />
test and re-evaluate the predictive value <strong>of</strong> groups <strong>of</strong><br />
genes or SNPs to ultimately arrive at the cluster <strong>of</strong> the<br />
greatest predictive power. The defined cluster can then be<br />
validated prospectively.<br />
A totally nonclinical example <strong>of</strong> this modeling is illustrative.<br />
Let’s say you are the new owner <strong>of</strong> the Chicago ASCOs,<br />
a pr<strong>of</strong>essional baseball team with a very limited budget. You<br />
thus decide to take a novel approach to select the players<br />
for your team. You have read Moneyball, 37 so rather than<br />
looking for individuals with excellent batting percentages<br />
548<br />
PETERSON, KEEFE, AND SONIS<br />
you recruit individuals with mediocre individual statistics.<br />
These players, however, have contributed to winning teams<br />
by adding synergies that collectively affect team performance.<br />
You eliminate any requirements for people who want<br />
to play for your team; 1,000 individuals report for tryouts. To<br />
determine the best team, your selection mechanism consists<br />
<strong>of</strong> taking nine random players at a time and having them<br />
play three innings with a team <strong>of</strong> former major leaguers.<br />
You observe the interaction and production <strong>of</strong> the first group<br />
<strong>of</strong> nine candidates. Two seem to work well together and are<br />
effective at getting on base (walks are as good as hits as well<br />
as avoiding making outs), so you keep them on the field and<br />
put the other seven back into the pool. You next randomly<br />
pick seven candidates and repeat the process. You repeat<br />
this process over and over until you have not only evaluated<br />
each candidate, but you have also evaluated every combination<br />
permutation. (You may not have much money, but you<br />
have all the time in the world.) At the end <strong>of</strong> the process, you<br />
produce a team that, taken together, is optimal.
MUCOSAL INJURY CAUSED BY CANCER THERAPIES<br />
Organization (Alphabetical Order) URL<br />
The SNP algorithm follows a similar pattern with a couple<br />
<strong>of</strong> caveats. First, the number <strong>of</strong> SNPs far exceeds 1,000, so<br />
developing the “team” requires billions <strong>of</strong> computergenerated<br />
calculations. Second, we have the ability to compare<br />
predictive accuracy against a clinical observation<br />
(almost like being able to see how each performed in the real<br />
world). Third, we are not limited to only nine players being<br />
on the team. Understanding that nongenetically controlled<br />
factors (e.g., epigenetic pathways, concomitant disease) also<br />
have potential to influence risk imposes an additional challenge,<br />
but the analytic algorithms have the capacity to<br />
overlay and integrate these elements into a risk-prediction<br />
model.<br />
This approach has been recently applied to the practical<br />
clinical problem <strong>of</strong> predicting oral mucositis risk for patients<br />
with hematologic malignancies receiving standard conditioning<br />
regimens in preparation for hematopoietic stem cell<br />
transplant (HSCT). These treatments typically result in<br />
severe mucositis in almost 40% <strong>of</strong> treated patients. 4 Records<br />
<strong>of</strong> approximately 500 patients who had received HSCT were<br />
reviewed, and we identified groups <strong>of</strong> patients who had<br />
developed significant mucositis and patients who had no<br />
stomatotoxicity. With DNA extracted from the saliva <strong>of</strong> each<br />
patient we ran genome-wide SNP arrays. Using a Bayesian<br />
network framework in computer algorithms that were similar<br />
to the recruitment <strong>of</strong> the Chicago ASCOs, we were able<br />
to identify a subset <strong>of</strong> SNPs that predicted mucositis (unpublished<br />
data; ST Sonis).<br />
Applied genomics also provides a platform for the differentiation<br />
<strong>of</strong> patients who respond to a particular drug from<br />
those who do not. Using RNA obtained from subjects before<br />
and after their participation in a clinical trial, a cluster <strong>of</strong><br />
genes was derived that was associated with drug response<br />
as measured clinically by its ability to modify the course <strong>of</strong><br />
chemoradiation-associated oral mucositis. Knowing this information<br />
in a phase II study could be transformative to<br />
subsequent study design. Inclusion criteria could be specifically<br />
adopted to reflect known responder genomics, resulting<br />
in the need for fewer subjects to power the study such<br />
that the trial was efficient, economical, and optimized for a<br />
successful outcome. This application <strong>of</strong> personalized medicine<br />
to determination <strong>of</strong> risk and selection <strong>of</strong> effective<br />
therapies for mucosal toxicities is no longer hypothetical. Its<br />
continued development represents a new frontier that will<br />
continue to expand as new treatments become available. 38<br />
Table 2. Examples <strong>of</strong> Mucositis Guidelines<br />
ASCO http://jco.ascopubs.org/content/27/1/127.full<br />
ESMO http://annonc.oxfordjournals.org/content/22/suppl_6/vi78.full<br />
MASCC/ISOO http://www.mascc.org/mc/page.do?sitePageId�88037<br />
NCCN http://www.nccn.org/JNCCN/PDF/mucositis_2008.pdf<br />
ONS http://www.ons.org/Research/PEP/Mucositis<br />
RTOG http://www.onlinecancereducationforum.com/OCEF/Oral%20mucositis%20in%20head%20and%20neck%20cancer.pdf<br />
Atlantic Provinces Pediatric<br />
Hematology <strong>Oncology</strong> Network<br />
http://www.apphon-rohppa.com/en/guidelines/mucositis-guidelines<br />
Meta-analysis: Cochrane review<br />
(prevention)<br />
http://summaries.cochrane.org/CD000978/interventions-for-preventing-oral-mucositis-for-patients-with-cancer-receiving-treatment<br />
Meta-analysis: Cochrane review<br />
(treatment)<br />
http://summaries.cochrane.org/CD001973/interventions-for-treating-oral-mucositis-for-patients-with-cancer-receiving-treatment<br />
Abbreviations: ASCO, <strong>American</strong> <strong>Society</strong> <strong>of</strong> <strong>Clinical</strong> <strong>Oncology</strong>; ESMO, European <strong>Society</strong> for Medical <strong>Oncology</strong>; MASCC/ISOO, Mucositis Study Group <strong>of</strong> Multinational<br />
Association for Supportive Care in Cancer/International <strong>Society</strong> <strong>of</strong> Oral <strong>Oncology</strong>; NCCN, National Comprehensive Cancer Network; ONS, <strong>Oncology</strong> Nursing <strong>Society</strong>;<br />
RTOG, Radiation Therapy <strong>Oncology</strong> Group.<br />
New Frontier 3: New Types <strong>of</strong> Cancer Therapy<br />
A number <strong>of</strong> targeted anticancer therapies are now in<br />
regular use or on the cusp <strong>of</strong> being clinically available. Still<br />
more are in development. With them come a new battery <strong>of</strong><br />
mucosal toxicities, some unique to particular agents and<br />
others synergistic with other forms <strong>of</strong> conventional therapy.<br />
27 The mucosa has not escaped as a tissue at risk and, in<br />
fact is actually a common site for toxicities. 3<br />
Toxicity data thus continue to accumulate at a rapid rate.<br />
In general, the mechanism underlying the pathogenesis <strong>of</strong><br />
mucosal injury has not been well studied, although it likely<br />
occurs through different pathways than those described for<br />
cytotoxic drugs or radiation. Consequently, clinical course<br />
may vary with respect to onset, duration, cycle-associated<br />
risk, and clinical presentation. Diarrhea is common, as is<br />
exacerbation <strong>of</strong> mouth ulcers.<br />
The mTOR inhibitors illustrate the unique presentation<br />
<strong>of</strong> targeted therapies. Of patients receiving this drug class,<br />
approximately 40% will develop severe oral mucosal lesions.<br />
39 These differ from conventional mucositis in time to<br />
onset, duration, cycle-related dependence, and clinical<br />
presentation. 39-41 Interestingly, concomitant toxicities typically<br />
seen with oral mucositis (other gastrointestinal toxicity)<br />
are uncommon, whereas rash is more frequent. 39-42<br />
These differences are likely attributable to variations in<br />
biologic pathways that contrast between the two types <strong>of</strong><br />
treatment.<br />
As targeted therapies continue to evolve, more aggressive<br />
investigation to understand their cellular etiologies will be<br />
critical to develop effective interventions.<br />
New Frontier 4: Studies Involving Nonalimentary Tract Mucosa<br />
The mucosal surfaces share similarities <strong>of</strong> development<br />
and structure, but have important differences relating to<br />
their individual functions. Normal tissue has a limited<br />
number <strong>of</strong> ways in which it can respond to insult, including<br />
response to cancer therapy. The cancer community can learn<br />
from other experts in mucosal health and disease such as<br />
gastroenterologists, 43 and it is also likely that members <strong>of</strong><br />
the oncology community can provide key insights to the<br />
modeling <strong>of</strong> nonalimentary tract mucosal lesions. For example,<br />
and as noted previously, regimen-related mucosal injury<br />
does not occur with equal frequency across other types <strong>of</strong><br />
mucosa. Why is it that patients receiving levels <strong>of</strong> chemo-<br />
549
therapy that are mucotoxic to the gastrointestinal tract do<br />
not develop injury <strong>of</strong> the conjunctiva or vaginal mucosa?<br />
What characteristics <strong>of</strong> these tissues differentiate them from<br />
other forms <strong>of</strong> mucosal stratified squamous epithelium to<br />
reduce susceptibility to toxicity? Clearly the external environment<br />
is different, although both have a defined micr<strong>of</strong>lora<br />
and a secretory component as noted previously.<br />
To further advance this research agenda, a new Gordon<br />
Research Conference 44 will take place in June 2013. It is<br />
designed to foster new and impactful research directions<br />
regarding mucosal health and disease. Mucosal injury in<br />
patients with cancer will be highlighted in this context.<br />
New Frontier 5: New Strategies for Mucositis Guideline Production<br />
and Utilization (A Proposed Opportunity for ASCO)<br />
The evolution <strong>of</strong> guidelines for mucositis management is<br />
even more recent in origin than is the creation <strong>of</strong> the<br />
contemporary pathobiologic model conceptualized in the<br />
1998. 45 Beginning in 2001, the Mucositis Study Group <strong>of</strong><br />
Multinational Association for Supportive Care in Cancer/<br />
International <strong>Society</strong> <strong>of</strong> Oral <strong>Oncology</strong> (MASCC/ISOO) conducted<br />
an evidence-based review <strong>of</strong> the mucositis literature,<br />
incorporating ASCO-level scoring criteria. The original version<br />
<strong>of</strong> the guidelines was published in 2004 46 and the<br />
revised version in 2007. 47 The modeling has been subsequently<br />
developed by several other health pr<strong>of</strong>essional<br />
groups such that there is currently a relatively robust<br />
portfolio that is largely complementary (Table 2).<br />
However, variability can occur in guideline production<br />
for several reasons including scope <strong>of</strong> literature review as<br />
well as methodology utilized to evaluate the quality <strong>of</strong> the<br />
literature. However, perhaps the ultimate challenge is how<br />
best to navigate the multiple challenges <strong>of</strong> incorporating<br />
guideline-based interventions into clinical practice. As with<br />
all guidelines, these challenges include the following:<br />
● deciding on strategies and then implementing the modeling<br />
to integrate the collective guidelines into a<br />
user-friendly, efficient series <strong>of</strong> suggestions and recommendations<br />
for clinicians;<br />
● customizing the guidelines for country- and/or regionspecific<br />
practices, including language translations and<br />
access to drug and device technology;<br />
● assessing health and economic outcomes in association<br />
with guideline use.<br />
Authors’ Disclosures <strong>of</strong> Potential Conflicts <strong>of</strong> Interest<br />
Author<br />
Employment or<br />
Leadership<br />
Positions<br />
Consultant or<br />
Advisory Role<br />
As noted in Table 2, many academic societies produce<br />
guidelines for the clinical management <strong>of</strong> toxicity. For example,<br />
MASCC/ISOO produces guidelines for alimentary<br />
tract mucositis, the Cochrane collaboration produces a metaanalysis<br />
review <strong>of</strong> oral mucositis, 48,49 and ASCO includes<br />
oral mucositis in its oral health guidelines. The European<br />
<strong>Society</strong> for Medical <strong>Oncology</strong> (ESMO) utilizes a system<br />
based on the MASCC/ISOO guidelines, and NCCN publishes<br />
recommendations as well. Similar guidelines for other toxicities<br />
such as nausea/vomiting or neutropenia are also<br />
produced by these socieities, although there has been one<br />
antiemetic guideline published jointly by MASCC/ASCO/<br />
ESMO. Unfortunately, however, the latest round was once<br />
again separate.<br />
It would seem important and logical for a single combined<br />
effort to review the literature regarding each toxicity, and<br />
move to a more modern methodology such as use <strong>of</strong> a Wiki<br />
platform. 50 Management <strong>of</strong> clinical practice guidelines via<br />
such technology could ultimately improve the standard and<br />
consistency <strong>of</strong> clinical practice according to the best and<br />
most recent scientific evidence available. Unlike written<br />
guidelines, Wiki guidelines are continually updated as new<br />
evidence becomes available; they are also linked to source<br />
abstracts and other evidence-based sites that add to the<br />
value <strong>of</strong> this technology.<br />
A Proposal for ASCO<br />
It thus appears to be an unprecedented time in science<br />
and clinical practice to capitalize on the multiple important<br />
opportunities ahead. The collective broad and deep expertise<br />
<strong>of</strong> the ASCO organization represents an ideal venue to<br />
further delineate, address, and overcome these clinical barriers.<br />
We thus respectfully propose the following strategy for<br />
consideration by ASCO membership.<br />
It is proposed that ASCO’s Patient and Survivor Care<br />
Committee, in partnership with MASCC/ISOO, assume a<br />
leadership role in advancing new opportunities for mucositis<br />
guideline development, dissemination, and measurement <strong>of</strong><br />
impact on clinical and economic outcomes in patients with<br />
cancer.<br />
This model could facilitate discussion across the multiple<br />
pr<strong>of</strong>essional organizations that have produced mucositis<br />
guidelines over the past decade. Creating a coordinated<br />
approach beginning in <strong>2012</strong> could strategically enhance care<br />
for patients with cancer worldwide in the years to come.<br />
Stock<br />
Ownership Honoraria<br />
Douglas E. Peterson Alder; Amgen;<br />
Merck<br />
Dorothy M. Keefe Helsinn Merck KGaA;<br />
Pfizer<br />
Stephen T. Sonis Biomodels (L) ActoGeniX;<br />
Axaxia<br />
Biologicals;<br />
Galera<br />
Therapeutics (U);<br />
Inform<br />
Genomics; Izun<br />
Pharmaceuticals;<br />
Merck; Novartis;<br />
Pfizer; Polymedix<br />
(U); ProCertus;<br />
SciClone<br />
550<br />
Research<br />
Funding<br />
GlaxoSmithKline;<br />
Nestec<br />
PETERSON, KEEFE, AND SONIS<br />
Expert<br />
Testimony<br />
Other<br />
Remuneration
MUCOSAL INJURY CAUSED BY CANCER THERAPIES<br />
1. Logan RM, Stringer AM, Bowen JM, et al. Is the pathobiology <strong>of</strong><br />
chemotherapy-induced alimentary tract mucositis influenced by the type <strong>of</strong><br />
mucotoxic drug administered? Cancer Chemother Pharmacol. 2009;63:239-<br />
251.<br />
2. Raber-Durlacher JE, Elad S, Barasch A. Oral mucositis. Oral Oncol.<br />
2010;46:452-456.<br />
3. Keefe DM, Gibson RJ. Mucosal injury from targeted anti-cancer therapy.<br />
Supportive Care Cancer. 2007;15:483-490.<br />
4. Naidu MU, Ramana GV, Rani PU, et al. Chemotherapy-induced and/or<br />
radiation therapy-induced oral mucositis-complicating the treatment <strong>of</strong> cancer.<br />
Neoplasia. 2004;6:423-431.<br />
5. Peterson DE, Bensadoun RJ, Roila F. Management <strong>of</strong> oral and gastrointestinal<br />
mucositis: ESMO <strong>Clinical</strong> Practice Guidelines. Ann Oncol. 2011;22<br />
(suppl 6):vi78-84.<br />
6. Sonis ST. Oral mucositis. Anti-Cancer Drugs. 2011;22:607-612.<br />
7. Crown JP, Burris HA, 3rd, Boyle F, et al. Pooled analysis <strong>of</strong> diarrhea<br />
events in patients with cancer treated with lapatinib. Breast Cancer Res<br />
Treat. 2008;112:317-325.<br />
8. Porta C, Paglino C, Imarisio I, et al. Safety and treatment patterns <strong>of</strong><br />
multikinase inhibitors in patients with metastatic renal cell carcinoma at a<br />
tertiary oncology center in Italy. BMC Cancer. 2011;11:105. Epub 2011 March<br />
24.<br />
9. Stein A, Voigt W, Jordan K. Chemotherapy-induced diarrhea. Pathophysiology,<br />
frequency and guideline-based management. Therap Adv Med<br />
Oncol. 2010;2:51-63.<br />
10. Peterson DE, Lalla RV. Oral mucositis: The new paradigms. Current<br />
Opin Oncol. 2010;22:318-322.<br />
11. Oral complications <strong>of</strong> chemotherapy and head/neck radiation. NCI PDQ.<br />
<strong>2012</strong>. www.cancer.gov/cancertopics/pdq/supportivecare/oralcomplications/<br />
HealthPr<strong>of</strong>essional. Accessed February 10, <strong>2012</strong>.<br />
12. Aprile G, Ramoni M, Keefe D, et al. Links between regimen-related<br />
toxicities in patients being treated for colorectal cancer. Current Opin Support<br />
Palliat Care. 2009;3:50-54.<br />
13. Nishimura N, Nakano K, Ueda K, et al. Prospective evaluation <strong>of</strong><br />
incidence and severity <strong>of</strong> oral mucositis induced by conventional chemotherapy<br />
in solid tumors and malignant lymphomas. Support Care Cancer. Epub<br />
2011 Nov 25.<br />
14. Lalla RV, Sonis ST, Peterson DE. Management <strong>of</strong> oral mucositis in<br />
patients who have cancer. Dental Clin N Am. 2008;52:61-77.<br />
15. Saad ED. Endpoints in advanced breast cancer: methodological aspects<br />
& clinical implications. Indian J Med Res. 2011;134:413-418.<br />
16. Bateman E, Keefe D. Patient-reported outcomes in supportive care.<br />
Semin Oncol. 2011;38:358-361.<br />
17. Cheng KK, Leung SF, Liang RH, et al. Severe oral mucositis associated<br />
with cancer therapy: impact on oral functional status and quality <strong>of</strong> life.<br />
Support Care Cancer. 2010;18:1477-1485.<br />
18. Elting LS, Keefe DM, Sonis ST, et al. Patient-reported measurements<br />
<strong>of</strong> oral mucositis in head and neck cancer patients treated with radiotherapy<br />
with or without chemotherapy: demonstration <strong>of</strong> increased frequency, severity,<br />
resistance to palliation, and impact on quality <strong>of</strong> life. Cancer. 2008;113:<br />
2704-2713.<br />
19. Abernethy AP, Ahmad A, Zafar SY, et al. Electronic patient-reported<br />
data capture as a foundation <strong>of</strong> rapid learning cancer care. Med Care.<br />
2010;48:S32-8.<br />
20. Barasch A, Epstein JB. Management <strong>of</strong> cancer therapy-induced oral<br />
mucositis. Dermatol Ther. 2011;24:424-431.<br />
21. Rodriguez ML, Martin MM, Padellano LC, et al. Gastrointestinal<br />
toxicity associated to radiation therapy. Clin Transl Oncol. 2010;12:554-561.<br />
22. Patterson WK, Sage RE, Keefe DM. Haemorrhagic gastritis in two<br />
patients treated with all-trans-retinoic acid in acute promyelocytic leukaemia.<br />
Austral New Zealand J Med. 1994;24:314-315.<br />
23. Gibson RJ, Keefe DM. Cancer chemotherapy-induced diarrhoea and<br />
constipation: mechanisms <strong>of</strong> damage and prevention strategies. Support Care<br />
Cancer. 2006;14:890-900.<br />
24. Shvartsbeyn M, Edelman MJ. Pemetrexed-induced typhlitis in nonsmall<br />
cell lung cancer. J Thor Oncol. 2008;3:1188-1190.<br />
25. Bowen JM, Keefe DM. New pathways for alimentary mucositis. J Oncol.<br />
2008; 907892. Epub 2008 Sept 23.<br />
26. Sonis S, Haddad R, Posner M, et al. Gene expression changes in<br />
peripheral blood cells provide insight into the biological mechanisms associated<br />
with regimen-related toxicities in patients being treated for head and<br />
neck cancers. Oral Oncol. 2007;43:289-300.<br />
REFERENCES<br />
27. Keefe DM, Bateman EH. Tumor control versus adverse events with<br />
targeted anticancer therapies. Nature Rev Clin Oncol. 2011;9:98-109.<br />
28. Von Roenn J. Patient and survivor care. <strong>Clinical</strong> Cancer Advances<br />
2011: ASCO’s Annual Report on Progress Against Cancer. 2011;34.<br />
29. Sonis ST, Scherer J, Phelan S, et al. The gene expression sequence <strong>of</strong><br />
radiated mucosa in an animal mucositis model. Cell Prolif. 2002;35 (suppl<br />
1):93-102.<br />
30. Logan RM, Stringer AM, Bowen JM, et al. Serum levels <strong>of</strong> NFkappaB<br />
and pro-inflammatory cytokines following administration <strong>of</strong> mucotoxic drugs.<br />
Cancer Biol Ther. 2008;7:1139-1145.<br />
31. Ong ZY, Gibson RJ, Bowen JM, et al. Pro-inflammatory cytokines play<br />
a key role in the development <strong>of</strong> radiotherapy-induced gastrointestinal<br />
mucositis. Rad Oncol. 2010;5:22. Epub 2010 March 16.<br />
32. Murphy CK, Fey EG, Watkins BA, et al. Efficacy <strong>of</strong> superoxide<br />
dismutase mimetic M40403 in attenuating radiation-induced oral mucositis<br />
in hamsters. Clin Cancer Res. 2008;14:4292-297.<br />
33. Hahn T, Zhelnova E, Sucheston L, et al. A deletion polymorphism in<br />
glutathione-S-transferase mu (GSTM1) and/or theta (GSTT1) is associated<br />
with an increased risk <strong>of</strong> toxicity after autologous blood and marrow transplantation.<br />
Biol Blood Marrow Transpl. 2010;16:801-808.<br />
34. Pratesi N, Mangoni M, Mancini I, et al. Association between single<br />
nucleotide polymorphisms in the XRCC1 and RAD51 genes and clinical<br />
radiosensitivity in head and neck cancer. Radiother Oncol. 2011;99:356-361.<br />
35. Bogunia-Kubik K, Mazur G, Urbanowicz I, et al. Lack <strong>of</strong> association<br />
between the TNF-alpha promoter gene polymorphism and susceptibility to<br />
B-cell chronic lymphocytic leukaemia. Int J Immunogenet. 2006;33:21-24.<br />
36. Sebastiani P, Ramoni MF, Nolan V, et al. Genetic dissection and<br />
prognostic modeling <strong>of</strong> overt stroke in sickle cell anemia. Nat Genet. 2005;37:<br />
435-440.<br />
37. Lewis M. Moneyball: The Art <strong>of</strong> Winning an Unfair Game. New York:<br />
WW Norton and Co; 2003.<br />
38. Sharma R, Tobin P, Clarke SJ. Management <strong>of</strong> chemotherapy-induced<br />
nausea, vomiting, oral mucositis, and diarrhoea. Lancet Oncol. 2005;6:93-102.<br />
39. Sonis S, Treister N, Chawla S, et al. Preliminary characterization <strong>of</strong><br />
oral lesions associated with inhibitors <strong>of</strong> mammalian target <strong>of</strong> rapamycin in<br />
cancer patients. Cancer. 2010;116:210-215.<br />
40. Campistol JM, de Fijter JW, Flechner SM, et al. mTOR inhibitorassociated<br />
dermatologic and mucosal problems. Clin Transpl. 2010;24:149-456.<br />
41. De Masson A, Fouchard N, Mery-Bossard L, et al. Cutaneous and<br />
mucosal aphthosis during temsirolimus therapy for advanced renal cell<br />
carcinoma: review <strong>of</strong> cutaneous and mucosal side effects <strong>of</strong> mTOR inhibitors.<br />
Dermatol. 2011;223:4-8.<br />
42. Gomez-Fernandez C, Garden BC, Wu S, et al. The risk <strong>of</strong> skin rash and<br />
stomatitis with the mammalian target <strong>of</strong> rapamycin inhibitor temsirolimus:<br />
a systematic review <strong>of</strong> the literature and meta-analysis. Eur J Cancer.<br />
<strong>2012</strong>;48:340-346.<br />
43. Kozarek RA. The <strong>Society</strong> for Gastrointestinal Intervention. Are we, as<br />
an organization <strong>of</strong> disparate disciplines, cooperative or competitive? Gut<br />
Liver. 2010;4 (suppl 1):S1-8.<br />
44. Gordon Research Conference. Mucosal Health & Disease 9-14 June<br />
2013. www.grc.org/programs.aspx?year�2013&program�mucosal. Accessed<br />
February 10, <strong>2012</strong>.<br />
45. Sonis ST. Mucositis as a biological process: a new hypothesis for the<br />
development <strong>of</strong> chemotherapy-induced stomatotoxicity. Oral Oncol. 1998;34:<br />
39-43.<br />
46. Rubenstein EB, Peterson DE, Schubert M, et al. <strong>Clinical</strong> practice<br />
guidelines for the prevention and treatment <strong>of</strong> cancer therapy-induced oral<br />
and gastrointestinal mucositis. Cancer. 2004;100:2026-2046.<br />
47. Keefe DM, Schubert MM, Elting LS, et al. Updated clinical practice<br />
guidelines for the prevention and treatment <strong>of</strong> mucositis. Cancer. 2007;109:<br />
820-831.<br />
48. Worthington HV, Clarkson JE, Bryan G, et al. Interventions for<br />
preventing oral mucositis for patients with cancer receiving treatment.<br />
Cochrane Database Syst Rev. 2011;13 Apr (4):CD000978.<br />
49. Clarkson JE, Worthington HV, Furness S, et al. Interventions for<br />
treating oral mucositis for patients with cancer receiving treatment. Cochrane<br />
Database Syst Rev. 2010;4 Aug (8):CD001973.<br />
50. Cancer Council Australia. <strong>Clinical</strong> practice guidelines for the treatment<br />
and management <strong>of</strong> endometrial cancer. wiki.cancer.org.au/australia/Guide<br />
lines:Endometrial_cancer/Treatment/Early_stage/Development_process. Accessed<br />
February 10, <strong>2012</strong>.<br />
551
SHORT- AND LONG-TERM CARDIOVASCULAR<br />
COMPLICATIONS OF CANCER TREATMENT:<br />
OVERVIEW FOR THE PRACTICING ONCOLOGIST<br />
CHAIR<br />
Gretchen Kimmick, MD, MS<br />
Duke University Medical Center<br />
Durham, NC<br />
SPEAKERS<br />
Dawn L. Hershman, MD, MS<br />
Columbia University Medical Center<br />
New York, NY<br />
Marianne Ryberg, MD<br />
Herlev University Hospital<br />
Herlev, Denmark<br />
Doug Sawyer, MD, PhD<br />
Vanderbilt University<br />
Nashville, TN
Short- and Long-term Cardiovascular<br />
Complications <strong>of</strong> Cancer Treatment:<br />
Overview for the Practicing Oncologist<br />
By Chetan Shenoy, MBBS, and Gretchen Kimmick, MD, MS<br />
Overview: As new therapies improve survival from cancer,<br />
attention to comorbid illness and complications <strong>of</strong> therapy—<br />
both short- and long-term—become much more important to<br />
improving not only quality <strong>of</strong> life but also overall survival.<br />
Recognized for its importance as the leading cause <strong>of</strong> death in<br />
the United States, heart disease <strong>of</strong>ten coexists with cancer,<br />
and cancer treatment may increase risk and/or severity. In<br />
AN ESTIMATED 1.6 million new cases <strong>of</strong> invasive<br />
cancer are diagnosed in the United States each year. 1<br />
Modern cancer therapies have improved survival, such that<br />
some cancers that were deemed uncontrollable a few years<br />
ago are now considered “chronic diseases” and some are<br />
cured. In other words, there are more people living with and<br />
living after cancer. In these patients, we have recognized the<br />
importance <strong>of</strong> not only the short-term, but also the long-term<br />
side effects <strong>of</strong> our therapies. For instance, many cancer<br />
survivors may have a higher risk <strong>of</strong> cardiovascular disease<br />
than <strong>of</strong> cancer recurrence; in a study <strong>of</strong> 63,566 women with<br />
breast cancer age 66 and older from the Surveillance,<br />
Epidemiology and End Results (SEER)-Medicare linked<br />
database with 12 years follow-up, survivors were equally<br />
likely to die as a result <strong>of</strong> cardiovascular disease as they<br />
were from breast cancer. 2<br />
In oncology, the complex business <strong>of</strong> preventing cardiovascular<br />
disease, treating predisposing factors for cardiovascular<br />
disease, ongoing management <strong>of</strong> the disease itself, and<br />
minimizing cardiovascular complications <strong>of</strong> cancer therapy<br />
needs multidisciplinary attention—and requires us to be<br />
acutely aware <strong>of</strong> these issues.<br />
Both cardiovascular disease and cancer are more prevalent<br />
with increasing age. An estimated 82.6 million <strong>American</strong><br />
adults (more than one in three) have one or more types<br />
<strong>of</strong> cardiovascular disease—hypertension, coronary artery<br />
disease, heart failure, or cerebrovascular disease. 3 It is,<br />
therefore, not surprising that patients diagnosed with cancer<br />
might also have cardiovascular disease or may be more<br />
vulnerable to cardiovascular complications <strong>of</strong> therapy. In<br />
one report, 38% <strong>of</strong> patients with cancer had hypertension. 4<br />
An analysis <strong>of</strong> the SEER-Medicare database found that<br />
among breast cancer survivors, 1.7% had myocardial infarction,<br />
6.7% had congestive heart failure, 2.6% had peripheral<br />
vascular disease, and 4.3% had cerebrovascular disease at<br />
the time <strong>of</strong> diagnosis. 5 Furthermore, pre-existing cardiovascular<br />
disease, such as hypertension, coronary artery disease,<br />
or cardiomyopathy, is a well-recognized risk factor for cardiovascular<br />
complications from chemotherapy, such as anthracyclines<br />
6,7 and trastuzumab. 8,9 Some <strong>of</strong> our newer<br />
cancer therapies also have specific adverse effects on the<br />
cardiovascular system, regardless <strong>of</strong> age. Cardiotoxicity related<br />
to anticancer treatment, therefore, may have an important<br />
effect on the overall prognosis and survival <strong>of</strong><br />
patients with cancer. 10<br />
Recognized cardiovascular complications <strong>of</strong> cancer treatment<br />
are many. The more commonly recognized include<br />
addition, there are well-recognized cardiovascular toxicities<br />
<strong>of</strong> cancer treatment, including not only cardiomyopathy, but<br />
also hypertension, hypercholesterolemia, and others. Oncologists<br />
and cardiologists are working closely to learn more<br />
about the complex interaction and to improve management<br />
and outcome for patients.<br />
KEY POINTS<br />
● The high prevalence <strong>of</strong> cardiovascular diseases and<br />
cancer requires attention to their interactions and<br />
prevention.<br />
● Many cancer survivors have a higher risk <strong>of</strong> cardiovascular<br />
disease than <strong>of</strong> cancer recurrence.<br />
● Cardiovascular complications <strong>of</strong> cancer treatment<br />
have long been recognized and we are now beginning<br />
to understand the mechanisms and design logical<br />
preventive strategies.<br />
● Preexisting cardiovascular diseases, including hypertension,<br />
coronary artery disease, cardiomyopathy,<br />
and others, are well recognized risk factors for cardiovascular<br />
complications <strong>of</strong> chemotherapy.<br />
heart failure or left ventricular dysfunction, myocardial<br />
ischemia/infarction, hypertension, and thromboembolism.<br />
Less commonly seen are arrhythmias, myocarditis, and<br />
pericarditis. Also important and recently recognized are<br />
changes in cholesterol pr<strong>of</strong>iles, which may increase risk <strong>of</strong><br />
adverse cardiovascular outcomes. These range in severity<br />
and chronicity according to the patient and the agent (type<br />
<strong>of</strong> chemotherapy or radiation) and likely require the attention<br />
<strong>of</strong> our cardiology colleagues.<br />
Management <strong>of</strong> cardiovascular complications <strong>of</strong> cancer<br />
treatment is likely to remain a significant challenge for both<br />
cardiologists and oncologists in the future, as a result <strong>of</strong> the<br />
growing size <strong>of</strong> the aging population <strong>of</strong> patients with cancer<br />
and the introduction <strong>of</strong> new cancer therapies. Identifying<br />
and understanding these effects is therefore crucial to the<br />
successful treatment <strong>of</strong> patients with cancer. Unfortunately,<br />
strong scientific evidence for the treatment <strong>of</strong> patients with<br />
cardiovascular complications from cancer treatment is lack-<br />
From the Divisions <strong>of</strong> Cardiology and Medical <strong>Oncology</strong>, Duke University Medical<br />
Center, Durham, NC.<br />
Authors’ disclosures <strong>of</strong> potential conflicts <strong>of</strong> interest are found at the end <strong>of</strong> this article.<br />
Address reprint requests to Gretchen G. Kimmick, MD, MS, Division <strong>of</strong> Medical <strong>Oncology</strong>,<br />
Duke University Medical Center, Box 3204, Suite 3800 Duke South, Durham, NC 27710;<br />
email: gretchen.kimmick@duke.edu.<br />
© <strong>2012</strong> by <strong>American</strong> <strong>Society</strong> <strong>of</strong> <strong>Clinical</strong> <strong>Oncology</strong>.<br />
1092-9118/10/1-10<br />
553
ing since studies <strong>of</strong> cardiovascular disease have generally<br />
excluded patients with cancer, and studies <strong>of</strong> patients with<br />
Authors’ Disclosures <strong>of</strong> Potential Conflicts <strong>of</strong> Interest<br />
Employment or<br />
Leadership Consultant or Stock<br />
Research<br />
Author<br />
Chetan Shenoy*<br />
Positions Advisory Role Ownership Honoraria Funding<br />
Gretchen Kimmick Wyeth<br />
*No relevant relationships to disclose.<br />
1. Siegel R, Ward E, Brawley O, et al. Cancer statistics, 2011: the impact <strong>of</strong><br />
eliminating socioeconomic and racial disparities on premature cancer deaths.<br />
CA Cancer J Clin. 2011;61:212-236.<br />
2. Patnaik JL, Byers T, DiGuiseppi C, et al. Cardiovascular disease<br />
competes with breast cancer as the leading cause <strong>of</strong> death for older females<br />
diagnosed with breast cancer: a retrospective cohort study. Breast Cancer Res.<br />
2011;13:R64.<br />
3. Roger VL, Go AS, Lloyd-Jones DM, et al. Heart disease and stroke<br />
statistics—2011 update: a report from the <strong>American</strong> Heart Association.<br />
Circulation. 2011;123:e18-e209.<br />
4. Piccirillo JF, Tierney RM, Costas I, et al. Prognostic importance <strong>of</strong><br />
comorbidity in a hospital-based cancer registry. JAMA. 2004;291:2441-2447.<br />
5. Patnaik JL, Byers T, Diguiseppi C, et al. The influence <strong>of</strong> comorbidities<br />
on overall survival among older women diagnosed with breast cancer. J Natl<br />
Cancer Inst. 2011;103:1101-1111.<br />
554<br />
REFERENCES<br />
cancer have generally excluded patients with concomitant<br />
heart disease.<br />
Expert<br />
Testimony<br />
SHENOY AND KIMMICK<br />
Other<br />
Remuneration<br />
6. Doyle JJ, Neugut AI, Jacobson JS, et al. Chemotherapy and cardiotoxicity<br />
in older breast cancer patients: a population-based study. J Clin Oncol.<br />
2005;23:8597-8605.<br />
7. Pinder MC, Duan Z, Goodwin JS, et al. Congestive heart failure in older<br />
women treated with adjuvant anthracycline chemotherapy for breast cancer.<br />
J Clin Oncol. 2007;25:3808-3815.<br />
8. Perez EA, Suman VJ, Davidson NE, et al. Cardiac safety analysis <strong>of</strong><br />
doxorubicin and cyclophosphamide followed by paclitaxel with or without<br />
trastuzumab in the North Central Cancer Treatment Group N9831 adjuvant<br />
breast cancer trial. J Clin Oncol. 2008;26:1231-1238.<br />
9. Guarneri V, Lenihan DJ, Valero V, et al. Long-term cardiac tolerability<br />
<strong>of</strong> trastuzumab in metastatic breast cancer: the M.D. Anderson Cancer<br />
Center experience. J Clin Oncol. 2006;24:4107-4115.<br />
10. Shenoy C, Klem I, Crowley AL, et al. Cardiovascular complications <strong>of</strong><br />
breast cancer therapy in older adults. Oncologist. 2011;16:1138-1143.
Recent Advances in Cardiotoxicity <strong>of</strong><br />
Anticancer Therapies<br />
Overview: The treatment <strong>of</strong> two major diseases in the Western<br />
world, cancer and heart disease, has improved significantly<br />
in recent years. Today, many more cancers are curable<br />
than in previous years. Cancer treatment <strong>of</strong>ten consists <strong>of</strong><br />
chemotherapy, radiation therapy, and now also targeted therapy.<br />
All three types <strong>of</strong> treatment can lead to an increased risk<br />
<strong>of</strong> developing or <strong>of</strong> worsening a pre-existent cardiovascular<br />
disease either during the treatment, immediately afterward,<br />
or several years after cessation <strong>of</strong> therapy. Anthracyclines, a<br />
class <strong>of</strong> drugs that are also known as anthracycline antibiotics,<br />
and the drug cisplatin have contributed to the success<br />
<strong>of</strong> cancer treatment. However, these agents can cause cardiovascular<br />
disease during treatment, and studies have shown<br />
CYTOTOXIC DRUGS, targeted therapy, and radiation<br />
therapy have changed the final outcome for many<br />
patients with cancer, many <strong>of</strong> whom are now cured <strong>of</strong><br />
cancers there were considered fatal just a few years ago.<br />
Instead <strong>of</strong> a death sentence, their diagnosis is comparable to<br />
that <strong>of</strong> a chronic disease with periods <strong>of</strong> remission, exacerbation,<br />
and re-treatment. This means that treating patients<br />
with cancer involves new challenges because modern cancer<br />
treatments can have severe side effects, not only during<br />
treatment but also in the long run. One side effect is the<br />
increased risk <strong>of</strong> cardiovascular disease, either transient or<br />
permanent depending on the nature <strong>of</strong> the injury to the<br />
cardiovascular system. 1-3 Some <strong>of</strong> the types <strong>of</strong> cardiovascular<br />
disease that presumably result from cancer treatment<br />
are heart failure, ischemia, hypertension, hypotension, arrhythmias,<br />
conductive disorder, thromboembolic events, and<br />
QTc prolongation. The risk <strong>of</strong> developing cardiovascular<br />
disease from cancer treatment may be underestimated because<br />
<strong>of</strong> the process drugs undergo to be approved for<br />
general use. To be approved for general use, positive phase<br />
I through phase III trials must be conducted. Patients<br />
participating in these trials are not always representative<br />
<strong>of</strong> the average patient with cancer because patients with<br />
significant comorbidity—including cardiovascular disease,<br />
children, and the elderly—are excluded from the trials. This<br />
means that the risk <strong>of</strong> cardiac disease is apparently greater<br />
than shown in the trials. In recognition <strong>of</strong> these issues,<br />
cooperation between oncologists and cardiologists is crucial.<br />
Anthracyclines<br />
Some <strong>of</strong> the classic cytostatic drugs, such as anthracyclines,<br />
cyclophosphamid, 5-fluorouracil, and cisplatin, are<br />
known to cause cardiovascular disease. Anthracyclines are<br />
one <strong>of</strong> the most feared <strong>of</strong> these treatments because <strong>of</strong> their<br />
ability to cause heart failure, not only in connection with the<br />
treatment but also in the long term. Because they are also<br />
highly active, widely used drugs, many long-term survivors<br />
<strong>of</strong> cancer treatment (including children) have been successfully<br />
treated with anthracycline-based chemotherapy. Anthracylines<br />
remain an indispensable option in treating<br />
cancer today, both for children and adults. Doxorubicin, an<br />
anthracycline and one <strong>of</strong> the most active anticancer drugs<br />
used today, was quickly recognized as having serious side<br />
effects, including heart failure. In a retrospective study <strong>of</strong><br />
By Marianne Ryberg, MD<br />
that the risk <strong>of</strong> disease persists for many years after treatment<br />
stops. Irradiation contributes significantly to this risk when<br />
the cardiovascular system is part <strong>of</strong> the radiation field. If<br />
the targeted therapy also inhibits the genes responsible for<br />
maintaining the function <strong>of</strong> the cardiovascular system, development<br />
<strong>of</strong> cardiovascular symptoms is inevitable. Therefore, it<br />
is essential to have a cardiovascular endpoint in trials with<br />
targeted therapy. When treatment stops, however, the effect<br />
on the cardiovascular system appears to cease, but it is not<br />
known whether the long-term risk <strong>of</strong> developing cardiovascular<br />
disease increases. Combined, these factors indicate that<br />
close cooperation between oncologists and cardiologists is<br />
essential to optimally benefit patients with cancer.<br />
cardiotoxicity, Van H<strong>of</strong>f 4 showed that the risk <strong>of</strong> developing<br />
cardiac failure increased exponentially when the doxorubicin<br />
dose was increased. Another anthracycline, epirubicin,<br />
apparently has a linearly increased risk <strong>of</strong> approximately<br />
40% for each 100 mg/m 2 . 5 This indicates that the cardiac<br />
myocytes will be affected as <strong>of</strong> the very first treatment with<br />
epirubicin. However, several studies show that risk increased<br />
with serious comorbidity, the various ages <strong>of</strong> the<br />
different patients, previous irradiation (including <strong>of</strong> the<br />
heart), and concomitant antitumor therapy. 4-6 Furthermore<br />
the upper limit <strong>of</strong> dose for doxorubicin and epirubicin is<br />
based on retrospective studies in patients with metastatic<br />
diseases. The problem with using this subset <strong>of</strong> patients to<br />
establish upper limits <strong>of</strong> dose is that some <strong>of</strong> these patients<br />
will die from cancer before they develop symptoms <strong>of</strong> heart<br />
failure. Furthermore, a predisposing genetic variant involved<br />
in the oxidative stress or metabolism and transport <strong>of</strong><br />
anthracycline can contribute to heart failure. 7 One paradox<br />
is that an increase <strong>of</strong> the dose <strong>of</strong> the anthracycline appears<br />
to improve the survival rate for those with metastatic<br />
diseases. 5 This calls for a more personalized approach for<br />
treatment <strong>of</strong> patients with metastatic disease using an<br />
anthracycline. Furthermore, the risk <strong>of</strong> developing cardiotoxicity<br />
in the long term after an anthracyline-based adjuvant<br />
treatment is not addressed by the aforementioned<br />
studies.<br />
A prominent feature <strong>of</strong> the risk <strong>of</strong> developing cardiotoxicity<br />
is subcellular damage or even necrosis in the heart<br />
muscle cells during treatment, which seems to continue even<br />
after treatment has stopped. 8 This is believed to be caused<br />
by an increase in oxidative stress in cells. Reactive oxygen<br />
species (ROS) are formed when the quinine moiety <strong>of</strong> anthracyclines<br />
is reduced to semiquinone, thus initiating a<br />
cascade <strong>of</strong> free-radical formation. Furthermore a highly<br />
reactive ROS is created when an anthracycline uncouples<br />
From the Department <strong>of</strong> <strong>Oncology</strong>, Herlev Hospital, University <strong>of</strong> Copenhagen, Denmark.<br />
Author’s disclosure <strong>of</strong> potential conflicts <strong>of</strong> interest are found at the end <strong>of</strong> this article.<br />
Address reprint requests to Marianne Ryberg MD, Department <strong>of</strong> <strong>Oncology</strong>, Herlev<br />
Hospital, University <strong>of</strong> Copenhagen, Herlev Ringvej 75, DK-2730 Herlev, Denmark; email:<br />
mary@heh.regionh.dk.<br />
© <strong>2012</strong> by <strong>American</strong> <strong>Society</strong> <strong>of</strong> <strong>Clinical</strong> <strong>Oncology</strong>.<br />
1092-9118/10/1-10<br />
555
the electron transport chain in the mitochondria. ROS have<br />
a deleterious effect on the cells, thus leading to cell death<br />
and to organ damage. 9,10 A recent experimental study in<br />
rats demonstrates that anthracycline-caused cardiomyopathy<br />
can be mediated by depletion <strong>of</strong> the cardiac stem cell<br />
pool. When exposed to an anthracycline, cardiac progenitor<br />
cells (CPCs) appeared to be even more vulnerable to damaged<br />
ROS. Additionally, CPC cells were arrested in the<br />
G2/M cell cycle, resulting in a depletion <strong>of</strong> the CPC cell pool<br />
in the myocardium. The final result is heart failure in the<br />
animal. To establish whether the delivery <strong>of</strong> syngeneic<br />
contaminating tumor cells (CTCs) could oppose the progression<br />
<strong>of</strong> anthracycline cardiotoxicity, the syngeneic CTCs<br />
were injected in the failing myocardium. This improved the<br />
regeneration <strong>of</strong> cardiac myocytes and <strong>of</strong> vascular structure,<br />
consequently improving ventricular performance and the<br />
rate <strong>of</strong> animal survival. 11 These encouraging results can<br />
perhaps lead to the possibility <strong>of</strong> implementation <strong>of</strong> stem cell<br />
transplantation as part <strong>of</strong> treatment with anthracyclines.<br />
Timing <strong>of</strong> Combination Chemotherapy<br />
The story behind the introduction <strong>of</strong> taxanes in the treatment<br />
<strong>of</strong> patients with breast cancer is a lesson in how<br />
important timing is in order to avoid cardiotoxicity. 12 In an<br />
early phase II study doxorubicin and paclitaxel were given<br />
simultaneously for metastatic breast cancer, resulting in<br />
an impressive response but with approximately one-fifth <strong>of</strong><br />
the patients experiencing cardiotoxicity. 12 Pharmacokinetics<br />
studies show that the presence <strong>of</strong> paclitaxel increases the<br />
plasma level area under the curve (AUC) <strong>of</strong> doxorubicin by<br />
30%, compared with a 24-hour delay in administration <strong>of</strong> a<br />
paclitaxel infusion. A later phase III study that separated<br />
the administration <strong>of</strong> doxorubicin and paclitaxcel by 24<br />
hours did not show an increase in cardiotoxicity. 13 These<br />
important results have been translated to benefit the adjuvant<br />
treatment <strong>of</strong> patients with breast cancer. Part <strong>of</strong> the<br />
adjuvant anthracycline-based treatment has been replaced<br />
with a taxane-based treatment so that the cumulative dose<br />
<strong>of</strong> an anthracycline recommended in adjuvant treatment has<br />
been reduced.<br />
Cisplatin<br />
During treatment, some chemotherapy agents can cause<br />
ischemia syndrome (i.e., chest pains), palpitations, and in<br />
rare cases a lethal myocardial infarction (MI). This can<br />
happen with cisplatin, for example, where the syndrome can<br />
KEY POINTS<br />
● The risk for development <strong>of</strong> cardiac disease is substantial<br />
after treatment with chemotherapy, radiation<br />
therapy, or with targeted therapy.<br />
● In order to avoid cardiotoxicity, timing <strong>of</strong> combination<br />
therapies is essential.<br />
● The cardiac risk in the long term for patients treated<br />
with targeted therapy is not known.<br />
● Preclinical and clinical trials must be designed to<br />
evaluate a cardiac risk.<br />
● Cooperation between oncologists and cardiologists is<br />
crucial.<br />
556<br />
occur not only during treatment but also subsequently. Our<br />
knowledge about the risk <strong>of</strong> cardiac disease after cessation <strong>of</strong><br />
treatment is obtained by studies <strong>of</strong> testicular cancer survivors.<br />
The treatment consisted <strong>of</strong> a cisplatin-based chemotherapy<br />
for patients generally diagnosed between 20 to 40<br />
years <strong>of</strong> age, and nearly all <strong>of</strong> them were cured. In all<br />
likelihood, they will survive for years to come, but they will<br />
have an increased risk <strong>of</strong> developing hypertension, hypercholesterolemia,<br />
myocardial ischemia, and MI. 14 The increased<br />
risk for MI was approximately 5.7% in a median<br />
observation time <strong>of</strong> 19 years (13–19 years) for 990 long-term<br />
survivors treated with cisplatin-based chemotherapy. One<br />
reason for this delayed risk is that the patients had an<br />
increased plasma level <strong>of</strong> endothelia and inflammatory<br />
marker protein years later that might eventually progress to<br />
more severe endothelia dysfunction and overt atherosclerosis.<br />
15 This increase may be because cisplatin can still be<br />
measured in the blood 20 years after treatment. 16<br />
5-Fluorouracil/Capecitabine<br />
5-Fluorouracil and its prodrug, capecitabine, can cause<br />
ischemic syndrome during treatment, but the symptoms<br />
disappear when treatment stops. The mechanism is not well<br />
established, but may be a result <strong>of</strong> coronary vasospasm. It<br />
has also been reported to cause Takotsubo cardiomyopathy.<br />
17 These two drugs are the foundation <strong>of</strong> the adjuvant<br />
treatment <strong>of</strong> colorectal cancer. When needed, the treatment<br />
can be repeated despite the occurrence <strong>of</strong> previous cardiac<br />
events. The treatment must be given in close cooperation<br />
with cardiologists and at lower doses. Capecitabine is also<br />
used in palliative setting for many types <strong>of</strong> cancer, including<br />
breast cancer. The true incidence <strong>of</strong> cardiac events is not<br />
known but is believed to be approximately 1% to 18%. 18,19<br />
Compared with cisplatin, no increased risk <strong>of</strong> cardiotoxicity<br />
has been reported subsequently.<br />
Radiotherapy<br />
MARIANNE RYBERG<br />
Radiotherapy has been and is still an essential modality in<br />
the curative treatment <strong>of</strong> many types <strong>of</strong> cancer. This applies<br />
to both children and adults with cancer. Radiation therapy<br />
can cause coronary artery disease (CAD), diffuse myocardial<br />
fibrosis, and the more seldom-occurring pericarditis and<br />
pericardial fibrosis. The most common side effect, however,<br />
is CAD. A radiation dose <strong>of</strong> greater than 30 GY is definitively<br />
known to cause damage to the heart. However, it is unknown<br />
whether a threshold dose exists that does not incur<br />
risk to the heart. 20 Because <strong>of</strong> a prolonged latency <strong>of</strong> 10 to 15<br />
years before the occurrence <strong>of</strong> cardiac disease, it is difficult<br />
to be certain as to whether the increased incidence is<br />
because <strong>of</strong> the irradiation or whether it is caused by a<br />
comorbidity due to aging. A study <strong>of</strong> 4,122 survivors treated<br />
for cancer in their childhoods has increased understanding<br />
<strong>of</strong> the extent <strong>of</strong> the risk. Compared with the general population,<br />
patients in this study were generally shown to have 5<br />
times greater risk <strong>of</strong> dying <strong>of</strong> cardiac disease. The authors<br />
have retrospectively estimated that the mean radiation dose<br />
delivered to the heart and the risk <strong>of</strong> dying <strong>of</strong> cardiac disease<br />
increased linearly when the average radiation dose exceeded<br />
5 Gy. 21<br />
Patients with breast cancer (BC) who underwent irradiation<br />
to the right breast had a lower risk for MI than those<br />
patients treated for left-sided BC. 20 This was demonstrated
CARDIOTOXICITY OF ANTICANCER THERAPIES<br />
by doing a retrospective estimation <strong>of</strong> the radiation dose to<br />
the heart in patients with BC. Researchers have defined<br />
hotspots for radiation, which included the proximal right<br />
coronary artery, the mid- and distal descending artery, and<br />
the distal diagonal artery. 20 A Swedish group examined the<br />
distribution <strong>of</strong> coronary stenosis in a coronary angiography<br />
in patients treated with adjuvant radiotherapy for BC. The<br />
patients were compared with otherwise healthy woman<br />
referred for a coronary angiography, and the incidence <strong>of</strong><br />
coronary stenosis was similar. The anatomic location <strong>of</strong> the<br />
stenosis was substantially different. In patients who were<br />
treated for left-sided BC, stenosis increased substantially<br />
in the areas deemed to be hotspots. The relative risk for<br />
high-risk compared with low-risk patients (no radiation<br />
therapy) was 1.90 for stenosis, grades 3–5 (95% CI: 1.11 to<br />
3.24). 22<br />
The biologic processes set in motion by radiation therapy<br />
continue after the end <strong>of</strong> therapy, and side effects might not<br />
be visible until decades later. Radiation decreases the capillary<br />
density with the passage <strong>of</strong> time, thus causing a loss <strong>of</strong><br />
the flow reserves territory <strong>of</strong> myocardium in the radiation<br />
portal. This leads to ischemia <strong>of</strong> the myocardium. Radiation<br />
also seems to accelerate the development <strong>of</strong> age-related<br />
atherosclerosis in patients, thus causing stenosis/thrombosis<br />
in the arteries in already weakened areas. This also<br />
explains the additive effect <strong>of</strong> anthracycline-based treatment,<br />
because <strong>of</strong> its ability to weaken the myocardium.<br />
In order to decrease the radiation dose to the heart, use <strong>of</strong><br />
radiation-reducing techniques are essential. The introduction<br />
<strong>of</strong> a three-dimensional computerized planning system<br />
makes dose delivery estimation possible. By defining specific<br />
cardiac substructures, such as arteries and the entire heart,<br />
the dose can be estimated. As a result, a dose relationship<br />
can be established between the risk <strong>of</strong> cardiac disease and<br />
the dose delivered. In the case <strong>of</strong> hotspots, another field<br />
technique can be considered, if possible. A promising new<br />
technique is breath holding, where the dose is delivered<br />
when the heart is out <strong>of</strong> the irradiation field. Modern<br />
techniques such as this one are important because cancer<br />
survival rates are improving, and awareness <strong>of</strong> the longterm<br />
effects <strong>of</strong> irradiation can help decrease risks.<br />
Targeted Therapies<br />
A highly promising new way to treat cancer is targeted<br />
therapy. The main principle is to use drugs designed to<br />
inhibit the dysregulated genes that drive malignant transformation.<br />
Targeted cancer therapies currently available are<br />
monoclonal antibodies (mAbs), which target growth factor<br />
receptors, and other small molecules, which are inhibitors <strong>of</strong><br />
kinases (mostly <strong>of</strong> tyrosine kinase inhibitors, TKIs). Kinases<br />
are enzymes that catalyze the transfer <strong>of</strong> phosphate from<br />
ATP, usually to a serine, threonine, or to tyrosine residue on<br />
protein substrates. These reactions act as critical mediators<br />
<strong>of</strong> cellular signal transduction, and thereby regulate cellular<br />
processes including cell cycle, progression, metabolism,<br />
transcription, and apoptosis. Mutation in their genes contributes<br />
to the malignant transformation in normal cells in<br />
many cancers. Inhibition <strong>of</strong> the mutated genes can also<br />
interfere with the function <strong>of</strong> the genes in a normal cell,<br />
including the function <strong>of</strong> the cells in the cardiovascular<br />
system. 23 Cardiac toxicity caused by targeted therapies can<br />
be divided into on-target and <strong>of</strong>f-target toxicity. On-target<br />
cardiac toxicity means that the inhibition <strong>of</strong> the mutation in<br />
the specific kinase responsible for the malignant transformation<br />
also has a decisive influence on the maintenance <strong>of</strong><br />
normal function in cardiac myocytes. As a result, the risk <strong>of</strong><br />
cardiotoxicity is inevitable when this therapy is used.<br />
An example <strong>of</strong> on-target toxicity involves trastuzumab,<br />
which is a human monoclonal antibody directed against the<br />
extracellular region <strong>of</strong> the HER2/ErbB2 receptor. The normal<br />
HER2 gene is involved in cell proliferation, angiogenesis,<br />
and cell survival. Thus, a deficiency in the HER2 gene<br />
will result in the downstreaming <strong>of</strong> these processes and<br />
cause a malignant transformation <strong>of</strong> the cells. Furthermore,<br />
an overexpression or amplification <strong>of</strong> tyrosine kinase epidermal<br />
growth factor HER2/ErbB2 is associated with an aggressive<br />
clinical phenotype <strong>of</strong> BC with a poor survival rate. 24<br />
The monoclonal antibody trastuzumab was introduced in<br />
the treatment <strong>of</strong> HER2-positive BC in both the adjuvant and<br />
metastatic setting, and it produced impressive results. 25<br />
The preclinical trials showed no sign <strong>of</strong> cardiotoxicity, but it<br />
soon became clear that the drug could cause cardiotoxicity.<br />
Five percent to 7% <strong>of</strong> the patients developed cardiac dysfunction<br />
in monotherapy, and 28% did so when trastuzumab<br />
was given concomitantly with an anthracycline. 24 Trastuzumab<br />
was then given sequentially with a remarkable<br />
reduction <strong>of</strong> the cardiac dysfunction in both the metastatic<br />
and adjuvant settings. The HER2 gene and its ligand<br />
neuregulin are also involved in normal cardiomyocyte proliferation<br />
during development and survival throughout<br />
life. 26 Thus, trastuzumab, by inhibiting HER2, also inhibits<br />
the myocardial survival pathway; its influence on the function<br />
<strong>of</strong> the cardiac myocyte is inevitable. This may result in<br />
cardiac dysfunction and thus increases the risk for heart<br />
failure. Meanwhile, the myocardiac biopsies show that the<br />
myocyte structure is normal without my<strong>of</strong>ibrillar loss or<br />
ultra-structural disturbance, which is in contrast to anthracyline.<br />
24 There are indications that the biologic process that<br />
trastuzumab sets in motion may be reversible. The ability <strong>of</strong><br />
cardiac myocytes to regenerate may decrease when an<br />
anthracycline and trastuzumab are given concomitantly. 13<br />
The HERA trial, which had a 3.6-year follow-up, showed<br />
that when trastuzumab was given after treatment with an<br />
anthracycline, the majority <strong>of</strong> cardiac events occurred during<br />
the treatment. The incidence was found to be 5%. The<br />
cardiac dysfunction improved in 80% <strong>of</strong> these patients, and<br />
only 20% had progressive dysfunction during follow-up. 27 A<br />
meta-analysis <strong>of</strong> 11,882 patients from 10 randomized trials<br />
has been published and showed that the addition <strong>of</strong> trastuzumab<br />
to anthracycline-based chemotherapy significantly<br />
increased the risk <strong>of</strong> cardiac dysfunction (relative risk 4.27;<br />
CI, 2.75–6.61, p � 0.00001). 28 The risk seems to increase<br />
with old age, hypertension, and obesity. In contrast to<br />
anthracycline treatment, treatment with trastuzumab can<br />
be repeated, but data on long-term follow-up is not available<br />
yet.<br />
Antiangiogenese Inhibitors<br />
Antiangiogenese inhibitors are designed to target the<br />
vascular endothelial growth factor (VEGF), which may lead<br />
to an increased risk <strong>of</strong> developing hypertension, cardiotoxicity,<br />
and to the occurrence <strong>of</strong> thomboembolic events. The<br />
most prominent types <strong>of</strong> this kind <strong>of</strong> inhibitor are bevacizumab,<br />
sunitinib, and sorafenib. Bevacizumab is a recombinant<br />
humanized monoclonal antibody that binds VEGF and<br />
prevents it from binding to its receptors (VEGFR-1 and 2).<br />
557
The drug has been approved in treatment <strong>of</strong> many types <strong>of</strong><br />
cancer (e.g., colorectal, lung, and breast cancers). Hypertension<br />
is the adverse event correlated most strongly to bevacizumab,<br />
with an incidence ranging from 16% to 47%. 29 From<br />
the results <strong>of</strong> prospective trials, it seems that higher doses<br />
<strong>of</strong> bevacizumab result in higher incidence <strong>of</strong> hypertension.<br />
However, the exact dose that causes hypertension is not<br />
known. 30 This effect appears to be dose dependent, but the<br />
exact dose that causes hypertension is not known. Given in<br />
combination with sorafenib, the incidence <strong>of</strong> hypertension<br />
was 67%. 31 This is because VEGF/VEGFR signaling affects<br />
the adaptive response <strong>of</strong> the heart to blood pressure stress.<br />
In addition, signal pressure is inhibited, which leads to<br />
heart failure. The incidence <strong>of</strong> ischemic heart disease and<br />
arterial thromboembolic events is as high as 3.3%. 30 Approved<br />
for the treatment <strong>of</strong> patients with metastatic renal<br />
cancer, the TKIs sunitinib and sorafenib are multikinases<br />
inhibitors. Sunitinib is also used in the treatment <strong>of</strong> gastrointestinal<br />
stromal tumors, and sorafenib is used in the<br />
treatment <strong>of</strong> advanced hepatocellular carcinoma. They are<br />
inhibitors <strong>of</strong> growth factor receptors, the most important <strong>of</strong><br />
which are VEGF, platelet-derived growth factor, the stem<br />
cell factor KIT receptor, and the Von Hippel-Lindau<br />
hypoxia-inducible gene pathway. Furthermore, sorafenib<br />
also affects the RAS-RAF-MEK-ERK pathway by inhibiting<br />
the intracellular kinases RAF and BRAF. This inhibition<br />
substantially affects tumor angiogenesis and/or cell proliferation.<br />
These kinases also play an important role in the<br />
maintenance <strong>of</strong> the normal function <strong>of</strong> the vascular system.<br />
Therefore, inhibition may also lead to dysfunction in the<br />
cardiovascular system. An observational single-center study<br />
Author’s Disclosure <strong>of</strong> Potential Conflicts <strong>of</strong> Interest<br />
Author<br />
Marianne Ryberg*<br />
*No relevant relationships to disclose.<br />
Employment or<br />
Leadership<br />
Positions<br />
Consultant or<br />
Advisory Role<br />
1. Albini A, Pennesi G, Donatelli F, et al. Cardiotoxicity <strong>of</strong> Anticancer<br />
drugs: The need for Cardio-Oncological Prevention. J Natl Cancer Inst.<br />
2010;102(1):14-25.<br />
2. Carver JR, Shapiro CL, Ng A, et al. <strong>American</strong> <strong>Society</strong> <strong>of</strong> <strong>Clinical</strong><br />
<strong>Oncology</strong> <strong>Clinical</strong> Evidence Review on the Ongoing Care <strong>of</strong> Adult Cancer<br />
Survivors: Cardiac and Pulmonary Late Effect. J Clin Oncol. 2007;25:3991-<br />
4008.<br />
3. Minotti G, SalvatorelliE, Menna P. Pharmacological Foundations <strong>of</strong><br />
Cardio-<strong>Oncology</strong>. J Pharmacol Exp Ther. 2010;334:2-8.<br />
4. von H<strong>of</strong>f DD, Layard MW, Basa P, et al. Risk Factors for Doxorubicininduced<br />
congestive heart failure. Ann Intern Med. 1970;91:710-717.<br />
5. Ryberg M, Nielsen D, Cortese G, et al. New insight to epirubicin cardiac<br />
toxicity. Competing risks analysis <strong>of</strong> 1097 breast cancer patients. J Natl<br />
Cancer Inst. 2008;100:1-10.<br />
6. Swain SM, Whaley FS, Ewer MS. Congestive Heart Failure in Patients<br />
Treated with Doxorubicin. A retrospective Analysis <strong>of</strong> Three Trials. Cancer.<br />
2003;97:2869-2879.<br />
7. Blanco JG, Leisenring WM, Gonzalez-Covarrubias VM, et al. Genetic<br />
Polymorphisms in the carbonyl reductase 3 gene CBR3 and the NAD(P)H:<br />
Quinone Oxioreductase 1 gene NQO1 in patients who developed anthracycline-related<br />
congestive heart failure after childhood cancer. Cancer. 2008;<br />
112:2789-2895.<br />
8. Zhou S, Palmeira CM, Wallace KB. Doxorubicin-induced persistent<br />
oxidative stress to cardiac myocytes. Toxicol Lett. 2011;121:151-157.<br />
9. Minotti G, Menna P, Salvatorelli E, et al. Anthracyclines: Molecular<br />
advances and pharmacologic development in antitumor activity and cardiotoxicity.<br />
Pharmacol Rev. 2004;56:185-229.<br />
558<br />
analyzed 86 patients who received TKI treatment for metastatic<br />
renal carcinoma and found that 18% had experienced<br />
life-threatening cardiovascular events. The cancer therapy<br />
was interrupted and treatment <strong>of</strong> the cardiovascular symptoms<br />
was initiated, after which all patients were able to<br />
continue treatment. This highlights the reversibility <strong>of</strong> cardiac<br />
symptoms and the importance <strong>of</strong> cardiac monitoring<br />
and intervention for these patients. 32<br />
The previously mentioned inhibitors are only the beginning<br />
<strong>of</strong> a new era, in which many more are in development<br />
for the treatment <strong>of</strong> cancer and other diseases. As a result,<br />
developing preclinical systems that can detect whether a<br />
potential drug is cardiotoxic is imperative. It is equally<br />
important that clinical trials be designed to assess the risk<br />
<strong>of</strong> cardiotoxicity, if there is any suspicion <strong>of</strong> it during<br />
preclinical trials.<br />
Conclusion<br />
The reason why cardiovascular symptoms develop during<br />
and after the treatment <strong>of</strong> cancer depends on the treatment<br />
modality and on the patient’s general status. As a result,<br />
increasing physician knowledge about the various treatment<br />
regimens and their short- and long-term effects on the heart<br />
is crucial. Furthermore, as many heart diseases are treatable,<br />
close collaboration between cardiologists and oncologists<br />
is imperative. The number <strong>of</strong> new therapies available<br />
for treating patients with cancer will continue to grow,<br />
which is why cataloging and updating side effects, including<br />
long-term side effects, are essential to keeping individual<br />
specialists abreast <strong>of</strong> the newest treatment developments.<br />
Stock<br />
Ownership Honoraria<br />
REFERENCES<br />
Research<br />
Funding<br />
Expert<br />
Testimony<br />
MARIANNE RYBERG<br />
Other<br />
Remuneration<br />
10. Wallace KB. Adriamycin-induced interference with cardiac mitochondrial<br />
calcium homeostasis. Cardiovasc Toxicol. 2007;7:101-107.<br />
11. De Angelis A, Piegari E, Cappetta D, et al. Anthracycline cardiomyopathy<br />
is mediated by depletion <strong>of</strong> the cardiac stem cell pool and is rescued by<br />
restoration <strong>of</strong> progenitor cell function. Circulation. 2010;121:276-292.<br />
12. Gianni L, Munzone E, Capri G, et al. Paclitaxel by 3-hour infusion in<br />
combination with bolus doxorubicin in women with untreated metastatic<br />
breast cancer: high antitumor efficacy and cardiac effects in a dose-finding<br />
and sequence-finding study. J Clin Oncol. 1995;23:2688-2699.<br />
13. Gianni L, Salvatorelli E, Minotti G. Anthracycline cardiotoxicity in<br />
breast cancer patients: Synergism with trastuzumab and taxanes. Cardiovasc<br />
Toxicol. 2007;7:67-71.<br />
14. Hauges HS, Wethal T, Aass N, et al. Cardiovascular Risk Factors and<br />
Morbidity in Long-Term Survivors. J Clin Oncol. 2010;28:4649-4657.<br />
15. Meinardi MT, Gietema JA, van der Graaf WT, et al. Cardiovascular<br />
Morbidity in Long-Term Survivors <strong>of</strong> Metastatic Testicular cancer. J Clin<br />
Oncol. 2000;18:1725-1732.<br />
16. Giordano SH, Kuo YF, Freeman JL, et al. Risk <strong>of</strong> cardiac death after<br />
adjuvant radiotherapy for breast cancer. J Natl Cancer Inst. 2005;97:419-424.<br />
17. Grunwald MR, Howie L, Diaz LA Jr. Takotsubo Cardiomyopathy and<br />
Fluorouracil: Case Report and Review <strong>of</strong> the Literature. J Clin Oncol.<br />
<strong>2012</strong>;30:e11-e14. Epub 2011 Dec 5.<br />
18. Dalzell JR, Samuel LM. The spectrum <strong>of</strong> 5-fluorouracil cardiotoxicity.<br />
Anti-cancer Drugs. 2009;20:79-80.<br />
19. Jensen SA, Sørensen JB. Risk Factors and prevention <strong>of</strong> cardiotoxicity<br />
induced by 5-fluorouracil or capecitabine. Cancer Chemother Pharmacol.<br />
2006; 58:487-493. Epub 2006 Jan 18.
CARDIOTOXICITY OF ANTICANCER THERAPIES<br />
20. Taylor CW, Nisbet A, McGale P, et al. Cardiac exposures in breast<br />
cancer radiotherapy: 1950s-1990s. Int J Radiat Oncol Biol Phys. 2007;69:<br />
1489-1495.<br />
21. Tukenova M, Guibout C, Oberlin O, et al. Role <strong>of</strong> Cancer Treatment in<br />
Long-Term Overall and Cardiovascular Mortality After Childhood Cancer.<br />
J Clin Oncol. 2010;28:1308-1315. Epub 2010 Feb 8.<br />
22. Nilsson G, Holmberg L, Garmo H, et al. Distribution <strong>of</strong> Coronary Artery<br />
Stenosis After Radiation for Breast Cancer. J Clin Oncol. <strong>2012</strong>;30:380-386.<br />
Epub 2011 Dec 27.<br />
23. Force T, Kerkelä R. Cardiotoxicity <strong>of</strong> the new cancer therapeutics—<br />
mechanisms <strong>of</strong>, and approaches to, the problem. Drug Discov Today. 2008;<br />
13:778-784. Epub 2008 Sep 2.<br />
24. Ewer SM, Ewer MS. Cardiotoxicity Pr<strong>of</strong>ile <strong>of</strong> Trastuzumab. Drug Saf.<br />
2008;31:409-467.<br />
25. Slamon DJ, Leyland-Jones B, Shak S, et al. Use <strong>of</strong> chemotherapy plus<br />
a monoclonal antibody against HER2 for metastatic breast cancer that over<br />
expresses HER2. N Engl J Med. 2001;344:783-792.<br />
26. Stortecky S, Suter T. Insights into cardiovascular side-effects <strong>of</strong> modern<br />
anticancer therapeutics. Curr Opin Oncol. 2010;22:312-317.<br />
27. Procter M, Suter T, de Azambuja E, et al. Longer-Term Assessment <strong>of</strong><br />
Trastuzumab-Related Cardiac adverse Events in the Herceptin Adjuvant<br />
(HERA) Trial. J Clin Oncol. 2010;28:3422-3428.<br />
28. Chen T, Xu T, Li Y, et al. Risk <strong>of</strong> cardiac dysfunction with trastuzumab<br />
in breast cancer patients: A meta-analysis. Cancer Treat Rev. 2011;37:312-<br />
320. Epub 2010 Oct 16.<br />
29. Zambelli A, Della Porta MG, Eleutri E, et al. Predicting and preventing<br />
cardiotoxicity in the era <strong>of</strong> breast cancer targeted therapies. Novel molecular<br />
tools for clinical issues. Breast. 2011;20:170-80. Epub 2010 Dec 13.<br />
30. Vaklavas C, Lenihan D, Kurzrock R, et al. Anti-vascular Endothelial<br />
Growth Factor Therapies and Cardiovascular Toxicity: What are the Important<br />
<strong>Clinical</strong> Markers to Target? Oncologist. 2010;15:130-141. Epub 2010<br />
Feb 5.<br />
31. Azad NS, Posadas EM, Kwirkowski VE, et al. Combination targeted<br />
therapy with sorafenib and bevacizumab results in enhanced toxicity and<br />
antitumor activity. J Clin Oncol. 2008;26:3709-3714.<br />
32. Schmidinger M, Zielinski CC, Vogl UM, et al. Cardiac Toxicity <strong>of</strong><br />
Sunitinib and Sorafenib in patients with metastatic renal cell carcinoma.<br />
J Clin Oncol. 2008;26:5204-5212. Epub 2008 Oct 6.<br />
559
WHEN CANCER BECOMES PERSONAL: THE<br />
PHYSICIAN’S VIEW OF PATIENT CARE<br />
CHAIR<br />
Teresa Gilewski, MD<br />
Memorial Sloan-Kettering Cancer Center<br />
New York, NY<br />
SPEAKERS<br />
Peter Bach, MD<br />
Memorial Sloan-Kettering Cancer Center<br />
New York, NY<br />
Martin Raber, MD<br />
University <strong>of</strong> Texas M. D. Anderson Cancer Center<br />
Houston, TX<br />
George W. Sledge Jr., MD<br />
Indiana University Simon Cancer Center<br />
Indianapolis, IN
The Oncologist as the Patient with Cancer or<br />
Relative<br />
By Teresa Gilewski, MD, Martin Raber, MD, and George W. Sledge Jr., MD<br />
Overview: To an extent, physicians are familiar with the<br />
consequences <strong>of</strong> illness through their interactions with patients.<br />
However, when cancer becomes personal, the physician<br />
has an opportunity to gain greater insight into the<br />
intricacies <strong>of</strong> medical care, including its humanistic elements.<br />
Physicians who encounter cancer in themselves or in a relative<br />
may deepen their understanding <strong>of</strong> the patient experience.<br />
PERSONAL EXPERIENCES may influence an individual’s<br />
thoughts and actions. It is plausible that a physician’s<br />
perspective <strong>of</strong> patient care may be affected by an<br />
encounter with illness on a personal level. Any human being<br />
who is sick or who has experienced illness along with a<br />
family member or close friend has the opportunity to gain<br />
considerable insight into the complexities <strong>of</strong> medical care.<br />
This may be especially true for physicians who have the<br />
unique position <strong>of</strong> viewing patient care from the perspective<br />
<strong>of</strong> the provider as well as the recipient. In particular, the<br />
humanistic aspects <strong>of</strong> medicine may become more apparent<br />
and more meaningful. Human interactions form the core <strong>of</strong><br />
patient care, and optimal patient care requires adequate<br />
observation and self-reflection <strong>of</strong> these interfaces.<br />
Teresa Gilewski, MD: The Humanistic Side <strong>of</strong><br />
Medicine—The Impact from a Personal<br />
Cancer Experience<br />
The practice <strong>of</strong> medicine is a complex interplay <strong>of</strong> science,<br />
humanism, business and social policy. During different<br />
periods in history, the emphasis on these various elements<br />
has shifted. For example, many <strong>of</strong> the basic moral and<br />
ethical tenets <strong>of</strong> medicine were established by Hippocrates<br />
and others in ancient Greece at a time when there was<br />
limited scientific knowledge. 1 With the onset <strong>of</strong> new scientific<br />
discoveries, a greater focus on technology and science<br />
was inevitable. In 1910, the noted educator Abraham Flexner<br />
emphasized the importance <strong>of</strong> both a scientific and<br />
clinical research basis in medical education. 2 His report<br />
helped to restructure medical school curriculum to one that<br />
was more research focused. However, 15 years later he felt<br />
that “scientific medicine” was “. . . sadly deficient in cultural<br />
and philosophic background.” 3 Finding a balance between<br />
the scientific and the human components <strong>of</strong> medicine remains<br />
a challenge to the present.<br />
In addition, now there is an ever-increasing presence, on a<br />
daily basis, <strong>of</strong> the business and social aspects <strong>of</strong> medical<br />
care. The humanistic aspects <strong>of</strong> medicine have <strong>of</strong>ten become<br />
secondary to these other more easily measurable facets <strong>of</strong><br />
medicine. However, in the last decade medical groups, such<br />
as the <strong>American</strong> Board <strong>of</strong> Internal Medicine and Institute <strong>of</strong><br />
Medicine, have underscored the importance <strong>of</strong> practicing<br />
medicine in a humanistic manner. 4,5 The association between<br />
pr<strong>of</strong>essionalism and humanism has become clearer. 6,7<br />
Yet, there are many challenges, both on institutional and<br />
personal levels, that may hinder the incorporation <strong>of</strong> a<br />
humanistic approach into patient care. These include the<br />
emphasis on research and greater clinical productivity as<br />
well as the stresses associated with caring for ill patients. 3,8<br />
Their views provide a unique perspective, on the basis <strong>of</strong> the<br />
convergence <strong>of</strong> their medical knowledge and personal reaction<br />
to illness. They also confront distinct challenges specific<br />
to their work environment. An enhanced recognition <strong>of</strong> their<br />
viewpoints provides valuable information in the quest to<br />
alleviate patient suffering and explore the fundamentals <strong>of</strong><br />
patient care.<br />
The medical literature contains heartfelt essays that focus<br />
on the importance <strong>of</strong> human interactions in medicine. 9,10 In<br />
particular, the sections “A Piece <strong>of</strong> My Mind” in the Journal<br />
<strong>of</strong> the <strong>American</strong> Medical Association and the “Art <strong>of</strong> <strong>Oncology</strong>”<br />
in the Journal <strong>of</strong> <strong>Clinical</strong> <strong>Oncology</strong> provide physicians<br />
an opportunity to contemplate various aspects <strong>of</strong> the personal<br />
impact <strong>of</strong> illness. 11-13 Narrative medicine utilizes the<br />
practice <strong>of</strong> writing about patients’ experiences and the<br />
writer’s reflection on those experiences to foster empathy. 14<br />
Physicians usually develop a general awareness <strong>of</strong> the<br />
difficulties that patients confront, but not necessarily a true<br />
understanding <strong>of</strong> those struggles. However, physicians who<br />
encounter cancer in themselves or in a family member<br />
develop a new intimacy with illness. 15-17 This altered relationship<br />
with disease may result in a heightened awareness<br />
<strong>of</strong> the consequences <strong>of</strong> sickness. 18 Once there is a personal<br />
connection with cancer, the day-to-day routine <strong>of</strong> the medical<br />
system may assume a fresh look. What may have seemed<br />
trivial and <strong>of</strong> little significance may become momentous.<br />
Some <strong>of</strong> the usual inconsequential basic human interactions<br />
between a patient and the physician suddenly become memorable<br />
for their compassion or lack there<strong>of</strong>.<br />
Experiencing cancer on a personal level may provide the<br />
physician with new points <strong>of</strong> reference that have the potential<br />
to influence one’s perspective <strong>of</strong> patient care. These<br />
experiences may originate not only in adulthood but in<br />
childhood as well. In childhood, the effect <strong>of</strong> caring for a sick<br />
relative and observing the interactions <strong>of</strong> adult relatives and<br />
physicians coping with illness may be long remembered.<br />
This child who later becomes a physician may use some <strong>of</strong><br />
these observations toward the care <strong>of</strong> patients. Of course,<br />
each physician brings a unique personality and other life<br />
experiences to every situation. However, exposure to personal<br />
illness has the potential to enhance the physician’s<br />
appreciation <strong>of</strong> the fragility and uncertainty <strong>of</strong> life as well as<br />
the value <strong>of</strong> kindness.<br />
The physician’s appreciation <strong>of</strong> the “human” experience<br />
surrounding illness may be heightened, not only in regards<br />
to patients but in regards to colleagues. Perhaps on the basis<br />
<strong>of</strong> this greater awareness, a deeper recognition <strong>of</strong> the essence<br />
<strong>of</strong> physicians may develop. Specifically, that they are<br />
From the Memorial Sloan-Kettering Cancer Center, New York, NY; M. D. Anderson<br />
Cancer Center, Houston, TX; University <strong>of</strong> Indiana, Indianapolis, IN.<br />
Authors’ disclosures <strong>of</strong> potential conflicts <strong>of</strong> interest are found at the end <strong>of</strong> this article.<br />
Address reprint requests to Teresa Gilewski, MD, Memorial Sloan-Kettering Cancer<br />
Center, 300 East 66 th St., New York, NY 10065; email: gilewskt@mskcc.org.<br />
© <strong>2012</strong> by <strong>American</strong> <strong>Society</strong> <strong>of</strong> <strong>Clinical</strong> <strong>Oncology</strong>.<br />
1092-9118/10/1-10<br />
561
imperfect human beings, who in their relationships with<br />
patients may at times fall short <strong>of</strong> their own best intentions<br />
and at other times reach pr<strong>of</strong>ound moments <strong>of</strong> harmony.<br />
Does experiencing illness personally make one a more<br />
humanistic physician? Perhaps for some the answer is yes,<br />
for others no, but for all it is a possibility. Regardless, it is<br />
important to generate dialogue about the humanistic element<br />
<strong>of</strong> medicine because it is fundamental to patient care.<br />
As the noted physician William Osler (1849–1919) eloquently<br />
stated, “Care more for the individual patient than<br />
for the special features <strong>of</strong> the disease ...Putyourself in his<br />
place ...Thekindly word, the cheerful greeting, the sympathetic<br />
look—these the patient understands.”<br />
Martin Raber, MD: The Oncologist as the Patient<br />
Without question, being a physician has influenced my<br />
experience as a patient and being a patient has influenced<br />
my experience as a physician. I am not sure that it has made<br />
me a better physician, but I know that it has made me a<br />
different physician. It has also given me a different view <strong>of</strong><br />
the clinic, the hospital, and the medical teams. At the same<br />
time that I have been a patient, I have also tried to maintain<br />
a semblance <strong>of</strong> an academic career and continue to see<br />
patients in my area <strong>of</strong> expertise. This has been possible only<br />
because my institution and my colleagues have been willing<br />
to make substantial accommodations to limit my responsibilities<br />
and to cover me whenever I was unable to work.<br />
I don’t think anything prepares you for being sick, and the<br />
experience <strong>of</strong> entering the hospital as a patient is so incredibly<br />
different from that <strong>of</strong> entering it as a physician that, at<br />
first, one is totally disoriented. As physicians, we tend to<br />
think <strong>of</strong> the hospital and clinic experience as centered on us.<br />
Although it is true that the physician’s decisions and comments<br />
are what drives the delivery <strong>of</strong> care, in many respects<br />
the patient experience is driven by the ancillary personnel.<br />
KEY POINTS<br />
● A personal experience with illness provides a unique<br />
opportunity for the physician to gain considerable<br />
insight into the humanistic elements <strong>of</strong> medicine that<br />
are at the core <strong>of</strong> patient care.<br />
● The patient experience is very much dependent on<br />
interactions with the ancillary staff; how they treat<br />
each other and how you treat them is as important as<br />
how they treat the patient.<br />
● Physician-patient communication is more difficult<br />
than it seems, particularly in areas in which we do<br />
not have lab tests, images, or clinical findings to<br />
explain the situation.<br />
● Caring for patients while also being a patient is a<br />
challenge; it requires substantial accommodation on<br />
the part <strong>of</strong> one’s colleagues, team members, and<br />
patients.<br />
● A physician who is the relative <strong>of</strong> a patient with<br />
cancer may struggle with distinctive challenges that<br />
relate to the intersection <strong>of</strong> the physician’s medical<br />
knowledge and personal emotions with family<br />
dynamics.<br />
562<br />
GILEWSKI, RABER, AND SLEDGE<br />
Patients spend most <strong>of</strong> their time with clinic and chemotherapy<br />
nurses, clerks, radiology and laboratory technologists,<br />
patient transportation personnel, and others that we tend<br />
not to consider as central to the patient experience. If they<br />
don’t deal with the patient in a positive way, the patient<br />
experience is not good. This has to do with not only how they<br />
deal with the patient but also with each other, and how you<br />
as the physician deal with them. Good communication skills<br />
are important not only for the physician but also for everyone<br />
who interacts with the patient.<br />
Patients see and hear everything that goes on. We sit in<br />
the waiting rooms and observe the interactions and process<br />
what we see. Even in the examining room we hear the<br />
conversations in the hallway between physician and nurse or<br />
trainee. Sitting in a waiting room (which is what we patients<br />
do a lot <strong>of</strong>) I am always amazed at how perceptive my fellow<br />
patients are about the staff and the clinic operations.<br />
Another surprise to me as a physician was the difficulty<br />
that I have had at times communicating with my medical<br />
team. This related most <strong>of</strong>ten to symptoms or situations for<br />
which there was not a good laboratory or radiologic corollary.<br />
I have come to believe that the doctor-patient visit in<br />
the clinic is much more stereotyped than we think. Although<br />
the physician wants to find out how the patient is feeling,<br />
assess the patient’s condition, and give the patient information<br />
about his disease and plans, and the patient wants to<br />
tell the physician how he feels and understand his condition<br />
and the plans, there is tremendous opportunity for misunderstanding.<br />
Patient and physician have somewhat different<br />
goals and expectations <strong>of</strong> the clinic visit. They also have a<br />
conversation in which many words are used without prior<br />
agreement as to their meaning. “Good,” “bad,” “fair,” “tired,”<br />
and “alright” are examples <strong>of</strong> words that may mean very<br />
different things to the patient and the physician. That<br />
realization caused me to substantially change the way that<br />
I interview patients in my clinic, and the way I speak to my<br />
physicians in their clinic.<br />
Trying to maintain a medical career at the same time one<br />
is facing serious illness, and undergoing cancer therapy, is a<br />
challenge. Early in the illness I had what I call “doctorpatient<br />
confusion,” and was unable to practice at all. Later,<br />
after I recovered from some devastating complications, I<br />
found that I had passed through that phase, and once again<br />
began to see patients. In this period I have come to realize<br />
that my illness has had a major effect on my colleagues who<br />
have <strong>of</strong>ten been simultaneously my colleagues and physicians,<br />
on my team, and on my patients, all <strong>of</strong> whom have had<br />
to adjust to the realities <strong>of</strong> my health problems. As have I.<br />
George Sledge, MD: The Oncologist as the Relative<br />
<strong>of</strong> a Patient with Cancer<br />
Because cancer is common, and because cancer doctors<br />
have relatives with cancer, cancer pr<strong>of</strong>essionals regularly<br />
come face to face with cancer in relatives. Their unique<br />
perspective on cancer—a perspective that has both intellectual<br />
and emotional components derived from years <strong>of</strong> caring<br />
for patients with cancer—clearly affects how they interact<br />
with those relatives, both for better and worse.<br />
Physicians are routinely taught to avoid taking care <strong>of</strong><br />
relatives, an admonition that is both wise and rarely<br />
completely respected. Wise, because physicians need to<br />
maintain emotional distance from their patients. Every<br />
relationship with a relative is fraught with family history,
THE ONCOLOGIST AS THE PATIENT OR RELATIVE<br />
and that history is not always either happy or uncomplicated.<br />
That history, and ongoing family dynamics, can<br />
complicate recommendations, and the recommendations<br />
themselves can easily become the source <strong>of</strong> future family<br />
discord as well as personal feelings <strong>of</strong> grief or guilt.<br />
At the same time, physician-relatives <strong>of</strong> patients with<br />
cancer are in the enviable position <strong>of</strong> being able to point the<br />
patient with cancer in the direction <strong>of</strong> the best care, or at<br />
least <strong>of</strong> the best caregiver, available. Although much <strong>of</strong><br />
cancer care is standard, not all cancer pr<strong>of</strong>essionals are<br />
equivalent in either expertise, competence, or compassion, a<br />
fact that physicians are well aware <strong>of</strong>: a “doctor’s doctor” is<br />
<strong>of</strong>ten defined in terms <strong>of</strong> whom a physician is willing to have<br />
care for a relative.<br />
A physician faces additional challenges as the relative <strong>of</strong> a<br />
patient with cancer. Physicians are routinely asked “what<br />
would you do if I was your ...[sister, brother, daughter, son,<br />
mother, father]? This question is more difficult to dodge or<br />
ignore when it comes from a relative, and the responsibility<br />
attached to the answer differs qualitatively from that encountered<br />
in usual practice, rightly or wrongly.<br />
Physician-relatives also face challenges related to their<br />
interactions with caregivers and hospital systems. If a<br />
physician disagrees with the advice given to a relative, what<br />
is that physician to do? If an interaction with the health care<br />
system is a negative one, does the physician-relative claim<br />
special attention for the patient? Under normal circumstances,<br />
most physicians accept that their colleagues deserve<br />
substantial autonomy in decision making and<br />
therapeutic recommendations, and recognize that the health<br />
care system is imperfect. Indeed, no health care system<br />
would long survive intrusive oversight by colleagues.<br />
Yet this dynamic frequently changes when a physician’s<br />
relative enters the health care system, particularly that part<br />
<strong>of</strong> the system in which the physician-relative practices. This<br />
Authors’ Disclosures <strong>of</strong> Potential Conflicts <strong>of</strong> Interest<br />
Author<br />
Teresa Gilewski*<br />
Martin Raber*<br />
George W. Sledge Jr.*<br />
*No relevant relationships to disclose.<br />
Employment or<br />
Leadership<br />
Positions<br />
Consultant or<br />
Advisory Role<br />
1. Greek Medicine, History <strong>of</strong> Medicine, U.S. National Library <strong>of</strong> Medicine,<br />
National Institutes <strong>of</strong> Health. http://www.nlm.nih.gov/hmd/. Accessed March<br />
7, <strong>2012</strong>.<br />
2. Flexner A. Medical education in the United States and Canada: A report<br />
to the Carnegie Foundation for the Advancement <strong>of</strong> Teaching. New York:<br />
Carnegie Foundation for the Advancement <strong>of</strong> Teaching, 1910.<br />
3. Cooke M, Irby DM, Sullivan W, et al. <strong>American</strong> medical education 100<br />
years after the Flexner report. N Engl J Med. 2006;355:1339-1344.<br />
4. <strong>American</strong> Board <strong>of</strong> Internal Medicine. Project Pr<strong>of</strong>essionalism. Philadelphia,<br />
Pa, America Board <strong>of</strong> Internal Medicine, 2001;4.<br />
5. Hewitt M, Herdman R, Holland J (eds). Meeting Psychosocial Needs <strong>of</strong><br />
Women with Breast Cancer. Institute <strong>of</strong> Medicine and National Research<br />
Council. Washington, DC, National Academies Press, 2004.<br />
6. Cohen JJ. Linking pr<strong>of</strong>essionalism to humanism: What it means, why it<br />
matters. Acad Med. 2007;82:1029-1032.<br />
7. Swick HM. Pr<strong>of</strong>essionalism and humanism beyond the academic health<br />
center. Acad Med. 2007;82:1022-1028.<br />
8. Whippen DA, Canellos GP. Burnout syndrome in the practice <strong>of</strong> oncology:<br />
Results <strong>of</strong> a random survey <strong>of</strong> 1,000 oncologists. J Clin Oncol. 1991;9:<br />
1916-1921.<br />
may result in excessive diagnostic testing and overtreatment<br />
<strong>of</strong> the relative, <strong>of</strong>ten to that patient’s detriment, a<br />
situation akin to the defensive medicine practiced by physicians<br />
concerned with malpractice. Physician-relatives who<br />
are aware <strong>of</strong> this dilemma may be caught between the Scylla<br />
and Charybdis <strong>of</strong> this dynamic, wishing the best care for a<br />
relative but also wishing to make the best use <strong>of</strong> a colleague’s<br />
expertise and wisdom.<br />
The physician is also faced with the problem <strong>of</strong> knowing<br />
too much. This is frequently the case when the physician’s<br />
relative has a poor-prognosis cancer. In this setting, relatives<br />
may explicitly encourage false hope, leaving the physician<br />
in the emotionally precarious position <strong>of</strong> balancing<br />
reality and optimism. These discussions may have long-term<br />
consequences, not just for the patient, but for the family as<br />
a whole: emotional wounds that never completely heal.<br />
Under normal circumstances, the physician can go home<br />
and unwind after a hard day at work, but when home is the<br />
source <strong>of</strong> stress there may be no place to turn.<br />
With all patients with advanced disease, there comes a<br />
time when active therapy is no longer appropriate, and in<br />
which a focus on quality-<strong>of</strong>-life measures, advanced care<br />
planning, and hospice care become reasonable. What is the<br />
role <strong>of</strong> the physician-relative in selecting that moment?<br />
Indeed, is there a role?<br />
Finally, the physician-relative must deal with his or her<br />
own emotional needs, both during the relative’s treatment<br />
arc and after a relative’s death. Physicians are <strong>of</strong>ten excellent<br />
and compassionate communicators when dealing with<br />
patients to whom they are unrelated; but it is the rare<br />
physician who is capable <strong>of</strong> expressing his or her own emotional<br />
trauma, or who is capable <strong>of</strong> instituting the healing<br />
process we all deserve in our own most pr<strong>of</strong>ound times <strong>of</strong> grief<br />
and loss. Dealing with a relative’s death reminds us <strong>of</strong> our own<br />
mortality, <strong>of</strong> that end to which we all must go.<br />
Stock<br />
Ownership Honoraria<br />
REFERENCES<br />
Research<br />
Funding<br />
Expert<br />
Testimony<br />
Other<br />
Remuneration<br />
9. Walshe FMR. Humanism, history, and natural science in medicine.<br />
BMJ. 1950;August;12:379-384.<br />
10. Pickering WG. Kindness, prescribed and natural in medicine. J Medical<br />
Ethics. 1997;23:116-118.<br />
11. Young RK. A Piece <strong>of</strong> My Mind. John Wiley and Sons, Inc. Hoboken,<br />
New Jersey, 2000.<br />
12. Loprinzi CL. A new addition to the J Clin Oncol: The Art <strong>of</strong> <strong>Oncology</strong>—<br />
When the Tumor is Not the Target. J Clin Oncol. 2000;18:3.<br />
13. Steensma DP. Art <strong>of</strong> <strong>Oncology</strong>: New voices wanted. J Clin Oncol.<br />
2011;29:3343-3344.<br />
14. Charon R. Narrative medicine-a model for empathy, reflection, pr<strong>of</strong>ession,<br />
and trust. JAMA. 2001;286:1897-1902.<br />
15. Biro D. One Hundred Days: My Unexpected Journey from Doctor to<br />
Patient. New York: Random House; 2000.<br />
16. Mullan F. Seasons <strong>of</strong> survival: reflections <strong>of</strong> a physician with cancer.<br />
N Engl J Med. 1985;313:270-273.<br />
17. Liberman L. I Signed As the Doctor: Memoir <strong>of</strong> a Cancer Doctor<br />
Surviving Cancer. Port Charlotte, FL: Booklocker.com, Inc.; 2009.<br />
18. Gilewski T. The physician as the patient [video]. New York: Memorial<br />
Sloan Kettering Cancer Center; 2007.<br />
563
ADVANCES IN INFECTION MANAGEMENT IN<br />
PEDIATRIC ONCOLOGY<br />
CHAIR<br />
Andrew Y. Koh, MD<br />
University <strong>of</strong> Texas Southwestern Medical Center<br />
Dallas, TX<br />
SPEAKERS<br />
Lillian Sung, MD, PhD<br />
The Hospital for Sick Children<br />
Toronto, ON, Canada<br />
Theoklis Zaoutis, MD<br />
Children’s Hospital <strong>of</strong> Philadelphia<br />
Philadelphia, PA
Prolonged Febrile Neutropenia in the<br />
Pediatric Patient with Cancer<br />
Overview: Infectious diseases continue to be major causes<br />
<strong>of</strong> morbidity and mortality in pediatric patients with cancer.<br />
Yet not all pediatric patients with cancer with fever and<br />
neutropenia are at equal risk for substantial morbidity or<br />
mortality from infection. Patients at highest risk for developing<br />
infectious complications are those with severe and prolonged<br />
neutropenia, substantial medical comorbidity, and<br />
hematologic malignancy, or recipients <strong>of</strong> stem-cell transplantation.<br />
These “high-risk” patients also have concomitant host<br />
immune deficits as well: severe mucositis, lymphopenia, hypogammaglobulinemia,<br />
and gut microbial dysbiosis. Because<br />
bacterial and fungal infections are the most common infectious<br />
complications, continuation <strong>of</strong> empirical antibacterial<br />
antibiotics that were initiated at the onset <strong>of</strong> febrile neutropenia<br />
and prompt initiation <strong>of</strong> empirical antifungal therapy in<br />
ALTHOUGH THE advances <strong>of</strong> infectious diseases supportive<br />
care during the last several decades have<br />
permitted patients to successfully recover from the impact <strong>of</strong><br />
cytotoxic cancer chemotherapy, infectious diseases continue<br />
to be major causes <strong>of</strong> morbidity and mortality in pediatric<br />
patients with cancer. This review will address the host<br />
immune deficits in pediatric patients with cancer, riskstratification<br />
<strong>of</strong> patients with febrile neutropenia, general<br />
management guidelines for patients with prolonged neutropenia,<br />
and potential strategies for preventing infectious<br />
diseases in patients with prolonged neutropenia.<br />
Host Defenses: Innate and Adaptive Immunity<br />
To best manage or prevent infectious complications in the<br />
patient with cancer with prolonged neutropenia, it is essential<br />
to understand the host immune deficits that result from<br />
the underlying disease and/or cancer chemotherapy.<br />
Mucocutaneous Barriers and Commensal Gut Microbiota<br />
Pediatric oncology patients sustain disruptions <strong>of</strong> the<br />
mucocutaneous integrity because <strong>of</strong> their underlying cancer<br />
and its treatment. The disruption <strong>of</strong> mucocutaneous barriers<br />
provides a key portal <strong>of</strong> entry for bacterial and fungal<br />
pathogens (Sidebar 1). Colonization by normal bacterial<br />
flora (aerobic gram-positive bacteria and a variety <strong>of</strong> anaerobic<br />
bacteria with relatively low virulence) provides a competitive<br />
microbiologic barrier to colonization by extrinsically<br />
acquired bacterial and fungal organisms. However, within<br />
24 hours <strong>of</strong> hospitalization, seriously ill patients undergo a<br />
change in their indigenous micr<strong>of</strong>lora toward one <strong>of</strong> aerobic<br />
gram-negative organisms. 1 Approximately one-half <strong>of</strong> the<br />
responsible pathogens causing infections are acquired by<br />
oncology patients after initial admission to the hospital, and<br />
more than 80% <strong>of</strong> the microbiologically documented infections<br />
that occur in adult patients with acute myelogenous<br />
leukemia (AML) are caused by organisms that were a<br />
component <strong>of</strong> the endogenous mucosal organisms. 1<br />
Phagocytic Cells<br />
By far the most important risk factor in the development<br />
<strong>of</strong> infections in children with cancer remains neutropenia.<br />
By Andrew Y. Koh, MD<br />
the setting <strong>of</strong> prolonged fever and neutropenia continue to be<br />
the standard <strong>of</strong> care. In high-risk patients, antibiotic therapy<br />
should be maintained until neutrophil counts have recovered.<br />
Adjunctive therapies have been shown to be ineffective (e.g.,<br />
colony-stimulating factors) or necessitate further study (e.g.,<br />
granulocyte infusions or keratinocyte growth factor treatment<br />
to heal mucositis). Prophylactic use <strong>of</strong> antibacterial and<br />
antifungal antibiotics in high-risk patients has shown promise<br />
but the fear <strong>of</strong> inducing antimicrobial-resistant strains remains<br />
a deterrent. Finally, the novel concepts <strong>of</strong> manipulating<br />
the host gut microbiota and/or augmenting GI mucosal immunity<br />
to prevent invasive bacterial and fungal infections in<br />
pediatric patients with cancer <strong>of</strong>fers great promise, but more<br />
definitive studies need to be performed.<br />
The relationship between neutrophil numbers and the risk<br />
<strong>of</strong> infectious complications in patients with leukemia was<br />
first described in the seminal work <strong>of</strong> Bodey and colleagues<br />
in 1966. 2 The investigators concluded that 1) the risk <strong>of</strong><br />
infection was directly related to the absolute neutrophil<br />
count (ANC), severe infections being more prevalent when<br />
the ANC fell below 100 cells/mm 3 ; 2) relapse <strong>of</strong> leukemia<br />
was associated with higher rates <strong>of</strong> infection than was<br />
remission; and 3) duration <strong>of</strong> neutropenia was the single<br />
most important factor in predicting risk <strong>of</strong> infection, with<br />
severe neutropenia that lasted longer than 3 weeks being<br />
associated with 100% risk <strong>of</strong> infection and the highest<br />
mortality rates. The rate <strong>of</strong> decline <strong>of</strong> circulating polymorphonuclear<br />
lymphocytes (PMNs) is also critical. For example,<br />
patients with rapidly declining ANCs after therapy for<br />
acute leukemia seem to be at greater risk for infectious<br />
complications than are patients with chronic neutropenia<br />
(e.g., aplastic anemia). 3<br />
Risk Assessment in Patients with Cancer with Fever<br />
and Neutropenia<br />
Not all patients with fever and neutropenia are at equal<br />
risk for substantial morbidity or mortality resulting from<br />
infection. The identification <strong>of</strong> a low-risk subset may allow<br />
for modifications <strong>of</strong> therapy in this group, with goals <strong>of</strong><br />
reduced therapy-related toxicity, improved quality <strong>of</strong> life,<br />
and decreased cost <strong>of</strong> treatment. Although there is no<br />
definitive consensus about the criteria used to prospectively<br />
distinguish high risk from low risk, the following key factors<br />
that may raise the risk <strong>of</strong> infectious complications can be<br />
surmised from studies that have been conducted: (1) anticipated<br />
duration <strong>of</strong> neutropenia 4 ; (2) substantial medical<br />
comorbidity 4 ; (3) cancer status and cancer type; (4) docu-<br />
From the University <strong>of</strong> Texas Southwestern Medical Center, Dallas, TX.<br />
Author’s disclosures <strong>of</strong> potential conflicts <strong>of</strong> interest are found at the end <strong>of</strong> this article.<br />
Address reprint requests to Andrew Koh, MD, University <strong>of</strong> Texas Southwestern Medical<br />
Center, 5323 Harry Hines Blvd, Dallas, TX 75390-9063; email: Andrew.koh@<br />
utsouthwestern.edu.<br />
© <strong>2012</strong> by <strong>American</strong> <strong>Society</strong> <strong>of</strong> <strong>Clinical</strong> <strong>Oncology</strong>.<br />
1092-9118/10/1-10<br />
565
mented infection on presentation; (5) evidence <strong>of</strong> bone marrow<br />
recovery 5 ; and (6) magnitude <strong>of</strong> fever (Sidebar 2). 5<br />
By using these risk criteria, patients with acute lymphoblastic<br />
leukemia (ALL) or AML in induction, relapsed ALL<br />
or AML, Burkitt’s lymphoma in induction, and hematopoietic<br />
stem cell transplantation would all be considered highrisk<br />
patients. These patients not only have severe and<br />
prolonged neutropenia but also have pr<strong>of</strong>ound lymphopenia,<br />
hypogammaglobulinemia, severe mucositis, and gut microbial<br />
dysbiosis. Thus, these high-risk patients might benefit<br />
from implementation <strong>of</strong> prophylactic strategies to prevent<br />
infectious complications.<br />
KEY POINTS<br />
● Infectious diseases, particularly bacterial and fungal<br />
infections, continue to be major causes <strong>of</strong> morbidity<br />
and mortality in pediatric patients with cancer.<br />
● Pediatric patients with cancer at highest risk for<br />
developing infectious complications are those with<br />
severe and prolonged neutropenia; substantial medical<br />
comorbidity, and specific cancer types and status<br />
(i.e., hematologic malignancy, relapse).<br />
● Besides severe and prolonged neutropenia, “highrisk”<br />
patients also have concomitant host immune<br />
deficits: severe mucositis, lymphopenia, hypogammaglobulinemia,<br />
and gut microbial dysbiosis.<br />
● Continuation <strong>of</strong> empirical antibacterial antibiotics<br />
that were initiated at the onset <strong>of</strong> febrile neutropenia<br />
and prompt initiation <strong>of</strong> empirical antifungal therapy<br />
in the setting <strong>of</strong> prolonged fever and neutropenia<br />
continue to be the standard <strong>of</strong> care for pediatric<br />
patients with cancer with prolonged neutropenia.<br />
● Adjunctive therapies (i.e., granulocyte transfusion or<br />
keratinocyte growth factor treatment for mucositis)<br />
and prophylactic use <strong>of</strong> antibacterial and antifungal<br />
antibiotics in high-risk patients have shown promise,<br />
but more definitive studies need to be done.<br />
566<br />
Sidebar 1. Predominant Microbial Pathogens<br />
Infecting Pediatric Patients with Cancer and<br />
Prolonged Neutropenia<br />
Bacteria<br />
Gram-negative enteric organisms<br />
Escherichia coli, Klebsiella pneumoniae, Enterobacter<br />
spp., Citrobacter spp., Pseudomonas<br />
aeruginosa, Bacteroides spp.<br />
Gram-positive<br />
Staphylococci: coagulase-negative, coagulasepositive<br />
Streptococci: group D, �-hemolytic, anaerobic<br />
Clostridia<br />
Fungi<br />
Candida spp. (C. albicans, C. tropicalis, other<br />
species)<br />
Aspergillus spp. (A. fumigatus, A. flavus)<br />
ANDREW Y. KOH<br />
Sidebar 2. Risk Assessment for Cancer Patients<br />
with Fever and Neutropenia<br />
High Risk<br />
1) Anticipated prolonged (� 7 days in duration)<br />
and pr<strong>of</strong>ound neutropenia (absolute neutrophil<br />
count � 100 cells/�L after cytotoxic chemotherapy)<br />
4,6<br />
2) Significant medical comorbid conditions, including<br />
hypotension, pneumonia, new-onset<br />
abdominal pain, or neurologic changes 4,6<br />
3) Cancer type (i.e., hematologic malignancy,<br />
Burkitt’s lymphoma)<br />
4) Cancer status (e.g., relapse)<br />
5) Documented infection on presentation<br />
Low Risk<br />
1) Anticipated brief (� 7 days in duration) neutropenic<br />
periods 4<br />
2) No or few comorbidities 4<br />
Standard Management and Presumptive Therapy<br />
The single most important advance in infectious diseases<br />
oncology supportive care leading to improved survival has<br />
been the prompt initiation <strong>of</strong> empirical antibacterial antibiotics<br />
when the neutropenic patient with cancer becomes<br />
febrile. Before this approach was instituted in the early<br />
1970s, the mortality rate from gram-negative infections,<br />
especially that <strong>of</strong> P. aeruginosa, E. coli, and K. pneumoniae,<br />
approached 80%, but with the widespread use <strong>of</strong> effective<br />
empirical antibiotics, the overall mortality rate has declined<br />
to approximately 10% to 40%.<br />
Approximately 85% to 90% <strong>of</strong> pathogens that are documented<br />
to be associated with new fevers in neutropenia<br />
patients are gram-positive and gram-negative bacteria. Several<br />
empirical regimens have been devised over the last<br />
several decades. Many <strong>of</strong> the regimens studied are equivalent<br />
in their efficacy, although study designs and definitions<br />
<strong>of</strong> success have not been uniform. The Infectious Diseases<br />
<strong>Society</strong> <strong>of</strong> America has published general guidelines for use<br />
<strong>of</strong> antimicrobial agents in neutropenic patients with unexplained<br />
fever. 6<br />
Duration <strong>of</strong> Therapy<br />
The management <strong>of</strong> high-risk patients with prolonged<br />
neutropenia is addressed in a series <strong>of</strong> prospective clinical<br />
studies that stratified according to those who defervesced<br />
after the initiation <strong>of</strong> broad-spectrum antibiotics or who<br />
remained persistently febrile. 7 Among patients with prolonged<br />
neutropenia who defervesced while undergoing<br />
therapy, 41% again became febrile within 3 days <strong>of</strong> stopping<br />
antibiotics on day 7; new bacterial isolates from those<br />
with documented infections were susceptible to the antibiotics<br />
that had been withdrawn. No subsequent infections<br />
were observed among patients who continued receiving<br />
antibiotics.<br />
The duration <strong>of</strong> antibiotic therapy depends on several<br />
factors, including isolation <strong>of</strong> the presumed pathogen, clinical<br />
identification <strong>of</strong> a presumed infectious process, duration<br />
<strong>of</strong> both fever and neutropenia, and the patient’s schedule for<br />
chemotherapy. Traditionally, broad-spectrum antibiotic
FEBRILE NEUTROPENIA IN PEDIATRIC CANCER<br />
therapy is continued until the ANC has returned to � 500<br />
cells/mm 3 . For patients in whom infection has been documented<br />
and resolution <strong>of</strong> fever and neutropenia has been<br />
prompt, the course <strong>of</strong> antibiotic treatment should be appropriate<br />
for the infection identified, parenteral antibiotics<br />
being preferred for patients with serious infections. For<br />
patients without a defined site <strong>of</strong> infection who show signs <strong>of</strong><br />
bone marrow recovery, antibiotics generally can be discontinued,<br />
even before the ANC reaches 500/mm 3 . But even<br />
when a bacterial pathogen is isolated from a febrile, neutropenic<br />
child, there is evidence that broad-spectrum antibiotic<br />
therapy should be continued. 8<br />
Empirical Antifungal Therapy<br />
Fungi have emerged as an important cause <strong>of</strong> superinfections<br />
in patients with prolonged neutropenia, and may affect<br />
9% to 31% <strong>of</strong> this population. In a randomized clinical trial,<br />
56% <strong>of</strong> patients with unexplained fever who remained febrile<br />
after receiving empirical antibiotics developed complications<br />
within 3 days <strong>of</strong> stopping therapy. 9 Strikingly, 31%<br />
<strong>of</strong> these patients eventually developed invasive fungal infections.<br />
Neutropenic patients who remain febrile despite a 4to<br />
7-day trial <strong>of</strong> broad-spectrum antimicrobial therapy are<br />
particularly prone to fungal disease. 9<br />
Traditionally, amphotericin B was the only available systemic<br />
antifungal agent. Its use in empirical regimens for<br />
prolonged or recurrent fever in neutropenic patients reduced<br />
the incidence <strong>of</strong> documented fungal infections and attributable<br />
mortality. 9 Considerable interest has been generated<br />
by preparations <strong>of</strong> liposomal amphotericin because <strong>of</strong> its<br />
lower toxicity. In the only randomized, double-blind trial<br />
comparing liposomal amphotericin with conventional amphotericin<br />
B as empirical antifungal therapy, the outcomes<br />
were similar with respect to survival, resolution <strong>of</strong> fever, and<br />
discontinuation <strong>of</strong> study drug because <strong>of</strong> toxic effects or lack<br />
<strong>of</strong> efficacy. 10 Liposomal amphotericin B was associated with<br />
fewer breakthrough fungal infections, less infusion-related<br />
toxicity, and less nephrotoxicity. 10<br />
The search for alternative antifungal agents has been<br />
prompted by the potential toxicity <strong>of</strong> amphotericin B and<br />
emergence <strong>of</strong> fungi resistant to it. The azoles represent a<br />
less toxic class <strong>of</strong> antifungal agents. Although fluconazole<br />
has been reported to be as effective as amphotericin in the<br />
treatment <strong>of</strong> candidemia in patients without neutropenia or<br />
other major immunodeficiency condition, 11 the picture is<br />
less clear in febrile and neutropenic patients with cancer.<br />
The azoles, however, may be less active than amphotericin B<br />
against some species. Voriconazole compared favorably with<br />
liposomal amphotericin B in adult patients with cancer with<br />
fever and neutropenia. 12 In 2006, the U.S. Food and Drug<br />
Administration (FDA) approved the use <strong>of</strong> posaconazole for<br />
prophylaxis against the development <strong>of</strong> invasive Aspergillus<br />
and Candida infections in immunocompromised patients 13<br />
years <strong>of</strong> age and older. Posaconazole is distinct in having<br />
been successfully used in salvage treatment <strong>of</strong> infections<br />
caused by zygomycetes. But posaconazole’s efficacy as an<br />
empirical antifungal agent in febrile and neutropenic patients<br />
with cancer has not been evaluated.<br />
The newest class <strong>of</strong> antifungal agents are the echinocandins:<br />
large lipopeptide molecules that are inhibitors <strong>of</strong><br />
�-(1,3)-glucan synthesis, which is essential for fungal cell<br />
wall synthesis. Both in vitro and in vivo, the echinocandins<br />
are rapidly fungicidal against most Candida spp and fungi-<br />
static against Aspergillus spp., but they are not active<br />
against Zygomycetes, Cryptococcus ne<strong>of</strong>ormans,orFusarium<br />
spp. Because the drug target is not present in mammalian<br />
cells, adverse events are generally mild, including (for casp<strong>of</strong>ungin)<br />
local phlebitis, fever, abnormal liver function tests,<br />
and mild hemolysis. Oral bioavailability is suboptimal,<br />
thereby limiting use to the intravenous route. In a prospective<br />
randomized, double-blind trial comparing the efficacy<br />
and safety <strong>of</strong> casp<strong>of</strong>ungin with that <strong>of</strong> liposomal amphotericin<br />
B, casp<strong>of</strong>ungin was found to be as effective and generally<br />
better tolerated than liposomal amphotericin B when<br />
administered as empirical antifungal therapy in patients<br />
with persistent fever and neutropenia. 13 The newer echinocandins,<br />
micafungin and anidulafungin, also appear to be<br />
well-tolerated in pediatric oncology patients.<br />
Thus, amphotericin B, voriconazole, and casp<strong>of</strong>ungin have<br />
been well characterized for empirical antifungal therapy for<br />
persistent fever in high-risk neutropenic patients. For patients<br />
who remain neutropenic, antifungal therapy should<br />
be continued until the resolution <strong>of</strong> neutropenia. Persistence<br />
or recrudescence <strong>of</strong> fever should prompt a meticulous investigation<br />
for nonfungal infectious causes (e.g., bacterial or<br />
viral superinfections) or for a fungus that is resistant to<br />
initial empirical antifungal coverage.<br />
Recombinant Human Colony-Stimulating Factors<br />
The use <strong>of</strong> hematopoietic growth factors, such as<br />
granulocyte-macrophage colony-stimulating factor (GM-<br />
CSF) and granulocyte colony-stimulating factor (G-CSF),<br />
has been implemented widely in the management <strong>of</strong> neutropenia<br />
in patients who have undergone cancer and bone<br />
marrow transplantation. Both G-CSF and GM-CSF have<br />
been evaluated as adjuncts to chemotherapy to assist bone<br />
marrow reconstitution and to prevent neutropenia and reduce<br />
infectious complications. Primary use <strong>of</strong> G-CSF has<br />
been shown to lower the incidence <strong>of</strong> febrile neutropenic<br />
episodes by approximately 50% in three randomized studies<br />
in which the incidence <strong>of</strong> fever and neutropenia in the<br />
control group was greater than 40%. Primary administration<br />
<strong>of</strong> CSFs should be reserved for patients who are expected<br />
to experience rates <strong>of</strong> fever and neutropenia (� 40%)<br />
that are comparable with or greater than those seen in the<br />
control patients in these randomized trials.<br />
Use <strong>of</strong> GM-CSF has been less consistently helpful and has<br />
been associated with more adverse effects than those <strong>of</strong><br />
G-CSF. No study to date has demonstrated an advantage in<br />
rates <strong>of</strong> tumor response, fatal infections, or overall survival.<br />
Secondary use <strong>of</strong> G-CSF or GM-CSF in subsequent cycles <strong>of</strong><br />
chemotherapy has not demonstrated disease-free or overall<br />
survival benefits when the dose <strong>of</strong> chemotherapy was maintained.<br />
Therefore, reduction in chemotherapy dose should be<br />
considered the primary therapeutic option in a patient who<br />
experiences neutropenic fever or severe or prolonged neutropenia<br />
after the previous cycle <strong>of</strong> treatment. Finally, if G-CSF<br />
is administered within the first few days <strong>of</strong> chemotherapy<br />
for the initial induction or first postremission course for<br />
acute lymphoblastic leukemia (ALL), the duration <strong>of</strong> neutropenia<br />
<strong>of</strong> less than 1,000/mm 3 can be shortened by approximately<br />
1 week. These studies and guidelines for the use <strong>of</strong><br />
hematopoietic growth factors are summarized in a consensus<br />
paper by the <strong>American</strong> <strong>Society</strong> <strong>of</strong> <strong>Clinical</strong> <strong>Oncology</strong>. 14<br />
<strong>Clinical</strong> trials have reported conflicting results when attempting<br />
to evaluate whether the addition <strong>of</strong> CSFs to<br />
567
antibiotics in the treatment <strong>of</strong> febrile neutropenia improves<br />
patient outcomes. A meta-analysis <strong>of</strong> 13 studies revealed<br />
that the use <strong>of</strong> CSFs in patients with established fever and<br />
neutropenia reduces the amount <strong>of</strong> time spent in the hospital<br />
and the neutrophil recovery period. Overall mortality<br />
was not influenced by use <strong>of</strong> CSFs, but a marginally significant<br />
decrease in infection-related mortality was noted. 15<br />
Granulocyte Transfusions<br />
Although the principal <strong>of</strong> granulocyte transfusions in<br />
neutropenic patients with refractory infections is physiologically<br />
sound, the data supporting this clinical practice have<br />
been the subject <strong>of</strong> considerable controversy. Some early<br />
randomized studies demonstrated a clinical benefit for the<br />
PMN transfusion group compared with controls, 16 whereas<br />
others showed no overall benefit although demonstrating<br />
efficacy in certain subgroups <strong>of</strong> patients, 17 and still others<br />
reported no benefit at all. 18 Furthermore, the administration<br />
<strong>of</strong> granulocyte transfusions carries risks <strong>of</strong> alloimmunization<br />
and respiratory distress.<br />
Recent technical advances in the ability to collect substantially<br />
larger numbers <strong>of</strong> PMNs per pheresis have opened<br />
new potential for the therapeutic benefit <strong>of</strong> this adjunctive<br />
modality. The use <strong>of</strong> G-CSF to mobilize granulocytes in<br />
normal healthy donors has been shown to be safe and<br />
effective, allowing for the collection <strong>of</strong> substantially more<br />
PMNs per cycle <strong>of</strong> pheresis. 19 In addition to total number <strong>of</strong><br />
cells transfused, human leukocyte antigen (HLA) match is<br />
also thought to be an important factor in the efficacy <strong>of</strong><br />
granulocyte transfusions. Given the technical advances in<br />
G-CSF mobilization, WBC collection, and donor matching, a<br />
randomized trial is warranted to investigate the use <strong>of</strong><br />
granulocyte transfusions in neutropenic patients with refractory<br />
infections. However, the challenges posed by standardization<br />
<strong>of</strong> collection methods, HLA matching, dose <strong>of</strong><br />
granulocytes, and type <strong>of</strong> patients enrolled are daunting to a<br />
multicenter trial. Until such studies are available, selection<br />
<strong>of</strong> candidate patients to receive granulocyte transfusions<br />
must be made by an individual assessment <strong>of</strong> risk and<br />
benefit.<br />
Prophylactic Antimicrobial Therapy<br />
Many clinical trials have focused on use <strong>of</strong> prophylactic<br />
antibiotics to prevent infections in neutropenic patients. The<br />
oral fluoroquinolone antibiotics have been investigated because<br />
<strong>of</strong> their good bioavailability and broad activity against<br />
aerobic bacteria. A meta-analysis <strong>of</strong> published randomized<br />
controlled trials <strong>of</strong> quinolone prophylaxis (18 trials with<br />
1,408 patients) found that quinolone prophylaxis substantially<br />
reduced the incidence <strong>of</strong> gram-negative bacterial infections,<br />
microbiologically documented infections, total<br />
infections, and fevers, but did not alter the incidence <strong>of</strong><br />
gram-positive infections or infection-related deaths. 20 Two<br />
recent distinct studies evaluating the use <strong>of</strong> prophylactic<br />
oral lev<strong>of</strong>loxacin (500 mg daily) in patients receiving chemotherapy<br />
for either solid tumors or lymphoma 21 or hematologic<br />
malignancies 22 both showed a reduction in documented<br />
infections. Finally, a recent meta-analysis <strong>of</strong> antibiotic prophylaxis<br />
in neutropenic patients with cancer (95 trials<br />
performed between 1973 and 2004) concluded that antibiotic<br />
prophylaxis (various antibiotic regimens) substantially decreased<br />
the risk for death compared with placebo or no<br />
568<br />
treatment. 23 When trials that used quinolone prophylaxis<br />
(52 trials) were separately analyzed, there was a substantial<br />
reduction in the risk for all-cause mortality, as well as<br />
infection-related mortality, fever, clinically documented infections,<br />
and microbiologically documented infections. Although<br />
these results are very encouraging, many <strong>of</strong> these<br />
studies reported increasing rates <strong>of</strong> antimicrobial resistance<br />
to quinolones. Thus, if quinolone prophylaxis is implemented,<br />
vigilant monitoring <strong>of</strong> the incidence <strong>of</strong> bacteremia<br />
(specifically gram-negative bacteremia) is mandatory to first<br />
detect the loss <strong>of</strong> efficacy <strong>of</strong> fluoroquinolone prophylaxis.<br />
The use <strong>of</strong> fluconazole as antifungal prophylaxis has been<br />
conducted in adult patients with leukemia who have undergone<br />
bone marrow transplantation. 24 Although a decrease in<br />
fungal colonization and superficial infections was noted, a<br />
reduction in systemic fungal infections and associated mortality<br />
was identified only in the patients undergoing bone<br />
marrow transplantation.<br />
Augmentation <strong>of</strong> Mucocutaneous Barriers and<br />
Commensal Gut Microbiota<br />
Keratinocyte growth factor, a member <strong>of</strong> the family <strong>of</strong><br />
fibroblast growth factors, has been shown to reduced the<br />
duration and severity <strong>of</strong> oral mucositis after intensive chemotherapy.<br />
25 A preliminary analysis <strong>of</strong> blood-borne infections<br />
showed a lower incidence among patients receiving<br />
palifermin than among those receiving placebo. Future<br />
investigations, however, are merited to assess the ability <strong>of</strong><br />
palifermin therapy to decrease the complications caused by<br />
systemic infection.<br />
Recent studies have shown that the GI microbiota plays a<br />
crucial role in preventing overgrowth <strong>of</strong> pathogenic microbes—by<br />
directly inhibiting the growth <strong>of</strong> specific pathogens<br />
and/or by stimulation <strong>of</strong> GI mucosal effectors (e.g.,<br />
antimicrobial proteins) that act to clear invading pathogens.<br />
Although probiotic bacterial strains have been used to treat<br />
inflammatory bowel disease, necrotizing enterocolitis, and<br />
enteric infections human patients, no studies have been<br />
performed in pediatric patients with cancer. The major<br />
concern is that introduction <strong>of</strong> enteric bacteria may result in<br />
dissemination <strong>of</strong> the introduced strain. The importance <strong>of</strong><br />
the microbiota in driving protective immune responses during<br />
GI infection is best illustrated by the finding that<br />
restoring signaling through innate immune receptors in<br />
antibiotic-treated mice can protect from intestinal infections<br />
(i.e., lipopolysaccharide activation <strong>of</strong> TLR4 or flagellin stimulation<br />
<strong>of</strong> TLR5 to prevent vancomycin-resistant Enterococci<br />
dissemination.). 26<br />
Conclusion<br />
ANDREW Y. KOH<br />
Pediatric patients with cancer with prolonged neutropenia<br />
have increased risk for severe, recurrent, or new bacterial<br />
and fungal infections. Although prompt initiation <strong>of</strong> empirical<br />
antibacterial antibiotics when the neutropenic patient<br />
with cancer becomes febrile has lead to substantial improvement<br />
in morbidity and mortality associated with bacterial<br />
and fungal infections, infectious complications still persist.<br />
Currently, appropriate administration <strong>of</strong> antibacterial and<br />
antifungal antibiotics remains standard <strong>of</strong> care. Adjunctive<br />
use <strong>of</strong> agents that promote healing <strong>of</strong> mucosal damage, use<br />
<strong>of</strong> probiotics or prebiotics to reestablish gut microbiota<br />
homeostasis, and granulocyte infusions <strong>of</strong>fer promise, but<br />
more definitive studies are necessary.
FEBRILE NEUTROPENIA IN PEDIATRIC CANCER<br />
Author’s Disclosures <strong>of</strong> Potential Conflicts <strong>of</strong> Interest<br />
Author<br />
Andrew Y. Koh*<br />
*No relevant relationships to disclose.<br />
Employment or<br />
Leadership<br />
Positions<br />
Consultant or<br />
Advisory Role<br />
1. Schimpff SC, Young VM, Greene WH, et al. Origin <strong>of</strong> infection in acute<br />
nonlymphocytic leukemia. Significance <strong>of</strong> hospital acquisition <strong>of</strong> potential<br />
pathogens. Ann Intern Med. 1972;77:707-714.<br />
2. Bodey GP, Buckley M, Sathe YS, et al. Quantitative relationships<br />
between circulating leukocytes and infection in patients with acute leukemia.<br />
Ann Intern Med. 1966;64:328-340.<br />
3. Pizzo PA. After empiric therapy: what to do until the granulocyte comes<br />
back. Rev Infect Dis. 1987;9:214-219.<br />
4. Talcott JA, Siegel RD, Finberg R, et al. Risk assessment in cancer<br />
patients with fever and neutropenia: a prospective, two-center validation <strong>of</strong> a<br />
prediction rule. J Clin Oncol. 1992;10:316-322.<br />
5. Rack<strong>of</strong>f WR, Gonin R, Robinson C, et al. Predicting the risk <strong>of</strong> bacteremia<br />
in childen with fever and neutropenia. J Clin Oncol. 1996;14:919-924.<br />
6. Freifeld AG, Bow EJ, Sepkowitz KA, et al. <strong>Clinical</strong> practice guideline for<br />
the use <strong>of</strong> antimicrobial agents in neutropenic patients with cancer: 2010<br />
Update by the Infectious Diseases <strong>Society</strong> <strong>of</strong> America. Clin Infect Dis.<br />
52:427-431.<br />
7. Pizzo PA, Robichaud KJ, Gill FA, et al. Duration <strong>of</strong> empiric antibiotic<br />
therapy in granulocytopenic patients with cancer. Am J Med. 1979;67:194-<br />
200.<br />
8. Pizzo PA, Ladisch S, Ribichaud K. Treatment <strong>of</strong> gram-positive septicemia<br />
in cancer patients. Cancer. 1980;45:206-207.<br />
9. Pizzo PA, Robichaud KJ, Gill FA, et al. Empiric antibiotic and antifungal<br />
therapy for cancer patients with prolonged fever and granulocytopenia. Am J<br />
Med. 1982;72:101-111.<br />
10. Walsh TJ, Finberg RW, Arndt C, et al. Liposomal amphotericin B for<br />
empirical therapy in patients with persistent fever and neutropenia. National<br />
Institute <strong>of</strong> Allergy and Infectious Diseases Mycoses Study Group. N Engl<br />
J Med. 1999;340:764-771.<br />
11. Rex JH, Bennett JE, Sugar AM, et al. A randomized trial comparing<br />
fluconazole with amphotericin B for the treatment <strong>of</strong> candidemia in patients<br />
without neutropenia. Candidemia Study Group and the National Institute.<br />
N Engl J Med. 1994;331:1325-1330.<br />
12. Walsh TJ, Pappas P, Winston DJ, et al. Voriconazole compared with<br />
liposomal amphotericin B for empirical antifungal therapy in patients with<br />
neutropenia and persistent fever. N Engl J Med. 2002;346:225-234.<br />
13. Walsh TJ, Teppler H, Donowitz GR, et al. Casp<strong>of</strong>ungin versus liposomal<br />
amphotericin B for empirical antifungal therapy in patients with<br />
persistent fever and neutropenia. N Engl J Med. 2004;351:1391-1402.<br />
Stock<br />
Ownership Honoraria<br />
REFERENCES<br />
Research<br />
Funding<br />
Expert<br />
Testimony<br />
Other<br />
Remuneration<br />
14. Smith TJ, Khatcheressian J, Lyman GH, et al. 2006 update <strong>of</strong> recommendations<br />
for the use <strong>of</strong> white blood cell growth factors: an evidence-based<br />
clinical practice guideline. J Clin Oncol. 2006;24:3187-3205.<br />
15. Clark OA, Lyman GH, Castro AA, et al. Colony-stimulating factors for<br />
chemotherapy-induced febrile neutropenia: a meta-analysis <strong>of</strong> randomized<br />
controlled trials. J Clin Oncol. 2005;23:4198-4214.<br />
16. Herzig RH, Herzig GP, Graw RG, Jr., et al. Successful granulocyte<br />
transfusion therapy for gram-negative septicemia: a prospectively randomized<br />
controlled study. N Engl J Med. 1977;296:701-705.<br />
17. Alavi JB, Root RK, Djerassi I, et al. A randomized clinical trial <strong>of</strong><br />
granulocyte transfusions for infection in acute leukemia. N Engl J Med.<br />
1977;296:706-711.<br />
18. Winston DJ, Ho WG, Gale RP. Therapeutic granulocyte transfusions<br />
for documented infections. A controlled trial in ninety-five infectious granulocytopenic<br />
episodes. Ann Intern Med. 1982;97:509-515.<br />
19. Bensinger WI, Price TH, Dale DC, et al. The effects <strong>of</strong> daily recombinant<br />
human granulocyte colony-stimulating factor administration on normal<br />
granulocyte donors undergoing leukapheresis. Blood. 1993;81:1883-1888.<br />
20. Engels EA, Lau J, Barza M. Efficacy <strong>of</strong> quinolone prophylaxis in<br />
neutropenic cancer patients: a meta-analysis. J Clin Oncol. 1998;16:1179-<br />
1187.<br />
21. Cullen M, Steven N, Billingham L, et al. Antibacterial prophylaxis after<br />
chemotherapy for solid tumors and lymphomas. N Engl J Med. 2005;353:988-<br />
998.<br />
22. Reuter S, Kern WV, Sigge A, et al. Impact <strong>of</strong> fluoroquinolone prophylaxis<br />
on reduced infection-related mortality among patients with neutropenia<br />
and hematologic malignancies. Clin Infect Dis. 2005;40:1087-1093.<br />
23. Gafter-Gvili A, Fraser A, Paul M, et al. Meta-analysis: Antibiotic<br />
prophylaxis reduces mortality in neutropenic patients. Ann Intern Med.<br />
2005;142:979-995.<br />
24. Goodman JL, Winston DJ, Greenfield RA, et al. A controlled trial <strong>of</strong><br />
fluconazole to prevent fungal infections in patients undergoing bone marrow<br />
transplantation. N Engl J Med. 1992;326:845-851.<br />
25. Spielberger R, Stiff P, Bensinger W, et al. Palifermin for oral mucositis<br />
after intensive therapy for hematologic cancers. N Engl J Med. 2004;351:<br />
2590-2598.<br />
26. Kinnebrew MA, Ubeda C, Zenewicz LA, et al. Bacterial flagellin<br />
stimulates Toll-like receptor 5-dependent defense against vancomycinresistant<br />
Enterococcus infection. J Infect Dis. 2010;201:534-543.<br />
569
Initial Management <strong>of</strong> Low-Risk Pediatric<br />
Fever and Neutropenia: Efficacy and<br />
Safety, Costs, Quality-<strong>of</strong>-Life Considerations,<br />
and Preferences<br />
Overview: Initial management options for pediatric low-risk<br />
fever and neutropenia (FN) include outpatient compared with<br />
inpatient management and oral compared with intravenous<br />
therapy. Single-arm and randomized trials have been conducted<br />
in children. Meta-analyses provide support for the<br />
equivalence <strong>of</strong> outpatient and inpatient approaches. Outpatient<br />
oral management may be associated with a higher risk <strong>of</strong><br />
readmission compared with outpatient intravenous management<br />
in children with FN, although other outcomes such as<br />
treatment failure and discontinuation <strong>of</strong> the regimen because<br />
<strong>of</strong> adverse effects were similar. Importantly, there have been<br />
no reported deaths among low-risk children treated as outpatients<br />
or with oral antibiotics. Costs, whether derived directly<br />
or through cost-effectiveness analysis, are consistently reduced<br />
when an outpatient approach is used. Quality <strong>of</strong> life<br />
FEVER AND neutropenia (FN) is a common and potentially<br />
fatal complication <strong>of</strong> intensive chemotherapy in<br />
children with cancer. 1 Over the last 4 decades, outcomes <strong>of</strong><br />
FN have improved dramatically related to hospitalization<br />
and prompt initiation <strong>of</strong> empiric antibiotic therapy. 2,3 However,<br />
children with FN are heterogeneous. 4 Based on characteristics<br />
at presentation, some <strong>of</strong> these children are<br />
predicted to be at low risk <strong>of</strong> mortality and adverse events,<br />
and these children may be managed less intensively. 5 Currently,<br />
there are several validated prediction rules designed<br />
to identify the low-risk child with FN. 6 These rules vary in<br />
terms <strong>of</strong> their specific characteristics, but all rules identify<br />
children with an anticipated short duration <strong>of</strong> neutropenia.<br />
The ability to identify such patients has led to the consideration<br />
and design <strong>of</strong> clinical trials to evaluate less intensive<br />
strategies for low-risk FN and, ultimately, to their implementation<br />
in clinical practice.<br />
The most relevant lesser intensity strategies relate to site<br />
<strong>of</strong> care (outpatient compared with inpatient) and mode <strong>of</strong><br />
antibiotic administration (oral compared with intravenous).<br />
In adult patients with low-risk FN, oral therapy is associated<br />
with similar outcomes to intravenous treatment, 7-9 and<br />
there is increasing evidence that outpatient management <strong>of</strong><br />
this population is a safe approach. 10,11 As a result, adultbased<br />
guidelines for the management <strong>of</strong> FN now recommend<br />
outpatient management and oral antibiotics for selected<br />
low-risk patients. 12,13 However, there has been reluctance to<br />
adopt these strategies for children with FN. 14 The following<br />
sections summarize the existing literature related to outpatient<br />
management and oral antibiotic therapy in children<br />
with FN with respect to efficacy and safety, costs, and QoL<br />
and preferences.<br />
Efficacy and Safety <strong>of</strong> Outpatient Management and<br />
Oral Antibiotic Administration<br />
Outpatient Compared with Inpatient Management<br />
Successful outpatient management <strong>of</strong> children with FN is<br />
predicated on several important considerations related to<br />
570<br />
By Lillian Sung, MD, PhD<br />
(QoL) and preferences should be considered in order to<br />
evaluate different strategies, plan programs, and anticipate<br />
uptake <strong>of</strong> outpatient programs. Using parent-proxy report,<br />
child QoL is consistently higher with outpatient approaches,<br />
although research evaluating child self-report is limited. Preferences<br />
incorporate estimated QoL, but, in addition, factor in<br />
issues such as costs, fear, anxiety, and logistical issues. Only<br />
approximately 50% <strong>of</strong> parents prefer outpatient management.<br />
Future research should develop tools to facilitate outpatient<br />
care and to measure caregiver burden associated with this<br />
strategy. Additional work should also focus on eliciting child<br />
preferences for outpatient management. Finally, studies <strong>of</strong><br />
effectiveness <strong>of</strong> an ambulatory approach in the real-world<br />
setting outside <strong>of</strong> clinical trials are important.<br />
the child, family, and local infrastructure. First, the child<br />
must be identified as having low-risk features using one <strong>of</strong><br />
the validated clinical prediction rules. 6 Family considerations<br />
include a history <strong>of</strong> compliance with other medical<br />
procedures, rapid accessibility to hospital, ability to communicate<br />
reliably by telephone, and ability to provide continual<br />
observation and assessment <strong>of</strong> the child. The local health<br />
care team must also have the infrastructure to support<br />
outpatient management, which includes clinics that can<br />
assess outpatients and pr<strong>of</strong>essionals who can be in contact<br />
with families on a regular basis and who are readily available<br />
if problems arise.<br />
Outpatient therapy can be initiated at the onset <strong>of</strong> FN<br />
or after a short period <strong>of</strong> inpatient treatment followed by<br />
discharge to the home (step-down management). Outpatient<br />
programs may deliver antibiotics by different routes <strong>of</strong><br />
administration, including entirely oral administration, entirely<br />
intravenous administration, or step-down management<br />
in which treatment begins with intravenous and then<br />
transitions to oral therapy. In the outpatient setting,<br />
follow-up assessments may occur in clinics, in the home, by<br />
telephone, or a combination <strong>of</strong> approaches. The frequency <strong>of</strong><br />
follow-up assessments may vary considerably depending on<br />
the specific outpatient model <strong>of</strong> care.<br />
There has been a single meta-analysis <strong>of</strong> randomized<br />
controlled trials (RCTs) that compared outpatient compared<br />
with inpatient management <strong>of</strong> FN. 15 Six studies were included:<br />
four were adult studies and two were pediatric<br />
studies. Two studies consisted <strong>of</strong> an entirely outpatient<br />
approach whereas four consisted <strong>of</strong> a step-down approach in<br />
From the Division <strong>of</strong> Haematology/<strong>Oncology</strong>, and Program in Child Health Evaluative<br />
Sciences, The Hospital for Sick Children, Toronto, ON, Canada.<br />
Author’s disclosure <strong>of</strong> potential conflicts <strong>of</strong> interest are found at the end <strong>of</strong> this article.<br />
Address reprint requests to Lillian Sung MD, PhD, Division <strong>of</strong> Haematology/<strong>Oncology</strong>,<br />
The Hospital for Sick Children, 555 University Avenue, Toronto, ON M5G1X8; email:<br />
lillian.sung@sickkids.ca<br />
© <strong>2012</strong> by <strong>American</strong> <strong>Society</strong> <strong>of</strong> <strong>Clinical</strong> <strong>Oncology</strong>.<br />
1092-9118/10/1-10
TREATMENT APPROACHES IN CHILDREN<br />
which the patient was admitted to hospital and then discharged<br />
early while still receiving empiric antibiotics for FN.<br />
Outpatient management was associated with a similar risk<br />
<strong>of</strong> treatment failure compared with inpatient management<br />
(rate ratio [RR] 0.81, 95% CI 0.55 to 1.28, p � 0.28) where<br />
RR � 1 favored inpatient care. In interpreting this RR, it is<br />
important to realize that failure was biased against outpatient<br />
care because readmission, a criterion for failure, is only<br />
applicable to outpatients. There was no difference in mortality<br />
(RR 1.11, 95% CI 0.41 to 3.05, p � 0.83). In a stratified<br />
analysis <strong>of</strong> the two pediatric studies, 16,17 results were similar<br />
to the overall analysis.<br />
Although this review provides evidence for the safety <strong>of</strong><br />
outpatient management, it includes only two studies in<br />
children, and, in total, 278 pediatric subjects were studied.<br />
Consequently, this analysis may have had insufficient power<br />
to evaluate outpatient care in pediatric FN. Thus, we subsequently<br />
conducted a systematic review in which we evaluated<br />
data from all prospective trials in pediatric FN that<br />
studied a homogeneous, initial antibiotic regimen. 18 There<br />
were 16 trials in which either outpatient or inpatient management<br />
were established within the first 24 hours <strong>of</strong><br />
treatment-initiation in low-risk FN. There was no increase<br />
in treatment failure (including modification <strong>of</strong> antibiotics)<br />
with outpatient management in comparison with inpatient<br />
management (15% compared with 27%, p � 0.04). The rate<br />
<strong>of</strong> adverse events leading to antibiotic discontinuation was<br />
similar (1% compared with 2%, p � 0.39). Among the 953<br />
outpatients, there were no infection-related deaths.<br />
In summary, when combining data from prospective observational<br />
and randomized trials in pediatric patients,<br />
similar outcomes were observed between those treated as<br />
outpatients or inpatients. However, outpatient management<br />
should only be instituted if the child can be confidently<br />
classified as low-risk, and if the social circumstances and<br />
local infrastructure support ambulatory management.<br />
Consequently, if the child, family, and health care facility<br />
characteristics support outpatient care, then ambulatory<br />
management may be <strong>of</strong>fered. Outpatient FN programs in<br />
KEY POINTS<br />
● Prospective single-arm and randomized trials have<br />
provided evidence for the efficacy and safety <strong>of</strong> outpatient<br />
management and oral antibiotic administration<br />
for low-risk children with fever and neutropenia.<br />
● Outpatient oral management may be associated with<br />
a higher risk <strong>of</strong> readmission in children, although<br />
other outcomes such as treatment failure and discontinuation<br />
<strong>of</strong> the regimen because <strong>of</strong> adverse effects<br />
are similar.<br />
● Outpatient management is cost-saving when compared<br />
to an inpatient or an early discharge strategy.<br />
● Anticipated child quality <strong>of</strong> life as reported by parents<br />
is higher with outpatient strategies.<br />
● When comparing inpatient with outpatient management<br />
<strong>of</strong> low-risk fever and neutropenia, the most<br />
preferred option is variable, and frequently, parents<br />
and children will prefer inpatient care.<br />
children vary considerably in terms <strong>of</strong> frequency <strong>of</strong> follow up<br />
and nature <strong>of</strong> follow up (for example, telephone contact<br />
compared with a clinic visit). Future work should consider<br />
minimal and optimal approaches to follow outpatient children<br />
with FN.<br />
Oral versus Parenteral Antibiotic Administration<br />
Regardless <strong>of</strong> whether low-risk children with FN are<br />
treated in the outpatient or inpatient setting, another question<br />
relates to the optimal mode <strong>of</strong> antibiotic administration.<br />
Oral antibiotic administration may be desirable because it<br />
facilitates outpatient management, is more convenient, and<br />
is usually less expensive compared with intravenous antibiotic<br />
administration. However, there are several unique issues<br />
with oral medication administration in young children.<br />
Issues to consider include the likelihood that the child can<br />
and will accept oral therapy in available formulations given<br />
the child’s level <strong>of</strong> cooperation. First, very young children<br />
may not be able to swallow pills or capsules, and, thus, oral<br />
antibiotics are only feasible in this age group if the medication<br />
is available in a liquid formulation. Second, taste <strong>of</strong> oral<br />
medication is a much more important issue in young children<br />
compared to adults, as an unpalatable drug may be<br />
refused. Third, children may refuse all oral medications<br />
regardless <strong>of</strong> taste, particularly if they feel unwell. Furthermore,<br />
the likelihood that the child will accept oral therapy<br />
may change if nausea, vomiting, or mucositis are present.<br />
Oral antibiotic administration may be initiated at the onset<br />
<strong>of</strong> the FN episode or following a short period <strong>of</strong> parenteral<br />
administration before transitioning to oral administration<br />
(step-down management).<br />
Two meta-analyses <strong>of</strong> RCTs evaluated oral and intravenous<br />
antibiotic administration for FN. One study <strong>of</strong> 2,770<br />
patients included both outpatients and inpatients 7 whereas<br />
the second study <strong>of</strong> 1,595 patients was restricted to outpatients.<br />
15 Neither study restricted their review to low-risk FN<br />
patients. Both reviews showed similar results with no difference<br />
in treatment failure (including modification), overall<br />
mortality, or adverse effects <strong>of</strong> antibiotics. These findings<br />
were demonstrated both among combined adult and pediatric<br />
analyses, as well as in stratified analyses <strong>of</strong> pediatric<br />
studies alone. However, a stratified analysis <strong>of</strong> five pediatric<br />
RCTs demonstrated that intravenous outpatient management<br />
was associated with a lower rate <strong>of</strong> readmission<br />
compared with oral outpatient management (RR 0.52, 95%<br />
CI 0.24 to 1.09, p � 0.08). 15<br />
More information about the efficacy and safety <strong>of</strong> oral<br />
antibiotic administration was derived from an analysis <strong>of</strong><br />
prospective pediatric trial data comparing oral and parenteral<br />
antibiotic therapy initiated within 24 hours <strong>of</strong> treatment<br />
initiation in low-risk FN. 18 Oral antibiotics used in<br />
these studies were fluoroquinolone monotherapy (7 studies<br />
with 581 patients), fluoroquinolone and amoxicillinclavulanate<br />
(3 studies with 159 patients), and cefixime<br />
(1 study with 45 patients). The review found no difference in<br />
treatment failure (including modification) among children<br />
who received oral compared with intravenous antibiotic<br />
therapy (20% compared with 22%, p � 0.68). The rate <strong>of</strong><br />
antibiotic discontinuation because <strong>of</strong> adverse events was<br />
similar between the oral and intravenous regimens (2%<br />
compared with 1%, p � 0.73). There were no infectionrelated<br />
deaths among the 676 children given oral antibiotics.<br />
To summarize, there were more readmissions among chil-<br />
571
dren treated with oral therapy compared with intravenous<br />
therapy in the outpatient setting, although other outcomes,<br />
including treatment failure and adverse events, were similar;<br />
and there were no infection-related deaths among lowrisk<br />
pediatric patients.<br />
Consequently, if oral antibiotics can confidently be administered,<br />
oral antibiotic therapy may be used in selected<br />
low-risk children with FN, either in the outpatient or inpatient<br />
settings. One approach to outpatient oral antibiotic<br />
administration may be to administer one dose <strong>of</strong> oral antibiotics<br />
during the initial health care encounter and only<br />
discharge home with oral therapy if that dose can be<br />
ingested. Costs, feasibility, and patient/family preferences<br />
should influence the route <strong>of</strong> antibiotic administration for<br />
children with low-risk FN in both the outpatient and inpatient<br />
settings.<br />
Costs<br />
The adult literature has consistently demonstrated that<br />
outpatient management is substantially cheaper than inpatient<br />
management <strong>of</strong> FN, primarily related to the costs <strong>of</strong><br />
hospitalization. 19-21 Limited data are available on costs in<br />
the pediatric setting. 16,17 However, the literature supports a<br />
similar cost-saving associated with ambulatory management<br />
in children. 22,23<br />
A pediatric cost-utility model has been created that includes<br />
event probabilities, costs, and QoL estimates. 24 Outpatient<br />
management was the most cost-effective approach,<br />
with outpatient intravenous antibiotic administration being<br />
more cost-effective than outpatient oral antibiotic administration<br />
because <strong>of</strong> the higher rate <strong>of</strong> readmission among<br />
those who receive oral antibiotics and better QoL seen<br />
with intravenous therapy (see below). However, within the<br />
plausible range <strong>of</strong> some inputs into the model, outpatient<br />
management with oral therapy could become the most<br />
cost-effective strategy. Entire treatment in hospital with<br />
intravenous antibiotics was the least cost-effective strategy;<br />
it was more expensive and less effective than an early<br />
discharge approach.<br />
These data suggest that from a cost perspective, outpatient<br />
management with either intravenous or oral antibiotics<br />
should be used if the infrastructure is in place to support<br />
this model <strong>of</strong> care, and, if the child, family, and system<br />
factors permit an ambulatory approach.<br />
QoL and Preferences<br />
QoL considerations and preferences become more important<br />
when strategies such as outpatient management <strong>of</strong> FN<br />
have advantages and trade-<strong>of</strong>fs. The advantages <strong>of</strong> outpatient<br />
therapy include substantial cost-saving for the health<br />
care system and reduction in nosocomial infections. However,<br />
the trade-<strong>of</strong>fs include increased responsibility and<br />
burden for parents and other care providers, the need to<br />
monitor children remotely from the health care center, and<br />
the need for follow-up visits. It is important to evaluate QoL<br />
and preferences for several reasons. First, this knowledge<br />
allows strategies to be compared using approaches such as<br />
cost-effectiveness analyses. Second, this information can<br />
help to develop outpatient programs and may influence how<br />
programs are structured. In other words, this information<br />
may be used to determine whether or which outpatient FN<br />
programs might be successful in a given context or practice<br />
setting.<br />
572<br />
LILLIAN SUNG<br />
QoL and preferences are two distinct constructs, although<br />
QoL may influence preferences. QoL is complex in pediatric<br />
medicine because QoL considerations may include the index<br />
child, siblings, and parents/providers. Outpatient management<br />
<strong>of</strong> children with FN may be associated with improved<br />
QoL for children 25 although the effect on parents is unclear.<br />
26 In a cross-sectional study in which we interviewed<br />
parents <strong>of</strong> children with cancer and health care providers<br />
who care for these children, we asked respondents to compare<br />
just two options, namely oral outpatient management<br />
compared with inpatient intravenous management. 26 Both<br />
parents and health care pr<strong>of</strong>essionals believed child QoL<br />
would be better at home compared with the hospital setting.<br />
However, health care pr<strong>of</strong>essionals, when compared to parents,<br />
overestimated QoL for children at home and underestimated<br />
QoL for parents in the hospital setting.<br />
In order to gain more insight, we asked parents <strong>of</strong> children<br />
with cancer and the children themselves to anticipate child<br />
QoL in four different scenarios, namely outpatient oral<br />
therapy, outpatient intravenous therapy, early discharge,<br />
and inpatient intravenous management. 27 The outpatient<br />
regimen in these studies always included three times weekly<br />
clinic visits as the standard scenario because almost all<br />
published studies <strong>of</strong> outpatient management in pediatric<br />
cancer include at least this frequency <strong>of</strong> clinic visits. Among<br />
parent respondents, early discharge and outpatient management<br />
with intravenous antibiotics were associated with<br />
higher anticipated child QoL (score <strong>of</strong> 5.9 each on a 10-point<br />
visual analog scale), whereas outpatient oral therapy was<br />
associated with the lowest anticipated child QoL (score <strong>of</strong><br />
4.7). In other words, parents were concerned about the effect<br />
<strong>of</strong> oral medication administration on the child’s QoL. For<br />
child self-respondents, all <strong>of</strong> whom were 12 years or older,<br />
early discharge was anticipated to be associated with the<br />
highest child QoL. It is important to note that there are very<br />
little data about child self-report <strong>of</strong> QoL in different FN<br />
health states.<br />
Preference will likely include QoL considerations but<br />
also may include factors such as fear and anxiety, costs,<br />
and logistics. We have taken several approaches to the<br />
elicitation <strong>of</strong> preferences. The simplest approach is one in<br />
which we have asked parents and children their preferred<br />
management strategy. Using this approach, consistently,<br />
approximately 50% <strong>of</strong> parents preferred outpatient management<br />
compared with inpatient management. 26,27 When<br />
asked to rank the most preferred strategy among the four<br />
options, we found that the most common top-ranked choice<br />
was inpatient intravenous management among both parents<br />
and children. 27<br />
We have also used other techniques to gain insight into<br />
parent decision making in the context <strong>of</strong> management options<br />
for low-risk FN. More specifically, we evaluated preferences<br />
using a threshold technique and conjoint analysis.<br />
With the threshold technique, one measures strength <strong>of</strong><br />
preference for a treatment option by systematically changing<br />
one attribute <strong>of</strong> a treatment option (for example, increasing<br />
or decreasing the number <strong>of</strong> clinic visits per week<br />
associated with outpatient care) until the respondent gives<br />
up their initially preferred option. Using this technique, we<br />
described strength <strong>of</strong> preferences for inpatient parenteral<br />
compared with outpatient oral management. 26 We found<br />
that, in general, in order to accept outpatient oral management,<br />
parents would not tolerate clinic visits more than
TREATMENT APPROACHES IN CHILDREN<br />
three times per week or a rate <strong>of</strong> readmission <strong>of</strong> more than<br />
15%. We also found that stronger preference for outpatient<br />
therapy was associated with higher anticipated QoL for the<br />
parent and child at home relative to hospital. 26<br />
We also used a conjoint analysis technique to assess<br />
preferences and decision making for outpatient management<br />
<strong>of</strong> low-risk FN. 28 Conjoint analysis is an emerging<br />
approach to the measurement <strong>of</strong> preferences in the face <strong>of</strong><br />
multiple trade-<strong>of</strong>fs in health care. In contrast to the threshold<br />
technique, conjoint analysis allows evaluation <strong>of</strong> multiple<br />
attributes concurrently or conjointly. 29 Using this<br />
technique, we quantified the relative importance <strong>of</strong> attributes<br />
associated with two treatment options—outpatient<br />
oral compared with inpatient intravenous management.<br />
Parents would be willing to accept only 2.1 (95% CI 1.1 to<br />
3.2) clinic visits weekly in order to accept outpatient management.<br />
With clinic visits three times weekly and a 7.5%<br />
chance <strong>of</strong> readmission, the probability <strong>of</strong> parents accepting<br />
an outpatient approach was only 43% (95% CI 39% to<br />
48%). 28<br />
Finally, we used a qualitative approach to better understand<br />
parental perspectives and how this can influence<br />
preferences. 30 The major themes identified when choosing<br />
between outpatient oral and inpatient intravenous therapy<br />
included convenience/disruptiveness for the family, concerns<br />
related to physical health <strong>of</strong> the child, emotional well-being<br />
for the child, and modifiers <strong>of</strong> parental decision making.<br />
Thus, we demonstrated that many parents and children<br />
prefer inpatient management, although anticipated QoL on<br />
the aggregate level was higher with early discharge and<br />
outpatient parenteral strategies. It is relatively straightforward<br />
to explain these differences. Preferences for a treatment<br />
option in this context incorporate considerations other<br />
than child QoL, such as parent QoL, convenience, costs,<br />
safety, and anxiety.<br />
Implementation <strong>of</strong> Outpatient FN<br />
We have used these findings to develop an outpatient<br />
low-risk FN program in our own health care setting. We<br />
have selected a very low-risk population, based on previ-<br />
Author’s Disclosure <strong>of</strong> Potential Conflicts <strong>of</strong> Interest<br />
Author<br />
Lillian Sung*<br />
*No relevant relationships to disclose.<br />
Employment or<br />
Leadership<br />
Positions<br />
Consultant or<br />
Advisory Role<br />
1. Bodey GP, Buckley M, Sathe YS, et al. Quantitative relationships<br />
between circulating leukocytes and infection in patients with acute leukemia.<br />
Ann Intern Med. 1966;64:328-340.<br />
2. Schimpff S, Satterlee W, Young VM, et al. Empiric therapy with<br />
carbenicillin and gentamicin for febrile patients with cancer and granulocytopenia.<br />
N Engl J Med. 1971;284:1061-1065.<br />
3. Pizzo PA. Management <strong>of</strong> fever in patients with cancer and treatmentinduced<br />
neutropenia. N Engl J Med. 1993;328:1323-1332.<br />
4. Rondinelli PI, Ribeiro Kde C, de Camargo B. A proposed score for<br />
predicting severe infection complications in children with chemotherapyinduced<br />
febrile neutropenia. J Pediatr Hematol Oncol. 2006;28:665-670.<br />
5. Ammann RA, Bodmer N, Hirt A, et al. Predicting adverse events in<br />
children with fever and chemotherapy-induced neutropenia: the prospective<br />
multicenter SPOG 2003 FN study. J Clin Oncol. 2010;28:2008-2014.<br />
6. Phillips B, Wade R, Stewart LA, et al. Systematic review and meta-<br />
ously validated clinical prediction rules, to ensure that the<br />
risk <strong>of</strong> readmission is very low and to build confidence in the<br />
outpatient program among families and health care pr<strong>of</strong>essionals.<br />
We administer empiric antibiotics orally because we<br />
have found it difficult to initiate empiric antibiotics intravenously<br />
in a timely fashion. We administer a single dose <strong>of</strong><br />
oral antibiotic in the emergency room or clinic to ensure that<br />
the child can ingest the oral antibiotic to be used in the<br />
outpatient setting before discharge to home. The current<br />
structure includes clinic visits two or three times weekly. A<br />
system <strong>of</strong> home visits by nursing staff has not shown to be<br />
feasible, and, consequently, the patients are monitored between<br />
clinic visits with daily phone calls by a health care<br />
pr<strong>of</strong>essional. There are plans for evaluation <strong>of</strong> the program<br />
on a regular basis, although it is too soon to provide such a<br />
report.<br />
Conclusion<br />
In summary, a series <strong>of</strong> prospective observational studies<br />
and RCTs have been conducted in pediatrics that support<br />
the efficacy and safety <strong>of</strong> outpatient care and oral antibiotic<br />
administration as initial treatment for children with lowrisk<br />
FN. Costs are clearly lower with an ambulatory approach.<br />
However, QoL is more complicated because child<br />
and parent QoL considerations are important, and incremental<br />
QoL at home compared with the hospital may not be<br />
the same for both respondent types. Preferences are also<br />
important to evaluate because this information may be used<br />
to plan outpatient programs and to anticipate uptake <strong>of</strong><br />
these programs.<br />
Future work may include the development <strong>of</strong> tools to ease<br />
care in the outpatient setting and to measure caregiver<br />
burden associated with this therapy. Additional work should<br />
also focus on eliciting child QoL and preferences for outpatient<br />
management <strong>of</strong> low-risk FN given that these may differ<br />
substantially from parent-proxy responses. 31,32 Finally, the<br />
study <strong>of</strong> the effectiveness <strong>of</strong> an ambulatory approach in the<br />
real-world setting outside <strong>of</strong> clinical trials is important to<br />
fully understand the effect <strong>of</strong> different management strategies<br />
for initial treatment <strong>of</strong> low-risk pediatric FN.<br />
Stock<br />
Ownership Honoraria<br />
REFERENCES<br />
Research<br />
Funding<br />
Expert<br />
Testimony<br />
Other<br />
Remuneration<br />
analysis <strong>of</strong> the discriminatory performance <strong>of</strong> risk prediction rules in febrile<br />
neutropaenic episodes in children and young people. Eur J Cancer. 2010;46:<br />
2950-2964.<br />
7. Vidal L, Paul M, Ben-Dor I, et al. Oral versus intravenous antibiotic<br />
treatment for febrile neutropenia in cancer patients. Cochrane Database Syst<br />
Rev. 2004;4:CD003992.<br />
8. Freifeld A, Marchigiani D, Walsh T, et al. A double-blind comparison <strong>of</strong><br />
empirical oral and intravenous antibiotic therapy for low-risk febrile patients<br />
with neutropenia during cancer chemotherapy. N Engl J Med. 1999;341:305-<br />
311.<br />
9. Kern WV, Cometta A, De Bock R, et al.Oral versus intravenous empirical<br />
antimicrobial therapy for fever in patients with granulocytopenia who are<br />
receiving cancer chemotherapy. International Antimicrobial Therapy Cooperative<br />
Group <strong>of</strong> the European Organization for Research and Treatment <strong>of</strong><br />
Cancer. N Engl J Med. 1999;341:312-318.<br />
573
10. Carstensen M, Sorensen JB. Outpatient management <strong>of</strong> febrile neutropenia:<br />
time to revise the present treatment strategy. J Support Oncol.<br />
2008;6:199-208.<br />
11. Talcott JA, Yeap BY, Clark JA, et al. Safety <strong>of</strong> early discharge for<br />
low-risk patients with febrile neutropenia: a multicenter randomized controlled<br />
trial. J Clin Oncol. 2011;29:3977-3983.<br />
12. Klastersky J, Paesmans M, Rubenstein EB, et al. The Multinational<br />
Association for Supportive Care in Cancer risk index: A multinational scoring<br />
system for identifying low-risk febrile neutropenic cancer patients. J Clin<br />
Oncol. 2000;18:3038-3051.<br />
13. Freifeld AG, Bow EJ, Sepkowitz KA, et al. <strong>Clinical</strong> practice guideline<br />
for the use <strong>of</strong> antimicrobial agents in neutropenic patients with cancer: 2010<br />
update by the Infectious Diseases <strong>Society</strong> <strong>of</strong> America. Clin Infect Dis.<br />
2011;52:e56-93.<br />
14. Ammann RA, Simon A, de Bont ES. Low risk episodes <strong>of</strong> fever and<br />
neutropenia in pediatric oncology: Is outpatient oral antibiotic therapy the<br />
new gold standard <strong>of</strong> care? Pediatr Blood Cancer. 2005;45:244-247.<br />
15. Teuffel O, Ethier MC, Alibhai SM, et al. Outpatient management <strong>of</strong><br />
cancer patients with febrile neutropenia: a systematic review and metaanalysis.<br />
Ann Oncol. 2011;22:2358-2365.<br />
16. Ahmed N, El-Mahallawy HA, Ahmed IA, et al. Early hospital discharge<br />
versus continued hospitalization in febrile pediatric cancer patients with<br />
prolonged neutropenia: A randomized, prospective study. Pediatr Blood<br />
Cancer. 2007;49:786-792.<br />
17. Santolaya ME, Alvarez AM, Aviles CL, et al. Early hospital discharge<br />
followed by outpatient management versus continued hospitalization <strong>of</strong><br />
children with cancer, fever, and neutropenia at low risk for invasive bacterial<br />
infection. J Clin Oncol. 2004;22:3784-3789.<br />
18. Manji A, Beyene J, Dupuis LL, et al. Outpatient and oral antibiotic<br />
management <strong>of</strong> low-risk febrile neutropenia are effective and safe in children—a<br />
systematic review <strong>of</strong> prospective trials. Poster presented at: 2011<br />
Annual Pediatric <strong>Oncology</strong> Group <strong>of</strong> Ontario Symposium; November 2011;<br />
Toronto, ON, Canada.<br />
19. Liou SY, Stephens JM, Carpiuc KT, et al. Economic burden <strong>of</strong> haematological<br />
adverse effects in cancer patients: a systematic review. Clin Drug<br />
Investig. 2007;27:381-396.<br />
20. Lathia N, Mittmann N, DeAngelis C, et al. Evaluation <strong>of</strong> direct medical<br />
costs <strong>of</strong> hospitalization for febrile neutropenia. Cancer. 2010;116:742-748.<br />
574<br />
LILLIAN SUNG<br />
21. Hendricks AM, Loggers ET, Talcott JA. Costs <strong>of</strong> home versus inpatient<br />
treatment for fever and neutropenia: analysis <strong>of</strong> a multicenter randomized<br />
trial. J Clin Oncol. 2011;29:3984-3989.<br />
22. Klaassen RJ, Allen U, Doyle JJ. Randomized placebo-controlled trial <strong>of</strong><br />
oral antibiotics in pediatric oncology patients at low-risk with fever and<br />
neutropenia. J Pediatr Hematol Oncol. 2000;22:405-411.<br />
23. Wiernikowski JT, Rothney M, Dawson S, et al. Evaluation <strong>of</strong> a home<br />
intravenous antibiotic program in pediatric oncology. Am J Pediatr Hematol<br />
Oncol. 1991;13:144-147.<br />
24. Teuffel O, Amir E, Alibhai SM, et al. Cost-effectiveness <strong>of</strong> outpatient<br />
management for febrile neutropenia in children with cancer. Pediatrics.<br />
2011;127:e279-286.<br />
25. Speyer E, Herbinet A, Vuillemin A, et al. Agreement between children<br />
with cancer and their parents in reporting the child’s health-related quality <strong>of</strong><br />
life during a stay at the hospital and at home. Child Care Health Dev.<br />
2009;35:489-495.<br />
26. Sung L, Feldman BM, Schwamborn G, et al. Inpatient versus outpatient<br />
management <strong>of</strong> low-risk pediatric febrile neutropenia: measuring parents’<br />
and healthcare pr<strong>of</strong>essionals’ preferences. J Clin Oncol. 2004;22:3922-<br />
3929.<br />
27. Cheng S, Teuffel O, Ethier MC, et al. Health-related quality <strong>of</strong> life<br />
anticipated with different management strategies for paediatric febrile neutropaenia.<br />
Br J Cancer. 2011;105:606-611.<br />
28. Sung L, Alibhai SM, Ethier MC, et al. Discrete choice experiment<br />
produced estimates <strong>of</strong> acceptable risk <strong>of</strong> therapeutic options in cancer patients<br />
with febrile neutropenia. J Clin Epidemiol. <strong>2012</strong>. In press.<br />
29. Ryan M, Farrar S. Using conjoint analysis to elicit preferences for<br />
health care. BMJ. 2000;320:1530-1533.<br />
30. Diorio C, Martino J, Boydell KM, et al. Parental perspectives on<br />
inpatient versus outpatient management <strong>of</strong> pediatric febrile neutropenia.<br />
J Pediatr Oncol Nurs. 2011;28:355-362.<br />
31. Ungar WJ. Economic Evaluation in Child Health. Oxford: Oxford<br />
University Press; 2010.<br />
32. Upton P, Lawford J, Eiser C. Parent-child agreement across child<br />
health-related quality <strong>of</strong> life instruments: a review <strong>of</strong> the literature. Qual Life<br />
Res. 2008;17:895-913.
GENETIC COUNSELING OF THE PATIENT WITH<br />
PEDIATRIC CANCER<br />
CHAIR<br />
Joshua D. Schiffman, MD<br />
University <strong>of</strong> Utah<br />
Salt Lake City, UT<br />
SPEAKERS<br />
Kim E. Nichols, MD<br />
Children’s Hospital <strong>of</strong> Philadelphia<br />
Philadelphia, PA<br />
Sara Knapke, MS<br />
Cincinnati Children’s Hospital Medical Center<br />
Cincinnati, OH
Identification, Management, and Evaluation<br />
<strong>of</strong> Children with Cancer-Predisposition<br />
Syndromes<br />
By Sara Knapke, MS, Kristin Zelley, MS, Kim E. Nichols, MD, Wendy Kohlmann, MS,<br />
and Joshua D. Schiffman, MD<br />
Overview: A substantial proportion <strong>of</strong> childhood cancers are<br />
attributable to an underlying genetic syndrome or inherited<br />
susceptibility. Recognition <strong>of</strong> affected children allows for<br />
appropriate cancer risk assessment, genetic counseling, and<br />
testing. Identification <strong>of</strong> individuals who are at increased risk<br />
to develop cancers during childhood can guide cancer surveil-<br />
MORE THAN 12,060 children and adolescents will be<br />
diagnosed with cancer this year in North America. 1<br />
Of these patients, it is estimated that up to 10% will develop<br />
their cancer because <strong>of</strong> the presence <strong>of</strong> an underlying cancerpredisposing<br />
condition. 2-5 However, this 10% estimate has<br />
been extrapolated primarily from adults whose cancers were<br />
associated with germline mutations 6 and the actual percentage<br />
<strong>of</strong> childhood cancers caused by underlying genetic mutations<br />
remains unclear. In fact, a recent study reported that<br />
up to 29% <strong>of</strong> children seen in a pediatric oncology survivorship<br />
program qualified for consideration <strong>of</strong> referral to a<br />
cancer genetics clinic based on a significant family history<br />
<strong>of</strong> cancer, a history <strong>of</strong> specific tumors in the surviving child,<br />
or the presence <strong>of</strong> unique physical findings. 7 These data<br />
suggest that more than one in four children or adolescents<br />
with a history <strong>of</strong> cancer may have a genetic cancerpredisposing<br />
condition and could therefore potentially benefit<br />
from further evaluation and management. 8 Mounting<br />
evidence now demonstrates that early cancer detection in<br />
this population may lead to improved survival and treatment<br />
outcomes. 5,6,9-11 In this chapter, we will outline for the<br />
practicing oncologist how to identify these high-risk pediatric<br />
patients based on specific tumor presentation and family<br />
history patterns, and discuss the issues involved in managing<br />
children and adolescents with a known genetic predisposition<br />
to cancer. Ethical and psychosocial considerations<br />
related to genetic testing for cancer risk in minors and<br />
suggestions for family-centered and multidisciplinary care<br />
will also be explored.<br />
Identification <strong>of</strong> Children with Cancer Predisposition<br />
There are several ways in which a patient with an underlying<br />
cancer-predisposing condition can be identified. Specifically,<br />
tumor type and laterality, family cancer history,<br />
and presence <strong>of</strong> other physical features or medical conditions<br />
must be taken into consideration when determining<br />
whether a child has a possible genetic predisposition to<br />
cancer.<br />
Presence <strong>of</strong> Specific Tumor Patterns<br />
As demonstrated in Table 1, a hereditary pediatric cancer<br />
syndrome can be suggested by the presence <strong>of</strong> a specific<br />
pattern <strong>of</strong> cancer presentation or tumor type in an individual.<br />
Bilateral tumors in paired organs, multifocal tumors,<br />
and multiple primary cancers in a single individual <strong>of</strong>ten<br />
indicate an underlying or inherited genetic cause <strong>of</strong> cancer.<br />
6,10,12,13 In addition, an earlier age <strong>of</strong> diagnosis can also<br />
576<br />
lance and clinical management, which may improve outcomes<br />
for both the patient and other at-risk relatives. The information<br />
provided through this article will focus on the current<br />
complexities involved in the evaluation and management <strong>of</strong><br />
children with cancer-predisposing genetic conditions and<br />
highlight remaining questions for discussion.<br />
be a feature <strong>of</strong> some syndromes. For example, over 70% <strong>of</strong><br />
patients with early or bilateral Wilms Tumor may have<br />
an inherited or de novo mutation in the WT1 gene. 14,15 An<br />
example <strong>of</strong> specific tumor types associated with an underlying<br />
cancer-predisposing gene mutation include adrenocortical<br />
tumors and choroid plexus carcinomas, which may be<br />
caused by germline TP53 mutations in 80% 16 or 35% to<br />
100% 17-19 <strong>of</strong> patients, respectively. Based on these data, it is<br />
now recommended that all children with adrenocortical<br />
tumors and choroid plexus tumors be <strong>of</strong>fered genetic testing<br />
for TP53 mutations (Li-Fraumeni Syndrome [LFS]). 20 In<br />
addition, children with rhabdomyosarcoma younger than<br />
age 3 should be considered for TP53 mutation screening. 21<br />
(Some oncologists also will consider testing for TP53 mutations<br />
in children with osteosarcoma younger than ages 5 to<br />
10 at presentation based on the strong association between<br />
LFS and osteosarcoma 22 ). All children with bilateral retinoblastoma<br />
and up to 20% <strong>of</strong> children with unilateral retinoblastoma<br />
will have heritable retinoblastoma and most will<br />
have a detectable germline mutation <strong>of</strong> the RB1 gene. 23<br />
More than 70% <strong>of</strong> pediatric patients with malignant paragangliomas<br />
(PGL) or pheochromocytomas (PCC) have underlying<br />
SDHB mutations (Familial PGL/PCC Syndrome),<br />
and the percentage <strong>of</strong> affected patients is even higher when<br />
considering other PGL- or PCC-related genes. 24,25 Nearly a<br />
third <strong>of</strong> patients with rhabdoid tumor (s<strong>of</strong>t tissue or brain)<br />
will harbor a SMARCB1/INI1 mutation (Rhabdoid Tumor<br />
Syndrome). 26 At least 12 percent <strong>of</strong> patients with pediatric<br />
(i.e., succinate dehydrogenase [SDH]-deficient) gastrointestinal<br />
stromal tumors (GISTs) lacking mutations in the KIT<br />
and PDGFRA genes have germline mutations in the SDHassociated<br />
genes. 27 Even patients with hepatoblastoma have<br />
a reported 10% risk <strong>of</strong> having a germline APC mutation<br />
(Familial Adenomatous Polyposis [FAP] Syndrome). 28 Unusual<br />
cancers diagnosed in children such as colorectal or<br />
thyroid cancer also can indicate an inherited cancerpredisposition<br />
syndrome and should trigger the pediatric<br />
From the Cincinnati Children’s Hospital Medical Center, Cincinnati, OH; Children’s<br />
Hospital <strong>of</strong> Philadelphia, Philadelphia, PA; Huntsman Cancer Institute, Salt Lake City,<br />
UT; Center for Children’s Cancer Research (C3R), Huntsman Cancer Institute, Salt Lake<br />
City, UT.<br />
Authors’ disclosures <strong>of</strong> potential conflicts <strong>of</strong> interest are found at the end <strong>of</strong> this article.<br />
Address reprint requests Joshua Schiffman, MD, Division <strong>of</strong> Pediatric Hematology/<br />
<strong>Oncology</strong>, Center for Children’s Cancer Research (C3R), Huntsman Cancer Institute, 2000<br />
Circle <strong>of</strong> Hope, Salt Lake City, UT 84112.<br />
© <strong>2012</strong> by <strong>American</strong> <strong>Society</strong> <strong>of</strong> <strong>Clinical</strong> <strong>Oncology</strong>.<br />
1092-9118/10/1-10
CANCER PREDISPOSITION IN CHILDHOOD<br />
oncologist to consider a cancer genetics referral. 29-32 Pediatric<br />
patients presenting with the tumors listed in Table 1<br />
should be considered for referral to a cancer genetics clinic<br />
where they can be seen by a team <strong>of</strong> specialists including<br />
genetic counselors, oncologists, and geneticists with expertise<br />
in cancer predisposition. We believe that the more an<br />
oncologist looks for these presentations, the more pediatric<br />
hereditary cancer syndromes will be identified in patients. 8<br />
If recognized, at-risk patients and their families can benefit<br />
from genetic testing, cancer screening, and enrollment in<br />
research protocols.<br />
KEY POINTS<br />
● There exists a growing number <strong>of</strong> hereditary cancer<br />
syndromes with implications in childhood.<br />
● Identification <strong>of</strong> cancer predisposition in childhood<br />
can substantially influence the choice <strong>of</strong> appropriate<br />
screening and management options for the child and<br />
other relatives.<br />
● Genetic risk assessment, counseling, and testing are<br />
critical elements in the care <strong>of</strong> children and families<br />
with cancer-predisposing conditions.<br />
Table 1. Differential Genetic Diagnosis by Tumor Type<br />
Tumor Type Differential Genetic Diagnosis Gene(s)<br />
Adrenocortical carcinoma Li-Fraumeni syndrome (LFS) TP53<br />
Atypical teratoid and malignant rhabdoid tumor Rhabdoid tumor syndrome SMARCB1/INI1<br />
Basal cell carcinoma Nevoid Basal Cell Carcinoma Syndrome (NBCCS)/Gorlin PTCH1<br />
Choroid plexus carcinoma LFS TP53<br />
Cystic nephroma (familial) Pleuropulmonary blastoma (PPB) family tumor and dysplasia syndrome DICER1<br />
Desmoid tumors Familial Adenomatous Polyposis (FAP) APC<br />
Endolymphatic sac tumors Von-Hippel Lindau syndrome VHL<br />
Glioblastoma multiforme Turcot syndrome/Lynch syndrome MLH1, MSH2, MSH6, PMS2<br />
LFS TP53<br />
Hemangioblastoma (retinal/cerebellar) Von-Hippel Lindau syndrome VHL<br />
Hepatoblastoma FAP APC<br />
Beckwith-Weidemann syndrome CDKN1C/11p15<br />
Medulloblastoma Turcot syndrome/FAP syndrome APC<br />
NBCCS/Gorlin syndrome PTCH1<br />
Medullary thyroid cancer Multiple Endocrine Neoplasia type 2 RET<br />
Neuroblastoma (bilateral or multifocal) Familial Neuroblastoma ALK, PHOX2B<br />
Optic pathway tumor Neur<strong>of</strong>ibromatosis type 1 NF1<br />
Ovarian sex cord-stromal tumors PPB family tumor and dysplasia syndrome DICER1<br />
Peutz-Jeghers syndrome STK11<br />
Paraganglioma/Pheochromocytoma Familial PGL/PCC syndrome SDHB/C/D/AF2<br />
Von Hippel Lindau syndrome VHL<br />
Multiple Endocrine Neoplasia type 2 RET<br />
Neur<strong>of</strong>ibromatosis type 1 NF1<br />
Retinoblastoma Hereditary Retinoblastoma RB1<br />
Pleuropulmonary blastoma<br />
Sarcomas<br />
PPB family tumor and dysplasia/DICER1 syndrome DICER1<br />
Rhabdomyosarcoma LFS TP53<br />
PPB family tumor and dysplasia/DICER1 syndrome DICER1<br />
Osteosarcoma LFS TP53<br />
Schwannoma (acoustic or vestibular) Neur<strong>of</strong>ibromatosis type 2 NF2<br />
Sertoli-Leydig cell tumor PPB family tumor and dysplasia/DICER1 syndrome DICER1<br />
Wilms tumor (bilateral) Wilms tumor syndrome WT1<br />
This report focuses on predisposition to solid tumors,<br />
where surveillance is most likely to influence treatment and<br />
outcome. However, it is important to recognize that a number<br />
<strong>of</strong> genetic conditions predispose the patient to the<br />
development <strong>of</strong> hematopoietic malignancies, myelodysplastic<br />
syndrome, or both. The reader is directed to the Seif<br />
reference 33 for an in-depth discussion <strong>of</strong> these conditions.<br />
Family History as a Screening Tool<br />
Evaluation <strong>of</strong> all children with cancer should include<br />
assessment <strong>of</strong> the family history. Features <strong>of</strong> the family<br />
medical history that may indicate the presence <strong>of</strong> a hereditary<br />
cancer syndrome include, but are not limited to (1)<br />
multiple family members on the same side <strong>of</strong> the family with<br />
the same or related type <strong>of</strong> cancers, or individuals with<br />
distinct cancer types that are known to group together in<br />
specific cancer-predisposition syndromes (e.g., rhabdomyosarcoma,<br />
osteosarcoma, adrenocortical carcinoma, brain tumors,<br />
leukemia, and early-onset breast cancer, as is seen in<br />
families with LFS); (2) close relatives with early, multiple, or<br />
bilateral cancers; and (3) a pattern suggestive <strong>of</strong> generationto-generation<br />
transmission (i.e., autosomal dominant inheritance).<br />
Therefore, clinicians should collect information on<br />
age <strong>of</strong> cancer onset, type <strong>of</strong> cancer and laterality in at least<br />
first- (parents and siblings), second- (grandparents, aunts,<br />
and uncles), and third- (cousins and great grandparents)<br />
577
degree relatives and consider referral for any concerning<br />
features in the family history.<br />
There are a number <strong>of</strong> potential limitations to the use <strong>of</strong><br />
family history as a screening tool, such as adoption <strong>of</strong> the<br />
patient or another family member without information<br />
about biologic relatives, small family size, limited or no<br />
contact with relatives, and early death from noncancer<br />
related causes. In addition, for some hereditary cancer<br />
syndromes, there are a number <strong>of</strong> de novo or new dominant<br />
mutations or alternative inheritance patterns (i.e., autosomal<br />
recessive) and, in these cases, family history would<br />
not be fully informative. As family histories evolve over<br />
time, it is also possible that members may not have yet<br />
demonstrated features <strong>of</strong> a particular hereditary cancer<br />
syndrome. This is especially true in pediatrics where parents<br />
will be younger than those <strong>of</strong> adult patients. For this<br />
reason, frequent updating <strong>of</strong> the family history is indicated.<br />
Many cancer syndromes are also characterized by reduced<br />
penetrance and may not exhibit a clear inheritance pattern.<br />
Medical History and Physical Features Concerning for a Hereditary<br />
Cancer Syndrome<br />
Medical history and physical examination are important<br />
components in the evaluation <strong>of</strong> children for a cancerpredisposing<br />
syndrome. Several hereditary cancer syndromes<br />
are characterized not only by the development <strong>of</strong><br />
cancer, but also by the presence <strong>of</strong> certain physical or<br />
developmental manifestations that might provide important<br />
clues to the diagnosis (Table 2). These may include congenital<br />
anomalies or dysmorphic features, along with developmental<br />
delays, intellectual disabilities, autism, or autism<br />
spectrum disorders. Other associated features may include<br />
skin findings, macrocephaly, and benign tumors or polyps.<br />
In many cases, one or more <strong>of</strong> these features precede the<br />
development <strong>of</strong> cancer and serve as an early clue to the<br />
diagnosis. The presence <strong>of</strong> one or more <strong>of</strong> these features in a<br />
child diagnosed with cancer should raise suspicion for a<br />
hereditary cancer syndrome and prompt referral to a geneticist<br />
or genetic counselor. Referral to other specialties may<br />
also be necessary to evaluate and manage the physical or<br />
neurocognitive issues related to these genetic conditions.<br />
Management <strong>of</strong> Children with Cancer Predisposition<br />
Surveillance. Recognition <strong>of</strong> hereditary cancer syndromes<br />
in children is necessary to facilitate appropriate surveillance<br />
and management <strong>of</strong> individuals who are predisposed to<br />
malignancies. Surveillance and management guidelines exist<br />
or are being developed for several hereditary cancer<br />
syndromes with onset in childhood (Table 2). The primary<br />
goal <strong>of</strong> cancer surveillance is to detect cancer at the earliest<br />
and most curable stage with the least amount <strong>of</strong> complications<br />
and late effects from treatment. For this reason, cancer<br />
surveillance protocols are most suited for solid tumors, such<br />
as hepatoblastoma and Wilms tumor, where outcome is<br />
directly linked to stage at diagnosis. By detecting cancer at<br />
an early stage, cancer surveillance protocols aim to improve<br />
overall survival and decrease morbidity in individuals at<br />
increased risk for tumor development. Cancer surveillance<br />
has the potential benefits <strong>of</strong> increasing the cure rate for<br />
cancers and reducing or eliminating the need for chemotherapy,<br />
radiation therapy, or both, which can have substantial<br />
side effects. However, cancer surveillance also can result in<br />
increased worry about cancer, an increased rate <strong>of</strong> biopsies<br />
578<br />
or invasive procedures because <strong>of</strong> false-positive results, and<br />
increased cost <strong>of</strong> care. Benefits and disadvantages <strong>of</strong> surveillance<br />
may differ from one genetic syndrome to another, and<br />
the benefits <strong>of</strong> each syndrome-specific protocol must outweigh<br />
the disadvantages. We recommend an open discussion<br />
with patients and their families about both the advantages<br />
and the risks involved in early cancer screening.<br />
Several factors must be considered in the development<br />
and implementation <strong>of</strong> an effective cancer surveillance protocol.<br />
First, there must be a benefit to early cancer detection<br />
in individuals with cancer-predisposing conditions. Second,<br />
the age-specific cancer risks for the hereditary cancer syndrome<br />
must be known and considered high enough to<br />
warrant surveillance. This information is needed to know to<br />
whom surveillance should be <strong>of</strong>fered, when to initiate surveillance,<br />
and the duration <strong>of</strong> surveillance. The surveillance<br />
method and interval must also be carefully considered in<br />
light <strong>of</strong> the specific cancer risks associated with a given<br />
syndrome. Ideally, surveillance methods should be readily<br />
available, safe, and have high sensitivity and specificity.<br />
Whenever possible, screening tools that involve no or minimal<br />
radiation should be used, as patients with hereditary<br />
cancer syndromes may be at increased risk for developing<br />
radiation-induced cancers. Finally, there must be effective<br />
treatment methods available for the cancer(s) identified<br />
through screening.<br />
For some hereditary cancer syndromes, such as hereditary<br />
retinoblastoma, surveillance and management guidelines<br />
are well established and have been shown to improve outcomes<br />
for affected individuals. 34,35 For other syndromes,<br />
such as LFS, there is ongoing debate about the optimal<br />
surveillance protocol. The development <strong>of</strong> an effective surveillance<br />
protocol for individuals diagnosed with or at risk<br />
for LFS has been complicated by the variability in type,<br />
location, and age <strong>of</strong> onset <strong>of</strong> tumors, as well as the increased<br />
risk for multiple primary tumors and radiation-induced<br />
cancers. 36 This raises many questions about the optimal<br />
method(s), interval, and duration <strong>of</strong> surveillance. Several<br />
groups have begun to <strong>of</strong>fer a combined modality surveillance<br />
protocol for LFS utilizing rapid whole-body magnetic resonance<br />
imaging (MRI), brain MRI, abdominal ultrasound,<br />
complete blood count, and biochemical markers. Recent data<br />
have shown that this surveillance protocol may indeed be<br />
effective and improves the survival <strong>of</strong> patients with LFS<br />
through early tumor detection. 11 Further prospective studies<br />
will be necessary to validate the effectiveness <strong>of</strong> this<br />
surveillance protocol in both children and adults.<br />
It is important to note that surveillance protocols may<br />
change over time as new evidence becomes available. Therefore,<br />
clinicians should refer to the literature and resources<br />
such as the National Comprehensive Cancer Network<br />
(NCCN) for the most up-to-date surveillance guidelines.<br />
Risk-Reduction Strategies<br />
KNAPKE ET AL<br />
For some pediatric cancer syndromes, early identification<br />
<strong>of</strong> high-risk children allows for the elimination or dramatic<br />
reduction <strong>of</strong> risk through prophylactic surgery. For example,<br />
multiple endocrine neoplasia type 2 (MEN2) and FAP are<br />
classic examples <strong>of</strong> when surgery to remove the at-risk<br />
organ may be considered early in the patient’s lifetime to<br />
prevent the development <strong>of</strong> cancer. MEN2 is caused by<br />
mutations in RET and is associated with nearly a 100%<br />
lifetime risk for medullary thyroid cancer (MTC), as well as
CANCER PREDISPOSITION IN CHILDHOOD<br />
Table 2. Hereditary Cancer Syndromes with Solid-Tumor Risk and Manifestations in Childhood: Genetics, Risks, Features, and Surveillance<br />
Syndrome Gene(s) Inheritance Cancer/Tumor Risks Other Features Cancer Surveillance<br />
Beckwith-Wiedemann<br />
syndrome/idiopathic<br />
hemihypertrophy 52,59<br />
Chromosome 11p15 Imprinting/AD Hepatoblastoma<br />
Hemihypertrophy<br />
- Abdominal ultrasound every 3 mo from birth until age<br />
methylation defects;<br />
Wilms tumor<br />
Macroglossia<br />
4y<br />
UPD; CDKN1C<br />
Omphalocele<br />
- Serum AFP level every 6–12 wk from birth until<br />
Umbilical hernia<br />
age4y<br />
Neonatal hypoglycemia<br />
Ear pits/creases<br />
- Renal ultrasound every 3 mo from 4–8 y<br />
Biallelic Mismatch Repair<br />
Gene syndrome 53<br />
Familial Adenomatous<br />
Polyposis 39<br />
MLH1, MSH2, MSH6,<br />
PMS2<br />
AR Lymphoma<br />
Leukemia<br />
Brain tumors<br />
Colorectal<br />
Small bowel<br />
Associated benign tumors:<br />
- GI polyps (adenomas)<br />
APC AD Colon<br />
Small bowel<br />
Pancreatic<br />
Thyroid - papillary<br />
Hepatoblastoma<br />
CNS–medulloblastoma<br />
Bile duct<br />
Gastric<br />
FAP-associated benign tumors:<br />
- S<strong>of</strong>t tissue tumors (Gardner’s<br />
fibromas, desmoids tumors)<br />
- Osteomas<br />
- GI polyps (adenomas)<br />
Café au lait macules<br />
Axillary and inguinal<br />
freckling<br />
Lisch nodules<br />
Epidermoid cysts<br />
Supernumerary/missing<br />
teeth<br />
Congenital hypertrophy <strong>of</strong><br />
the retinal pigmented<br />
epithelium (CHRPE)<br />
Familial Neuroblastoma ALK, PHOX2B AD Neuroblastoma Hirschsprung disease<br />
(PHOX2B only)<br />
Familial PGL/PCC<br />
Syndrome 54<br />
Heritable<br />
Retinoblastoma 55<br />
Juvenile Polyposis<br />
syndrome 39<br />
Li-Fraumeni syndrome 56<br />
Multiple Endocrine<br />
Neoplasia type 1 50,51<br />
SDHB/C/D/AF2 AD Paraganglioma<br />
Pheochromocytoma<br />
Possible associations: Renal,<br />
Thyroid<br />
RB1 AD Retinoblastoma<br />
Pinealomas<br />
Sarcomas<br />
Melanoma<br />
SMAD4, BMPR1A AD Colon<br />
Gastric<br />
Small intestine<br />
Pancreatic<br />
Associated benign tumors:<br />
- Juvenile GI polyps<br />
TP53 AD Adrenocortical carcinoma<br />
Choroid plexus carcinoma<br />
Bone and s<strong>of</strong>t tissue sarcomas<br />
Leukemia<br />
Breast<br />
Numerous other<br />
MEN1 AD Parathyroid adenoma<br />
Gastrinoma<br />
Insulinoma<br />
Anterior pituitary<br />
Forgut carcinoid<br />
Numerous other<br />
MEN1-associated benign tumors:<br />
- angi<strong>of</strong>ibromas<br />
- collagenomas<br />
- lipomas<br />
- Colonoscopy annually beginning at age 3 or at<br />
diagnosis<br />
- Upper endoscopy (EGD) and video capsule endoscopy<br />
annually<br />
- Ultrasound <strong>of</strong> head at birth and MRI <strong>of</strong> brain every<br />
6 mos<br />
- CBC, erythrocyte sedimentation rate, lactate<br />
dehydrogenase every 4 mos<br />
- Urinary and uterine ultrasound annually in adulthood<br />
- Colonoscopy every12–24 mo from age 10–12 y until<br />
colectomy<br />
- Upper endoscopy (EGD) every 12–36 mo starting at<br />
age 18, lifelong<br />
- Thyroid exam and consider thyroid ultrasound every<br />
12 mo starting at age 18, lifelong<br />
- Consider hepatoblastoma screening:<br />
- Abdominal ultrasound every 3 mo from birth until age<br />
4y<br />
- Serum AFP level every 6–12 wk from birth until<br />
age4y<br />
No published guidelines available<br />
None - Urine or plasma metanephrines and catecholemines<br />
every 12 mo starting at age 10 y, lifelong<br />
- MRI <strong>of</strong> neck, chest, abdomen, and pelvis every 12 mo<br />
starting at age 10 y, lifelong<br />
None - Brain MRI every 6 mo from birth until age 5 y<br />
- Eye exam, frequency determined by ophthalmologist,<br />
from birth until age 5 y<br />
- Thorough annual physical exam and careful attention<br />
to development <strong>of</strong> any lumps or lesions (because <strong>of</strong><br />
the high rate <strong>of</strong> second cancers, particularly among<br />
those who had RX treatment)<br />
Arteriovascular<br />
malformations (SMAD4<br />
only)<br />
- Monitor for rectal bleeding, lifelong<br />
- Colonoscopy/EGD every 12–36 mo starting at<br />
age 15 or if symptoms<br />
- CBC every 12 mo starting in early childhood<br />
-IfSMAD4 mutation, surveillance for Hereditary<br />
Hemorrhagic Telangiectasia<br />
None - Physical exam (with careful skin and neurologic exam)<br />
every 12 mo, lifelong<br />
- Self breast exam every 1 mo starting at age 18 y,<br />
lifelong<br />
- <strong>Clinical</strong> breast exam every 6–12 mo starting at<br />
age 20–25 y, lifelong<br />
- Mammogram and breast MRI every 12 mo starting at<br />
age 20–25 y, lifelong<br />
- Colonoscopy (suggested) every 2–5 y starting no later<br />
than age 25 y, lifelong<br />
- Consider additional screening with abdominal<br />
ultrasound, brain MRI, whole body MRI, and<br />
biochemical markers 11<br />
None - Annual serum concentration <strong>of</strong> prolactin from<br />
age5y<br />
- Annual fasting total serum calcium concentration<br />
(corrected for albumin) and/or ionized-serum calcium<br />
concentration from age 8 y<br />
- Annual fasting serum gastrin concentration from<br />
age 20 y<br />
579
Table 2. Hereditary Cancer Syndromes with Solid-Tumor Risk and Manifestations in Childhood: Genetics, Risks, Features, and Surveillance (Cont’d)<br />
Syndrome Gene(s) Inheritance Cancer/Tumor Risks Other Features Cancer Surveillance<br />
- To be considered: annual fasting serum<br />
concentration <strong>of</strong> intact (full-length) PTH<br />
- Head MRI from age 5 y every 3–5 y<br />
- Abdominal CT or MRI from age 20 y every 3–5 y<br />
- To be considered: yearly chest CT, SRS octreotide<br />
scan<br />
Multiple Endocrine<br />
Neoplasia type 2 37<br />
Neur<strong>of</strong>ibromatosis<br />
type 1 51<br />
Neur<strong>of</strong>ibromatosis<br />
type 2 51<br />
Nevoid Basal Cell<br />
Carcinoma (Gorlin)<br />
syndrome 51<br />
Peutz-Jeghers syndrome 56<br />
Pleuropulmonary blastoma<br />
(PPB) Family Tumor<br />
and Dysplasia<br />
syndrome/DICER1<br />
syndrome<br />
580<br />
RET AD Thyroid–medullary<br />
Pheochromocytoma<br />
MEN2-associated benign<br />
tumors (MEN2B only):<br />
- Mucosal neuromas<br />
- Ganglioneuromas<br />
NF1 AD Schwannoma<br />
Pheochromocytoma<br />
Optic pathway tumor<br />
Neur<strong>of</strong>ibromas<br />
JMML<br />
AML<br />
NF2 AD Vestibular schwannoma<br />
Meningioma<br />
Schwannoma<br />
Glioma<br />
Neur<strong>of</strong>ibroma<br />
PTCH1 AD Basal cell carcinoma<br />
Medulloblastoma<br />
STK11 AD Colorectal<br />
Gastric<br />
Small intestine<br />
Pancreatic<br />
Breast<br />
Ovarian<br />
Cervix<br />
Uterus<br />
Testes<br />
Lung<br />
PJS-associated benign tumors:<br />
- Peutz-Jeghers GI polyps<br />
- Sex cord tumors with annular<br />
tubules (SCTAT)<br />
DICER1 AD PPB<br />
Rhabdomyosarcoma<br />
Thyroid<br />
Ovarian Sertoli-Leydig cell<br />
tumors and dysgerminoma<br />
Testicular seminoma<br />
Other gonadal germ cell tumors<br />
Leukemia<br />
Associated benign tumors:<br />
- Multinodular goiter<br />
- Cystic nephroma<br />
Hyperparathyroidism<br />
Marfanoid habitus (MEN2B only)<br />
Café au lait macules<br />
Axillary and inguinal<br />
freckling<br />
Lisch nodules<br />
Tibial bowing<br />
Developmental delay/<br />
intellectual disability/<br />
autism<br />
Posterior subcapsular<br />
lenticular opacities<br />
Cataract<br />
Hearing loss<br />
Focal weakness<br />
Tinnitus<br />
Balance dysfunction<br />
Seizure<br />
Focal sensory loss<br />
Blindness<br />
Jaw keratocysts<br />
Macrocephaly<br />
Palmar/plantar pits<br />
Bifid ribs<br />
Calcification <strong>of</strong> the falx<br />
- Serum calcitonin every 12 mo<br />
- PTH every 12 mo<br />
- Urine/plasma metanephrines and catecholamines<br />
every 12 mo<br />
- Age at which screening is recommended to begin<br />
varies based on the specific genetic mutation. See<br />
<strong>American</strong> Thyroid Association Guidelines for<br />
Management <strong>of</strong> Medullary Thyroid Cancer for<br />
detailed recommendations.<br />
- Annual physical exam by a physician who is familiar<br />
with the individual and with the disease<br />
- Annual ophthalmologic exam in early childhood, less<br />
frequent exam in older children and adults<br />
- Regular developmental assessment by screening<br />
questionnaire (in childhood)<br />
- Regular blood pressure monitoring<br />
- Other studies only as indicated based on clinically<br />
apparent signs or symptoms<br />
- Monitoring <strong>of</strong> those who have abnormalities <strong>of</strong> CNS,<br />
skeletal system, or cardiovascular system by an<br />
appropriate specialist<br />
- Cranial MRI annually beginning at age 10–12 y until<br />
at least fourth decade <strong>of</strong> life<br />
- Routine complete eye exam<br />
- Hearing evaluation including BAER testing<br />
- Monitoring <strong>of</strong> head circumference through childhood<br />
- Developmental assessment and physical exam every<br />
6mo<br />
- Orthopantogram every 12–18 mo starting at � 8y<br />
- Skin exam at least annually<br />
Mucocutaneous pigmentation - <strong>Clinical</strong> breast exam every 6 mo starting at age 25 y,<br />
lifelong<br />
- Mammogram and breast MRI every 12 mo starting at<br />
age 25 y, lifelong<br />
- Colonoscopy and EGD every 2–3 y starting in late<br />
teenage years, lifelong<br />
- MRCP and/or endoscopic ultrasound every 12–24<br />
mo, starting at age 25–30 y, lifelong<br />
- CA 19–9 every 12–24 mo starting at age 25–30 y,<br />
lifelong<br />
- Small bowel visualization, baseline at age 8–10 y,<br />
follow-up based on findings<br />
- Pelvic exam and pap smear every 12 mo starting at<br />
age 18–20 y<br />
- Consider transvaginal ultrasound starting at age<br />
18–20 y<br />
- Testicular exam every 12 mo starting at age 10 y<br />
- Education about symptoms <strong>of</strong> lung cancer and<br />
smoking cessation<br />
Pulmonary cysts No published guidelines available<br />
KNAPKE ET AL
CANCER PREDISPOSITION IN CHILDHOOD<br />
Table 2. Hereditary Cancer Syndromes with Solid-Tumor Risk and Manifestations in Childhood: Genetics, Risks, Features, and Surveillance (Cont’d)<br />
Syndrome Gene(s) Inheritance Cancer/Tumor Risks Other Features Cancer Surveillance<br />
PTEN Hamartoma Tumor<br />
syndrome 56<br />
PTEN AD Breast<br />
Thyroid<br />
Endometrial<br />
Renal<br />
Associated benign tumors:<br />
- Multinodular goiter<br />
- Cystic nephroma<br />
PTEN-associated benign tumors:<br />
- Lipomas<br />
- Thyroid nodules/goiter<br />
- Hamartomatous GI polyps<br />
Rhabdoid syndrome SMARCB1/INI1 AD Rhabdoid tumors<br />
Schwannomatosis<br />
von-Hippel Lindau<br />
syndrome 57<br />
VHL AD Hemangioblastoma (retinal/<br />
cerebellar)<br />
Renal Cell Carcinoma<br />
Pancreatic–neuroendocrine<br />
Pheochromocytoma<br />
Endolymphatic sac tumors<br />
Epididymal tumors<br />
an increased risk for hyperparathyroidism and pheochromocytoma.<br />
Although all individuals with a RET mutation have<br />
an increased risk for MTC, there are direct genotypephenotype<br />
correlations that predict the age <strong>of</strong> onset and<br />
determine timing for prophylactic thyroidectomy. 37 Studies<br />
have found that genetic identification <strong>of</strong> children at risk for<br />
MEN2 and prophylactic thyroidectomy has greatly reduced<br />
the likelihood <strong>of</strong> developing MTC in this population. 38<br />
Similarly, the identification <strong>of</strong> the APC gene as the cause<br />
<strong>of</strong> FAP has allowed for genetic testing to identify children<br />
with this condition. Individuals with APC mutations associated<br />
with a classic FAP phenotype develop hundreds to<br />
thousands <strong>of</strong> colonic polyps beginning in adolescence and<br />
have a risk <strong>of</strong> colon cancer that approaches 100% if untreated.<br />
Current guidelines recommend beginning surveillance<br />
for colon polyps at age 10. When polyps become too<br />
numerous to follow endoscopically, colectomy is recommended.<br />
39 Use <strong>of</strong> genetic testing to evaluate at-risk children<br />
for a familial APC mutation is more cost-effective than<br />
relying on colon examinations to determine whether a child<br />
has inherited this condition. 40 Because <strong>of</strong> the morbidities<br />
associated with colectomy, alternatives to surgery are being<br />
Macrocephaly<br />
Arteriovascular malformations<br />
Developmental delay/<br />
intellectual disability/<br />
autism<br />
Wilms tumor syndromes 58 WT1 AD Wilms tumor Aniridia (WAGR syndrome)<br />
WAGR<br />
Denys-Drash<br />
Familial Wilms<br />
Frasier<br />
- Physical exam every 12 mo starting at age 18 y,<br />
lifelong<br />
- Thyroid ultrasound every 12 mo starting at age<br />
18 y, lifelong<br />
- Self breast exam every 1 mo starting at age 18 y,<br />
lifelong<br />
- <strong>Clinical</strong> breast exam every 6–12 mo starting at<br />
age 25 y, lifelong<br />
- Mammogram and breast MRI every 12 mo starting<br />
at age 30–35 y (or 5–10 y before earliest age <strong>of</strong><br />
diagnosis in family, whichever comes first)<br />
- Patient education about signs and symptoms <strong>of</strong><br />
endometrial cancer and encourage prompt followup<br />
if issues arise<br />
- Counsel about risk-reducing mastectomy and<br />
hysterectomy on a case-by-case basis<br />
- Colonoscopy (suggested) every 5–10 y, more<br />
frequently if symptoms or polyps, starting at age<br />
35 y, lifelong<br />
None No published guidelines available<br />
Renal and Pancreatic cysts - Ophthalmologic screening every 12 mo starting at<br />
age 1 y, lifelong<br />
- Physical exam, blood pressure monitoring, and<br />
neurologic assessment every12 mo starting at age<br />
2 y, lifelong<br />
- Urine/plasma catecholamine metabolites every<br />
12 mo starting at age 2 y, lifelong<br />
- Abdominal ultrasound every 12 mo from age<br />
8–20 y<br />
- Abdominal CT or MRI every 24 mo, to be<br />
alternated with annual abdominal ultrasound<br />
starting at age 20 y<br />
- Brain/spine MRI every 12 mo starting at puberty<br />
- Audiology assessment every 2–3 y from ages<br />
2–10 y, and then as symptoms arise<br />
Genitourinary defects<br />
Developmental delay/<br />
intellectual disability/<br />
autism<br />
- Renal ultrasound every 3 mo from birth until age<br />
8y<br />
Abbreviations: AD, autosomal dominant; AML, Acute myeloid leukemia; AR, autosomal recessive; AFP, alpha-fetoprotein; BAER, brainstem auditory evoked response;<br />
CA 19-9, carbohydrate antigen 19-9; CBC, complete blood count; CHRPE, congenital hypertrophy <strong>of</strong> the retinal pigmented epithelium; CNS, central nervous system;<br />
CT, computed tomography; EGD, esophagogastroduodenoscopy; GI, gastrointestinal; JMML, juvenile myelomonocytic leukemia; MEN2, multiple endocrine neoplasia<br />
type 2; mo, month (s); MRCP, magnetic resonance cholangiopancreatography; MRI, magnetic resonance imaging; PPB, pleuropulmonary blastoma; PTH, parathyroid<br />
hormone; SCTAT, sex cord tumors with annular tubules; SRS, somatostatin receptor scintigraphy; WAGR, Wilms tumor, aniridia, genitourinary anomalies, and mental<br />
retardation syndrome; wk, week(s); XRT, radiation; y, year(s).<br />
sought. Nonsteroidal anti-inflammatory inhibitors, such as<br />
sulindac, and Cox-2 inhibitors have been found to reduce<br />
polyp development in adults with FAP. 41,42 Studies have<br />
now begun to look at the potential utility <strong>of</strong> these medications<br />
in children 43 and an international clinical trial is<br />
currently enrolling children with a molecular diagnosis <strong>of</strong><br />
FAP in a 5-year trial comparing the rate <strong>of</strong> polyp development<br />
with celecoxib compared with placebo. 44<br />
Genetic Risk Assessment and Counseling<br />
Elements <strong>of</strong> an appropriate cancer genetic evaluation<br />
include collection <strong>of</strong> a thorough personal and family medical<br />
history, genetic risk assessment through pedigree analysis<br />
and published literature, genetic testing when appropriate<br />
for specific cancer syndromes, informed consent, results<br />
disclosure, and psychosocial assessment. 45 This process is<br />
<strong>of</strong>ten complex and may also require other essential components<br />
including medical records ascertainment and review,<br />
health insurance preauthorization for testing, and facilitating<br />
communication with other at-risk family members about<br />
complex results.<br />
Benefits <strong>of</strong> cancer genetic risk assessment, counseling,<br />
581
and testing include identification <strong>of</strong> at-risk individuals and<br />
families. In cases where a causative gene mutation can be<br />
identified, mutation-specific testing can identify those individuals<br />
in the family who have inherited the genetic risk<br />
factor and warrant high-risk screening and management.<br />
Single-site mutation testing can also identify those individuals<br />
in the family who did not inherit the condition and,<br />
therefore, are not predicted to be at increased risk and can<br />
forego additional measures. Many family members <strong>of</strong>ten<br />
have increased anxiety and worry about the risk for cancer<br />
in the family, and appropriate risk assessment and identification<br />
<strong>of</strong> a specific cause can provide accurate information<br />
and, in some cases, help to empower family members and<br />
alleviate emotional burden.<br />
There are also potential disadvantages <strong>of</strong> and obstacles to<br />
cancer genetics evaluations. For example, although the<br />
sensitivity and utility <strong>of</strong> genetic testing continues to improve,<br />
a causative gene mutation cannot be identified in<br />
some cases even when there is a high suspicion <strong>of</strong> a specific<br />
diagnosis. Therefore, clinical judgment and expertise must<br />
be applied in these cases to develop an acceptable screening<br />
and management plan for the patient as well as at-risk<br />
family members. In addition, there are also some cases with<br />
striking features <strong>of</strong> a hereditary cancer syndrome that do<br />
not fit a specific diagnosis or may represent a previously<br />
undescribed syndrome. Research studies as well as advances<br />
in gene finding and exome sequencing may be beneficial in<br />
such cases. However, these newer genomic technologies may<br />
lead to the identification <strong>of</strong> more variants <strong>of</strong> unknown<br />
significance for which it can be difficult to counsel the<br />
patient and his or her family. Advancements in genetic<br />
information and testing continue to change at a rapid pace.<br />
Therefore, there needs to be a clear expectation in the<br />
pediatric cancer genetics clinic for periodic follow-up and<br />
recontact with families in the event that new information is<br />
obtained. The results <strong>of</strong> genetic testing should not stand<br />
alone in risk assessment but rather be one tool in the genetic<br />
cancer risk assessment process.<br />
Ethical and psychosocial considerations remain critical in<br />
the assessment and care <strong>of</strong> children with potential cancerpredisposition<br />
syndromes. For example, with respect to the<br />
informed consent process, clinicians need to consider the<br />
child’s capacity for autonomy as well as participation in<br />
assent/consent. In addition, much historic debate has occurred<br />
about genetic testing in minors. 46-49 A distinction<br />
between diagnostic and presymptomatic or predictive testing<br />
is relevant. Although diagnostic testing is generally<br />
acceptable in children with features <strong>of</strong> a genetic condition,<br />
predictive testing is generally reserved for those conditions<br />
for which clinical management would be altered during<br />
childhood. 48 Regarding the psychosocial effects on the family<br />
experience, it is important to explore implications on emotional<br />
well-being, family dynamics, risk perception, influ-<br />
582<br />
Table 3. Cancer Genetic Services Resources<br />
Resource Web site<br />
KNAPKE ET AL<br />
National <strong>Society</strong> <strong>of</strong> Genetic Counselors<br />
Find a Genetic Counselor Tool<br />
www.nsgc.org<br />
National Cancer Institute http://cancer.gov/cancertopics/genetics/<br />
Cancer Genetic Services Directory directory<br />
ence on other siblings, reproductive decision-making, and<br />
financial consequences.<br />
Because <strong>of</strong> all <strong>of</strong> these potential risks and benefits, it is<br />
important that “discussions on genetic testing are done in a<br />
sensitive, comprehensive, and inclusive manner by fully<br />
trained specialist health pr<strong>of</strong>essionals, such as genetic counselors<br />
and clinical geneticists, in a relaxed and comfortable<br />
environment.” 48 After identifying individuals and families<br />
who might be at increased risk for cancer, referral to a<br />
program with expertise in childhood cancer predisposition is<br />
indicated. Many health care systems may have genetic<br />
counselors or other specialists with genetic expertise on site.<br />
Others may need to seek out and establish appropriate<br />
referral practices to another organization in their area.<br />
Resources for finding local genetic specialists can be found in<br />
Table 3. If cancer genetics services are not available nearby,<br />
an increasing number <strong>of</strong> programs also <strong>of</strong>fer their services<br />
through a telemedicine service model. When possible, it also<br />
may be beneficial to establish relationships with cancer<br />
genetics programs that practice through a multidisciplinary<br />
approach to care. A growing number <strong>of</strong> cancer genetics<br />
programs have established specific clinics related to pediatric<br />
cancer predisposition and integrate expertise from clinical<br />
geneticists, pediatric oncologists, and other relevant<br />
subspecialists.<br />
Conclusion<br />
In an era <strong>of</strong> personalized medicine, identification <strong>of</strong> disease<br />
susceptibility is no longer solely for academic interest<br />
but is becoming an accepted and clinically relevant element<br />
in the current management <strong>of</strong> patients. Therefore, it is<br />
imperative for clinicians to recognize those children and<br />
families who will benefit most from a cancer genetics referral,<br />
and assist in the follow-up and management <strong>of</strong> these<br />
individuals. Although a great deal <strong>of</strong> knowledge about<br />
cancer-predisposing conditions affecting children now exists,<br />
the scope <strong>of</strong> genomic information is expanding at a<br />
rapid pace and the future <strong>of</strong> this field will become increasingly<br />
complex. These new genetic data must be carefully<br />
examined through clinical, translational, and basic research<br />
protocols to ensure their effective translation to the optimized<br />
care <strong>of</strong> children at increased genetic risk for cancer.<br />
Acknowledgements<br />
K.E.N. acknowledges support in part by the Grundy Vision <strong>of</strong><br />
Life Fund. W.K. and J.D.S. acknowledge the use <strong>of</strong> the Genetic<br />
Counseling Shared Resource supported by P30 CA042014<br />
awarded to Huntsman Cancer Institute.
CANCER PREDISPOSITION IN CHILDHOOD<br />
Authors’ Disclosures <strong>of</strong> Potential Conflicts <strong>of</strong> Interest<br />
Author<br />
Employment or<br />
Leadership<br />
Positions<br />
Consultant or<br />
Advisory Role<br />
Sara Knapke*<br />
Kristin Zelley*<br />
Kim E. Nichols*<br />
Wendy Kohlmann Myriad<br />
Joshua Schiffman*<br />
*No relevant relationships to disclose.<br />
1. <strong>American</strong> Cancer <strong>Society</strong>. Cancer Facts & Figures <strong>2012</strong>. Atlanta: <strong>American</strong><br />
Cancer <strong>Society</strong>, <strong>2012</strong>.<br />
2. Narod SA, Stiller C, Lenoir GM. An estimate <strong>of</strong> the heritable fraction <strong>of</strong><br />
childhood cancer. Br J Cancer. 1991;63:993-999.<br />
3. Pakakasama S, Tomlinson GE. Genetic predisposition and screening in<br />
pediatric cancer. Pediatr Clin North Am. 2002;49:1393-1413.<br />
4. Plon SE, Nathanson K. Inherited susceptibility for pediatric cancer.<br />
Cancer J. 2005;11:255-267.<br />
5. Strahm B, Malkin D. Hereditary cancer predisposition in children:<br />
genetic basis and clinical implications. Int J Cancer. 2006;119:2001-2006.<br />
6. Garber JE, Offit K. Hereditary cancer predisposition syndromes. J Clin<br />
Oncol. 2005;23:276-292.<br />
7. Knapke S, Nagarajan R, Correll J, et al. Hereditary cancer risk assessment<br />
in a pediatric oncology follow-up clinic. Pediatr Blood Cancer. <strong>2012</strong>;58:85-89.<br />
8. Schiffman JD. Hereditary cancer syndromes: if you look, you will find<br />
them. Pediatr Blood Cancer. <strong>2012</strong>;58:5-6.<br />
9. <strong>American</strong> <strong>Society</strong> <strong>of</strong> <strong>Clinical</strong> <strong>Oncology</strong> policy statement update: genetic<br />
testing for cancer susceptibility. J Clin Oncol. 2003;21:2397-2406.<br />
10. Eng C, Hampel H, de la Chapelle A. Genetic testing for cancer<br />
predisposition. Annu Rev Med. 2001;52:371-400.<br />
11. Villani A, Tabori U, Schiffman J, et al. Biochemical and imaging<br />
surveillance in germline TP53 mutation carriers with Li-Fraumeni syndrome:<br />
a prospective observational study. Lancet Oncol. 2011;12:559-567.<br />
12. Nichols KE, Malkin D, Garber JE, et al. Germ-line TP53 mutations<br />
predispose to a wide spectrum <strong>of</strong> early-onset cancers. Cancer Epidemiol<br />
Biomarkers Prev. 2001;10:83-87.<br />
13. Qualman SJ, Bowen J, Erdman SH. Molecular basis <strong>of</strong> the brain<br />
tumor-polyposis (Turcot) syndrome. Pediatr Dev Pathol. 2003;6:574-576.<br />
14. Royer-Pokora B, Beier M, Henzler M, et al. Am J Med Genet A.<br />
2004;127:249-257.<br />
15. Ruteshouser EC, Huff V. Familial Wilms tumor. Am J Med Genet C<br />
Semin Med Genet. 2004;129C:29-34.<br />
16. Varley JM, McGown G, Thorncr<strong>of</strong>t M, et al. Are there low-penetrance<br />
TP53 alleles? Evidence from childhood adrenocortical tumors. Am J Hum<br />
Genet. 1999;65:995-1006.<br />
17. Gonzalez KD, Noltner KA, Buzin CH, et al. Beyond Li Fraumeni<br />
syndrome: clinical characteristics <strong>of</strong> families with TP53 germline mutations.<br />
J Clin Oncol. 2009;27:1250-1256.<br />
18. Gozali AE, Britt B, Shane L, et al. Choroid plexus tumors; management,<br />
outcome, and association with the Li-Fraumeni syndrome: The Children’s<br />
Hospital Los Angeles (CHLA) experience, 1991-2010. Pediatr Blood<br />
Cancer. Epub 2011 Oct 11.<br />
19. Tabori U, Shlien A, Baskin B, et al. TP53 alterations determine clinical<br />
subgroups and survival <strong>of</strong> patients with choroid plexus tumors. J Clin Oncol.<br />
2010;28:1995-2001.<br />
20. Tinat J, Bougeard G, Baert-Desurmont S, et al. 2009 version <strong>of</strong> the<br />
Chompret criteria for Li Fraumeni syndrome. J Clin Oncol. 2009;27:e108-109.<br />
21. Diller L, Sexsmith E, Gottlieb A, et al. Germline p53 mutations are<br />
frequently detected in young children with rhabdomyosarcoma. J Clin Invest.<br />
1995;95:1606-1611.<br />
22. Calvert G, Randall LR, Jones KB, et al. At risk populations for<br />
osteosarcoma: the syndromes and beyond. Sarcoma. In press.<br />
23. Draper GJ, Sanders BM, Brownbill PA, et al. Patterns <strong>of</strong> risk <strong>of</strong><br />
hereditary retinoblastoma and applications to genetic counseling. Br J<br />
Cancer. 1992;66:211-219.<br />
24. King KS, Prodanov T, Kantorovich V, et al. Metastatic pheochromocytoma/paraganglioma<br />
related to primary tumor development in childhood or<br />
adolescence: significant link to SDHB mutations. J Clin Oncol. 2011;29:4137-<br />
4142.<br />
25. Schiffman JD. No child left behind in SDHB testing for paragangliomas<br />
and pheochromocytomas. J Clin Oncol. 2011;29:4070-4072.<br />
26. Eaton KW, Tooke LS, Wainwright LM, et al. Spectrum <strong>of</strong> SMARCB1/<br />
Stock<br />
Ownership Honoraria<br />
REFERENCES<br />
Research<br />
Funding<br />
Expert<br />
Testimony<br />
Other<br />
Remuneration<br />
INI1 mutations in familial and sporadic rhabdoid tumors. Pediatr Blood<br />
Cancer. 2011;56:7-15.<br />
27. Janeway KA, Kim SY, Lodish M, et al. Defects in succinate dehydrogenase<br />
in gastrointestinal stromal tumors lacking KIT and PDGFRA mutations.<br />
Proc Natl Acad Sci USA.2011;108:314-318.<br />
28. Aretz S, Koch A, Uhlhaas S, et al. Should children at risk for familial<br />
adenomatous polyposis be screened for hepatoblastoma and children with<br />
apparently sporadic hepatoblastoma be screened for APC germline mutations?<br />
Pediatr Blood Cancer. 2006;47:811-818.<br />
29. Dinauer C, Francis GL. Thyroid cancer in children. Endocrinol Metab<br />
Clin North Am. 2007;36:779-806.<br />
30. Moore SW, Appfelstaedt J, Zaahl MG. Familial medullary carcinoma<br />
prevention, risk evaluation, and RET in children <strong>of</strong> families with MEN2.<br />
J Pediatr Surg. 2007;42:326-332.<br />
31. Walker M, O’Sullivan B, Perakath B, et al. Selecting patients with<br />
young-onset colorectal cancer for mismatch repair gene analysis. Br J Surg.<br />
2007;94:1567-1571.<br />
32. Lwiwski N, Greenberg CR, Mhanni AA. Genetic testing <strong>of</strong> children at<br />
risk for adult onset conditions: when is testing indicated? J Genet Couns.<br />
2008;17:523-525.<br />
33. Seif AE. Pediatric leukemia predisposition syndromes: clues to understanding<br />
leukemogenesis. Cancer Genetics. 2011;204:227-244.<br />
34. Imh<strong>of</strong> SM, Moll AC, Schouten-van Meeteren AY. Stage <strong>of</strong> presentation<br />
and visual outcome <strong>of</strong> patients screened for familial retinoblastoma: nationwide<br />
registration in the Netherlands. Br J Ophthalmol. 2006;90:875-878.<br />
35. Kivelö T. Trilateral retinoblastoma: a meta-analysis <strong>of</strong> hereditary<br />
retinoblastoma associated with primary ectopic intracranial retinoblastoma.<br />
J Clin Oncol. 1999;17:1829-1837.<br />
36. Varley JM, Evans DG, Birch JM. Li-Fraumeni syndrome: A molecular<br />
and clinical review. Br J Cancer. 1997;76:1-14.<br />
37. Kloos RT, Eng C, Evans DB, et al. Medullary thyroid cancer: management<br />
guidelines <strong>of</strong> the <strong>American</strong> Thyroid Association. Thyroid. 2009;19:565-<br />
612.<br />
38. Skinner MA, Moley JA, Dilley WG, et al. Prophylactic thyroidectomy in<br />
multiple endocrine neoplasia type 2A. N Engl J Med. 2005;353:1105-1113.<br />
39. National Comprehensive Cancer Network. Colorectal cancer screening.<br />
NCCN <strong>Clinical</strong> Practice Guidelines in <strong>Oncology</strong>. Version 2.2011. www.nccn.<br />
org. Accessed February 6, 2011.<br />
40. Bapat B, Noorani H, Cohen Z, et al. Cost comparisons <strong>of</strong> predictive<br />
genetic testing versus conventional screening for familial adenomatous polyposis.<br />
Gut. 1999;44:698-703.<br />
41. Boon EM, Keller JJ, Wormhoudt TA, et al. Sulindac targets nuclear<br />
beta-catenin accumulation and Wnt signaling in adenomas <strong>of</strong> patients with<br />
familial adenomatous polyposis and in human colorectal cancer cell lines.<br />
Br J Cancer. 2004;90:224-229.<br />
42. Steinbach G, Lynch PM, Phillips RK, et al. The effect <strong>of</strong> celecoxib, a<br />
cyclooxygenase-2 inhibitor, in familial adenomatous polyposis. N Engl J Med.<br />
2000;342:1946-1952.<br />
43. Lynch PM, Ayers GD, Hawk E, et al. The safety and efficacy <strong>of</strong> celecoxib<br />
in children with familial adenomatous polyposis. Am J Gastroenterol. 2010;<br />
105:1437-1443.<br />
44. <strong>Clinical</strong>Trials.gov. Accessed March 23, <strong>2012</strong>.<br />
45. Riley BD, Culver JO, Skrzynia C, et al. Essential elements <strong>of</strong> genetic<br />
cancer risk assessment, counseling and testing: updated recommendations <strong>of</strong><br />
the National <strong>Society</strong> <strong>of</strong> Genetic Counselors. J Genetic Counsel. Epub 2011 Dec<br />
2.<br />
46. Clayton EW. Genetic testing in children. J Med Philos. 1997;22:233-<br />
251.<br />
47. Malpas PJ. Predictive genetic testing <strong>of</strong> children for adult-onset diseases<br />
and psychological harm. J Med Ethics. 2008;34:275-278.<br />
48. Tischkowitz M, Rosser E. Inherited cancer in children: practical/ethical<br />
problems and challenges. Eur J Cancer. 2004;40:2459-2470.<br />
583
49. Nelson RM, Botkin JR, Kodish ED, et al. Ethical issues with genetic<br />
testing in pediatrics. Pediatrics. 2001;107:1451-1455.<br />
50. Brandi ML, Gagel RF, Angeli A, et al. Guidelines for diagnosis and<br />
therapy <strong>of</strong> MEN type 1 and type 2. J Clin Endocrinol Metab. 2001;86:5658-5671.<br />
51. GeneReviews. www.genetests.org. Accessed March 23, <strong>2012</strong>.<br />
52. Debaun MR. Screening protocol for Beckwith-Wiedemann syndrome.<br />
http://www.beckwith-wiedemann.info/protocol.html. Accessed February 6,<br />
<strong>2012</strong>.<br />
53. Durno CA, Aronson M, Tabori U, et al. Oncologic surveillance for<br />
subjects with biallelic mismatch repair gene mutations: 10-year follow-up <strong>of</strong> a<br />
kindred. Pediatr Blood Cancer. Epub 2011 Dec 16.<br />
54. Timmers HJ, Gimenez-Roqueplo AP, Mannelli M, et al. <strong>Clinical</strong> aspects<br />
<strong>of</strong> SDHx-related pheochromocytoma and paraganglioma. Endocr Relat Cancer.<br />
2009;16:391-400.<br />
584<br />
KNAPKE ET AL<br />
55. National Cancer Institute. Retinoblastoma treatment (Physician Data<br />
Query). http://www.cancer.gov/cancertopics/pdq/treatment/retinoblastoma/<br />
HealthPr<strong>of</strong>essional.Accessed January 30, <strong>2012</strong>.<br />
56. National Comprehensive Cancer Network. Genetic/familial high-risk<br />
assessment: breast and ovarian. NCCN <strong>Clinical</strong> Practice Guidelines in<br />
<strong>Oncology</strong>. Version 1.2011. www.nccn.org. Accessed February 6, <strong>2012</strong>.<br />
57. VHL Family Alliance. Suggested protocol for screening. www.vhl.org.<br />
Accessed February 6, <strong>2012</strong>.<br />
58. Scott RH, Walker L, Olsen OE, et al. Surveillance for Wilms tumor in<br />
at-risk children: pragmatic recommendations for best practice. Arch Dis<br />
Child. 2006;91:995-999.<br />
59. Choyke PL, Siegel MJ, Craft AW, et al. Screening for Wilms tumor in<br />
children with Beckwith-Wiedemann syndrome or idiopathic hemihypertrophy.<br />
Med Pediatr Oncol. 1999;32:196-200.
HOW TO MANAGE VERY RARE PEDIATRIC CANCERS<br />
CHAIR<br />
Alberto S. Pappo, MD<br />
St. Jude Children’s Research Hospital<br />
Memphis, TN<br />
SPEAKERS<br />
Fariba Navid, MD<br />
St. Jude Children’s Research Hospital<br />
Memphis, TN<br />
Stephen Skapek, MD<br />
The University <strong>of</strong> Chicago<br />
Chicago, IL
Treating Rare Cancer in Children:<br />
The Importance <strong>of</strong> Evidence<br />
Overview: The study <strong>of</strong> pediatric rare cancers, which account<br />
for approximately 9% <strong>of</strong> all childhood malignancies, has been<br />
hindered by their histologic heterogeneity and by their preferential<br />
occurrence in adolescents, a population that has been<br />
underrepresented in clinical trials sponsored by the National<br />
Cancer Institute.<br />
The use <strong>of</strong> cooperative group and investigator-initiated<br />
registries can help improve our ability to identify and select<br />
populations <strong>of</strong> patients with rare cancers that can benefit from<br />
single-arm studies, and incorporation <strong>of</strong> biologic aims and<br />
tissue banking can accelerate our understanding <strong>of</strong> the biology<br />
<strong>of</strong> these cancers. These studies should be promoted<br />
further through expansion <strong>of</strong> international outreach efforts.<br />
THE STUDY <strong>of</strong> rare pediatric cancers is complicated by<br />
the fact that pediatric cancer itself is a rare occurrence,<br />
with only approximately 13,000 patients younger than age<br />
20 diagnosed yearly in the United States. 1 Several groups<br />
have tried to define a pediatric rare cancer; for example the<br />
European Italian Tumori Rari in Eta Pediatrica (TREP)<br />
(“the rare tumor project for pediatric patients”) project uses<br />
a definition <strong>of</strong> a cancer that has an annual incidence <strong>of</strong> up to<br />
2 per million and for whom there is no available clinical<br />
trial. 2 However, on the basis <strong>of</strong> data from the Surveillance,<br />
Epidemiology, and End Results (SEER) program <strong>of</strong> the<br />
National Cancer Institute, this definition would exclude<br />
tumors such as thyroid cancer and melanoma, which have<br />
an increasing incidence in adolescents and young adults, but<br />
are still considered rare. With the merger <strong>of</strong> the Pediatric<br />
<strong>Oncology</strong> Group and the Children’s Cancer Group in 2000,<br />
the newly formed Children’s <strong>Oncology</strong> Group (COG) was<br />
poised to increase our understanding <strong>of</strong> rare cancers in<br />
children. With this objective in mind, a Rare Tumor Committee<br />
was formed and the results <strong>of</strong> its initial experience<br />
within the context <strong>of</strong> a cooperative group have been published<br />
and summarized. 3 When initially conceived, the main<br />
objectives <strong>of</strong> the COG Rare Tumor Committee were to<br />
facilitate the study <strong>of</strong> rare cancers, using the infrastructure<br />
available at the time that the two Groups merged. To better<br />
estimate the actual numbers <strong>of</strong> rare tumors diagnosed in the<br />
United States, we opted to define them as those tumors<br />
(mostly carcinomas) mostly diagnosed in older patients.<br />
These subset <strong>of</strong> tumors closely fit the description <strong>of</strong> “other<br />
malignant epithelial neoplasms and melanomas” listed in<br />
the International Classification <strong>of</strong> Childhood Cancer subgroup<br />
XI <strong>of</strong> the SEER database. These histologies include<br />
nasopharyngeal carcinoma, adrenocortical carcinoma, melanoma,<br />
nonmelanoma skin cancers, and other unspecified<br />
carcinomas. We then used the available COG registry to<br />
capture cases <strong>of</strong> these rare cancers and compared the registration<br />
rates with the expected number <strong>of</strong> cases as calculated<br />
according to available through the SEER database. We<br />
were surprised to see that only 7% <strong>of</strong> expected cases <strong>of</strong><br />
selected histologies such as melanoma, thyroid carcinoma,<br />
adrenocortical carcinoma, and nasopharyngeal carcinoma<br />
were registered within COG. We also noted that the registration<br />
rates varied according to histologic subtype and age,<br />
586<br />
By Alberto S. Pappo, MD<br />
Well-designed preclinical models that accurately recapitulate<br />
human disease <strong>of</strong>fer an attractive alternative to the study<br />
<strong>of</strong> rare cancers and may accelerate the process <strong>of</strong> target<br />
identification and drug discovery and development.<br />
The concept <strong>of</strong> specialized clinics for selected rare cancers<br />
has proven to be very successful in pediatric gastrointestinal<br />
tumors. This paradigm should be further explored in other rare<br />
cancers because it <strong>of</strong>fers an unprecedented opportunity to<br />
collaborate closely with interested investigators. In addition,<br />
it <strong>of</strong>fers patients an opportunity to discuss their disease with<br />
specialists, allows these patients to provide tissue for further<br />
research, and ultimately can promote the development <strong>of</strong><br />
clinical trials that are unique for that specific disease.<br />
with very low registration rates in patients with melanoma<br />
and thyroid carcinoma and higher registration rates in<br />
patients with retinoblastoma, adrenocortical carcinoma, and<br />
nasopharyngeal carcinoma. These differences might be explained<br />
in part by referral patterns that are dependent on<br />
the age <strong>of</strong> the patient. Younger patients with adrenocrtiocal<br />
carcinoma and retinoblastoma require multidisciplinary<br />
care and are likely to be referred to tertiary academic<br />
institutions for treatment. In contrast, adolescents with<br />
thyroid carcinoma or melanoma are treated <strong>of</strong>ten with<br />
primary surgical resection by pr<strong>of</strong>essionals outside <strong>of</strong> a<br />
pediatric cancer program.<br />
Our initial mandate also included increasing the number<br />
<strong>of</strong> banked “rare tumor” biologic specimens available for<br />
future research through a COG banking specimen protocol,<br />
activated in October <strong>of</strong> 2003. The number <strong>of</strong> samples received<br />
by this repository over a 4-year period totaled 517<br />
snap-frozen specimens, <strong>of</strong> which only 56 (11%) were considered<br />
to belong to a rare cancer. Finally, the COG Rare<br />
Tumor Committee promoted the development <strong>of</strong> clinical<br />
trials, and two COG-sponsored trials for the treatment <strong>of</strong><br />
nasopharyngeal carcinoma and adrenocortical carcinoma<br />
were developed. The enrollment rates for both studies were<br />
less than initially projected, and the statistical sections had<br />
to be modified in order to ensure that these trials could be<br />
completed in a timely manner. 3 Thus our preliminary experience<br />
with the study <strong>of</strong> rare cancers within the context <strong>of</strong> a<br />
pediatric cooperative group was somewhat disappointing<br />
and highlighted the difficulties associated with the study<br />
<strong>of</strong> this very challenging group <strong>of</strong> pediatric tumors. Other<br />
opportunities or avenues for research that could potentially<br />
increase our knowledge in rare cancers, as well as methods<br />
for evidence gathering and interpretation must be considered.<br />
From the Division <strong>of</strong> <strong>Oncology</strong>, St. Jude Children’s Research Hospital, Memphis, TN.<br />
Author’s disclosure <strong>of</strong> potential conflicts <strong>of</strong> interest are found at the end <strong>of</strong> this article.<br />
Address reprint requests to Albert S. Pappo, MD, St. Jude Children’s Research Hospital,<br />
262 Danny Thomas Place, Memphis, TN 38105; email: alberto.pappo@stjude.org.<br />
© <strong>2012</strong> by <strong>American</strong> <strong>Society</strong> <strong>of</strong> <strong>Clinical</strong> <strong>Oncology</strong>.<br />
1092-9118/10/1-10
RARE CANCER IN CHILDREN<br />
Registries<br />
The use <strong>of</strong> registries to study rare cancers has become<br />
commonplace within the setting <strong>of</strong> national and international<br />
cooperative groups. 2-5 For example, TREP in Italy<br />
was launched in 2000 and accrued more than 300 eligible<br />
patients with a variety <strong>of</strong> histologies over a 6-year period. 6<br />
In their experience, thyroid carcinoma, carcinoma <strong>of</strong> the<br />
appendix, and gonadal germ cell tumors were the most<br />
common tumors registered, and significant underreporting<br />
<strong>of</strong> various rare cancers was also found among adolescents. 6<br />
The TREP project has published valuable information regarding<br />
the natural history, treatment, and outcome <strong>of</strong><br />
selected rare cancers and has conducted a prospective trial<br />
for the treatment <strong>of</strong> nasopharyngeal carcinoma. 7 In an effort<br />
to expand the registration efforts within the Children’s<br />
<strong>Oncology</strong> Group, the Children’s Cancer Research Network<br />
(CCRN) was launched in 2007. The purpose <strong>of</strong> this initiative<br />
was to integrate the registration process for patients with<br />
tumors at COG member institutions into the research <strong>of</strong> the<br />
group, providing a research resource in North America. The<br />
main purpose <strong>of</strong> this trial was to obtain informed consent<br />
from patients—infant and children, adolescent, and young<br />
adults with newly diagnosed cancers—and securely record<br />
patient identification and contact information. In addition,<br />
data concerning their cancer and consent for permission for<br />
future contact with the patients for possible epidemiologic<br />
studies is also obtained. In contrast with the CCRN, SEER<br />
does not ascertain all cases <strong>of</strong> childhood cancer and does not<br />
allow for future contact <strong>of</strong> research participants, limiting the<br />
availability <strong>of</strong> epidemiologic studies in these patients. A<br />
preliminary analysis <strong>of</strong> enrollments <strong>of</strong> rare cancers to the<br />
CCRN study demonstrates that the registration rates for<br />
adrenocortical carcinoma, thyroid carcinoma, nasopharyngeal<br />
carcinoma, and melanoma have not significantly<br />
changed, with approximately two-thirds <strong>of</strong> the estimated<br />
incidence cases <strong>of</strong> patients with adrenocortical carcinoma<br />
being enrolled but only approximately 5% <strong>of</strong> patients with<br />
melanoma being enrolled on the registry. In an attempt to<br />
more efficiently define the natural history and biology or<br />
rare cancers, various groups have developed independent<br />
registries that have provided meaningful insights into the<br />
biology <strong>of</strong> pediatric rare cancers. The International Pediatric<br />
Adrenocortical Tumor Registry was founded in 1990 and has<br />
KEY POINTS<br />
● Rare cancers in children are difficult to study.<br />
● Registries and single-arm trials can help develop<br />
standards <strong>of</strong> care and increase the numbers <strong>of</strong> biologic<br />
specimens, which could further facilitate study<br />
<strong>of</strong> these tumors.<br />
● Preclinical models <strong>of</strong>fer a unique mechanism for<br />
rapidly identifying novel druggable targets and for<br />
prioritizing new therapies.<br />
● Specialized clinics can foster collaborations that will<br />
improve the care and study <strong>of</strong> pediatric rare cancers.<br />
● International collaborative efforts must be expanded<br />
in order to speed up progress in the field <strong>of</strong> pediatric<br />
rare cancers.<br />
described the clinical characteristics and outcome <strong>of</strong> 254 <strong>of</strong><br />
children with this rare disease. 8 Their findings provided the<br />
basis for the development <strong>of</strong> a collaborative cooperative<br />
group trial between COG and Brazilian institutions. This<br />
trial assesses the efficacy <strong>of</strong> surgery in stage I disease, the<br />
role <strong>of</strong> retroperitoneal lymph node dissection in stage II<br />
disease, and the role <strong>of</strong> multimodal chemotherapy with<br />
mitotane and cisplatin-based regimens in advanced-stage<br />
disease. In addition, the trial will continue to expand on the<br />
initial observations <strong>of</strong> the registry that TP53 germ-line<br />
mutations appear to occur in up to 70% <strong>of</strong> cases (one-third<br />
being de novo) and that a unique TP53 Arg337 mutation<br />
affects children in Southern Brazil. The latter does not show<br />
the hereditary pattern seen in Li–Fraumeni syndrome and<br />
has different transactivation activity that may disrupt the<br />
function <strong>of</strong> p53 in a pH dependent manner. 9,10 In another<br />
independent effort, investigators <strong>of</strong> the pleuropulmonary<br />
blastoma registry have identified heterozygous germ-line<br />
mutations <strong>of</strong> DICER1, an endoribonuclease that is essential<br />
for processing micro RNAs in 10 families with familial<br />
pleuropulmonary blastoma. 11 More recently, the mutational<br />
spectrum <strong>of</strong> this gene has been expanded to include other<br />
tumors and associations including ovarian Sertoli-Leydig<br />
cell tumors, cystic nephroma, thyroid cysts, multinodular<br />
goiter, and embryonal rhabdomyosarcoma. 12-15<br />
Preclinical Models<br />
Well-designed comprehensive preclinical studies using<br />
genetically engineered and orthotopic xenografts that<br />
closely recapitulate the molecular, cellular, and genetic<br />
features <strong>of</strong> human tumors can be an invaluable tool for<br />
identifying novel targets and promising therapies for rare<br />
cancers. In adults, the use <strong>of</strong> preclinical models successfully<br />
identified everolimus and sunitinib as active agents for the<br />
treatment <strong>of</strong> pancreatic neuroendocrine tumors. 16,17 Retinoblastoma<br />
is a rare cancer <strong>of</strong> the retina that begins in utero<br />
and is usually diagnosed during the first few years <strong>of</strong> life.<br />
There are approximately 300 cases in the United States each<br />
year and approximately 5,000 diagnosed world-wide. Preclinical<br />
studies have determined that the p53 gene is intact<br />
but the pathway is silenced by overexpression <strong>of</strong> MDMX. 18<br />
Preclinical models have demonstrated that the subconjunctival<br />
administration <strong>of</strong> nutlin-3a, a small molecule inhibitor<br />
<strong>of</strong> the MDMX-p53 interaction in combination with systemic<br />
topotecan is effective in killing and reducing the tumor<br />
burden <strong>of</strong> retinoblastoma cells in culture, as well as in<br />
genetic and human orthotopic xenograft models <strong>of</strong> retinoblastoma.<br />
19 These studies provided direct evidence that<br />
targeting the p53 pathway in vivo with MDMX inhibitors is<br />
feasible and provides a potential novel therapy for this<br />
disease. Similarly, using an integrated epigenetic and gene<br />
expression analysis, SYK has been shown to be an important<br />
oncogene in retinoblastoma. Furthermore, the use <strong>of</strong> SYK<br />
inhibitors such as BAY-613606 and R406 cause cell death;<br />
when combined with topotecan, this combination significantly<br />
improves the outcome <strong>of</strong> mice. 20 These observations<br />
suggest that well-designed preclinical studies can identify<br />
novel targets and prioritize the development <strong>of</strong> new agents<br />
for rare cancers.<br />
Prospective <strong>Clinical</strong> Trials<br />
The small number <strong>of</strong> patients available for enrollment in<br />
pediatric rare cancer trials prevents the design <strong>of</strong> prospec-<br />
587
tive randomized trials. However, small, well-designed, and<br />
meaningful clinical trials can yield important information<br />
about the biology, clinical presentation, and treatment options<br />
<strong>of</strong> rare diseases. Additionally, limited trials taht explore<br />
novel therapies such as ipilimumab and vemurafenib<br />
for pediatric melanoma should be encouraged. Unfortunately,<br />
perhaps as a result <strong>of</strong> the current infrastructure<br />
constraints, even single-arm cooperative group studies for<br />
rare cancers have had suboptimal enrollment rates. 3,21<br />
Specialized Clinics<br />
Highly specialized clinics for rare diseases is not a new<br />
concept; however, rare cancer clinics that address specific<br />
tumors such as the one developed at the NIH for pediatric<br />
gastrointestinal tumors (GIST; www.pediatricgist.cancer.<br />
gov) are unique and provide a novel model for multidisciplinary<br />
care <strong>of</strong> pediatric patients with rare cancers. In this<br />
clinic, physicians and other health care pr<strong>of</strong>essionals from<br />
various disciplines including oncology, surgery, genetics,<br />
and nutrition gather twice a year to study patients with<br />
pediatric and wild-type GIST. Since its inception in 2008,<br />
more than 90 patients have been seen, 38% <strong>of</strong> whom are<br />
pediatric patients and 41% are young adults. Tissue has<br />
been obtained from 61 patients and fresh frozen tumor in<br />
Author’s Disclosure <strong>of</strong> Potential Conflicts <strong>of</strong> Interest<br />
Author<br />
Employment or<br />
Leadership<br />
Positions<br />
Consultant or<br />
Advisory Role<br />
Alberto S. Pappo Ziopharm<br />
<strong>Oncology</strong><br />
1. Howlader N, Krapcho M, Neyman N, et al (eds). SEER Cancer Statistics<br />
Review, 1975-2008: Updated November 10, 2011. SEER web site. http://<br />
seer.cancer.gov/csr/1975_2008/index.html. Accessed February 26, <strong>2012</strong>.<br />
2. Ferrari A, Bisogno G, De Salvo GL, et al. The challenge <strong>of</strong> very rare<br />
tumours in childhood: The Italian TREP project. Eur J Cancer. 2007;43(4):<br />
654-659.<br />
3. Pappo AS, Krailo M, Chen Z, et al. Infrequent tumor initiative <strong>of</strong> the<br />
Children’s <strong>Oncology</strong> Group: Initial lessons learned and their impact on future<br />
plans. J Clin Oncol. 2010;28(33):5011-5016.<br />
4. Bien E, Godzinski J, Dall’igna P, et al. Pancreatoblastoma: A report from<br />
the European cooperative study group for paediatric rare tumours (EXPeRT).<br />
Eur J Cancer. 2011; 47(15):2347-2352.<br />
5. Brecht IB, Graf N, Schweinitz D, et al. Networking for children and<br />
adolescents with very rare tumors: Foundation <strong>of</strong> the GPOH Pediatric Rare<br />
Tumor Group. Klin Padiatr. 2009;221(3):181-185.<br />
6. Pastore G, De Salvo GL, Bisogno G, et al. Evaluating access to pediatric<br />
cancer care centers <strong>of</strong> children and adolescents with rare tumors in Italy: The<br />
TREP project. Pediatr Blood Cancer. 2009;53(2):152-155.<br />
7. Casanova M, Bisogno G, Gandola L, et al. A prospective protocol for<br />
nasopharyngeal carcinoma in children and adolescents: The Italian Rare<br />
Tumors in Pediatric Age (TREP) project. Cancer. Epub 2011 Sep 14.<br />
8. Michalkiewicz E, Sandrini R, Figueiredo B, et al. <strong>Clinical</strong> and outcome<br />
characteristics <strong>of</strong> children with adrenocortical tumors: A report from the<br />
International Pediatric Adrenocortical Tumor Registry. J Clin Oncol. 2004;<br />
22(5):838-845.<br />
9. DiGiammarino EL, Lee AS, Cadwell C, et al. A novel mechanism <strong>of</strong><br />
tumorigenesis involving pH-dependent destabilization <strong>of</strong> a mutant p53 tetramer.<br />
Nat Struct Biol. 2002;9(1):12-16.<br />
10. Wasserman JD, Zambetti GP, Malkin D. Towards an understanding <strong>of</strong><br />
the role <strong>of</strong> p53 in adrenocortical carcinogenesis. Mol Cell Endocrinol. <strong>2012</strong>;<br />
351(1):101-110.<br />
588<br />
five. The clinic has facilitated the study <strong>of</strong> rare diseases and<br />
has contributed to the understanding <strong>of</strong> the unique biology<br />
<strong>of</strong> these tumors. For example, pediatric GIST most commonly<br />
affect females, have epithelioid or mixed morphology,<br />
arise in the stomach, have an indolent clinical course, and<br />
rarely have activating mutations <strong>of</strong> KIT or PDGFR. In<br />
addition the almost universal lack <strong>of</strong> SDHB expression in<br />
tumor samples from these patients suggests that defects in<br />
cellular respiration may play a pivotal role in the pathogenesis<br />
<strong>of</strong> the disease in younger patients. 22<br />
In summary, the study <strong>of</strong> rare pediatric cancers is challenging.<br />
Multiple mechanisms should be explored in order to<br />
advance our understanding <strong>of</strong> these diseases. The use <strong>of</strong><br />
international registries can help identify the numbers <strong>of</strong><br />
patients at risk for a specific rare cancer and can aid in the<br />
collection <strong>of</strong> biologic specimens in this population. Welldesigned,<br />
single-arm, collaborative clinical trials that incorporate<br />
banking and biologic endpoints can greatly advance<br />
our understanding <strong>of</strong> these diseases and establish standards<br />
<strong>of</strong> care for these patient. Preclinical models can more rapidly<br />
aid in the identification <strong>of</strong> novel druggable targets. Finally,<br />
specialized clinics <strong>of</strong>fer the opportunity to study large numbers<br />
<strong>of</strong> patients by interested individuals facilitating the<br />
study and specimen collection <strong>of</strong> pediatric rare cancers.<br />
Stock<br />
Ownership Honoraria<br />
REFERENCES<br />
Research<br />
Funding<br />
Expert<br />
Testimony<br />
ALBERTO S. PAPPO<br />
Other<br />
Remuneration<br />
11. Hill DA, Ivanovich J, Priest JR, et al. DICER1 mutations in familial<br />
pleuropulmonary blastoma. Science. 2009;325(5943):965.<br />
12. Doros L, Yang J, Dehner L, et al: DICER1 Mutations in embryonal<br />
rhabdomyosarcomas from children with and without familial PPB-tumor<br />
predisposition syndrome. Pediatr Blood Cancer. Epub 2011 Dec 16.<br />
13. Schultz KA, Pacheco MC, Yang J, et al. Ovarian sex cord-stromal<br />
tumors, pleuropulmonary blastoma and DICER1 mutations: A report from<br />
the International Pleuropulmonary Blastoma Registry. Gynecol Oncol. 2011;<br />
122(2):246-250. Epub 2011 Apr 17.<br />
14. Rio Frio T, Bahubeshi A, Kanellopoulou C, et al. DICER1 mutations in<br />
familial multinodular goiter with and without ovarian Sertoli-Leydig cell<br />
tumors. JAMA. 2011;305(1):68-77.<br />
15. Heravi-Moussavi A, Anglesio MS, Cheng SW, et al. Recurrent somatic<br />
DICER1 mutations in nonepithelial ovarian cancers. N Engl J Med. <strong>2012</strong>;<br />
366(3):234-242.<br />
16. Yao JC, Shah MH, Ito T, et al. Everolimus for advanced pancreatic<br />
neuroendocrine tumors. N Engl J Med. 2011;364(6):514-523.<br />
17. Raymond E, Dahan L, Raoul JL, et al. Sunitinib malate for the<br />
treatment <strong>of</strong> pancreatic neuroendocrine tumors. NEnglJM.2011;364(6):<br />
501-513.<br />
18. Laurie NA, Donovan SL, Shih CS, et al. Inactivation <strong>of</strong> the p53<br />
pathway in retinoblastoma. Nature. 2006;444:61-66.<br />
19. Brennan RC, Federico S, Bradley C, et al. Targeting the p53 pathway in<br />
retinoblastoma with subconjunctival Nutlin-3a. Cancer Res. 2011;71:4205-4213.<br />
20. Zhang J, Benavente CA, McEvoy J, et al. A novel retinoblastoma<br />
therapy from genomic and epigenetic analyses. Nature. <strong>2012</strong>;481:329-334.<br />
21. Casanova M, Bisogno G, Gandola L, et al. A prospective protocol for<br />
nasopharyngeal carcinoma in children and adolescents: The Italian Rare<br />
Tumors in Pediatric Age (TREP) project. Cancer. Epub 2011 Sep 14.<br />
22. Janeway KA, Kim SY, Lodish M, et al. Defects in succinate dehydrogenase<br />
in gastrointestinal stromal tumors lacking KIT and PDGFRA mutations.<br />
Proc Natl Acad Sci USA.2011;108:314-318.
Genetic Alterations in Childhood Melanoma<br />
Overview: Melanoma is rare in children. However, the clinical<br />
features <strong>of</strong> the disease in this population have been welldocumented<br />
through single-institution experiences and<br />
population-based analyses. Still, our understanding <strong>of</strong> the<br />
etiologic factors in children remains unclear and diagnosis <strong>of</strong><br />
DESPITE BEING the most common skin cancer in<br />
childhood, melanoma is a rare childhood cancer, making<br />
up less than 3% <strong>of</strong> all cancers seen in children. In the<br />
United States, approximately 300 to 400 new cases per year<br />
are diagnosed in patients younger than age 20; 80% <strong>of</strong> these<br />
patients are between the ages <strong>of</strong> 15 and 19. 1 The incidence <strong>of</strong><br />
melanoma in this older population is rising yearly. 2 Most<br />
information about the risk factors, natural history, treatment,<br />
and outcome <strong>of</strong> cutaneous melanoma in children has<br />
been derived from retrospective reports <strong>of</strong> single-institution<br />
experiences and population-based analyses (i.e., Surveillance<br />
Epidemiology and End Results [SEER] data). 3-8 The<br />
results <strong>of</strong> some <strong>of</strong> these studies seem to indicate that the<br />
natural history and response to therapy <strong>of</strong> melanoma in<br />
children and adolescents are similar to those in adults and<br />
are stage dependent; however, those <strong>of</strong> other studies refute<br />
these points and suggest that the biology <strong>of</strong> the disease<br />
might be different in children than in adults.<br />
Understanding the biologic similarities and differences<br />
between adult and pediatric melanoma may provide insight<br />
into the causative factors <strong>of</strong> the disease in children and<br />
influence treatment decisions. Such knowledge might also<br />
aid clinicians dealing with diagnostically challenging issues<br />
such as distinguishing between Spitz nevi and spitzoid<br />
melanoma or between malignant transformation and benign<br />
proliferation in congenital nevi. Emerging evidence suggests<br />
that molecular characterization <strong>of</strong> these lesions may be<br />
helpful. This article reviews common genetic alterations in<br />
pediatric melanoma and their potential utility in differentiating<br />
histologically challenging cases.<br />
Common Genetic Alterations<br />
During the past decade, several genetic alterations in<br />
adult melanomas have been described. 9-12 These changes<br />
are summarized in Table 1. Characterization <strong>of</strong> these aberrations<br />
suggests that various distinct molecular pathways<br />
are associated with the development <strong>of</strong> melanoma on the<br />
basis <strong>of</strong> the tumor site, host phenotype, and amount <strong>of</strong> sun<br />
exposure. 13,14 These findings highlight the heterogeneity <strong>of</strong><br />
adult melanomas and raise the possibility that the same<br />
genetic alterations may not be present in childhood melanoma.<br />
Large-scale genetic pr<strong>of</strong>iling studies <strong>of</strong> pediatric melanoma<br />
performed by using comparative genomic<br />
hybridization (CGH) or genome-wide association techniques<br />
have not been reported and may not be feasible because <strong>of</strong><br />
the tumor’s rarity. However, some genes have been evaluated<br />
in the pediatric population; findings about two that are<br />
mutated in this setting are summarized below.<br />
BRAF<br />
The RAS/RAF/MAPK signaling pathway is the one most<br />
commonly implicated in the pathogenesis <strong>of</strong> melanoma and<br />
By Fariba Navid, MD<br />
melanoma remains challenging in certain cases. This article<br />
reviews emerging evidence indicating that molecular characterization<br />
<strong>of</strong> these lesions in children may be <strong>of</strong> diagnostic and<br />
therapeutic value.<br />
is the focus <strong>of</strong> recent drug development efforts. This pathway<br />
is dysregulated in approximately 50% <strong>of</strong> adult melanomas,<br />
frequently having mutations in the BRAF gene (less<br />
frequently NRAS). The most common BRAF mutation in<br />
melanoma is a substitution <strong>of</strong> glutamic acid for valine at<br />
position 600 (BRAF V600E ). This mutation accounts for more<br />
than 90% <strong>of</strong> the BRAF mutations that occur in adult<br />
melanomas and drives cell proliferation, invasion, and metastasis<br />
in these tumors. The U.S. Food and Drug Administration<br />
recently approved vemurafenib, an oral tyrosine<br />
kinase inhibitor <strong>of</strong> mutated BRAF, because <strong>of</strong> its favorable<br />
response rates in adult patients with BRAF-mutated melanoma.<br />
15,16 BRAF V600E mutations in pediatric melanoma<br />
have only been reported from one small study showing the<br />
mutation in five <strong>of</strong> 10 pediatric tumors. 17 Finding<br />
BRAF V600E and other aberrations <strong>of</strong> this pathway in more<br />
samples would provide a rationale to test selective inhibitors<br />
targeting this mutation and other RAS/RAF/MAPK signaling<br />
molecules in pediatric melanoma.<br />
CDKN2A<br />
CDKN2A, located on chromosome 9p21, encodes two distinct<br />
tumor suppressor genes, p16/INK4a and p14/ARF,<br />
that play a key role in cell cycle regulation. The CDKN2A<br />
locus is deleted in approximately 50% <strong>of</strong> sporadic adult<br />
melanomas. 13 Daniotti and colleagues 17 found homozygous<br />
CDKN2A deletions in nine <strong>of</strong> 14 pediatric melanomas (patients<br />
aged 2 to 19). Alterations in CDKN2A are also<br />
implicated in the pathogenesis <strong>of</strong> 25% to 40% <strong>of</strong> familial<br />
cutaneous melanomas. 18 In these patients, melanoma tends<br />
to develop when they are younger but not commonly when<br />
they are younger than age 18. Whiteman and colleagues 19<br />
analyzed DNA from 31 children in the Queensland Cancer<br />
registry who were younger than age 15. One patient among<br />
the 10 with a family history <strong>of</strong> melanoma had a germ-line<br />
mutation in CDKN2A; she had a history <strong>of</strong> two primary<br />
melanomas before the age <strong>of</strong> 13. Among 147 adolescents in<br />
the same registry who were aged 15 to 19, two had germ-line<br />
alterations in CDKN2A; however, neither had a family<br />
history <strong>of</strong> melanoma. 20 In a retrospective review <strong>of</strong> 15<br />
Swedish families with known germ-line CDKN2A mutations,<br />
the youngest family member to develop melanoma<br />
was 18 years. 21 The results <strong>of</strong> these studies suggest that<br />
From the Division <strong>of</strong> Solid Malignancies, Department <strong>of</strong> <strong>Oncology</strong>, St. Jude Children’s<br />
Research Hospital, Memphis, TN.<br />
Authors’ disclosures <strong>of</strong> potential conflicts <strong>of</strong> interest are found at the end <strong>of</strong> this article.<br />
Address reprint requests to Fariba Navid, MD, Division <strong>of</strong> Solid Malignancies, Department<br />
<strong>of</strong> <strong>Oncology</strong>, St. Jude Children’s Research Hospital, 262 Danny Thomas Pl., Memphis,<br />
TN 38105-2794; email: fariba.navid@stjude.org.<br />
© <strong>2012</strong> by <strong>American</strong> <strong>Society</strong> <strong>of</strong> <strong>Clinical</strong> <strong>Oncology</strong>.<br />
1092-9118/10/1-10<br />
589
Gene<br />
Table 1. Common Genetic Alterations in Adult Melanoma<br />
Chromosomal<br />
Locus Alteration Frequency (%)<br />
CDKN2A alterations may play a role in a small subset <strong>of</strong><br />
melanomas that develop in those younger than 18 years.<br />
Distinguishing between Spitz Nevi and<br />
Spitzoid Melanoma<br />
A Spitz nevus usually manifests in childhood and is<br />
considered to be a benign melanocytic lesion. The “classic”<br />
Spitz is described as a small, symmetric dome-shaped lesion;<br />
it is typically nonpigmented and consists <strong>of</strong> epithelioid<br />
and/or spindle-shaped cells with evidence <strong>of</strong> maturation at<br />
the base. The presence <strong>of</strong> a high mitotic rate, pagetoid<br />
spread, lack <strong>of</strong> maturation, pleomorphism and nuclear<br />
atypia, ulceration, and giant cells have all been cited as<br />
discriminating features between Spitz nevi and melanomas;<br />
however, there is no consensus about these characteristics<br />
among pathologists. 22 Therefore, in certain cases, a specific<br />
diagnosis <strong>of</strong> benign Spitz nevi versus malignant melanoma<br />
is difficult to make on the basis <strong>of</strong> histologic features alone,<br />
and no immunohistochemical markers reliably differentiate<br />
KEY POINTS<br />
Implicated in<br />
Familial<br />
Melanoma?<br />
CDKN2A 9p21 Mutated, deleted, silenced 30–70 Yes<br />
CDK4 12q13 Amplified or mutated 1–2 Yes<br />
CCND1 11q13 Amplified 6–44 No<br />
BRAF 7q34 Mutated 50–70 No<br />
NRAS 1p13 Mutated 15–30 No<br />
MITF 3p14 Amplified or mutated 10–16 Yes<br />
MC1R 16q24 Mutated ? Yes<br />
PTEN 10q23 Deleted or mutated 5–20 No<br />
APAF1 12q22 Silenced 40 No<br />
TP53 17p13 Mutated, deleted 10 No<br />
C-KIT 4q11 Mutated 20–40 No<br />
Abbreviations: APAF1, apoptotic protease-activating factor 1; BRAF, v-raf<br />
murine sarcoma viral oncogene homolog B1; CCND1, cyclin D1; CDKN2A,<br />
cyclin-dependent kinase inhibitor 2A; CDK4, cyclin-dependent kinase 4; C-KIT,<br />
stem cell growth factor receptor; MC1R, melanocortin-1 receptor; MITF,<br />
microphthalmia-associated transcription factor; NRAS, v-ras neuroblastoma RAS<br />
viral oncogene homolog; PTEN, phosphatase and tensin homologue; TP53, tumor<br />
protein p53.<br />
● The results <strong>of</strong> molecular studies may aid clinicians in<br />
the diagnosis <strong>of</strong> melanocytic lesions that are histologically<br />
difficult to classify in children.<br />
● In contrast to malignant melanoma, Spitz nevi rarely<br />
contain chromosomal aberrations.<br />
● Benign proliferations arising in congenital nevi have<br />
a different pattern <strong>of</strong> chromosomal aberrations than<br />
do proliferations arising in malignant melanoma.<br />
● Development <strong>of</strong> melanoma in individuals harboring<br />
germ-line mutations in CDKN2A usually occurs in<br />
adulthood and is uncommon in children and adolescents.<br />
● A better understanding <strong>of</strong> the similarities and differences<br />
between genetic alterations in pediatric melanomas<br />
and those in adult melanoma may accelerate<br />
the use <strong>of</strong> promising or targeted therapies in the<br />
pediatric population.<br />
590<br />
between the two neoplasms. Therefore, lesions are <strong>of</strong>ten<br />
referred to as an “atypical Spitz nevus” or “spitzoid tumor <strong>of</strong><br />
uncertain malignant potential.” 23<br />
Chromosomal Aberrations<br />
To better distinguish between melanoma and benign<br />
melanocytic nevi, Bastian and colleagues 24 performed a<br />
CGH study <strong>of</strong> 54 benign nevi and 132 melanomas. Their<br />
findings showed that 127 melanomas (96%) had chromosomal<br />
aberrations, whereas only seven (13%) <strong>of</strong> the benign<br />
nevi had aberrations. All seven benign cases were Spitz nevi.<br />
In six <strong>of</strong> these, the chromosomal abnormality was restricted<br />
to a gain in the short arm <strong>of</strong> chromosome 11, an aberration<br />
that was not present in any <strong>of</strong> the melanoma samples. In a<br />
study <strong>of</strong> 102 Spitz nevi, fluorescent in situ hybridization<br />
results showed that 12 had a gain in chromosome 11.<br />
Furthermore, sequencing analysis showed that eight <strong>of</strong> the<br />
12 had a mutated HRAS allele on the gained copy <strong>of</strong><br />
chromosome 11. 25 Since this finding, two independent investigators<br />
have found that HRAS mutations are exclusive to<br />
Spitz and atypical Spitz nevi and are not present in melanoma.<br />
26,27<br />
RAS/RAF/MAPK Pathway Alterations<br />
The RAS/RAF/MAPK signaling pathway, which can lead<br />
to cell growth and division, is activated in Spitz nevi with or<br />
without HRAS mutations. Despite the growth stimulatory<br />
effects <strong>of</strong> this pathway, both lesions have a low proliferative<br />
index as assessed by MIB-1 staining. The growth inhibitory<br />
signal that prevents malignant transformation in these cells<br />
is hypothesized to be a high expression <strong>of</strong> intact p16. 28 In<br />
line with this hypothesis, Dhaybi and colleagues 29 detected<br />
no p16 expression in six cases <strong>of</strong> spitzoid melanoma from<br />
children aged 3 to 14; however, strong p16 expression was<br />
detected in 18 Spitz nevi excised from children aged 1 to 14<br />
and in 12 benign nevi.<br />
Researchers have also investigated the RAS/RAF/MAPK<br />
pathway genes BRAF and NRAS in Spitz nevi and spitzoid<br />
melanoma. Gill and colleagues 26 found no mutations in<br />
either <strong>of</strong> these genes in nine spitzoid melanomas and 10<br />
age-matched Spitz nevi from prepubescent children (aged 2<br />
to 10). A follow-up study <strong>of</strong> 33 spitzoid melanomas in varying<br />
age groups showed mutated BRAF in only one lesion and no<br />
NRAS-mutated lesions. 30 The authors concluded that the<br />
absence <strong>of</strong> the frequently mutated BRAF and NRAS genes<br />
in spitzoid melanoma suggests that these tumors are biologically<br />
different from conventional nonspitzoid melanomas.<br />
However, although others have also reported the absence <strong>of</strong><br />
BRAF mutations in most Spitz nevi, BRAF and NRAS<br />
mutations have been detected in spitzoid melanomas by<br />
other researchers. 27,31-33 The reasons for the conflicting<br />
results are unclear. In these studies and others, mutations<br />
in HRAS, NRAS, and BRAF appear to be mutually exclusive.<br />
Thus, what we can conclude from the mutational<br />
analysis <strong>of</strong> melanocytic lesions is that a lesion with a HRAS<br />
mutation is unlikely to be a melanoma; however, other<br />
features <strong>of</strong> a lesion with a NRAS or BRAF mutation must be<br />
taken into account to make a diagnosis <strong>of</strong> melanoma.<br />
Congenital Nevi and Malignant Transformation<br />
FARIBA NAVID<br />
Congenital nevi are pigmented lesions that are present at<br />
birth or appear shortly thereafter, having an overall preva-
GENETIC ALTERATIONS IN CHILDHOOD MELANOMA<br />
lence <strong>of</strong> approximately 1%. 34 The risk <strong>of</strong> malignant transformation<br />
<strong>of</strong> congenital nevi increases with the size <strong>of</strong> the<br />
nevus. In a review <strong>of</strong> 289 published cases <strong>of</strong> large congenital<br />
nevi (more than 20 cm in diameter), melanoma developed<br />
within a congenital nevus in 34 patients (12%). 35 The<br />
clinical or molecular factors that lead to the transformation<br />
<strong>of</strong> these lesions are unknown.<br />
Secondary Proliferations<br />
Secondary proliferations within a congenital nevus can<br />
occur and are <strong>of</strong>ten difficult to histologically classify as being<br />
malignant or benign. The results <strong>of</strong> CGH analysis <strong>of</strong> 29<br />
congenital nevi and associated benign and malignant proliferations<br />
showed no aberrations in the typical congenital<br />
nevi, in congenital nevi <strong>of</strong> increased cellularity, or in congenital<br />
nevi with benign proliferation. Additionally, seven <strong>of</strong><br />
nine cases <strong>of</strong> congenital melanocytic nevi with proliferations<br />
simulating nodular melanoma predominately contained either<br />
gains or losses <strong>of</strong> entire chromosomes. This finding<br />
differs from that in melanoma in which fragments <strong>of</strong> various<br />
chromosomes are detected. The pattern <strong>of</strong> chromosomal<br />
abnormalities between the nodular lesions and melanoma<br />
was also different. In contrast, the six cases <strong>of</strong> melanoma<br />
arising in congenital nevi had multiple cytogenetic aberrations<br />
in a pattern that was indistinguishable from that <strong>of</strong><br />
melanomas not associated with congenital nevi. 36 These<br />
findings suggest that the type <strong>of</strong> genomic instability in<br />
atypical nodular proliferations differs from that predominat-<br />
Author’s Disclosures <strong>of</strong> Potential Conflicts <strong>of</strong> Interest<br />
ing in melanoma and that CGH may be a useful tool in<br />
distinguishing between the two entities.<br />
RAS/RAF/MAPK Pathway Alterations<br />
Some genes reported to be associated with melanoma,<br />
including p53, p16, and CDK4, do not appear to be altered in<br />
these proliferative lesions within a congenital nevi. 37 NRAS<br />
mutations, however, are frequent in congenital nevi: 26 <strong>of</strong> 32<br />
cases <strong>of</strong> congenital nevi in one series and 10 <strong>of</strong> 17 cases in<br />
another had NRAS mutations. 37,38 Reports <strong>of</strong> BRAF mutations<br />
in congenital nevi are conflicting. For example, Yazdi<br />
and colleagues 32 and Pollock and colleagues 39 detected<br />
BRAF mutations in six <strong>of</strong> 13 congenital nevi and six <strong>of</strong> seven<br />
congenital nevi, respectively. In contrast, Bauer and colleagues<br />
38 found no BRAF mutations in a series <strong>of</strong> 32 cases <strong>of</strong><br />
congenital nevi. They speculate that the discrepancy between<br />
their findings and others may be due to selection bias<br />
because they chose only lesions present at birth, but others<br />
may have also considered nevi appearing in the first year <strong>of</strong><br />
life. It is well-documented that more than 80% <strong>of</strong> acquired<br />
nevi have BRAF mutations. 39 The absence <strong>of</strong> BRAF mutations<br />
would support the hypothesis that nevi that develop in<br />
utero, in the absence <strong>of</strong> ultraviolet exposure, are genetically<br />
distinct from nevi that develop after birth.<br />
A better understanding <strong>of</strong> the embryonic development <strong>of</strong><br />
congenital nevi and their transformation to benign and<br />
malignant lesions may provide some insight into the etiology<br />
<strong>of</strong> melanoma that develops in childhood in the absence <strong>of</strong><br />
these lesions.<br />
Employment or<br />
Leadership Consultant or Stock<br />
Research<br />
Author<br />
Positions Advisory Role Ownership Honoraria Funding<br />
Fariba Navid Schering-Plough<br />
1. Bernstein L, Gurney, JG. Carcinomas and other malignant epithelial<br />
neoplasms. In Ries LG, Smith M, Gurney J, et al (eds). Cancer Incidence and<br />
Survival among Children and Adolescents: United States SEER Program<br />
1975-1995. Bethesda, MD: National Cancer Institute, 1999;139-147.<br />
2. Herzog C, Pappo AS, Bondy M, et al. Malignant melanoma. In Bleyer A,<br />
O’Leary M, Barr R, et al (eds). Cancer Epidemiology in Older Adolescents and<br />
Young Adults 15 to 19 Years <strong>of</strong> Age, Including SEER Incidence and Survival:<br />
1975-2000. Bethesda, MD: National Cancer Institute, 2006;53-63.<br />
3. Moore-Olufemi S, Herzog C, Warneke C, et al. Outcomes in pediatric<br />
melanoma: comparing prepubertal to adolescent pediatric patients. Ann Surg.<br />
2011;253:1211-1215.<br />
4. Paradela S, Fonseca E, Prieto VG. Melanoma in children. Arch Pathol<br />
Lab Med. 2011;135:307-316.<br />
5. Aldrink JH, Selim MA, Diesen DL, et al. Pediatric melanoma: a singleinstitution<br />
experience <strong>of</strong> 150 patients. J Pediatr Surg. 2009;44:1514-1521.<br />
6. Strouse JJ, Fears TR, Tucker MA, et al. Pediatric melanoma: risk factor<br />
and survival analysis <strong>of</strong> the surveillance, epidemiology and end results<br />
database. J Clin Oncol. 2005;23:4735-4741.<br />
7. Lange JR, Palis BE, Chang DC, et al. Melanoma in children and<br />
teenagers: an analysis <strong>of</strong> patients from the National Cancer Data Base. J Clin<br />
Oncol. 2007;25:1363-1368.<br />
8. Livestro DP, Kaine EM, Michaelson JS, et al. Melanoma in the young:<br />
differences and similarities with adult melanoma: a case-matched controlled<br />
analysis. Cancer. 2007;110:614-624.<br />
9. Gaudi S, Messina JL. Molecular bases <strong>of</strong> cutaneous and uveal melanomas.<br />
Patholog Res Int. 2011;2011:159421.<br />
10. High WA, Robinson WA. Genetic mutations involved in melanoma: a<br />
summary <strong>of</strong> our current understanding. Adv Dermatol. 2007;23:61-79.<br />
11. Romano E, Schwartz GK, Chapman PB, et al. Treatment implications<br />
REFERENCES<br />
Expert<br />
Testimony<br />
Other<br />
Remuneration<br />
<strong>of</strong> the emerging molecular classification system for melanoma. Lancet Oncol.<br />
2011;12:913-922.<br />
12. Yokoyama S, Woods SL, Boyle GM, et al. A novel recurrent mutation in<br />
MITF predisposes to familial and sporadic melanoma. Nature. 2011;480:99-<br />
103.<br />
13. Curtin JA, Fridlyand J, Kageshita T, et al. Distinct sets <strong>of</strong> genetic<br />
alterations in melanoma. N Engl J Med. 2005;353:2135-2147.<br />
14. Bauer J, Büttner P, Murali R, et al. BRAF mutations in cutaneous<br />
melanoma are independently associated with age, anatomic site <strong>of</strong> the<br />
primary tumor, and the degree <strong>of</strong> solar elastosis at the primary tumor site.<br />
Pigment Cell Melanoma Res. 2011;24:345-351.<br />
15. Flaherty KT, Puzanov I, Kim KB, et al. Inhibition <strong>of</strong> mutated, activated<br />
BRAF in metastatic melanoma. N Engl J Med. 2010;363:809-819.<br />
16. Chapman PB, Hauschild A, Robert C, et al. Improved survival with<br />
vemurafenib in melanoma with BRAF V600E mutation. N Engl J Med.<br />
2011;364:2507-2516.<br />
17. Daniotti M, Ferrari A, Frigerio S, et al. Cutaneous melanoma in<br />
childhood and adolescence shows frequent loss <strong>of</strong> INK4A and gain <strong>of</strong> KIT.<br />
J Invest Dermatol. 2009;129:1759-1768.<br />
18. Kefford RF, Newton Bishop JA, Bergman W, et al. Counseling and DNA<br />
testing for individuals perceived to be genetically predisposed to melanoma: a<br />
consensus statement <strong>of</strong> the Melanoma Genetics Consortium. J Clin Oncol.<br />
1999;17:3245-3251.<br />
19. Whiteman DC, Milligan A, Welch J, et al. Germline CDKN2A mutations<br />
in childhood melanoma. J Natl Cancer Inst. 1997;89:1460.<br />
20. Youl P, Aitken J, Hayward N, et al. Melanoma in adolescents: a<br />
case-control study <strong>of</strong> risk factors in Queensland, Australia. Int J Cancer.<br />
2002;98:92-98.<br />
21. Magnusson S, Borg A, Krist<strong>of</strong>fersson U, et al. Higher occurrence <strong>of</strong><br />
591
childhood cancer in families with germline mutations in BRCA2, MMR and<br />
CDKN2A genes. Fam Cancer. 2008;7:331-337.<br />
22. Barnhill RL, Argenyi ZB, From L, et al. Atypical Spitz nevi/tumors:<br />
lack <strong>of</strong> consensus for diagnosis, discrimination from melanoma, and prediction<br />
<strong>of</strong> outcome. Hum Pathol. 1999;30:513-520.<br />
23. Barnhill RL. The Spitzoid lesion: rethinking Spitz tumors, atypical<br />
variants, ‘Spitzoid melanoma’ and risk assessment. Mod Pathol. 2006;19:S21-<br />
S33 (suppl 2).<br />
24. Bastian BC, Olshen AB, LeBoit PE, et al. Classifying melanocytic<br />
tumors based on DNA copy number changes. Am J Pathol. 2003;163:1765-<br />
1770.<br />
25. Bastian BC, LeBoit PE, Pinkel D. Mutations and copy number increase<br />
<strong>of</strong> HRAS in Spitz nevi with distinctive histopathological features. Am J<br />
Pathol. 2000;157:967-972.<br />
26. Gill M, Cohen J, Renwick N, et al. Genetic similarities between Spitz<br />
nevus and Spitzoid melanoma in children. Cancer. 2004;101:2636-2640.<br />
27. van Dijk MC, Bernsen MR, Ruiter DJ. Analysis <strong>of</strong> mutations in B-RAF,<br />
N-RAS, and H-RAS genes in the differential diagnosis <strong>of</strong> Spitz nevus and<br />
spitzoid melanoma. Am J Surg Pathol. 2005;29:1145-1151.<br />
28. Maldonado JL, Timmerman L, Fridlyand J, et al. Mechanisms <strong>of</strong><br />
cell-cycle arrest in Spitz nevi with constitutive activation <strong>of</strong> the MAP-kinase<br />
pathway. Am J Pathol. 2004;164:1783-1787.<br />
29. Al Dhaybi R, Agoumi M, Gagné I, et al. P16 expression: a marker <strong>of</strong><br />
differentiation between childhood malignant melanomas and Spitz nevi. JAm<br />
Acad Dermatol. 2011;65:357-363.<br />
30. Lee DA, Cohen JA, Twaddell WS, et al. Are all melanomas the same?<br />
592<br />
FARIBA NAVID<br />
Spitzoid melanoma is a distinct subtype <strong>of</strong> melanoma. Cancer. 2006;106:907-<br />
913.<br />
31. Palmedo G, Hantschke M, Rutten A, et al. The T1796A mutation <strong>of</strong> the<br />
BRAF gene is absent in Spitz nevi. J Cutan Pathol. 2004;31:266-270.<br />
32. Yazdi AS, Palmedo G, Flaig MJ, et al. Mutations <strong>of</strong> the BRAF gene in<br />
benign and malignant melanocytic lesions. J Invest Dermatol. 2003;121:1160-<br />
1162.<br />
33. Fullen DR, Poynter JN, Lowe L, et al. BRAF and NRAS mutations in<br />
spitzoid melanocytic lesions. Mod Pathol. 2006;19:1324-1332.<br />
34. Castilla EE, da Graca Dutra M, Orioli-Parreiras IM. Epidemiology <strong>of</strong><br />
congenital pigmented naevi: I. Incidence rates and relative frequencies. Br J<br />
Dermatol. 1981;104:307-315.<br />
35. DeDavid M, Orlow SJ, Provost N, et al. A study <strong>of</strong> large congenital<br />
melanocytic nevi and associated malignant melanomas: review <strong>of</strong> cases in the<br />
New York University Registry and the world literature. J Am Acad Dermatol.<br />
1997;36:409-416.<br />
36. Bastian BC, Xiong J, Frieden IJ, et al. Genetic changes in neoplasms<br />
arising in congenital melanocytic nevi: differences between nodular proliferations<br />
and melanomas. Am J Pathol. 2002;161:1163-1169.<br />
37. Papp T, Pemsel H, Zimmermann R, et al. Mutational analysis <strong>of</strong> the<br />
N-ras, p53, p16INK4a, CDK4, and MC1R genes in human congenital melanocytic<br />
naevi. J Med Genet. 1999;36:610-614.<br />
38. Bauer J, Curtin JA, Pinkel D, et al. Congenital melanocytic nevi<br />
frequently harbor NRAS mutations but no BRAF mutations. J Invest Dermatol.<br />
2007;127:179-182.<br />
39. Pollock PM, Harper UL, Hansen KS, et al. High frequency <strong>of</strong> BRAF<br />
mutations in nevi. Nat Genet. 2003;33:19-20.
Desmoid-Type Fibromatosis in Children:<br />
A Step Forward in the Cooperative<br />
Group Setting<br />
By Natalie Pounds, MD, and Stephen X. Skapek, MD<br />
Overview: Desmoid-type (aggressive) fibromatosis (desmoid<br />
tumor) is a s<strong>of</strong>t tissue neoplasm that can occur in both<br />
children and adults. Although it is formally classified as an<br />
intermediate-grade neoplasm because <strong>of</strong> its propensity for<br />
locally invasive growth, it can lead to severe and lifethreatening<br />
problems. Because metastases do not arise from<br />
desmoid tumor, therapeutic interventions have historically<br />
focused on surgery or radiation to achieve local tumor control.<br />
These approaches may be ineffective or impractical for some<br />
children. In those cases, systemic therapy with cytotoxic or<br />
noncytotoxic therapy has been used. Because <strong>of</strong> the relative<br />
DESMOID TUMOR, also known as aggressive or<br />
desmoid-type fibromatosis, represents a relatively rare<br />
neoplasm that affects both children and adults. The overall<br />
incidence <strong>of</strong> desmoid tumor is estimated to be two to four<br />
new diagnoses per 1 million people per year. 1 Desmoid<br />
tumor incidence peaks in individuals from 6 to 15 years <strong>of</strong><br />
age and again between puberty and 40 years <strong>of</strong> age in<br />
women. There seems to be a female predominance during<br />
adolescence. 2,3 Mortality from desmoid tumor is rare but has<br />
been reported in children. Nonetheless, the disease poses<br />
substantial problems related to diseases progression and<br />
consequences from therapy.<br />
The most common sites <strong>of</strong> origin <strong>of</strong> desmoid tumor are in<br />
the abdominal wall, intra-abdominal or mesenteric sites, but<br />
extra-abdominal sites include extremities, shoulder girdle,<br />
chest wall, and inguinal region. 1,4 Children tend to develop<br />
desmoid tumors in similar extra-abdominal locations, but<br />
both infants and young children have a higher propensity for<br />
tumors in the head and neck region, 2 a rare disease site for<br />
adults. Patients with intra-abdominal desmoid tumors typically<br />
are asymptomatic or have symptoms associated with<br />
an enlarging mass, including weight loss, cachexia, malaise,<br />
renal failure, and small bowel compression or perforation.<br />
Invasion <strong>of</strong> adjacent muscle, nerve, or vessels can cause<br />
pain, limitation in joint movement, or contractures and<br />
deformities. Desmoid tumor may be multifocal, but bona-fide<br />
metastasis does not occur. 1,4<br />
Desmoid tumor poses a substantial clinical problem because<br />
it carries a propensity to recur after seemingly complete<br />
surgical resection. The reason for the high recurrence<br />
rate is apparent from histologic studies. Desmoid tumor has<br />
a benign histologic appearance, lacking nuclear and cytoplasmic<br />
features <strong>of</strong> a malignant tumor. 1 Unlike most sarcomas,<br />
which are usually separated from adjacent structures<br />
by a pseudocapsule, desmoid tumor <strong>of</strong>ten displays an irregular,<br />
infiltrating border. Several histologic differences have<br />
been noted between desmoid tumors in children and adults.<br />
First, higher mitotic rates are seen in childhood tumors.<br />
Second, lesions tend to be more cellular. 5 Whether rare<br />
mitotic activity contributes to their relative resistance to<br />
cytotoxic therapy is not clear.<br />
Desmoid tumor is typically a sporadic disease with no<br />
identified cause. However, it is clear that there may be an<br />
underlying genetic component in disease pathogenesis. This<br />
rarity <strong>of</strong> this neoplasm in children, knowledge on the use <strong>of</strong><br />
chemotherapy is based largely on anecdotal reports or retrospective<br />
series. Limited conclusions can be drawn, though,<br />
from these types <strong>of</strong> reports. In the last 10 years, two prospective<br />
phase II clinical trials <strong>of</strong> chemotherapy for children with<br />
desmoid tumor have been conducted in cooperative clinical<br />
trials centered in North America. We review the results <strong>of</strong><br />
those clinical trials and suggest future directions for systematically<br />
approaching this disease to better define the role <strong>of</strong><br />
chemotherapy for children with desmoid tumor.<br />
genetic component is most evident in Gardner syndrome, a<br />
well-described variant <strong>of</strong> familial adenomatous polyposis<br />
(FAP), an autosomal-dominant disease characterized by the<br />
presence <strong>of</strong> innumerable adenomatous polyps in the colon<br />
(OMIM #175100). Patients with Gardner syndrome have a<br />
number <strong>of</strong> extracolonic manifestations, including osteoma<br />
and desmoid tumor. Both FAP and Gardner syndrome are<br />
associated with germline mutation in the adenomatous<br />
polyposis coli (APC) gene. 6 The APC protein acts, in part, to<br />
promote degradation <strong>of</strong> �-catenin, an intracellular protein<br />
that aids in the transduction <strong>of</strong> cell proliferation signals to<br />
the nucleus. A number <strong>of</strong> studies during the past decade<br />
have identified somatic mutations in APC and CTNNB1, the<br />
gene encoding �-catenin, in sporadic desmoid tumor cases 7 ;<br />
both mutations ultimately enhance �-catenin activity. Despite<br />
an association with Gardner syndome, heritable predisposition<br />
is clearly the less common form <strong>of</strong> desmoid tumor.<br />
Other causative factors include trauma and estrogens.<br />
The former is typically surgical in nature, and it may be<br />
most relevant to FAP patients. The high frequency <strong>of</strong> abdominal<br />
disease in this population seems to correlate with<br />
mesenteric fibromatosis, a putative precursor lesion the<br />
presence <strong>of</strong> which correlates with increased numbers <strong>of</strong><br />
abdominal operations. 8 The role <strong>of</strong> estrogen signaling was<br />
originally suggested by the increased incidence during pregnancy<br />
and enhanced tumor growth rate in pregnant women<br />
when compared with that in men or premenopausal or<br />
postmenopausal women. 4 This association is supported by<br />
evidence <strong>of</strong> strong estrogen receptor expression in more than<br />
80% <strong>of</strong> desmoid tumors. 9<br />
Surgery and Radiation as the Historical Standard<br />
<strong>of</strong> Care<br />
Although we make a case that optimal therapy for desmoid<br />
tumor in a child begins with a multidisciplinary<br />
From the University <strong>of</strong> Texas Southwestern Medical Center and Center for Cancer and<br />
Blood Disorders, Children’s Medical Center, Dallas, TX.<br />
Authors’ disclosures <strong>of</strong> potential conflicts <strong>of</strong> interest are found at the end <strong>of</strong> this article.<br />
Address reprint requests to Stephen X. Skapek, MD, University <strong>of</strong> Texas Southwestern<br />
Medical Center, 5323 Harry Hines Blvd, Mail Code 9063, Dallas, TX-9063; email: Stephen.<br />
Skapek@utsouthwestern.edu.<br />
© <strong>2012</strong> by <strong>American</strong> <strong>Society</strong> <strong>of</strong> <strong>Clinical</strong> <strong>Oncology</strong>.<br />
1092-9118/10/1-10<br />
593
evaluation, because desmoid tumor is a largely localized<br />
disease, treatment approaches have historically focused on<br />
focal therapies, such as surgery or external beam ionizing<br />
radiation.<br />
Most would agree that surgical resection represents the<br />
best therapy for children in whom the tumor can be completely<br />
removed without substantial functional or cosmetic<br />
consequences. The ability to obtain a complete surgical<br />
resection is generally believed to be an important factor that<br />
influences tumor control. However, other factors, such as<br />
whether the disease is primary or recurrent 10 and the<br />
presence <strong>of</strong> certain CTNNB1 mutations, 7 also seem to be<br />
important. Numerous issues regarding ongoing physical and<br />
emotional development in children add to the complexity<br />
regarding long-term surgical outcomes in children with<br />
desmoid tumor; and this has not been well studied in this<br />
patient population.<br />
Ionizing radiation therapy has been shown to be effective<br />
in many studies <strong>of</strong> adult patients. In particular, when<br />
applied in conjunction with surgery, ionizing radiation<br />
therapy improves local disease control. 11 The high radiation<br />
dose needed makes its use more complicated if the<br />
radiation port includes growing bones or joints. Furthermore,<br />
radiation therapy may be less effective in children<br />
than adults. This concern led to a retrospective review,<br />
which concluded that the role <strong>of</strong> radiation in the initial<br />
treatment <strong>of</strong> children with desmoid tumors should be reconsidered.<br />
12 In this series, five <strong>of</strong> 13 children received radiation<br />
therapy immediately after diagnosis, whereas the<br />
remaining eight patients received radiation therapy after<br />
recurrence. The median dose <strong>of</strong> radiation used was 50 Gy.<br />
Tumor recurred after radiation therapy in 11 <strong>of</strong> the 13<br />
patients, and three patients died <strong>of</strong> their disease. In those<br />
surviving, substantial morbidity associated with the radia-<br />
KEY POINTS<br />
● Desmoid-type fibromatosis (desmoid tumor) represents<br />
a locally aggressive s<strong>of</strong>t tissue neoplasm that<br />
has a propensity for locally aggressive growth and<br />
recurrence after attempted surgical resection.<br />
● Desmoid tumor is associated with loss <strong>of</strong> function<br />
mutations in the adenomatous polyposis coli tumor<br />
suppressor and gain <strong>of</strong> function mutations in<br />
�-catenin.<br />
● Surgical resection with or without radiation therapy<br />
has represented the historical standard for treating<br />
this disease.<br />
● A variety <strong>of</strong> cytotoxic and noncytotoxic chemotherapies<br />
have demonstrated the capacity to control desmoid<br />
tumor when surgery or radiation are not<br />
effective; however, most <strong>of</strong> this literature is in the<br />
form <strong>of</strong> case reports or retrospective studies, <strong>of</strong>ten<br />
including children and adults.<br />
● Two multicenter, prospective clinical trials conducted<br />
through the Pediatric <strong>Oncology</strong> Group and Children’s<br />
<strong>Oncology</strong> Group have demonstrated the feasibility <strong>of</strong><br />
this approach and the added value stemming from<br />
systematic, prospective study.<br />
594<br />
POUNDS AND SKAPEK<br />
Table 1. Event-Free Survival for Children Enrolled in POG 9650<br />
or COG ARST0321 and Treated With Vinblastine/Methotrexate or<br />
Tamoxifen/Sulindac, Respectively<br />
Chemotherapy Event-Free Survival at 1 Year (Range)<br />
Vinblastine/methotrexate 0.58 (0.29–0.57)<br />
Tamoxifen/sulindac 0.44<br />
Abbreviations: COG, Children’s <strong>Oncology</strong> Group; POG, Pediatric <strong>Oncology</strong><br />
Group.<br />
tion therapy was noted. A more recent study, though,<br />
supports the role that radiation therapy can play in this<br />
disease in children. 13<br />
Like surgery, radiation therapy holds the potential for<br />
consequences that can be particularly troubling in a growing<br />
child. Reported adverse effects include bone fractures, skeletal<br />
and s<strong>of</strong>t tissue growth retardation, tissue fibrosis, and<br />
lymphedema. 12 Jabbari and colleagues 13 also reported substantial<br />
morbidity in the form <strong>of</strong> peripheral neuropathy,<br />
pain, bowel obstruction, and the development <strong>of</strong> papillary<br />
cancer.<br />
Chemotherapy for Children with Desmoid Tumor<br />
For many children with desmoid tumor, surgery or radiation<br />
has proven ineffective or is believed to be associated<br />
with unacceptable consequences, which has led to the use <strong>of</strong><br />
chemotherapeutic agents that are believed to act by cytotoxic<br />
or noncytotoxic mechanisms. Regrettably, most <strong>of</strong> the<br />
reports stemming from their use represent retrospective,<br />
single-institution analyses <strong>of</strong> small groups <strong>of</strong> patients,<br />
thereby limiting the conclusions that one can draw.<br />
Cytotoxic regimens demonstrated to have some activity in<br />
retrospective analyses include liposomal doxorubicin 14 ;<br />
doxorubicin with dacarbazine 15 ; vincristine, dactinomycin,<br />
and cyclophosphamide 16 ; hydroxyurea 17 ; and vinblastine<br />
and methotrexate. 18 Perhaps the most striking report centered<br />
on the use <strong>of</strong> doxorubicin and dacarbazine by 96-hour<br />
infusion and followed by the nonsteroidal anti-inflammatory<br />
drug (NSAID) meloxicam. 19 The authors reported complete<br />
and partial responses in three and four, respectively, <strong>of</strong> 11<br />
adult patients, with a mean progression-free survival <strong>of</strong><br />
approximately 6 years. Beyond the biases inherent in retrospective<br />
analyses, the complex nature <strong>of</strong> desmoid tumor,<br />
which has been reported to undergo prolonged stabilization<br />
and even spontaneous regression, 20 limits the conclusions<br />
one can draw on the relative efficacy <strong>of</strong> any <strong>of</strong> these regimens.<br />
The same is true <strong>of</strong> reports <strong>of</strong> noncytotoxic regimens, most<br />
<strong>of</strong> which include NSAIDs, such as sulindac and meloxicam,<br />
and estrogen antagonists, such as tamoxifen. In some cases,<br />
tamoxifen 21 or NSAIDs 22 are used alone. In most, though,<br />
tamoxifen is combined with an NSAID, such as sulindac. 23<br />
In one series, 10 <strong>of</strong> 13 adults with FAP-associated disease<br />
had either a partial or complete response. 23 Although the<br />
initial rationale for using NSAIDs may have been based on<br />
incomplete understanding <strong>of</strong> the disease biology, more recent<br />
laboratory studies provide a potential mechanism: the<br />
transcription factor PPAR� is deregulated in the setting <strong>of</strong><br />
APC mutations, but NSAIDs can block PPAR� activity. 24<br />
Receptor tyrosine kinase inhibitors, particularly imatinib<br />
mesylate, also have shown the capacity to stabilize desmoid<br />
tumor and foster regression in a smaller subset. 25 Its activity<br />
may be based on expression <strong>of</strong> platelet-derived growth<br />
factor receptor �. 25
CHEMOTHERAPY FOR DESMOID TUMOR<br />
Table 2. Complete and Partial Responses for Children Enrolled<br />
in POG 9650 or COG ARST0321 and Treated With Vinblastine/<br />
Methotrexate or Tamoxifen/Sulindac, Respectively<br />
Chemotherapy<br />
Chemotherapy Lessons from Prospective, Cooperative<br />
Group Trials<br />
The reports discussed in this article are not fully satisfying,<br />
given the long-standing concept that desmoid tumor<br />
may remain stable or undergo spontaneous regression. 4<br />
Indeed, a recent retrospective report suggests that<br />
progression-free survival is the same in patients treated<br />
with medical therapy or a “wait and see” approach; nearly<br />
50% had no disease progression at 5 years. 26 The variable<br />
natural history <strong>of</strong> desmoid tumor spurred members <strong>of</strong> the<br />
former Pediatric <strong>Oncology</strong> Group (POG) to prospectively<br />
study desmoid tumor in the cooperative group setting.<br />
Vinblastine and Methotrexate<br />
Complete and Partial<br />
Responses, No. Total, No. CR/PR Rate, %<br />
Vinblastine/methotrexate 5 26 19<br />
Tamoxifen/sulindac 5 60 8<br />
Abbreviations: COG, Children’s <strong>Oncology</strong> Group; POG, Pediatric <strong>Oncology</strong><br />
Group; CR, complete response; PR, partial response.<br />
At the time this study was conceived, Weiss and Lackman<br />
18 had published a retrospective study on the use <strong>of</strong><br />
vinblastine and methotrexate in adult patients with recur-<br />
Fig 1. Potential strategy for risk-based treatment <strong>of</strong><br />
desmoid tumor in children (A) and how this type <strong>of</strong><br />
approach might be compared with chemotherapy in a<br />
clinical trial (B).<br />
Abbreviations: SD, Stable Disease; R, Response; PD,<br />
Progressive Disease.<br />
rent desmoid tumors. In this report <strong>of</strong> eight adult patients,<br />
two had complete and four had partial responses. A similar<br />
retrospective study <strong>of</strong> these two agents in 10 children<br />
showed a 50% response rate and 70% disease stabilization. 27<br />
Vinblastine and methotrexate were particularly appealing<br />
for prospective study in children because <strong>of</strong> a favorable<br />
late effects pr<strong>of</strong>ile when compared with doxorubin- and<br />
cyclophosphamide-based regimens.<br />
The POG 9650 study represented the first prospective,<br />
multi-institutional trial in children. 28 The primary goal was<br />
to estimate the safety and efficacy <strong>of</strong> vinblastine and<br />
methotrexate in patients younger than 19 years with<br />
desmoid-type fibromatosis that was recurrent or not amenable<br />
to surgery or radiation. In this single-arm, phase II<br />
study, both drugs were given weekly for 26 weeks and then<br />
every other week for an additional 26 weeks, at which point<br />
therapy was terminated. Response was assessed by bidimensional<br />
measurements on axial imaging using either<br />
computed tomography or magnetic resonance imaging,<br />
and responses were prospectively defined in the study.<br />
Of 26 patients, measurable response was evident in eight<br />
patients (one with complete response, four with partial<br />
response, and three with minor response), whereas an<br />
additional 10 patients had stable disease as the best response<br />
(Table 1). The 1-year event-free survival rate was<br />
58%, and the median time to progression was 15.9 months<br />
after therapy was discontinued (Table 2). Seven patients<br />
595
(27% <strong>of</strong> total) were free <strong>of</strong> disease progression 50 months<br />
after enrollment.<br />
This regimen was chosen because <strong>of</strong> the anticipated low<br />
toxicity rate. In fact, 18 patients (67%) reported National<br />
Cancer Institute (NCI) grade 2 or higher toxicity; neutropenia<br />
represented the most common grade 3 or 4 toxicity.<br />
28 Other toxicities included anemia, nausea, vomiting,<br />
and elevation <strong>of</strong> hepatic transaminases; all toxicities<br />
were reversed on ceasing chemotherapy. These results suggested<br />
that vinblastine and methotrexate administration<br />
arrested tumor progression or promoted frank regression<br />
in most children, but the treatment was associated with<br />
substantial myelosuppression. Perhaps most importantly,<br />
this study set a precedent for prospectively studying a<br />
relatively rare neoplasm in children in the cooperative group<br />
setting.<br />
Sulindac and High-Dose Tamoxifen<br />
While the results <strong>of</strong> POG 9650 were maturing, the Children’s<br />
<strong>Oncology</strong> Group (COG) supported a follow-up, singlearm,<br />
phase II trial <strong>of</strong> sulindac and high-dose tamoxifen<br />
based on the previously mentioned retrospective studies and<br />
case reports. Like the POG 9650 study, the primary aim <strong>of</strong><br />
the ARST0321 was to estimate the safety and efficacy <strong>of</strong><br />
sulindac and tamoxifen in the same patient population and<br />
using the same response criteria. 29 Secondary objectives<br />
investigated whether magnetic resonance imaging signal<br />
features correlate with response or nonresponse, as previously<br />
suggested, 27 and whether hormone receptor status<br />
influences the outcome. Patients received sulindac and tamoxifen<br />
in combination for 12 4-week cycles or until disease<br />
progression. After a complete response, patients were to<br />
receive an additional 1 month <strong>of</strong> the study drugs.<br />
A total <strong>of</strong> 70 patients were enrolled in the study, with an<br />
annual accrual rate <strong>of</strong> 13.3 patients per year. The 61 eligible<br />
patients included 22 with newly diagnosed disease and 39<br />
with recurrent disease. A response (complete response or<br />
partial response) was observed in five patients (8%), and<br />
Authors’ Disclosures <strong>of</strong> Potential Conflicts <strong>of</strong> Interest<br />
Author<br />
Natalie Pounds*<br />
Stephen X. Skapek*<br />
*No relevant relationships to disclose.<br />
Employment or<br />
Leadership<br />
Positions<br />
Consultant or<br />
Advisory Role<br />
1. Fletcher CDM, Unni KK, Mertens F (eds). World Health Organization<br />
Classification <strong>of</strong> Tumours. Pathology and Genetics <strong>of</strong> S<strong>of</strong>t Tissue and Bone.<br />
Lyons, France: IARC Press; 2002.<br />
2. Buitendijk S, van de Ven CP, Dumans TG, et al. Pediatric aggressive<br />
fibromatosis: a retrospective analysis <strong>of</strong> 13 patients and review <strong>of</strong> literature.<br />
Cancer. 2005;104:1090-1099.<br />
3. Faulkner LB, Hajdu SI, Kher U, et al. Pediatric desmoid tumor:<br />
retrospective analysis <strong>of</strong> 63 cases. J Clin Oncol. 1995;13:2813-2818.<br />
4. Häyry P, Scheinin TM. The desmoid (Reitamo) syndrome: etiology, manifestations,<br />
pathogenesis, and treatment. Curr Probl Surg. 1988;25:233-320.<br />
5. Ayala AG, Ro JY, Goepfert H, et al. Desmoid fibromatosis: a clinicopathologic<br />
study <strong>of</strong> 25 children. Semin Diagn Pathol. 1986;3:138-150.<br />
6. Rustgi AK. The genetics <strong>of</strong> hereditary colon cancer. Genes Dev. 2007;21:<br />
2525-2538.<br />
7. Lazar AJ, Tuvin D, Hajibashi S, et al. Specific mutations in the<br />
beta-catenin gene (CTNNB1) correlate with local recurrence in sporadic<br />
desmoid tumors. Am J Pathol. 2008;173:1518-1527.<br />
596<br />
progression-free survival at 1 year was 44% (Table 1 and<br />
Table 2). In contrast to vinblastine/methotrexate, myelosuppression<br />
was rarely observed; and patients rarely experienced<br />
NCI grade 3 or 4 toxicities. One notable exception was<br />
that 11 <strong>of</strong> the 30 eligible female patients (27%) had large<br />
ovarian cysts documented by ultrasonography; although not<br />
believed to be a medically severe consequence, the ovarian<br />
cysts interfered with treatment in most <strong>of</strong> the cases.<br />
Conclusion<br />
Although these two successive cooperative group studies<br />
have more accurately defined radiographic response rates<br />
and progression-free survival for children treated with<br />
vinblastine/methotrexate and tamoxifen/sulindac, the trials<br />
are not without their own pitfalls. Because these studies<br />
were single arm, phase II studies, we cannot confidently<br />
state whether one regimen is better than the other, even<br />
though the eligibility criteria and response definitions were<br />
the same. We must also be cautious about attributing a<br />
cause-effect relationship to either treatment regimen. Such<br />
caution is particularly important when one considers reports<br />
that desmoid tumor can remain stable for extended periods<br />
with only a “wait and see” approach. 20,26<br />
These reports have led to the concept that asymptomatic<br />
patients should initially by “treated” with a period <strong>of</strong><br />
observation. However, such an approach, illustrated in<br />
Fig. 1A, is unproven in children. Conceivably, one could<br />
prospectively test this concept while also designing a<br />
study to more definitively prove that chemotherapy plays a<br />
role in disease control. Options to consider might include<br />
studies randomizing two regimens or a study design in<br />
which an observation period is incorporated as a “window”<br />
for those with asymptomatic and non–life-threatening disease<br />
(Fig. 1B). Building on the POG and COG success<br />
performing prospective trials for this disease, we are optimistic<br />
that we can take the needed steps to better define the<br />
natural history and role <strong>of</strong> chemotherapeutics for children<br />
with desmoid tumor.<br />
Stock<br />
Ownership Honoraria<br />
REFERENCES<br />
Research<br />
Funding<br />
Expert<br />
Testimony<br />
POUNDS AND SKAPEK<br />
Other<br />
Remuneration<br />
8. Clark SK, Smith TG, Katz DE, et al. Identification and progression <strong>of</strong> a<br />
desmoid precursor lesion in patients with familial adenomatous polyposis.<br />
Br J Surg. 1998;85:970-973.<br />
9. Deyrup AT, Tretiakova M, Montag AG. Estrogen receptor-beta expression<br />
in extraabdominal fibromatoses: an analysis <strong>of</strong> 40 cases. Cancer. 2006;<br />
106:208-213.<br />
10. Gronchi A, Casali PG, Mariani L, et al. Quality <strong>of</strong> surgery and outcome<br />
in extra-abdominal aggressive fibromatosis: a series <strong>of</strong> patients surgically<br />
treated at a single institution. J Clin Oncol. 2003;21:1390-1397.<br />
11. Pritchard DJ, Nascimento AG, Petersen IA. Local control <strong>of</strong> extraabdominal<br />
desmoid tumors. J Bone Joint Surg Am. 1996;78:848-854.<br />
12. Merchant TE, Nguyen D, Walter AW, et al. Long-term results with<br />
radiation therapy for pediatric desmoid tumors. Int J Radiation <strong>Oncology</strong> Biol<br />
Phys. 2000;47:1267-1271.<br />
13. Jabbari S, Andolino D, Weinberg V, et al. Successful treatment <strong>of</strong> high<br />
risk and recurrent pediatric desmoids using radiation as a component <strong>of</strong><br />
multimodality therapy. Int J Radiat Oncol Biol Phys. 2009;75:177-182.
CHEMOTHERAPY FOR DESMOID TUMOR<br />
14. Constantinidou A, Jones RL, Scurr M, et al. Pegylated liposomal<br />
doxorubicin, an effective, well-tolerated treatment for refractory aggressive<br />
fibromatosis. Eur J Cancer. 2009;45:2930-2934.<br />
15. Patel SR, Evans HL, Benjamin RS. Combination chemotherapy in<br />
adult desmoid tumors. Cancer. 1993;72:3244-3247.<br />
16. Raney B, Evans A, Granowetter L, et al. Nonsurgical management <strong>of</strong><br />
children with recurrent or unresectable fibromatosis. Pediatrics. 1987;79:394-<br />
398.<br />
17. Meazza C, Casanova M, Trecate G, et al. Objective response to hydroxyurea<br />
in a patient with heavily pre-treated aggressive fibromatosis.<br />
Pediatr Blood Cancer. 2010;55:587-588.<br />
18. Weiss AJ, Lackman RD. Low-dose chemotherapy <strong>of</strong> desmoid tumors.<br />
Cancer. 1989;64:1192-1194.<br />
19. Gega M, Yanagi H, Yoshikawa R, et al. Successful chemotherapeutic<br />
modality <strong>of</strong> doxorubicin plus dacarbazine for the treatment <strong>of</strong> desmoid tumors<br />
in association with familial adenomatous polyposis. J Clin Oncol. 2006;24:<br />
102-105.<br />
20. Lewis JJ, Boland PJ, Leung DH, et al. The enigma <strong>of</strong> desmoid tumors.<br />
Ann Surg. 1999;229:866-873.<br />
21. Sportiello DJ, Hoogerland DL. A recurrent pelvic desmoid tumor<br />
successfully treated with tamoxifen. Cancer. 1991;67:1443-1446.<br />
22. Belliveau P, Graham AM. Mesenteric desmoid tumor in Gardner’s<br />
syndrome treated by sulindac. Dis Colon Rectum. 1984;27:53-54.<br />
23. Hansmann A, Adolph C, Vogel T, et al. High-dose tamoxifen and<br />
sulindac as first-line treatment for desmoid tumors. Cancer. 2003;100:612-<br />
620.<br />
24. He TC, Chan TA, Vogelstein B, et al. PPARdelta is an APC-regulated<br />
target <strong>of</strong> nonsteroidal anti-inflammatory drugs. Cell. 1999;99:335-345.<br />
25. Heinrich MC, McArthur GA, Demetri GD, et al. <strong>Clinical</strong> and molecular<br />
studies <strong>of</strong> the effect <strong>of</strong> imatinib on advanced aggressive fibromatosis (desmoid<br />
tumor). J Clin Oncol. 2006;24:1195-1203.<br />
26. Fiore M, Rimareix F, Mariani L, et al. Desmoid-type fibromatosis: a<br />
front-line conservative approach to select patients for surgical treatment. Ann<br />
Surg Oncol. 2009;16:2587-2593.<br />
27. Skapek SX, Hawk BJ, H<strong>of</strong>fer FA, et al. Combination chemotherapy<br />
using vinblastine and methotrexate for the treatment <strong>of</strong> progressive desmoid<br />
tumor in children. J Clin Oncol. 1998;16:3021-3027.<br />
28. Skapek SX, Ferguson WS, Granowetter L, et al. Vinblastine and<br />
methotrexate for desmoid fibromatosis in children: results <strong>of</strong> a Pediatric<br />
<strong>Oncology</strong> Group phase II trial. J Clin Oncol. 2007;25:501-506.<br />
29. Skapek SX, Anderson JR, Hill DA, et al. The safety and efficacy <strong>of</strong><br />
high-dose tamoxifen and sulindac for desmoid-type aggressive fibromatosis in<br />
children: results <strong>of</strong> a Children’s <strong>Oncology</strong> Group Phase II study. Paper<br />
presented at: Annual Meeting <strong>of</strong> the Connective Tissue <strong>Oncology</strong> <strong>Society</strong>;<br />
Chicago, IL; October 27, 2011.<br />
597
MOLECULAR PATHWAYS FOR THE PRACTICING<br />
PEDIATRIC ONCOLOGIST<br />
CHAIR<br />
Donald W. Parsons, MD, PhD<br />
Texas Children’s Hospital<br />
Houston, TX<br />
SPEAKERS<br />
Katia Scotlandi, PhD<br />
Istituto Ortopedico Rizzoli<br />
Bologna, Italy<br />
Tobey MacDonald, MD<br />
Emory University<br />
Atlanta, GA
Targeting the Insulin-Like Growth Factor<br />
(IGF) System Is Not as Simple as Just<br />
Targeting the Type 1 IGF Receptor<br />
By Katia Scotlandi, PhD, and Antonino Belfiore, MD<br />
Overview: Increased signaling <strong>of</strong> the insulin-like growth factor<br />
(IGF) system via alterations in expression levels <strong>of</strong> its<br />
components has been demonstrated in various tumor types.<br />
Numerous experimental studies have supported the involvement<br />
<strong>of</strong> the IGF system signaling axis in tumor initiation and<br />
progression. These studies, combined with data that link<br />
alterations in the levels <strong>of</strong> circulating IGFs with cancer risk<br />
and prognosis, have focused on the IGF-1 receptor (IGF-1R) as<br />
a therapeutic target for patients with cancer. As a consequence,<br />
most therapeutic strategies have been designed to<br />
specifically inhibit IGF-1R but have for the most part ignored<br />
the insulin receptor (IR), based on concerns that targeting IR<br />
would lead to unacceptable toxicity both because <strong>of</strong> its role in<br />
THE IGF system comprises a phylo-genetically ancient<br />
family <strong>of</strong> peptides involved in growth, development,<br />
and metabolism, as well as in cellular processes such as<br />
proliferation, survival, cell migration, and differentiation.<br />
The IGF system is composed <strong>of</strong> three ligands (IGF-1, IGF-2,<br />
and insulin), their receptors (the IGF-1 receptor [IGF-1R],<br />
the mannose 6-phosphate/IGF-2 receptor [M6P/IGF-2R], the<br />
insulin receptor [IR], and the hybrid IR/IGF-1R), at least six<br />
high-affinity binding proteins and binding protein proteases.<br />
IGF binding proteins modulate the activity <strong>of</strong> IGFs but also<br />
have a life <strong>of</strong> their own inducing cellular processes in an<br />
IGF-independent way. Molecular details <strong>of</strong> the IGF system<br />
have been excellently reviewed by Samani and colleagues. 1<br />
In this context, it is important to highlight the complexity<br />
<strong>of</strong> the system and the presence <strong>of</strong> several critical nodes that<br />
within the signaling networks control various cellular processes.<br />
A simple scheme <strong>of</strong> divergent pathways is usually<br />
sufficient to describe and explain IGF/insulin signaling (Fig.<br />
1). However, when examined in detail, the number <strong>of</strong> genes<br />
and protein is<strong>of</strong>orms involved in the activation <strong>of</strong> mitogenactivated<br />
protein kinase (MAPK) or AKT signaling pathways,<br />
the two main signaling mediators <strong>of</strong> the IGF system or<br />
in genes involved in the generation <strong>of</strong> proliferative, antiapoptotic,<br />
differentiating, or metabolic effects, it becomes<br />
clear that hundreds <strong>of</strong> molecules are involved in the IGF/<br />
insulin-signaling pathway. It is beyond the scope <strong>of</strong> this<br />
manuscript to describe in detail this molecular level <strong>of</strong><br />
complexity and interactions 2 but, to make the reader more<br />
aware <strong>of</strong> the peculiarities <strong>of</strong> this signaling axis, some examples<br />
<strong>of</strong> at least the best-defined critical nodes are described.<br />
For instance, the IR has two splice is<strong>of</strong>orms, IR-A that is<br />
highly expressed in fetal tissues and cancer, and IR-B that is<br />
mainly found in adult tissue (details to follow). Both are<br />
usually coexpressed in cells that also express IGF-1R. Insulin<br />
and IGFs bind with high affinity to their cognate receptors<br />
(e.g., insulin 3 IR; IGFs 3 IGF-1R), whereas IGF-2R<br />
serves mainly as a sink for the regulation <strong>of</strong> IGF2 levels.<br />
However, IGF-2 also binds IR-A with high affinity, 3 although<br />
at lower affinity, IGF-1R can also be activated by<br />
insulin, and IR can be activated by IGFs. This implies that<br />
whenever we study the effects <strong>of</strong> IGF-1R, IR, or both, we<br />
physiologic metabolism and because we frequently try to<br />
oversimplify biologic complexity whenever we are urged to<br />
find practical, friendly solutions for clinical practice. Although<br />
this is an understandable and necessary starting point in the<br />
complex and long-lasting processes that leads to translational<br />
biology, the crude reality <strong>of</strong> the results obtained from phase I<br />
and II studies suggest a need for researchers to be humble<br />
and go back to the drawing board. Cancer research has<br />
substantially neglected the role <strong>of</strong> IR, and it remains unclear<br />
whether and to what extent avoiding the inhibition <strong>of</strong> IR has<br />
compromised the efficacy <strong>of</strong> anti–IGF-1R therapy. Clarifying<br />
its role might also help us take advantage <strong>of</strong> older drugs that<br />
could <strong>of</strong>fer new perspectives in cancer care.<br />
should also pay attention to the most prevalent types and<br />
expression levels <strong>of</strong> the ligand(s) in that specific cellular<br />
context.<br />
Similar to IR, IGF-1R consists <strong>of</strong> two extracellular ligandbinding<br />
subunits (the alpha subunits) and <strong>of</strong> two transmembrane<br />
beta subunits, which are linked to alpha subunits by<br />
disulfide bonds and are composed <strong>of</strong> a transmembrane<br />
domain, an intracellular tyrosine kinase (TK) domain, and a<br />
C-terminal tail. IGF-1R has 70% homology to IR, with which<br />
it shares some signaling pathways. As a consequence <strong>of</strong> the<br />
close homology <strong>of</strong> IR and IGF-1R, hybrid receptors can be<br />
formed by an insulin alpha-beta hemireceptor and an IGF-1<br />
alpha-beta hemireceptor in cells expressing both. The biologic<br />
response elicited by these hybrid receptors can vary,<br />
depending on the ligands involved and the specific IR is<strong>of</strong>orms.<br />
4 These hybrid receptors appear to bind IGF-1 and<br />
IGF-2 with high affinity similar to IGF-1R. Ligand binding<br />
induces tyrosines within the TK domain to be transphosphorylated<br />
by the dimeric subunit partner. Phosphorylated<br />
residues serve as docking sites for other signaling molecules,<br />
such as IR substrates (IRS) and the adaptor protein Shc,<br />
which leads to the activation <strong>of</strong> the phosphatidylinositol-3<br />
kinase (PI3K)-Akt and MAPK pathways (Fig. 1). However,<br />
both IR and IGF-1R can phosphorylate at least six known<br />
substrate proteins (IRS1-6) that are capable <strong>of</strong> interacting<br />
with eight known forms <strong>of</strong> the PI3K regulatory subunit,<br />
which leads to the activation <strong>of</strong> three known is<strong>of</strong>orms <strong>of</strong> AKT<br />
signaling, besides being able to crosstalk with the MAPK<br />
signaling pathway. Moreover, these signaling pathways are<br />
shared with most other TK receptors and it is possible that<br />
other pathways that have yet to be identified are involved or<br />
From the CRS Development <strong>of</strong> Biomolecular Therapies, Experimental <strong>Oncology</strong> Lab,<br />
Orthopaedic Rizzoli Institute, Bologna, Italy; Department <strong>of</strong> Endocrinology, Department <strong>of</strong><br />
Health University Magna Graecia <strong>of</strong> Catanzaro, Catanzaro, Italy.<br />
Authors’ disclosures <strong>of</strong> potential conflicts <strong>of</strong> interest are found at the end <strong>of</strong> this article.<br />
Address reprint requests to Katia Scotlandi, PhD, Orthopaedic Rizzoli Institute, CRS<br />
Development <strong>of</strong> Biomolecular Therapies, Experimental <strong>Oncology</strong> Lab, Via di Barbiano<br />
1/10, 40126 Bologna, Italy; email: katia.scotlandi@ior.it.<br />
© <strong>2012</strong> by <strong>American</strong> <strong>Society</strong> <strong>of</strong> <strong>Clinical</strong> <strong>Oncology</strong>.<br />
1092-9118/10/1-10<br />
599
that subtle differences in the recruitment <strong>of</strong> certain docking<br />
proteins and intracellular mediators exist about which we<br />
still know little regarding their dynamics and biologic effects.<br />
Finally, in addition to the well-established signaling<br />
pathway from the cell membrane, recent reports have highlighted<br />
how IGF-1R or IR signaling, or both, may also be<br />
dependent on cell localization. Differential endocytosis and<br />
signaling dynamics have been reported for IGF-1R 5 as well<br />
as for IR-A and IR-B in relation to the mitogenic or metabolic<br />
activities <strong>of</strong> the two receptors. 6 Nuclear translocation <strong>of</strong><br />
IGF-1R has also been reported. 7 The modification <strong>of</strong> IGF-1R<br />
by small ubiquitin-like modifier (SUMO) proteins occurs in a<br />
ligand-dependent manner and is necessary for nuclear<br />
translocation <strong>of</strong> the receptor. As expected, nuclear IGF-1R<br />
has different biologic functions: it binds to genomic DNA and<br />
may act as a transcriptional enhancer.<br />
This level <strong>of</strong> complexity has been grossly ignored when<br />
targeted therapies against IGF-1R were developed. However,<br />
despite the difficulties <strong>of</strong> deeply understanding such<br />
complex signaling, steps have been recently taken and<br />
indeed some achievements have already been obtained.<br />
Why Was IGF-1R Chosen as a Therapeutic Target?<br />
The first inkling that IGF-1R played a crucial role in<br />
malignant transformation was provided by Sell and colleagues<br />
(1993), who found that R cells could not be transformed<br />
by the SV40 large T antigen. R cells are 3T3 cells<br />
originating from mouse embryos with a targeted disruption<br />
<strong>of</strong> the IGF-1R gene. These cells have a tendency to transform<br />
spontaneously in culture, and the SV40 T antigen is, by<br />
itself, a strong transforming agent in 3T3c cells. The failure<br />
<strong>of</strong> R cells to become transformed by SV40 T antigen indicated<br />
a role <strong>of</strong> the IGF-1R in transformation <strong>of</strong> cells in<br />
culture. This finding has since been confirmed with different<br />
viral and cellular oncogenes and in different laboratories. 8<br />
Reintroduction <strong>of</strong> an IGF-1R into R cells promptly renders<br />
KEY POINTS<br />
● The insulin-like growth factor (IGF) system is an<br />
important mediator <strong>of</strong> cancer pathogenesis and progression.<br />
Drug resistance to conventional or targeted<br />
therapies frequently involves components <strong>of</strong> the<br />
insulin-like growth factor system.<br />
● Researchers and pharmaceutical companies have<br />
focused on IGF-1 receptor (IGF-1R), which is clearly<br />
able to deliver a proliferative, antiapoptotic, and<br />
promigratory signal in cancer cells.<br />
● Several approaches inhibiting IGF-1R functions have<br />
shown very encouraging results in preclinical conditions,<br />
but only limited evidence <strong>of</strong> efficacy has been<br />
demonstrated in phase I and II clinical studies.<br />
● The IGF system is quite complex, with many players<br />
in the field. Insulin receptor function in cancer cells<br />
has been underestimated, but also little attention has<br />
been paid to the type <strong>of</strong> ligands that are mainly<br />
involved in each tumor type.<br />
● Strategies considering the IGF system in all its complexity<br />
are encouraged.<br />
600<br />
SCOTLANDI AND BELFIORE<br />
these cells susceptible to transformation. Thus, there should<br />
be a signal originating from the IGF-1R that facilitates and<br />
is quasi-necessary for the transformation by the usual<br />
agents (i.e., physical, chemical, and/or genetic). The level <strong>of</strong><br />
expression <strong>of</strong> IGF-1R does not need to be high. Even low<br />
levels <strong>of</strong> expression are sufficient to send the permissive<br />
signal that allows oncogenes to transform mammalian cells.<br />
Accordingly, IGF-1R is overexpressed in some malignant<br />
tissues but amplification and overexpression are less common<br />
for IGF-1R than for other oncogenetic receptors. Similarly,<br />
mutations have not been described as a way to<br />
increase receptor activity. Activation <strong>of</strong> IGF-1R is indeed<br />
mainly induced by either circulating or locally synthesized<br />
IGFs in an autocrine or paracrine manner. 9 Although some<br />
studies report no relationship between IGF-1 levels and<br />
cancer risk, many others report that individuals with IGF-1<br />
levels at the upper end <strong>of</strong> the normal range have an<br />
increased risk <strong>of</strong> developing certain cancers (e.g., colon,<br />
breast, and prostate). 10 Conversely, individuals with growth<br />
hormone receptor deficiency, also known as Laron syndrome,<br />
who have very low IGF-1 levels, appear to be protected<br />
from the development <strong>of</strong> cancer when compared with<br />
their relatives without hormonal deficiency. 11,12 Dietary<br />
factors and lifestyle have also been shown to have a substantial<br />
effect on the activation <strong>of</strong> the IGF system. In animal<br />
models, caloric restriction reduced circulating IGF-1 levels<br />
as well as bladder tumor growth, by increasing apoptosis<br />
and decreasing cell proliferation. 13<br />
In any case, if IGF-1R is quasi-obligatory for cell transformation,<br />
downregulation <strong>of</strong> IGF-1R in malignant cells ought<br />
to reverse the transformed phenotype. Downregulation or<br />
inhibition <strong>of</strong> IGF-1R functions, by neutralizing antibodies or<br />
small-molecule TK inhibitors (TKIs), by antisense strategies,<br />
or by developing agents to modulate IGF binding<br />
proteins, causes massive apoptosis <strong>of</strong> tumor cells in vitro<br />
and in vivo and results in the inhibition <strong>of</strong> tumorigenesis<br />
and metastasis formation. Preclinical data suggest that<br />
agents used to target the IGF system may be more effective<br />
when used in combination with chemotherapy compared<br />
with when used as monotherapy. In addition, the induction<br />
<strong>of</strong> IGF-1R signaling has been described to be involved in<br />
mediating resistance to both conventional and some targeted<br />
drugs. 14-16 It is therefore not surprising that targeting<br />
IGF-1R has become popular with pharmaceutical and biotechnology<br />
companies. Currently, most therapeutic agents,<br />
monoclonal antibodies (MAbs), or TKIs have been designed<br />
to specifically target IGF-1R while sparing IR, on the<br />
basis <strong>of</strong> the concern that cotargeting IR would lead to<br />
unacceptable toxicity. MAbs targeting IGF-1R were the<br />
furthest in development and had the benefit <strong>of</strong> inhibiting<br />
hybrid receptors besides IGF-1R. Recently, several phase I<br />
to III clinical trials have been conducted to evaluate the<br />
safety and efficacy <strong>of</strong> drugs targeting the IGF-1R. From<br />
these studies, we obtained some important indications:<br />
(1) anti–IGF-1R drugs have modest toxic effects, with mild<br />
and reversible hyperglycemia as the most common toxicity;<br />
and (2) anti-IGF-1R drugs show limited effectiveness. In<br />
particular, the best tumor responses have been observed<br />
in Ewing’s sarcoma, in which IGF/IGF-1R functions have<br />
been clearly associated with the pathogenesis <strong>of</strong> this tumor<br />
and in which few, if any, other TK receptors are fundamentally<br />
activated. 9 However, despite the presence <strong>of</strong> the target
TARGETING THE IGF SYSTEM<br />
Fig. 1. The IGF system. 5<br />
Abbreviations: ERK, extracellular regulated kinase; IGF, insulin-like growth factor; IRS, insulin receptor substrate; PI3-K, phosphoinositide-<br />
3-kinase; mTOR, mammalian target <strong>of</strong> rapamycin.<br />
in all tumors and ample preclinical evidence supporting<br />
the potential value <strong>of</strong> anti–IGF-1R agents, less than 10% <strong>of</strong><br />
patients respond to this therapy with extraordinary results.<br />
17<br />
On one side, this implies that the presence <strong>of</strong> the target<br />
is not sufficient to benefit from this targeted therapy and<br />
that other redundant pathways may be present to render<br />
IGF-1R–targeted cells resistant to anti–IGF-1R MAb.<br />
Recent studies in cell lines have demonstrated that knocking<br />
out, downregulating, or pharmacologically inhibiting<br />
IGF-1R can lead to a compensatory increase in IR<br />
signaling. 18-21 So, the take-home messages <strong>of</strong> these studies<br />
are: (1) the ratio IGF-1R/IR as well as the type <strong>of</strong> ligand(s)<br />
that are prevalent in the specific cellular context should be<br />
considered to identify patients that may benefit from anti–<br />
IGF-1R therapy; (2) we need to better identify the mechanisms<br />
<strong>of</strong> action <strong>of</strong> IR in cancer, viewing this receptor in a<br />
new light.<br />
Insulin/IR in Cancer<br />
As mentioned above, IR shares high homology to IGF-1R.<br />
However, unlike IGF-1R, the IR is characterized by the<br />
ability to alternatively splice a small exon (exon 11) encoding<br />
a 12–amino acid stretch contiguous to the CT<br />
peptide (encoded by the C-terminal sequence). The exclusion<br />
<strong>of</strong> exon 11 generates is<strong>of</strong>orm A (IR-A); its inclusion<br />
generates is<strong>of</strong>orm B (IR-B). 4 Although the current thinking<br />
is that IR primarily mediates the metabolic effects <strong>of</strong><br />
insulin through the activation <strong>of</strong> the PI3K pathway and<br />
IGF-1R mainly mediates the growth effects <strong>of</strong> IGFs via the<br />
activation <strong>of</strong> MAPK, it is now clearly established that in<br />
cancer cells IR, particularly IR-A, is <strong>of</strong>ten overexpressed<br />
and its signaling pathway deregulated with substantial<br />
crosstalk with the IGF-1R pathway. Several factors account<br />
for the loss <strong>of</strong> IR physiologic specificity in cancer. First,<br />
cancer cells predominantly overexpress the IR-A is<strong>of</strong>orm.<br />
601
Although IR-B is a specific receptor for insulin, IR-A also<br />
binds IGF-2 and at lower affinity IGF-1, and may induce<br />
biologic effects in response to both IGFs. Second, overexpressed<br />
IR enhances the effects <strong>of</strong> IGF-1 and IGF-2<br />
through the formation <strong>of</strong> IR/IGF-1R hybrid receptors,<br />
which bind both IGFs with high affinity. Third, cancers<br />
<strong>of</strong>ten produce both IGF-2 and IGF-1 in an autocrine/paracrine<br />
manner. 1 Finally, in patients with cancer affected by<br />
insulin resistance, the elevated levels <strong>of</strong> circulating insulin<br />
induce unbalanced IR activation, with predominant activation<br />
in the mitogenic pathway rather than the metabolic<br />
pathway. 22,23<br />
The importance <strong>of</strong> IR and insulin in tumor development<br />
and progression has been demonstrated in both animal<br />
models and clinical studies. 10 In humans, high levels <strong>of</strong><br />
insulin—but not blood glucose or obesity per se—are associated<br />
with increased risk for various malignancies. 24<br />
Women with breast cancer that also have insulin resistance<br />
show increased cancer-specific mortality. 25,26<br />
Obesity is a very important determinant for inducing or<br />
worsening insulin resistance. Obesity is also a predisposing<br />
factor <strong>of</strong> type 2 diabetes (T2DM), and several studies have<br />
now firmly established that both obesity and/or T2DM are<br />
associated with an increased risk <strong>of</strong> cancer. 23,27,28 Patients<br />
with T2DM carry an increased risk for almost every cancer<br />
histotype, except prostate cancer. Obese patients are at<br />
increased risk for a variety <strong>of</strong> malignancies, including most<br />
common cancers and hematologic malignancies. Metabolic<br />
syndrome, a disorder characterized by obesity, hypertension,<br />
dyslipidemia, and long-term insulin resistance, is also<br />
associated with worse cancer prognosis. 29 Conversely, body<br />
weight reduction decreases cancer risk. 22 Because <strong>of</strong> these<br />
findings and considering the studies suggesting that insulin<br />
analogs may promote tumorigenesis, 10 the effects <strong>of</strong> insulin/IR<br />
on tumor growth have recently received greater attention.<br />
New Opportunities for Cancer Prevention and<br />
Therapy Involving the IR pathway<br />
As long-term exposure to hyperinsulinemia is an important<br />
risk factor for cancer development and progression in<br />
patients with obesity, T2DM, or both, measures and drugs<br />
aimed at improving insulin resistance and reducing circulating<br />
insulin levels should contribute to prevent cancer in<br />
these patients and to ameliorate prognosis in patients with<br />
cancer. 25,26 Nonpharmacologic measures, such as lifestyle<br />
changes involving caloric restriction and physical exercise,<br />
may also be useful.<br />
Among drugs aimed at reducing insulin resistance and<br />
circulating insulin levels, biguanides and thiazolidinediones<br />
(TZDs) (collectively classified as insulin sensitizers) have<br />
received attention as potential anticancer agents. Metformin<br />
is the only biguanide used in the clinical setting and<br />
is currently recommended as first-line therapy in patients<br />
with T2DM for its excellent long-term safety pr<strong>of</strong>ile. Metformin<br />
impairs the production <strong>of</strong> adenosine 5�-triphosphate<br />
(ATP) by targeting complex I in the mitochondrial electron<br />
transport chain. This event activates AMP-activated protein<br />
kinase (AMPK), a kinase with a key role in the regulation <strong>of</strong><br />
cellular energy homeostasis and growth. AMPK causes, on<br />
one hand, downregulation <strong>of</strong> gluconeogenesis in the liver<br />
with reductions in blood glucose and insulin levels and, on<br />
602<br />
SCOTLANDI AND BELFIORE<br />
the other hand, direct reduction <strong>of</strong> cell growth through the<br />
inhibition <strong>of</strong> the mTOR/AKT pathway. Indeed, in vitro<br />
studies and animal models strongly suggest that metformin<br />
may have anticancer effects. 30,31 In humans, observational<br />
clinical studies have actually shown a decrease in cancer<br />
risk in patients with T2DM using metformin compared with<br />
those following other treatment regimens. In a case-control<br />
study, metformin use was associated with a reduced risk for<br />
breast cancer. 32 Moreover, the adjunct <strong>of</strong> metformin to<br />
insulin was reported to <strong>of</strong>fset the increased risk for colorectal<br />
or pancreatic cancer observed when insulin was used as<br />
monotherapy, and patients with T2DM treated with metformin<br />
were found to have a reduced cancer-specific mortality<br />
compared with those using insulin. From a therapeutic<br />
point <strong>of</strong> view, metformin may improve response rates in<br />
women with breast cancer that have T2DM who are receiving<br />
adjuvant chemotherapy, 33 as well as progression-free<br />
survival for chemotherapy-treated patients with advanced<br />
cell lung cancer that have diabetes. 34 Currently, pilot clinical<br />
trials are being conducted with women without diabetes<br />
to evaluate the possible effect <strong>of</strong> metformin on the outcome<br />
<strong>of</strong> breast cancer (clinical trials NCT00897884 and<br />
NCT01101438).<br />
TZDs, the second class <strong>of</strong> insulin sensitizers available for<br />
clinical use, belong to the group <strong>of</strong> peroxisome proliferatoractivated<br />
receptor gamma (PPARgamma) agonists. These<br />
drugs have shown substantial antitumoral effects in vitro<br />
and in some, but not in all, animal models. However to date,<br />
enthusiasm for the anticancer potential <strong>of</strong> the currently<br />
available TZDs has declined because <strong>of</strong> toxicity and an<br />
associated increased risk <strong>of</strong> tumors. 35 When new TZDs reach<br />
the market, more studies are warranted to explore the<br />
effects <strong>of</strong> these drugs for patients with cancer who have<br />
insulin resistance.<br />
As mentioned above, IR-A overexpressed by cancer cells<br />
may be stimulated not only by circulating insulin, but also<br />
by autocrine- or paracrine-produced IGF-2 and IGF-1. In<br />
patients with cancers overexpressing IR-A and IGFs, therefore,<br />
lowering insulin levels with the use <strong>of</strong> insulin sensitizers<br />
is not sufficient and direct inhibition <strong>of</strong> this IGFs/IR-A<br />
pathway should be pursued.<br />
These considerations, together with evidence indicating<br />
that selective IGF-1R inhibitors can favor the emergence <strong>of</strong><br />
cell clones with enhanced IGFs/IR-A loops 21,36 and worsen<br />
hyperinsulinemia, 37 bring about the concept that cotargeting<br />
IGF-1R and IR may be a suitable approach for patients<br />
with these malignancies. Small molecules with TK inhibitory<br />
activity appear to be the most promising drugs because<br />
<strong>of</strong> their ability to block the ATP-binding site <strong>of</strong> the kinase<br />
domain, which shares a high degree <strong>of</strong> homology between IR<br />
and IGF-1R. These drugs can be given orally and administered<br />
in combination with standard chemotherapy. Two<br />
currently available TKIs (BMS-754807 and OSI-906) share<br />
the ability to inhibit both IR and IGF-1R. BMS-754807<br />
inhibits both IGF-1R and IR with very similar activity (the<br />
half maximal inhibitory concentrations [IC 50] were 1.8<br />
nmol/L and 1.7 nmol/L, respectively), 38 but also elicits substantial<br />
inhibition toward other TK receptors (e.g., Met,<br />
recepteur d’origine nantais [RON], TrkA, TrkB) and Aurora<br />
A and B. BMS-754807 is currently being evaluated in<br />
several clinical trials as a single agent and in combination<br />
with other drugs in patients with advanced or metastatic
TARGETING THE IGF SYSTEM<br />
malignancies (clinical trials NCT00898716, NCT00569036,<br />
NCT00908024, NCT00788333, NCT01225172, and NCT-<br />
00793897). OSI-906 is a selective dual-inhibitor <strong>of</strong> IR and<br />
IGF-1R (IC 50 <strong>of</strong> 19 nmol/L to 35 nmol/L). 39 Preliminary<br />
studies with OSI-906 have yielded encouraging results and<br />
this drug is now being evaluated in phase I escalation<br />
studies as a single agent (clinical trials NCT00514306 and<br />
NCT00514007) and in phase I to III trials (clinical trials<br />
NCT01101906, NCT00924989, NCT01387386, and others)<br />
for various malignancies generally characterized by an activated<br />
IGF-2/1R-A loop.<br />
These studies will hopefully provide pro<strong>of</strong>-<strong>of</strong>-concept that<br />
IR inhibition, in addition to IGF-1R inhibition, may be<br />
clinically relevant for patients with cancer.<br />
Other Controversial Issues<br />
In addition, IGF-1R was shown to mediate differentiation<br />
in some cancers. Specific experimental models indicated<br />
that the balance between mitogenesis and differentiation<br />
is strongly influenced by the relative level <strong>of</strong> expression<br />
<strong>of</strong> the two main IGF-1R mediators, Shc and IRS1. 40<br />
Unfortunately, this is not a general rule and the exact<br />
mechanism that shifts the message from IGF-1R is still<br />
unknown. In any case, evidence indicates that the IGF<br />
system is an important mediator <strong>of</strong> mesenchymal or neural<br />
differentiation, an aspect that we need to consider for<br />
sarcomas and brain tumors. We need to be aware that<br />
IGF-1R and its substrates can also send contradictory signals,<br />
signals that can actually lead to growth inhibition or to<br />
the inhibition <strong>of</strong> metastatic spread. These contradictions<br />
ought to become fertile areas <strong>of</strong> investigation for both basic<br />
and applied research.<br />
There is no doubt that anti–IGF-1R therapy should be<br />
combined with conventional or other targeted drugs. Each<br />
tumor requires a unique cocktail <strong>of</strong> drugs and dedicated<br />
studies. This concept is true for most targeted therapies<br />
and further effort, time, and resources to be translated<br />
into effective treatments are needed. The results achieved<br />
to date are not satisfactory to justify routine clinical use<br />
<strong>of</strong> IGF-1R–targeting agents. Nevertheless, for some heavily<br />
pretreated patients with refractory rare tumors, responses<br />
and clinical benefit in combination with chemotherapy<br />
have been observed. Unfortunately, rare tumors seem<br />
to be most sensitive to these targeted therapies, which<br />
does not inspire pharmaceutical companies. However, we<br />
Authors’ Disclosures <strong>of</strong> Potential Conflicts <strong>of</strong> Interest<br />
Author<br />
Katia Scotlandi*<br />
Antonino Belfiore*<br />
*No relevant relationships to disclose.<br />
Employment or<br />
Leadership<br />
Positions<br />
Consultant or<br />
Advisory Role<br />
1. Samani AA, Yakar S, LeRoith D, et al. The role <strong>of</strong> the IGF system in<br />
cancer growth and metastasis: overview and recent insights. Endocr Rev.<br />
2007;28:20-47.<br />
2. Taniguchi CM, Emanuelli B, Kahn CR. Critical nodes in signalling<br />
pathways: insights into insulin action. Nat Rev Mol Cell Biol. 2006;7:85-96.<br />
3. Frasca F, Pandini G, Scalia P, et al. Insulin receptor is<strong>of</strong>orm A, a newly<br />
strongly believe that joint efforts between academia and<br />
industry are in the interests <strong>of</strong> both. We have a good level<br />
<strong>of</strong> knowledge in the field and several drugs already developed.<br />
Just put them together and take another step toward<br />
light.<br />
Conclusion<br />
Recent phase I to II clinical studies with selective anti–<br />
IGF-1R MAbs together with epidemiologic data have shifted<br />
attention from IGF-1R to IR, and to a more comprehensive<br />
view <strong>of</strong> the IGF system. To date, small molecules acting as<br />
dual inhibitors <strong>of</strong> IGF-1R and IR appear to be the most<br />
promising approaches to deprive cancer cells <strong>of</strong> this important<br />
signaling axis. Unfortunately, hyperglycemia and hyperinsulinemia<br />
are important adverse effects <strong>of</strong> dual IGF-1R<br />
and IR inhibitors, although hyperglycemia seems to be<br />
reversible after the cessation <strong>of</strong> treatment. It can be hypothesized,<br />
therefore, that insulin sensitizers (e.g., metformin)<br />
should be given together with these inhibitors to limit these<br />
adverse effects.<br />
Moreover, we need new biomarkers to select patients<br />
suitable for IR and IGF-1R dual inhibition and to monitor<br />
therapeutic efficacy. Recently, it has been reported that<br />
response to a dual anti–IR/IGF-1R inhibitor may be correlated<br />
with an IGF expression signature 41 or with lack <strong>of</strong><br />
epithelial-mesenchymal transition, 42 but we clearly need<br />
more extensive studies and validated biomarkers.<br />
Another possible approach involves the use <strong>of</strong> antibodies<br />
recognizing both IGF-2 and IGF-1. Such antibodies have<br />
been described 43 and, in animal models, show promising<br />
results in IGFs-driven malignancies. 44 Also in this case,<br />
more studies are needed regarding the applicability <strong>of</strong> this<br />
approach in humans.<br />
Overall, recent experimental evidence has shed light on to<br />
some new players, like IR, which has been substantially<br />
neglected in the field <strong>of</strong> cancer research as a mediator <strong>of</strong><br />
tumor progression. We are now well aware <strong>of</strong> the complexity<br />
<strong>of</strong> the pathway and have some new potentially promising<br />
drugs in our hands. Further efforts are needed to learn how<br />
to maximize their efficacy in patients with cancer.<br />
Acknowledgment<br />
The studies cited in this review were supported by grants<br />
from the Italian Association for Cancer Research (IG-10452<br />
[Katia Scotlandi] and IG-10625 [Antonino Belfiore]).<br />
Stock<br />
Ownership Honoraria<br />
REFERENCES<br />
Research<br />
Funding<br />
Expert<br />
Testimony<br />
Other<br />
Remuneration<br />
recognized, high-affinity insulin-like growth factor II receptor in fetal and<br />
cancer cells. Mol Cell Biol. 1999;19:3278-3288.<br />
4. Belfiore A, Frasca F, Pandini G, et al. Insulin receptor is<strong>of</strong>orms and<br />
insulin receptor/insulin-like growth factor receptor hybrids in physiology and<br />
disease. Endocr Rev. 2009;30:586-623.<br />
5. Martins AS, Ordonez JL, Amaral AT, et al. IGF1R signaling in Ewing<br />
603
sarcoma is shaped by clathrin-/caveolin-dependent endocytosis. PLoS One.<br />
2011;6:e19846.<br />
6. Giudice J, Leskow FC, Arndt-Jovin DJ, et al. Differential endocytosis<br />
and signaling dynamics <strong>of</strong> insulin receptor variants IR-A and IR-B. J Cell Sci.<br />
2011;124:801-811.<br />
7. Sehat B, T<strong>of</strong>igh A, Lin Y, et al. SUMOylation mediates the nuclear<br />
translocation and signaling <strong>of</strong> the IGF-1 receptor. Sci Signal. 2010;3:ra10<br />
8. Baserga R. The insulin-like growth factor-I receptor as a target for<br />
cancer therapy. Expert Opin Ther Targets. 2005;9:753-768.<br />
9. Scotlandi K, Picci P. Targeting insulin-like growth factor 1 receptor in<br />
sarcomas. Curr Opin Oncol. 2008;20:419-427.<br />
10. Gallagher EJ, LeRoith D. Minireview: IGF, insulin, and cancer. Endocrinology.<br />
2011;152:2546-2451.<br />
11. Steuerman R, Shevah O, Laron Z. Congenital IGF1 deficiency tends to<br />
confer protection against post-natal development <strong>of</strong> malignancies. Eur J<br />
Endocrinol. 2011;164:485-489.<br />
12. Guevara-Aguirre J, Balasubramanian P, Guevara-Aguirre M, et al.<br />
Growth hormone receptor deficiency is associated with a major reduction in<br />
pro-aging signaling, cancer, and diabetes in humans. Sci Transl Med. 2011;<br />
3:70ra13.<br />
13. Dunn SE, Kari FW, French J, et al. Dietary restriction reduces<br />
insulin-like growth factor I levels, which modulates apoptosis, cell proliferation,<br />
and tumor progression in p53-deficient mice. Cancer Res. 1997;57:4667-<br />
4672.<br />
14. Benini S, Manara MC, Baldini N, et al. Inhibition <strong>of</strong> insulin-like growth<br />
factor I receptor increases the antitumor activity <strong>of</strong> doxorubicin and vincristine<br />
against Ewing’s sarcoma cells. Clin Cancer Res. 2001;7:1790-1797.<br />
15. Bodzin AS, Wei Z, Hurtt R, et al. Gefitinib resistance in HCC Mahlavu<br />
cells: upregulation <strong>of</strong> CD133 expression, activation <strong>of</strong> IGF-1R signaling<br />
pathway, and enhancement <strong>of</strong> IGF-1R nuclear translocation. J Cell Physiol.<br />
Epub 2011 Sep 29.<br />
16. Rosenzweig SA. Acquired resistance to drugs targeting receptor tyrosine<br />
kinases. Biochem Pharmacol. Epub 2011 Dec 26.<br />
17. Olmos D, Postel-Vinay S, Molife LR, et al. Safety, pharmacokinetics,<br />
and preliminary activity <strong>of</strong> the anti-IGF-1R antibody figitumumab (CP-<br />
751,871) in patients with sarcoma and Ewing’s sarcoma: a phase 1 expansion<br />
cohort study. Lancet Oncol. 2010;11:129-135.<br />
18. Buck E, Gokhale PC, Koujak S, et al. Compensatory insulin receptor<br />
(IR) activation on inhibition <strong>of</strong> insulin-like growth factor-1 receptor (IGF-1R):<br />
rationale for cotargeting IGF-1R and IR in cancer. Mol Cancer Ther. 2010;9:<br />
2652-2664.<br />
19. Ulanet DB, Ludwig DL, Kahn CR, et al. Insulin receptor functionally<br />
enhances multistage tumor progression and conveys intrinsic resistance to<br />
IGF-1R targeted therapy. Proc Natl Acad Sci USA. 2010;107:10791-10798.<br />
20. Dinchuk JE, Cao C, Huang F, et al. Insulin receptor (IR) pathway<br />
hyperactivity in IGF-IR null cells and suppression <strong>of</strong> downstream growth<br />
signaling using the dual IGF-IR/IR inhibitor, BMS-754807. Endocrinology.<br />
2010;151:4123-4132.<br />
21. Gar<strong>of</strong>alo C, Manara MC, Nicoletti G, et al. Efficacy <strong>of</strong> and resistance to<br />
anti-IGF-1R therapies in Ewing’s sarcoma is dependent on insulin receptor<br />
signaling. Oncogene. 2011;30:2730-2740.<br />
22. Kaaks R, Lukanova A. Effects <strong>of</strong> weight control and physical activity in<br />
cancer prevention: role <strong>of</strong> endogenous hormone metabolism. Ann N Y Acad<br />
Sci. 2002;963:268-281.<br />
23. Vigneri P, Frasca F, Sciacca L, et al. Diabetes and cancer. Endocr Relat<br />
Cancer. 2009;16:1103-1123.<br />
604<br />
SCOTLANDI AND BELFIORE<br />
24. Pisani P. Hyper-insulinaemia and cancer, meta-analyses <strong>of</strong> epidemiological<br />
studies. Arch Physiol Biochem. 2008;114:63-70.<br />
25. Duggan C, Irwin ML, Xiao L, et al. Associations <strong>of</strong> insulin resistance<br />
and adiponectin with mortality in women with breast cancer. J Clin Oncol.<br />
2011;29:32-39.<br />
26. Irwin ML, Duggan C, Wang CY, et al. Fasting C-peptide levels and<br />
death resulting from all causes and breast cancer: The health, eating, activity,<br />
and lifestyle study. J Clin Oncol. 2011;29:47-53.<br />
27. Calle EE, Kaaks R. Overweight, obesity and cancer: epidemiological<br />
evidence and proposed mechanisms. Nat Rev Cancer. 2004;4:579-591.<br />
28. Rousseau MC, Parent ME, Pollak MN, et al. Diabetes mellitus and<br />
cancer risk in a population-based case-control study among men from Montreal,<br />
Canada. Int J Cancer. 2006;118:2105-2109.<br />
29. Stocks T, Borena W, Strohmaier S, et al. Cohort Pr<strong>of</strong>ile: The Metabolic<br />
syndrome and Cancer project (Me-Can). Int J Epidemiol. 2010;39:660-667.<br />
30. Jalving M, Gietema JA, Lefrandt JD, et al. Metformin: taking away the<br />
candy for cancer? Eur J Cancer. 2010;46:2369-2380.<br />
31. Bost F, Sahra IB, Le Marchand-Brustel Y, et al. Metformin and cancer<br />
therapy. Curr Opin Oncol. <strong>2012</strong>;24:103-108.<br />
32. Bosco JL, Antonsen S, Sorensen HT, et al. Metformin and incident<br />
breast cancer among diabetic women: a population-based case-control study<br />
in Denmark. Cancer Epidemiol Biomarkers Prev. 2011;20:101-111.<br />
33. Jiralerspong S, Palla SL, Giordano SH, et al. Metformin and pathologic<br />
complete responses to neoadjuvant chemotherapy in diabetic patients with<br />
breast cancer. J Clin Oncol. 2009;27:3297-3302.<br />
34. Tan BX, Yao WX, Ge J, et al. Prognostic influence <strong>of</strong> metformin as<br />
first-line chemotherapy for advanced nonsmall cell lung cancer in patients<br />
with type 2 diabetes. Cancer. 2011;117:5103-5111.<br />
35. Belfiore A, Genua M, Malaguarnera R. PPAR-gamma agonists and<br />
their effects on IGF-I receptor signaling: implications for cancer. PPAR Res.<br />
2009. Epub Jul 7.<br />
36. Zhang H, Pelzer AM, Kiang DT, et al. Down-regulation <strong>of</strong> type I<br />
insulin-like growth factor receptor increases sensitivity <strong>of</strong> breast cancer cells<br />
to insulin. Cancer Res. 2007;67:391-397.<br />
37. Pollak M. Targeting insulin and insulin-like growth factor signalling in<br />
oncology. Curr Opin Pharmacol. 2008;8:384-392.<br />
38. Carboni JM, Wittman M, Yang Z, et al. BMS-754807, a small molecule<br />
inhibitor <strong>of</strong> insulin-like growth factor-1R/IR. Mol Cancer Ther. 2009;8:3341-<br />
3349.<br />
39. Mulvihill MJ, Cooke A, Rosenfeld-Franklin M, et al. Discovery <strong>of</strong><br />
OSI-906: a selective and orally efficacious dual inhibitor <strong>of</strong> the IGF-1 receptor<br />
and insulin receptor. Future Med Chem. 2009;1:1153-1171.<br />
40. Valentinis B, Baserga R. IGF-I receptor signalling in transformation<br />
and differentiation. Mol Pathol. 2001;54:133-137.<br />
41. Litzenburger BC, Creighton CJ, Tsimelzon A, et al. High IGF-IR<br />
activity in triple-negative breast cancer cell lines and tumorgrafts correlates<br />
with sensitivity to anti-IGF-IR therapy. Clin Cancer Res. 2011;17:2314-2327.<br />
42. Gualberto A, Dolled-Filhart M, Gustavson M, et al. Molecular analysis<br />
<strong>of</strong> non-small cell lung cancer identifies subsets with different sensitivity to<br />
insulin-like growth factor I receptor inhibition. Clin Cancer Res. 2010;16:<br />
4654-4665.<br />
43. Feng Y, Zhu Z, Xiao X, et al. Novel human monoclonal antibodies to<br />
insulin-like growth factor (IGF)-II that potently inhibit the IGF receptor type<br />
I signal transduction function. Mol Cancer Ther. 2006;5:114-120.<br />
44. Gao J, Chesebrough JW, Cartlidge SA, et al. Dual IGF-I/II-neutralizing<br />
antibody MEDI-573 potently inhibits IGF signaling and tumor growth.<br />
Cancer Res. 2011;71:1029-1040.
Hedgehog Pathway in Pediatric Cancers:<br />
They’re Not Just for Brain Tumors Anymore<br />
Overview: The Hedgehog (HH) pathway regulates fundamental<br />
processes in embryonic development, including stem cell<br />
maintenance, cell differentiation, tissue polarity, and cell<br />
proliferation. In the vertebrate pathway, Sonic hedgehog<br />
(SHH) binds to Patched1 (PTCH1), which relieves its inhibition<br />
<strong>of</strong> Smoothened (SMO), allowing the GLI family <strong>of</strong> transcription<br />
factors to translocate to the nucleus and activate HH target<br />
genes such as GLI1, GLI2, PTCH1, CYCLIN D1, BCL-2, and<br />
MYCN. The HH pathway is also an active participant in<br />
tumorigenesis. In 1996, loss-<strong>of</strong>-function mutation in PTCH1<br />
was discovered to be the cause <strong>of</strong> nevoid basal cell carcinoma<br />
syndrome (NBCCS, or Gorlin syndrome), an autosomal dominant<br />
disease associated with increased rates <strong>of</strong> basal cell<br />
carcinoma (BCC), medulloblastoma (MB), and rarely, rhabdomyosarcoma.<br />
It is now estimated that 100% <strong>of</strong> sporadic BCC<br />
and up to 20% to 30% <strong>of</strong> MB also harbor activating HH<br />
THE HEDGEHOG pathway was first discovered in 1980<br />
by Nobel Laureates Nusslein-Volhard and Wieschaus<br />
after their isolation <strong>of</strong> mutations in genes that control the<br />
development <strong>of</strong> anterior-posterior body axis segmentation in<br />
Drosophila melanogaster. 1 In mammals, three genes, Desert<br />
Hedgehog (DHH), Indian Hedgehog (IHH), and Sonic<br />
Hedgehog (SHH), function as ligands for the receptor<br />
PTCH1. 2 In the unbound state, PTCH1 inhibits the sevenpass<br />
transmembrane protein SMO, and on HH ligand binding,<br />
PTCH1 undergoes internalization and degradation,<br />
resulting in the activation and release <strong>of</strong> SMO to enter the<br />
primary cilia where it promotes the dissociation <strong>of</strong> the<br />
Suppressor <strong>of</strong> fused (SUFU)–glioma-associated oncogene<br />
homolog (GLI) complex. 2,3 This results in nuclear translocation<br />
and activation <strong>of</strong> the GLI1 and GLI2 transcription<br />
factors, and degradation <strong>of</strong> the repressor forms <strong>of</strong> GLI<br />
(primarily GLI3). GLI proteins stimulate the transcription<br />
<strong>of</strong> HH pathway target genes, including GLI1, GLI2, PTCH1,<br />
CYCLIN D1, BCL-2, and MYCN, which in turn, function<br />
in a wide range <strong>of</strong> developmental signaling roles. 4 In some<br />
cases, HH ligands function as mitogens, whereas in others<br />
they promote differentiation. 2 SHH, the most common vertebrate<br />
homolog, is required for the correct patterning <strong>of</strong> the<br />
neural tube, the somites, and anterior-posterior positioning<br />
<strong>of</strong> the limb bud. 5 The importance <strong>of</strong> HH signaling in mammalian<br />
development is underscored by observations that<br />
mutations in SHH cause holoprosencephaly, a developmental<br />
disorder that affects the midline <strong>of</strong> the face and nervous<br />
system. 2-5 The HH pathway can be activated by mutations<br />
in PTCH1 or SUFU (loss <strong>of</strong> function), or SMO (gain <strong>of</strong><br />
function) that lead to ligand-independent, constitutive signaling.<br />
A clear genetic contribution <strong>of</strong> such HH pathway<br />
activation to oncogenesis was established with the discovery<br />
<strong>of</strong> loss-<strong>of</strong>-function mutations in germ-line PTCH1 (chromosome<br />
9q22.3) as the cause <strong>of</strong> familial NBCCS (or Gorlin’s<br />
syndrome), an autosomal-dominant disease in which patients<br />
have an increased tendency to develop basal cell<br />
carcinoma (BCC), medulloblastoma (MB), rhabdomyosarcoma,<br />
and ovarian neoplasia. 6 This review will further detail<br />
the known relationships <strong>of</strong> the HH signaling pathway in<br />
cancer development, especially medulloblastoma and other<br />
By Tobey J. MacDonald, MD<br />
pathway mutations. Together, these discoveries firmly established<br />
the linkage between HH pathway activation and cancer<br />
development. Intense research has since been focused on<br />
further defining the role <strong>of</strong> the HH pathway in BCC and MB and<br />
potential therapeutic strategies to inhibit HH signaling. Early<br />
clinical trials <strong>of</strong> SMO inhibitors have shown promising results<br />
in the treatment <strong>of</strong> adult BCC and SHH-driven MB. More<br />
recently, a number <strong>of</strong> other pediatric cancers have been<br />
reported to show HH activity, making these tumors potential<br />
candidates for HH inhibitor therapy. To date however, no HH<br />
pathway mutations have been identified in other pediatric<br />
cancers. This review will describe the HH pathway signaling in<br />
development and cancer with a focus on recent evidence for<br />
HH pathway activation in central nervous system (CNS) and<br />
non-CNS pediatric cancers.<br />
pediatric CNS tumors, and will focus on the most recent<br />
evidence implicating HH pathway signaling in non-CNS<br />
pediatric cancers.<br />
HH Pathway Mutations and HH Signaling in Cancer<br />
Following the discovery that germ-line mutations in<br />
PTCH1 are responsible for NBCCS, a number <strong>of</strong> studies<br />
subsequently confirmed the detection <strong>of</strong> HH pathway somatic<br />
mutations in sporadic BCC and MB. 7,8 It is estimated<br />
that almost all BCC tumors show evidence <strong>of</strong> constitutive<br />
HH pathway activity, with 90% exhibiting loss <strong>of</strong> PTCH1<br />
and 10% harboring activating mutations in SMO. 7 Medulloblastomas<br />
appear to be more heterogeneous, with up to 20%<br />
to 30% <strong>of</strong> tumors displaying a gene expression signature<br />
that is indicative <strong>of</strong> HH pathway activation. However, only<br />
50% <strong>of</strong> these tumors have confirmed loss <strong>of</strong> PTCH1, loss<br />
<strong>of</strong> SUFU, or gain-<strong>of</strong>-function SMO mutations. 3,9-11 Thus, a<br />
subset <strong>of</strong> medulloblastoma exhibits HH pathway activation<br />
through an alternative mechanism.<br />
Mice heterozygous for patched (ptc1) mutations, like<br />
heterozygous PTCH1 in humans, have a high rate <strong>of</strong> MB, as<br />
well as other tumors, and concomitant loss <strong>of</strong> Tp53 has<br />
been shown to further accelerate MB formation. 12 Subsequent<br />
studies have confirmed that Ptc1 deletion in lineagerestricted<br />
progenitor or stem cells is sufficient to initiate<br />
MB, and that MYCN is an essential downstream effector <strong>of</strong><br />
SHH in both normal and neoplastic cerebellar growth. 13,14<br />
Not only have these murine studies confirmed the role <strong>of</strong><br />
activating HH pathway mutations in the development <strong>of</strong> a<br />
subset <strong>of</strong> MB, and provide insight into the function <strong>of</strong> HH<br />
signaling in tumorigenesis, but also, these transgenic mice<br />
From the Pediatric Neuro-<strong>Oncology</strong> Program, Aflac Cancer Center and Blood Disorders<br />
Service, Children’s Healthcare <strong>of</strong> Atlanta, and Emory University School <strong>of</strong> Medicine, Emory<br />
Children’s Center, Atlanta, GA.<br />
Author’s disclosure <strong>of</strong> potential conflicts <strong>of</strong> interest are found at the end <strong>of</strong> this article.<br />
Address reprint requests to Tobey J. MacDonald, MD, Emory University School <strong>of</strong><br />
Medicine, Emory Children’s Center, 2015 Uppergate Drive NE, Suite 442, Atlanta, GA;<br />
email: tobey.macdonald@emory.edu.<br />
© <strong>2012</strong> by <strong>American</strong> <strong>Society</strong> <strong>of</strong> <strong>Clinical</strong> <strong>Oncology</strong>.<br />
1092-9118/10/1-10<br />
605
MB models have served as an invaluable preclinical tool for<br />
the testing <strong>of</strong> HH pathway inhibitors for the treatment <strong>of</strong><br />
SHH-driven MB.<br />
It has long been known that a constellation <strong>of</strong> midline<br />
developmental anomalies that are observed in embryos with<br />
primary congenital defects in SHH activation has similarly<br />
been observed in developing embryos exposed to naturally<br />
occurring teratogenic alkaloids such as cyclopamine, found<br />
in extracts from Veratrum Album, commonly known as<br />
White Hellebore, a plant within the lily family. 3 It has been<br />
demonstrated that these teratogens function as specific<br />
inhibitors <strong>of</strong> HH signaling by binding to SMO. As all HH<br />
signaling through the canonical pathway requires SMO,<br />
small molecules such as cyclopamine, which inhibit SMO<br />
function, completely block all HH pathway signaling regardless<br />
<strong>of</strong> the ligand. 3 HH pathway antagonists also provide<br />
valuable tools for dissecting the biochemistry and biology <strong>of</strong><br />
HH signaling and have enabled the current development <strong>of</strong><br />
unique molecularly targeted therapies for HH-driven cancer.<br />
For example, treatment with SMO inhibitors have been<br />
shown to completely eradicate HH-driven murine MB as<br />
well as markedly inhibit SHH in human BCC, and transiently<br />
induce full remission in adult human metastatic MB,<br />
indicating great promise for the clinical use <strong>of</strong> these agents<br />
for HH-driven disease. 15-17<br />
Although the HH pathway is activated by autocrine signaling<br />
by HH ligands, it can also initiate paracrine signaling<br />
with cells in the microenvironment. This creates a network<br />
<strong>of</strong> HH pathway signaling that determines the malignant<br />
behavior <strong>of</strong> the tumor cells. As a result <strong>of</strong> paracrine signal<br />
transmission, the effects <strong>of</strong> HH signaling most pr<strong>of</strong>oundly<br />
influence the stromal cells that constitute the tumor microenvironment.<br />
The stromal cells in turn produce factors<br />
that nurture the tumor. Thus, such cellular cross-talk can<br />
KEY POINTS<br />
● Hedgehog (HH) signaling has been investigated for<br />
its role in tumorigenesis because <strong>of</strong> its known function<br />
in embryonic stem cell maintenance, cell differentiation,<br />
tissue polarity, and cell proliferation.<br />
● Key constituents <strong>of</strong> the HH pathway include Sonic<br />
hedgehog (SHH) ligand and its receptor Patched1<br />
(PTCH1), which in the absence <strong>of</strong> SHH represses<br />
Smoothened (SMO) and prevents GLI transcription<br />
factor activation.<br />
● PTCH1 germ-line mutation results in nevoid basal<br />
cell carcinoma (BCC) syndrome, while somatic HH<br />
pathway mutations are found in sporadic BCC and<br />
medulloblastoma (MB), thereby establishing a linkage<br />
between HH activation and cancer development.<br />
● Encouraging results have been observed using SMO<br />
inhibitors to treat adult BCC and MB; however, the<br />
potential risks <strong>of</strong> SMO inhibitors in developing children<br />
remain a concern.<br />
● HH pathway activity, but not mutations, has recently<br />
been shown in other pediatric cancers, yet it remains<br />
to be seen whether SMO inhibition will be effective<br />
for this group <strong>of</strong> cancers.<br />
606<br />
greatly amplify HH signaling, resulting in the promotion <strong>of</strong><br />
tumor progression and metastasis. Ultimately, the linkage<br />
<strong>of</strong> aberrant HH signaling to tumorigenesis is thought to<br />
be mediated through cell-cycle dysregulation, protection <strong>of</strong><br />
cancer cells against apoptosis, and modulation <strong>of</strong> angiogenesis.<br />
2,3 Several lines <strong>of</strong> evidence support three mechanistic<br />
roles for HH signaling in cancer: (1) a cancer cellautonomous<br />
role, in which tumor growth is driven by<br />
activating mutations in the pathway; (2) a paracrine signaling<br />
role involving tumor and stromal interactions that<br />
promote tumor growth and invasion; and (3) an autocrine<br />
signaling role via cancer stem cells that promotes selfrenewal<br />
and proliferation. 2,3 The downstream effectors <strong>of</strong><br />
the HH signaling pathway are the GLI transcription factors,<br />
which promote cell proliferation, differentiation, and survival<br />
through induction <strong>of</strong> relevant target genes. 2,3 To date,<br />
different molecular lesions in PTCH1, SMO, and SUFU <strong>of</strong><br />
the HH pathway have been described in tumors. In each<br />
case, these alterations have resulted in increased transcriptional<br />
activity <strong>of</strong> the GLI1 and GLI2 transcription factors.<br />
Indeed, the first indication that genes in the HH pathway<br />
were associated with human cancer was the observation<br />
that GLI1 was amplified in glioblastoma, although it is now<br />
believed that this is not a common primary mechanism<br />
underlying glioblastoma formation. 3 Together with GLI and<br />
PTCH1, these effector targets are representative <strong>of</strong> the gene<br />
signature indicating SHH active tumors. Therefore, GLI1<br />
mRNA levels either in tumor tissue or relevant surrogate<br />
tissue is considered a reliable indicator <strong>of</strong> HH pathway<br />
activity. 2,3 Most recently, a number <strong>of</strong> common pediatric<br />
cancers have been shown to have expression and activation<br />
<strong>of</strong> the HH pathway through the measurement <strong>of</strong> these target<br />
effectors.<br />
HH Pathway in Pediatric Cancer<br />
The HH signaling pathway is believed to be active in<br />
early-onset pediatric tumors, both <strong>of</strong> CNS and non-CNS<br />
origin, because <strong>of</strong> the important role that HH plays in<br />
embryonic development. Indeed, GLI1 amplification has also<br />
been described in childhood sarcoma. In a more comprehensive<br />
study <strong>of</strong> a series <strong>of</strong> pediatric surgical tumor specimens,<br />
Oue and colleagues used expression <strong>of</strong> GLI1 as a marker <strong>of</strong><br />
HH pathway activation to demonstrate that almost 70% <strong>of</strong><br />
the pediatric tumors examined, including neuroblastoma,<br />
hepatoblastoma, high-grade glioma, and osteosarcoma, were<br />
positive for HH activity. 18 However, to date no HH mutations<br />
have been identified in these other pediatric cancers.<br />
Only descriptive studies have been performed, and thus it<br />
remains to be determined whether a valid functional relationship<br />
exists between HH activity and tumor progression<br />
in these cancers. The findings <strong>of</strong> these pediatric studies<br />
and others similarly reporting HH pathway expression and<br />
activation in both CNS and non-CNS pediatric cancers are<br />
further detailed below.<br />
CNS tumors<br />
TOBEY J. MACDONALD<br />
The role <strong>of</strong> HH pathway in pediatric MB is well established.<br />
However, more recent reports now suggest that the<br />
HH pathway may be implicated in other non-MB pediatric<br />
CNS tumors.
HEDGEHOG PATHWAY IN PEDIATRIC CANCER<br />
Diffuse Intrinsic Pontine Glioma<br />
In a report by Monje and colleagues, early postmortem<br />
tissue from patients with diffuse intrinsic pontine glioma<br />
(DIPG) was used to establish in vivo xenograft models <strong>of</strong><br />
DIPG. 19 Subsequent analysis <strong>of</strong> the established tumors<br />
showed that the HH signaling pathway is active in DIPG<br />
tumor cells and that DIPG self-renewal capacity in neurosphere<br />
culture is significantly reduced with inhibition <strong>of</strong> the<br />
HH signaling pathway.<br />
Juvenile Pilocytic Astrocytoma<br />
Rush and colleagues 20 demonstrated that mRNA expression<br />
levels <strong>of</strong> members <strong>of</strong> the HH pathway were elevated in<br />
45% <strong>of</strong> juvenile pilocytic astrocytoma specimens analyzed<br />
and that the expression <strong>of</strong> the HH pathway correlated<br />
inversely with patient age. Immunohistochemical (IHC)<br />
staining for PTCH1, GLI1, and the proliferation marker<br />
Ki67 demonstrated that patients diagnosed before the age <strong>of</strong><br />
10 years had an increased frequency and level <strong>of</strong> marker<br />
immunopositivity compared with those diagnosed after 10<br />
years <strong>of</strong> age. 20 A significant correlation was also observed<br />
between Ki67, PTCH1, and GLI1 positive staining, with 86%<br />
<strong>of</strong> Ki67-positive cells also expressing PTCH1.<br />
Non-CNS Pediatric Cancer<br />
Rhabdomyosarcoma<br />
Rhabdomyosarcoma (RMS) is the most common s<strong>of</strong>t tissue<br />
sarcoma in children and comprises two major histologic<br />
subtypes: alveolar rhabdomyosarcoma (ARMS) and embryonal<br />
rhabdomyosarcoma (ERMS). Although RMS is frequently<br />
said to be associated with activating mutations in<br />
the HH pathway, the role <strong>of</strong> the HH pathway in sporadic<br />
human RMS is not clear, and linkage to specific HH pathway<br />
mutations has not been confirmed. 3 Although one study<br />
reported mutations in PTCH1 and SUFU in RMS, most <strong>of</strong><br />
the analysis was restricted to loss <strong>of</strong> heterozygosity (LOH)<br />
analysis <strong>of</strong> PTCH1 and SMO in rhabdomyoma, and in only<br />
one case <strong>of</strong> RMS. 21 Since this region contains other potential<br />
tumor suppressor candidate genes, definitive conclusion<br />
from LOH analysis alone cannot be drawn. S<strong>of</strong>t tissue<br />
sarcomas that are very similar to human RMS are seen in<br />
Ptc1� mice, but the incidence <strong>of</strong> these tumors is strongly<br />
influenced by the murine genetic background. 3<br />
The initial descriptions <strong>of</strong> Gorlin syndrome identified an<br />
association <strong>of</strong> the disease with benign fetal rhabdomyoma,<br />
as well as rare instances <strong>of</strong> RMS. Tostar and colleagues<br />
reported overexpression <strong>of</strong> PTCH1 and GLI1 mRNA, as<br />
determined by in situ hybridization, in sporadic RMS. 21<br />
Using array comparative genomic hybridization to examine<br />
a specific subset <strong>of</strong> clinically defined intermediate-risk<br />
ERMS tumors, Paulson and colleagues similarly demonstrated<br />
that over 50% <strong>of</strong> tumors had low-level gains <strong>of</strong> a<br />
region containing GLI1 along with a gene expression signature<br />
consistent with HH-pathway activation. 22 Zibat and<br />
colleagues reported that PTCH1, GLI1, GLI3, and MYF5 are<br />
expressed at significantly higher levels in ERMS and that<br />
GLI1 expression consistently correlates with PTCH1 expression<br />
in ERMS, as well as a specific subset <strong>of</strong> ARMS. 23 This<br />
study also showed that high PTCH1 expression significantly<br />
correlated with reduced survival in a subset <strong>of</strong> RMS. Likewise,<br />
Pressey and colleagues demonstrated that expression<br />
<strong>of</strong> GLI1, with or without PTCH1, is detectable in substantial<br />
subsets <strong>of</strong> ERMS, but only rarely in ARMS tumors. 24 Importantly,<br />
neither PTCH1 mutations nor activating SMO mutations<br />
were detected in ERMS tumors with high GLI1<br />
expression, and in contrast to other reports, HH pathway<br />
activity in ERMS tumors did not correlate with a unique<br />
clinical phenotype. Furthermore, the gene expression patterns<br />
in ERMS indicated that approximately 29% exhibited<br />
evidence <strong>of</strong> HH pathway activity, yet this pattern was<br />
always coassociated with either p53 or RB pathway signatures.<br />
25 Finally, Oue and colleagues examined 18 RMS by<br />
IHC and showed that the majority <strong>of</strong> the tumors were<br />
immunopositive for SHH (78%), PTCH1 (100%), and GLI1<br />
(78%). 18 In contrast to the study by Pressey and colleagues,<br />
marker expression was higher in ARMS than in ERMS.<br />
Caution should be taken in interpreting the clinical significance<br />
<strong>of</strong> these results, and attempts to clinically translate<br />
these findings to therapeutic interventions would be premature<br />
as the presence <strong>of</strong> the HH pathway signature does<br />
not necessarily mean that the tumor cells are dependent on<br />
SMO activity. Cyclopamine studies in the Ptc1� mice are<br />
an example <strong>of</strong> this, whereby loss <strong>of</strong> ptc1 may contribute<br />
to tumor initiation, but is not required for tumor maintenance.<br />
26<br />
Neuroblastoma<br />
Neuroblastoma (NB) is a heterogeneous pediatric malignancy,<br />
with variable differentiation and growth potential,<br />
that shares a common origin arising from neural crest cells<br />
in the sympathetic nervous system. Using IHC, Souzaki and<br />
colleagues examined 82 NB and 10 ganglioneuroblastoma<br />
(GNB) and demonstrated tumor immunopositivity for SHH,<br />
GLI1, and PTCH1 in 67 (73%), 62 (67%), and 73 (79%),<br />
respectively. 27 Most NBs without MYCN amplification were<br />
positive for all three HH pathway markers. Only two (10%)<br />
<strong>of</strong> 20 cases with MYCN amplification were also positive for<br />
SHH and GLI1, and four (20%) were positive for PTCH1.<br />
The percentage <strong>of</strong> GLI1-positive cells without MYCN amplification<br />
was significantly higher than those with MYCN<br />
amplification, and the prognosis <strong>of</strong> the GLI1-positive cases<br />
was significantly better than that <strong>of</strong> the GLI1-negative<br />
cases. In tumors without MYCN amplification, high expression<br />
<strong>of</strong> GLI1 was significantly associated with early clinical<br />
stage, more differentiated tumors, and a good prognosis.<br />
Oue and colleagues examined 25 NB by IHC and similarly<br />
showed that 24 (96%), 17 (68%), and 25 (100%) stained<br />
positive for SHH, PTCH, and GLI1, respectively. 18 Likewise,<br />
all <strong>of</strong> the Gli1-negative tumors were poorly differentiated<br />
and exhibited advanced-stage disease. In this study, there<br />
was no significant relationship between GLI1 expression<br />
and MYCN amplification or prognosis. This study also<br />
showed that GLI1 transduction <strong>of</strong> NB cells inhibited proliferation<br />
in vitro and induced a gene expression pattern that<br />
resembled benign differentiated ganglioneuroma. 28 Notably,<br />
GLI1 transduction did not induce MYCN expression in NB<br />
cells.<br />
Renal Tumors<br />
IHC analysis <strong>of</strong> seven Wilms’ tumors showed that five<br />
(71%), seven (100%), and three (43%) stained positive for<br />
SHH, PTCH1, and GLI1, respectively. 18 Two renal clear cell<br />
sarcoma and three rhabdoid tumors <strong>of</strong> the kidney showed<br />
very high expression <strong>of</strong> HH pathway markers, suggesting<br />
607
that the HH pathway may contribute to the more unfavorable<br />
biologic behaviors <strong>of</strong> these renal tumors. 18<br />
Hepatic Tumors<br />
Eichenmuller and colleagues reported that HH signaling<br />
is active in pediatric hepatoblastoma and showed that blocking<br />
HH signaling with cyclopamine in hepatoblastoma<br />
cells leads to a significant decrease in cell viability and<br />
increased apoptosis. 29 In another study, IHC <strong>of</strong> 11 hepatoblastoma<br />
cases demonstrated that 100% <strong>of</strong> tumors stained<br />
positive for SHH and PTCH1, while eight (73%) were positive<br />
for GLI1. 18 One case <strong>of</strong> embryonal carcinoma and two<br />
hepatic tumor cell lines (Heh6 and Heh7) showed strong<br />
expression <strong>of</strong> all three HH pathway markers. No relationship<br />
was observed between GLI1 expression and histologic<br />
subtype, clinical stage, or prognosis.<br />
Osteosarcoma<br />
Recently published data also suggest HH signaling may<br />
play a role in the pathogenesis <strong>of</strong> osteosarcoma. A study <strong>of</strong><br />
osteosarcoma cell lines and biopsy specimens showed overexpression<br />
<strong>of</strong> HH target genes via real-time PCR. 30 Inhibition<br />
<strong>of</strong> the HH pathway with cyclopamine in vitro promoted<br />
G1 arrest and reduced the expression <strong>of</strong> positive cell-cycle<br />
regulators, including cyclins D1 and E1. In addition, SMO<br />
knockdown with SMO shRNA prevented the growth <strong>of</strong><br />
osteosarcoma, both in vitro and in vivo. 30<br />
HH Pathway Inhibitors in Pediatric <strong>Clinical</strong> Trials<br />
Inhibitors <strong>of</strong> HH signaling have recently been the focus <strong>of</strong><br />
intense research. It is noteworthy that many tumors that<br />
appear to lack specific HH pathway mutations have been<br />
reported to be sensitive to SMO inhibitors in vitro and in<br />
vivo. 3 This has led to estimates that as many as 25% <strong>of</strong><br />
human tumors may depend on HH pathway activity for<br />
growth. As a consequence, a wide range <strong>of</strong> tumors have been<br />
included in the early-adult clinical trials investigating SMO<br />
inhibitors. 3 Xenograft tumor models have shown that hu-<br />
Author’s Disclosure <strong>of</strong> Potential Conflicts <strong>of</strong> Interest<br />
Employment or<br />
Leadership Consultant or<br />
Author<br />
Positions Advisory Role<br />
Tobey J. MacDonald Novartis<br />
1. Nusslein-Volhard C, Wieschaus E. Mutations affecting segment number<br />
and polarity in Drosophila. Nature. 1980;287:795-801.<br />
2. Ingham PW, Nakano Y, Seger C. Mechanisms and functions <strong>of</strong> Hedgehog<br />
signalling across the metazoa. Nat Rev Genet. 2011;12:393-406.<br />
3. Ng JM, Curran T. The Hedgehog’s tale: Developing strategies for<br />
targeting cancer. Nat Rev Cancer. 2011;11:493-501.<br />
4. McMahon AP, Ingham PW, Tabin CJ. Developmental roles and clinical<br />
significance <strong>of</strong> hedgehog signaling. Curr Top Dev Biol. 2003;53:1-114.<br />
5. Lee J, Platt KA, Censullo P, et al. Gli1 is a target <strong>of</strong> Sonic hedgehog that<br />
induces ventral neural tube development. Development. 1997;124:2537-2552.<br />
6. Johnson RL, Rothman AL, Xie J, et al. Human homolog <strong>of</strong> patched, a<br />
candidate gene for the basal cell nevus syndrome. Science. 1996;272:1668-1671.<br />
7. Epstein EH. Basal cell carcinomas: Attack <strong>of</strong> the hedgehog. Nat Rev<br />
Cancer. 2008;8:743-754.<br />
8. Raffel C, Jenkins RB, Frederick L, et al. Sporadic medulloblastomas<br />
contain PTCH mutations. Cancer Res. 1997;57:842-845.<br />
9. Thompson MC, Fuller C, Hogg TL, et al. Genomics identifies medullo-<br />
608<br />
man tumors without HH pathway mutations that express<br />
HH ligands actually induce upregulation <strong>of</strong> HH pathway<br />
target genes in the stromal cells <strong>of</strong> mouse origin, rather than<br />
in the tumor cells. 3 Under these conditions, treatment with<br />
SMO inhibitors inhibited tumor growth to some degree, but<br />
did not eliminate the tumors. Some leukemias have also<br />
been reported to depend on SMO signaling for growth, but<br />
more recent studies indicate that survival was not dependent<br />
on HH ligands or SMO activity. 3 The pro<strong>of</strong> <strong>of</strong> concept<br />
for the clinical utility <strong>of</strong> SMO inhibitors has been established<br />
in patients with metastatic or locally advanced BCC<br />
or MB. 15-17 The more recent data suggest that this class <strong>of</strong><br />
agents may have broader utility in other adult and pediatric<br />
cancers. The dilemma is to identify tumors that would<br />
benefit most from HH inhibitor treatment, as the mere<br />
presence <strong>of</strong> HH ligands or a pathway signature does not<br />
guarantee a response. Because <strong>of</strong> the importance <strong>of</strong> the HH<br />
signaling pathway in the normal growth and development,<br />
the use <strong>of</strong> SMO inhibitors in children is a specific concern<br />
that will need to be carefully assessed in order to manage<br />
the potential adverse effects on bone growth plates, developing<br />
teeth, and the immature reproductive system. 3 At<br />
present, early phase I and II clinical trials with the SMO<br />
inhibitors LDE225 (Novartis) and GDC-0449 (Genentech)<br />
are being evaluated in pediatric medulloblastoma and other<br />
pediatric tumors that are potentially dependent on HH<br />
pathway signaling (Table 1).<br />
Stock<br />
Ownership Honoraria<br />
REFERENCES<br />
Table 1. Hedgehog Pathway Inhibitors in Current <strong>Clinical</strong> Trials<br />
Agent Target<br />
Research<br />
Funding<br />
Expert<br />
Testimony<br />
TOBEY J. MACDONALD<br />
Mechanism<br />
<strong>of</strong> Action Manufacturer<br />
GDC-0449* SMO Antagonist GDC-0449 (Genentech)<br />
LDE225* SMO Antagonist LDE225 (Novartis)<br />
LEQ506 SMO Antagonist LEQ506 (Novartis)<br />
PF-04449913 SMO Antagonist PF-04449913 (Pfizer)<br />
IPI-926 SMO Antagonist IPI-926 (Ifinity)<br />
BMS-833923 SMO Antagonist BMS-833923 (Bristol-Myers Squibb)<br />
* Currently being investigated in pediatric cancer. As <strong>of</strong> February 8, <strong>2012</strong><br />
(<strong>Clinical</strong>trials.gov).<br />
Other<br />
Remuneration<br />
blastoma subgroups that are enriched for specific genetic alterations. J Clin<br />
Oncol. 2006;24:1924-1931.<br />
10. Taylor MD, Liu L, Raffel C, et al. Mutations in SUFU predispose to<br />
medulloblastoma. Nature Genet. 2002;31:306-310.<br />
11. Pastorino, L. Ghiorzo P, Nasti S, et al. Identification <strong>of</strong> a SUFU<br />
germline mutation in a family with Gorlin syndrome. Am J Med Genet A.<br />
2009;149A:1539-1443.<br />
12. Wetmore C, Eberhart DE, Curran T. Loss <strong>of</strong> p53 but not ARF accelerates<br />
medulloblastoma in mice heterozygous for patched. Cancer Res. 2001;61:<br />
513-516.<br />
13. Yang ZJ, Ellis T, Markant SL, et al. Medulloblastoma can be initiated<br />
by deletion <strong>of</strong> Patched in lineage-restricted progenitors or stem cells. Cancer<br />
Cell. 2008;14:135-145.<br />
14. Hatton BA, Knoepfler PS, Kenney AM, et al. N-myc is an essential<br />
downstream effector <strong>of</strong> Shh signaling during both normal and neoplastic<br />
cerebellar growth. Cancer Res. 2006;66:8655-8661.<br />
15. Von H<strong>of</strong>f DD, LoRusso PM, Rudin CM, et al. Inhibition <strong>of</strong> the hedgehog
HEDGEHOG PATHWAY IN PEDIATRIC CANCER<br />
pathway in advanced basal-cell carcinoma. N Engl J Med. 2009;361:1164-<br />
1172.<br />
16. Romer JT, Kimura H, Magdaleno S, et al. Suppression <strong>of</strong> the Shh<br />
pathway using a small molecule inhibitor eliminates medulloblastoma in<br />
Ptc1(�/�)p53(�/�) mice. Cancer Cell. 2004;6:229-240.<br />
17. Rudin CM, Hann CL, Laterra J, et al. Treatment <strong>of</strong> medulloblastoma<br />
with hedgehog pathway inhibitor GDC-0449. N Engl J Med. 2009;361:1173-<br />
1178.<br />
18. Oue T, Yoneda A, Uehara S, et al. Increased expression <strong>of</strong> the hedgehog<br />
signaling pathway in pediatric solid malignancies. J Pediatr Surg. 2010;45:<br />
387-392.<br />
19. Monje M, Mitra SS, Freret ME, et al. Hedgehog-responsive candidate<br />
cell <strong>of</strong> origin for diffuse intrinsic pontine glioma. Proc Natl Acad Sci USA.<br />
2011;108:4453-4458.<br />
20. Rush SZ, Abel TW, Valadez, et al. Activation <strong>of</strong> the Hedgehog pathway<br />
in pilocytic astrocytomas. Neuro Oncol. 2010;12:790-798.<br />
21. Tostar U, Malm CJ, Meis-Kindblom JM, et al. Deregulation <strong>of</strong> the<br />
hedgehog signalling pathway: a possible role for the PTCH and SUFU genes<br />
in human rhabdomyoma and rhabdomyosarcoma development. J Pathol.<br />
2006;208:17-25.<br />
22. Paulson V, Chandler G, Rakheja D, et al. High-resolution array CGH<br />
identifies common mechanisms that drive embryonal rhabdomyosarcoma<br />
pathogenesis. Genes Chromosomes Cancer. 2011;50:397-408.<br />
23. Zibat A, Missiaglia E, Rosenberger A, et al. Activation <strong>of</strong> the hedgehog<br />
pathway confers a poor prognosis in embryonal and fusion gene-negative<br />
alveolar rhabdomyosarcoma. Oncogene. 2010;29:6323-6330.<br />
24. Pressey JG, Anderson JR, Crossman DK, et al. Hedgehog pathway<br />
activity in pediatric embryonal rhabdomyosarcoma and undifferentiated<br />
sarcoma: A report from the Children’s <strong>Oncology</strong> Group. Pediatr Blood Cancer.<br />
2011;57:930-938.<br />
25. Rubin BP, Nishijo K, Chen HI, et al. Evidence for an unanticipated<br />
relationship between undifferentiated pleomorphic sarcoma and embryonal<br />
rhabdomyosarcoma. Cancer Cell. 2011;19:177-191.<br />
26. Ecke I, Rosenberger A, Obenauer S, et al. Cyclopamine treatment <strong>of</strong><br />
full-blown Hh/Ptch-associated RMS partially inhibits Hh/Ptch signaling, but<br />
not tumor growth. Mol Carcinog. 2008;47:361-372.<br />
27. Souzaki R, Tajiri T, Souzaki M, et al. Hedgehog signaling pathway in<br />
neuroblastoma differentiation. J Pediatr Surg. 2010;45:2299-2304.<br />
28. Gershon TR, Shirazi A, Qin LX, et al. Enteric neural crest differentiation<br />
in ganglioneuromas implicates Hedgehog signaling in peripheral neuroblastic<br />
tumor pathogenesis. PLoS One. 2009;4(10):e7491.<br />
29. Eichenmüller M, Gruner I, Hagl B, et al. Blocking the hedgehog<br />
pathway inhibits hepatoblastoma growth. Hepatology. 2009;49:482-490.<br />
30. Hirotsu M, Setoguchi T, Sasaki H, et al. Smoothened as a new<br />
therapeutic target for human osteosarcoma. Molecular Cancer. 2010;9:5.<br />
609
PEDIATRIC ONCOLOGY <strong>EDUCATIONAL</strong> SESSION IN<br />
HONOR OF DR. JAMES NACHMAN<br />
CHAIR<br />
Stephen P. Hunger, MD<br />
University <strong>of</strong> Colorado Denver School <strong>of</strong> Medicine<br />
Aurora, CO<br />
SPEAKERS<br />
Melissa M. Hudson, MD<br />
St. Jude Children’s Research Hospital<br />
Memphis, TN<br />
Carola A. S. Arndt, MD<br />
Mayo Clinic<br />
Rochester, MN
Development and Refinement <strong>of</strong> Augmented<br />
Treatment Regimens for Pediatric High-Risk<br />
Acute Lymphoblastic Leukemia<br />
Overview: The 5-year survival rate for children and adolescents<br />
with acute lymphoblastic leukemia (ALL) is now at least<br />
90%. However, clinical features (age and initial white blood<br />
cell count [WBC]), early treatment response, and the presence/absence<br />
<strong>of</strong> specific sentinel genomic lesions can identify<br />
subsets <strong>of</strong> high-risk (HR) ALL patients with a much higher risk<br />
<strong>of</strong> treatment failure. Chemotherapy regimens used to treat HR<br />
ALL have been refined over the past 3 decades through<br />
randomized clinical trials conducted by the Children’s <strong>Oncology</strong><br />
Group (COG) in North America and the Berlin-Frankfurt-<br />
Muenster (BFM) group in Western Europe. Contemporary COG<br />
HR ALL treatment regimens were developed from the BFM-76<br />
regimen, with subsequent changes that led to development<br />
and refinement <strong>of</strong> a so-called augmented BFM (ABFM) regimen<br />
used today. Although contemporary COG and BFM treatment<br />
ALL IS the most common cancer that occurs in children<br />
and adolescents younger than age 20, comprising<br />
approximately 25% <strong>of</strong> malignancies occurring before age 15<br />
and 12% <strong>of</strong> those occurring in adolescents aged 15 to 20. 1<br />
ALL was virtually incurable until the early- to mid-1960s,<br />
with less than a 10% survival rate as recently as the late<br />
1960s. A recent large review <strong>of</strong> results <strong>of</strong> COG clinical trials<br />
showed 5-year survival rates <strong>of</strong> 90% for more than 7,000<br />
children diagnosed with ALL between 2000 and 2005, and it<br />
is anticipated that the long-term survival rate for those<br />
diagnosed between 2006 and 2010 will approach or exceed<br />
90%. 2 Despite these dramatic improvements in survival,<br />
clinical features, early-treatment response characteristics,<br />
and the presence/absence <strong>of</strong> specific favorable- or poor-risk<br />
sentinel genetic lesions in the leukemia cells can be used to<br />
identify patient subsets at higher risk <strong>of</strong> relapse. 3 Powerful<br />
predictive characteristics include the combination <strong>of</strong> age and<br />
WBC at initial diagnosis, which together are used to define<br />
the so-called NCI/Rome standard risk (SR; age 1 to 9.99<br />
years and WBC less than 50,000/microliter) and HR (age �<br />
1or�10 years and/or WBC � 50,000 microliter) groups. 4<br />
The second powerful predictor <strong>of</strong> outcome is initial treatment<br />
response, measured either by response to a 1-week<br />
prednisone (PRED) (plus a single dose <strong>of</strong> intrathecal methotrexate<br />
[IT MTX]) prephase (also termed prophase), bone<br />
marrow morphology after 1 to 2 weeks <strong>of</strong> multiagent chemotherapy,<br />
or measurements <strong>of</strong> minimal residual disease<br />
(MRD) at end <strong>of</strong> induction and/or consolidation therapy. The<br />
MRD response is the most powerful single prognostic factor<br />
and is now used by most groups in North America and<br />
Western Europe to modulate the intensity <strong>of</strong> postinduction<br />
therapy. 5,6 Leukemia genetics also provides critical prognostic<br />
information, with ETV6-RUNX1 (TEL-AML1) fusion and<br />
hyperdiploidy and/or favorable chromosome trisomies recognized<br />
as favorable features, and BCR-ABL1 fusion (Phpositive<br />
ALL) or MLL translocations generally considered to<br />
be adverse features that merit specific therapies and/or<br />
intensified treatment. 3<br />
HR ALL is generally treated with more intensive therapies<br />
than SR ALL. The COG uses an ABFM baseline<br />
By Stephen P. Hunger, MD<br />
regimens are not identical, there are many more similarities<br />
than differences. With improvements in survival, it has become<br />
clear that although the outcome <strong>of</strong> some patients with<br />
HR ALL can be improved by optimizing use <strong>of</strong> standard<br />
cytotoxic chemotherapy agents, this approach has had only<br />
limited success for other patient subsets. In contrast, introduction<br />
<strong>of</strong> the tyrosine kinase inhibitor imatinib has led to<br />
dramatic outcome improvements for children and adolescents<br />
with Philadelphia chromosome–positive ALL. Genomic studies<br />
are identifying new sentinel genomic lesions that can serve as<br />
potential therapeutic targets, which will likely lead to the<br />
testing <strong>of</strong> novel and/or targeted therapies in more children<br />
with HR ALL. Such studies will require increased collaboration<br />
between Western European and North <strong>American</strong> cooperative<br />
groups.<br />
treatment regimen 7 that bears significant similarity to, but<br />
has fundamental differences from, the BFM regimen developed<br />
in the early 1970s by Riehm and colleagues in Berlin. 8<br />
Table 1 provides a comparison <strong>of</strong> different contemporary<br />
BFM and COG treatment regimens. Different therapies are<br />
used for certain HR subtypes <strong>of</strong> ALL, most notably Phpositive<br />
ALL that is treated with intensive chemotherapy<br />
and tyrosine kinase inhibitors (TKI) such as imatinib or<br />
dasatinib. 9<br />
Development and Refinement <strong>of</strong> the BFM Treatment<br />
Regimen for ALL<br />
In 1977, Riehm and colleagues reported outstanding early<br />
results <strong>of</strong> treatment outcome with an intensive 8-week,<br />
eight-drug induction regimen, cranial irradiation, and maintenance<br />
therapy. 8 The 8-week induction was later termed<br />
protocol I and included two distinct phases. Protocol Ia is<br />
what the COG calls a four-drug induction and included<br />
PRED, asparaginase, vincristine, and daunorubicin. This<br />
was immediately followed by protocol Ib (the COG consolidation),<br />
which included cyclophosphamide, repeated low<br />
doses <strong>of</strong> ara-C, 6-mercaptopurine (6-MP), four weekly doses<br />
<strong>of</strong> IT MTX, and cranial irradiation (18 Gy).<br />
Subsequently, the BFM 76/79 study showed that outcome,<br />
particularly for patients with a high WBC, could be improved<br />
when protocol I was repeated with some modifications<br />
designed to minimize drug resistance (replacement <strong>of</strong><br />
PRED with dexamethasone [DEX], daunorubicin with doxorubicin,<br />
and 6-MP with 6-thioguanine). This was termed<br />
protocol II (delayed intensification [DI] in COG terminology).<br />
The 10-year disease-free survival rate was 67% for<br />
patients treated in the BFM 76/79 trial with protocols I and<br />
From the Children’s Hospital Colorado, Aurora, CO; Department <strong>of</strong> Pediatrics, University<br />
<strong>of</strong> Colorado School <strong>of</strong> Medicine, Aurora, CO.<br />
Author’s disclosures <strong>of</strong> potential conflicts <strong>of</strong> interest are found at the end <strong>of</strong> this article.<br />
Address reprint requests to Stephen P. Hunger, MD, Center for Cancer and Blood<br />
Disorders, Children’s Hospital Colorado, 13123 East 16th Ave., Box B115, Aurora, CO<br />
80045; email: stephen.hunger@childrenscolorado.org.<br />
© <strong>2012</strong> by <strong>American</strong> <strong>Society</strong> <strong>of</strong> <strong>Clinical</strong> <strong>Oncology</strong>.<br />
1092-9118/10/1-10<br />
611
II separated by an 8-week interim maintenance (IM) phase<br />
followed by maintenance therapy. 10<br />
Subsequently, the BFM has refined therapy to include the<br />
PRED prephase and replaced the 8-week IM phase with<br />
protocol M that consists <strong>of</strong> four courses <strong>of</strong> high-dose methotrexate<br />
(HD MTX; 5 g/m 2 over 24 hours followed by leucovorin<br />
rescue) given at 2-week intervals, with outstanding<br />
contemporary outcomes. 6 A fundamental difference between<br />
BFM and COG regimens is that the BFM regimens do not<br />
include steroid/vincristine pulses during maintenance. 11 Patients<br />
identified to be at HR <strong>of</strong> relapse based on a poor<br />
response to the PRED prephase (more than 1000 blasts/<br />
microliter at day 8), adverse cytogenetics, or more recently<br />
high levels <strong>of</strong> MRD at end induction/consolidation receive<br />
KEY POINTS<br />
Table 1. Comparison <strong>of</strong> Components <strong>of</strong> BFM and COG ALL Therapies<br />
Treatment Phase BFM BFM HR COG SR ALL COG HR ALL (hABFM)<br />
Induction (protocol Ia) 4 drugs/4 wks<br />
PRED, Vcr, ASNase, Dauno<br />
Consolidation (protocol Ib) 6 wks<br />
CPM, Ara-C, 6-MP<br />
4 drugs/4 wks<br />
PRED, Vcr, ASNase, Dauno<br />
6 wks<br />
CPM, Ara-C<br />
● Acute lymphoblastic leukemia (ALL) is the most<br />
common pediatric cancer.<br />
● Five-year survival rates for pediatric ALL now exceed<br />
90%, as compared to less than 10% in the 1960s.<br />
● Contemporary Children’s <strong>Oncology</strong> Group ALL treatment<br />
regimens are derived from regimens developed<br />
by the Berlin-Frankfurt-Muenster group in the 1970s<br />
but have been refined through a series <strong>of</strong> randomized<br />
clinical trials.<br />
● Optimizing the use <strong>of</strong> standard cytotoxic chemotherapy<br />
agents has led to significant outcome improvements<br />
for some subsets <strong>of</strong> high-risk ALL patients,<br />
whereas other subsets have benefitted little from this<br />
approach.<br />
● Novel and/or targeted therapies have benefitted children<br />
and adolescents with Philadelphia chromosome–positive<br />
ALL and will likely be tested in other<br />
high-risk ALL subsets.<br />
3 drugs/4 wks; DEX, Vcr,<br />
ASNase<br />
4 wks<br />
6-MP, MTX<br />
IM #1 HD MTX Three intensive 3-wk HR blocks 8 wks<br />
Capizzi MTX, Dex, Vcr<br />
DI #1 (protocol II) 8 wks<br />
DEX, Vcr, ASNase, Doxo, CPM,<br />
Ara-C, 6-TG<br />
IM #2 Not given 4 wks<br />
6-MP, MTX<br />
8 wks<br />
DEX, Vcr, ASNase, Doxo, CPM,<br />
Ara-C, 6-TG<br />
8 wks<br />
DEX, Vcr, ASNase, Doxo, CPM,<br />
Ara-C, 6-TG<br />
8 wks<br />
Capizzi MTX, DEX, Vcr<br />
DI #2 (protocol II) Not given 8 wks<br />
DEX, Vcr, ASNase, Doxo, CPM,<br />
Ara-C, 6-TG<br />
Maintenance 6-MP, MTX 6-MP, MTX 6-MP, MTX plus monthly<br />
DEX/Vcr pulses<br />
Not given Not given<br />
4 drugs/4 wks; PRED (age 10�) or<br />
DEX (age 1–9.99), Vcr, ASNase,<br />
Dauno<br />
8 wks (augmented)<br />
CPM, Ara-C, 6-MP, Vcr, ASNase<br />
8 wks<br />
HD MTX, Vcr<br />
8 wks (augmented)<br />
DEX, Vcr, ASNase, Doxo, CPM,<br />
Ara-C, 6-TG<br />
8 wks<br />
Capizzi MTX � ASNase, DEX, Vcr,<br />
(some not all patients)<br />
6-MP, MTX plus monthly PRED/Vcr<br />
pulses<br />
Treatment duration 30 mo 30 mo 27 mo, females; 39 mo, males 27 mo, females; 39 mo, males<br />
Abbreviations: ALL, acute lymphoblastic leukemia; ASNase, asparaginase; BFM, Berlin-Frankfurt-Muenster; COG, Children’s <strong>Oncology</strong> Group; CPM, cyclophosphamide;<br />
Dauno, daunorubicin; DEX, dexamethasone; DI, delayed intensification; Doxo, doxorubicin; hABFM, hemi-augmented BFM; HR, high risk; IM, interim<br />
maintenance; mo, months; MTX, methotrexate; PRED, prednisone; SR, standard risk; Vcr, vincristine; wks, weeks; 6-MP, 6-mercaptopurine; 6-TG, 6-thioguanine.<br />
612<br />
more intensive therapy in which protocol M is replaced by<br />
three intensive multiagent chemotherapy blocks (HR blocks)<br />
and protocol II is given twice. 12<br />
CCG/COG Adoption <strong>of</strong> BFM-76-Based Therapy<br />
STEPHEN P. HUNGER<br />
The Children’s Cancer Group (CCG) recognized in the<br />
1980s that outcomes reported by the BFM were superior to<br />
those obtained in CCG trials with less intensive therapy.<br />
This led to development <strong>of</strong> two pivotal trials that compared<br />
the CCG regimens <strong>of</strong> that time to the BFM-76 regimen,<br />
which did not include a PRED prephase or protocol M.<br />
Because <strong>of</strong> this, CCG and COG ALL treatment regimens<br />
resemble but do not recapitulate BFM therapies, although<br />
they are now becoming more similar, with COG adoption <strong>of</strong><br />
HD MTX (see below).<br />
The CCG 105 trial (1983 to 1989) was designed for<br />
children with intermediate-risk ALL, which is similar but<br />
not identical to NCI SR ALL. 13 Children enrolled in CCG<br />
105 were randomly selected to receive the standard CCG<br />
regimen <strong>of</strong> that time, which included a three-drug induction<br />
(PRED, vincristine, and asparaginase but no daunorubicin)<br />
followed by central nervous system (CNS) control and maintenance,<br />
or the BFM regimen that included protocols I and<br />
II. Two other randomly selected groups received protocol I<br />
followed by CCG postinduction therapy or the CCG induction/consolidation<br />
followed by protocol II. The results <strong>of</strong> this<br />
study (updated in 2005 with 16-year follow-up) showed that<br />
all three <strong>of</strong> the BFM-based regimens were superior to the<br />
CCG regimen. 13,14 There was little outcome difference between<br />
the three BFM-based regimens; thus, the CCG selected<br />
the CCG induction/consolidation followed by protocol<br />
II to bring forward in subsequent trials, as that regimen had<br />
less short-term toxicity than the other two BFM-based<br />
regimens. For this reason, all subsequent CCG and COG<br />
regimens for SR ALL have included a three-drug induction<br />
(without the intensive consolidation phase) and a DI phase.<br />
Based on the results <strong>of</strong> the CCG 1922 study, children with<br />
SR ALL enrolled in COG ALL trials now receive DEX rather
ABFM THERAPY FOR PEDIATRIC ALL<br />
than PRED during induction and in the steroid/vincristine<br />
pulses given every 4 weeks during maintenance therapy. 15<br />
The CCG 1991 trial showed that Capizzi escalating intravenous<br />
(IV) MTX without rescue (no asparaginase) was superior<br />
to oral 6-MP and MTX during the IM phase(s), 16 and<br />
this approach is now standard in contemporary COG SR<br />
ALL trials.<br />
In parallel to CCG 105, the CCG 106 trial (1983 to 1987)<br />
compared BFM therapy to the contemporary CCG HR regimen<br />
and to the more intensive “New York” regimen. 17<br />
Better outcomes were seen with both the BFM and New<br />
York regimens as compared to the CCG regimen, but the<br />
BFM regimen was associated with less toxicity and lower<br />
cumulative doses <strong>of</strong> chemotherapy than the New York regimen.<br />
Based on the results <strong>of</strong> CCG 106, the BFM-76-based<br />
regimen became the baseline CCG/COG regimen used for<br />
HR ALL, and subsequent trials included a four-drug induction,<br />
the intensive consolidation (protocol Ib), and one or<br />
more protocol II (DI) phases.<br />
Development <strong>of</strong> Refinement <strong>of</strong> ABFM Therapy<br />
by the CCG/COG<br />
The CCG recognized that a poor early response to chemotherapy<br />
was a strong adverse prognostic factor. 18 In CCG<br />
1882 (1991 to 1995), children with HR ALL and a slow early<br />
response to therapy (SER; more than 25% marrow blasts at<br />
day 8 <strong>of</strong> induction) were randomly selected to received the<br />
CCG-modified BFM regimen or an ABFM regimen that<br />
contained a number <strong>of</strong> changes to baseline treatment. 7<br />
These changes included intensifying consolidation (Ib) therapy<br />
by extending it to 8 weeks and including doses <strong>of</strong><br />
vincristine and asparaginase during the neutropenic phases<br />
that followed cyclophosphamide and ara-C, using Capizzi I<br />
escalating IV MTX without leucovorin rescue plus asparaginase<br />
during the 8-week IM #1 phase, giving second IM and<br />
DI phases, and giving prophylactic cranial irradiation to all<br />
patients. The ABFM regimen produced results that were<br />
significantly better than the standard CCG regimen, with<br />
5-year event-free survival (EFS) rates <strong>of</strong> 75% versus 55%<br />
(p � 0.001) and overall survival (OS) rates <strong>of</strong> 78% versus<br />
67% (p � 0.02). 7<br />
Based on the results <strong>of</strong> CCG 1882, the subsequent CCG<br />
1961 HR-ALL trial tested the ABFM regimen in patients<br />
with a rapid early response to therapy (day 8 marrow blasts<br />
less than or equal to 25%) and attempted to determine the<br />
most important components <strong>of</strong> ABFM therapy by randomly<br />
selecting patients in a2x2manner to receive the baseline<br />
regimen, the full ABFM regimen, the baseline regimen with<br />
2 IM and DI phases, or the ABFM regimen with only single<br />
IM and DI phases (termed hemi-ABFM, or hABFM). 19 The<br />
results <strong>of</strong> CCG 1961 showed that the augmented parts <strong>of</strong><br />
ABFM therapy were critical and improved 5-year EFS (81%<br />
vs. 72%; p � 0.001) and OS (89% vs. 83%; p � 0.003)<br />
significantly. Equally important, repeating the IM and DI<br />
phases did not improve outcome (5-year EFS: 76% vs. 76.8%<br />
for single compared with double IM/DI; p � 0.94). Based on<br />
these results, the hABFM regimen with single IM/DI phases<br />
became the standard regimen for children with HR ALL and<br />
a good response to induction therapy in COG trials, although<br />
a second IM and/or DI phase has been retained in some<br />
trials for those with a poor early response.<br />
COG AALL0232 Shows Superiority <strong>of</strong> HD MTX<br />
The COG AALL0232 (2003 to 2011) trial for children and<br />
adolescents with HR B-cell precursor (BCP) ALL used the<br />
hABFM regimen and had a2x2randomization to 28 days<br />
<strong>of</strong> PRED 60 mg/m 2 /day versus 14 days <strong>of</strong> DEX 10 mg/m 2 /day<br />
during induction and to HD MTX versus Capizzi MTX plus<br />
asparaginase during IM #1. Accrual to this trial was stopped<br />
in early 2011 when results for the MTX randomization<br />
crossed predefined monitoring boundaries. Eric Larsen, the<br />
COG AALL0232 study chair, presented results <strong>of</strong> the MTX<br />
randomization at the plenary session <strong>of</strong> the 2011 ASCO<br />
annual meeting. 20 Planned interim monitoring showed that<br />
the 5-year EFS for patients randomly selected to receive<br />
HD-MTX (1,209 patients) was 82% � 3.4% vs. 75.4% � 3.6%<br />
for the Capizzi MTX plus asparaginase (1,217 patients)<br />
regimen, p � 0.006. Based on these practice-changing results,<br />
the COG now considers HD MTX to be the standard <strong>of</strong><br />
care for BCP HR ALL, further increasing the symmetry<br />
between COG and BFM regimens.<br />
The results <strong>of</strong> the induction steroid randomization were<br />
more complex. In 2008, the COG stopped the steroid randomization<br />
in children age 10 or older because <strong>of</strong> an increased<br />
incidence <strong>of</strong> osteonecrosis (ON) among patients <strong>of</strong><br />
this age treated with DEX. Because the rate <strong>of</strong> ON was quite<br />
low in children younger than age 10 and there was no<br />
difference in ON rates between the two steroid arms, the<br />
randomization continued for younger patients. There was an<br />
interaction between the steroid and MTX regimens among<br />
patients younger than age 10, so when the outcome for<br />
children randomly selected to receive DEX versus PRED<br />
was compared, the analysis was limited to those randomly<br />
selected to receive HD MTX. 21 The younger children who<br />
received DEX/HD MTX had a superior outcome to those who<br />
received PRED/HD MTX (5-year EFS: 93.7% � 5.4% vs.<br />
81.2% � 7.7%; p � 0.03). Retrospective review <strong>of</strong> the<br />
patients age 10 or older who were randomly selected to<br />
receive DEX versus PRED during the initial 4 years <strong>of</strong> the<br />
study showed no differences in outcome (5-year EFS: 74.7%<br />
� 4.6% and 76.5% � 4.6%, respectively; p � 0.80) and<br />
confirmed the higher rate <strong>of</strong> ON seen in the patients age 10<br />
or older who were treated with DEX (24.3% vs. 15.1%; p �<br />
0.0007). Based on these results, 14 days <strong>of</strong> DEX 10 mg/m 2 /<br />
day on days 1 through 14 is now considered the standard<br />
regimen for children younger than age 10 enrolled in COG<br />
HR BCP ALL trials, whereas PRED 60 mg/m 2 /day on days 1<br />
through 28 is considered the standard for those age 10 or<br />
older.<br />
COG T-ALL Trials<br />
Because <strong>of</strong> differences between T-cell ALL (T-ALL) and<br />
BCP ALL in biology, treatment response, prognostic factors,<br />
and outcome, the COG is conducting separate trials for BCP<br />
and T-ALL, whereas the BFM group has a single trial that<br />
includes both T-ALL and BCP ALL. The backbone therapy<br />
for the COG AALL0434 T-ALL study (opened to patient<br />
accrual in 2007 and currently projected to complete accrual<br />
in 2014) is the PRED/Capizzi MTX plus asparaginase arm <strong>of</strong><br />
AALL0232 (see above). The study isa2x2randomized trial.<br />
The first comparison is the identical HD MTX versus Capizzi<br />
MTX plus asparaginase randomization conducted in<br />
AALL0232. Careful consideration was given to the results <strong>of</strong><br />
AALL0232 when they became available, with a decision to<br />
613
continue this randomization based on the different biology <strong>of</strong><br />
BCP and T-ALL (the reason that the question was asked<br />
separately in two parallel trials in the first place). Interestingly,<br />
the best reported contemporary results in pediatric<br />
T-ALL come from the UKALL 2003 trial, which used essentially<br />
a COG ABFM backbone regimen with Capizzi MTX<br />
plus asparaginase during IM. 22<br />
The second randomization in AALL0434 is to receive<br />
versus not receive six 5-day courses <strong>of</strong> nelarabine during the<br />
first �16 months <strong>of</strong> therapy. Nelarabine, a prodrug <strong>of</strong> ara-G,<br />
showed substantial clinical activity in phase I/II T-ALL<br />
trials, but was associated with significant, <strong>of</strong>ten severe, and<br />
sometimes fatal CNS toxicity. 23,24 The COG then conducted<br />
a pilot study <strong>of</strong> nelarabine plus BFM-86 type chemotherapy<br />
in children with newly diagnosed T-ALL and found encouraging<br />
activity but much lower rates <strong>of</strong> overall and severe<br />
toxicity than had been seen in relapsed patients treated in<br />
the phase I/II trials. 25 AALL0434 included a pilot safety<br />
phase during which only an HR subset <strong>of</strong> T-ALL patients<br />
with a poor response was randomly selected to be treated<br />
with/without nelarabine. This safety phase was concluded<br />
with no significant differences in toxicity among the patients<br />
treated with/without nelarabine, and AALL0434 is now in<br />
the efficacy phase in which �90% <strong>of</strong> patients are randomly<br />
selected to be treated with/without nelarabine (all except for<br />
a tightly defined low-risk patient subset). 26<br />
VHR ALL<br />
A small subset <strong>of</strong> pediatric ALL patients can be defined as<br />
being at very high risk (VHR) <strong>of</strong> relapse. The definition <strong>of</strong><br />
this subgroup is constantly evolving, but there is a general<br />
consensus that alternative treatment strategies may be<br />
indicated for these patients in comparison with treatment<br />
given to other patients with HR ALL. As discussed above,<br />
the BFM group has termed this group HR and uses intensive<br />
multiagent chemotherapy HR blocks, an additional protocol<br />
II for these patients, with hematopoietic stem cell transplant<br />
(HSCT) in CR1 for the subset with a poor treatment<br />
response at end <strong>of</strong> protocol Ib. The COG AALL0031 pilot<br />
study (2002 to 2006) tested an intensive multiagent chemotherapy<br />
regimen in children with VHR ALL, which included<br />
Ph-positive ALL, hypodiploidy (chromosome number less<br />
than 44), poor response to induction therapy (more than 25%<br />
marrow blasts at day 29 or 5% to 25% marrow blasts/more<br />
than 1% MRD at day 43 after 2 weeks <strong>of</strong> extended induction<br />
therapy), or MLL translocation and an SER (more than 5%<br />
blasts at day 15 <strong>of</strong> induction). 9,27 The results were encouraging<br />
for patients with Ph-negative VHR ALL, with a<br />
nonsignificant trend toward improved outcome for those<br />
treated with HSCT. 27 The study design with all patients<br />
receiving identical chemotherapy and those with Ph-positive<br />
ALL also receiving imatinib facilitated assessment <strong>of</strong> imatinib<br />
toxicity. Notably, the addition <strong>of</strong> imatinib was quite<br />
safe and did not lead to increased toxicity. 9 More importantly,<br />
adding imatinib led to major improvements in early<br />
614<br />
outcome for patients with Ph-positive ALL, with a 3-year<br />
EFS rate <strong>of</strong> 80% � 11% (95% CI, 64% to 90%), which was<br />
more than twice that <strong>of</strong> historic controls treated in the<br />
preimatinib era (35% � 4%; p � 0.0001). There was no<br />
apparent advantage to HSCT as compared with intensive<br />
chemotherapy plus imatinib. These results have been stable<br />
with longer follow-up 28 and have led to changes in clinical<br />
practice in treatment <strong>of</strong> children with Ph-positive ALL.<br />
The COG has now completed enrollment (2008 to <strong>2012</strong>) on<br />
a successor Ph-positive ALL trial (AALL0622) that utilized<br />
the AALL0031 chemotherapy backbone with imatinib replaced<br />
with dasatinib, a more potent second-generation Abl<br />
TKI. The study showed that it was safe to add continuous<br />
dasatinib treatment (60 mg/m 2 /day) to this intensive chemotherapy<br />
regimen; it is still much too early to assess treatment<br />
efficacy. In <strong>2012</strong>, the COG and European EsPhALL<br />
groups will commence enrollment on a joint pilot study (developed<br />
in conjunction with Bristol Myers Squibb) that will test<br />
the safety and efficacy <strong>of</strong> dasatinib added to the BFM HR<br />
treatment backbone, which has lower cumulative doses <strong>of</strong><br />
many chemotherapy agents than the AALL0031 regimen.<br />
Future Perspectives<br />
STEPHEN P. HUNGER<br />
The substantial improvements that have occurred in outcome<br />
for pediatric HR ALL over the past few decades have<br />
been accompanied by recognition that there are a number <strong>of</strong><br />
different ALL subsets that come under the umbrella term<br />
HR ALL. For some subsets, outcomes have been improved<br />
by optimizing delivery <strong>of</strong> chemotherapy agents that have<br />
been in widespread clinical use for the past 25 years, as<br />
exemplified by the results <strong>of</strong> COG AALL0232. For other<br />
subsets, such as Ph-positive ALL, optimizing traditional<br />
cytotoxic chemotherapy agents led to limited improvements<br />
in outcome, 29 but introduction <strong>of</strong> new and/or targeted therapies<br />
had a major effect on outcome. 9 It is expected that<br />
ongoing genomic studies will lead to recognition <strong>of</strong> additional<br />
VHR ALL subsets defined by the presence <strong>of</strong> specific<br />
sentinel genomic lesions that can potentially be targeted by<br />
novel therapies. 30 The rarity <strong>of</strong> these and other VHR ALL<br />
subsets will require increased collaborations between North<br />
<strong>American</strong> and Western European investigators in order to<br />
test the efficacy <strong>of</strong> novel/targeted therapies in these patient<br />
subsets.<br />
Acknowledgments<br />
The author dedicates this manuscript to the memory <strong>of</strong> Jim<br />
Nachman, who passed away unexpectedly, and far too early, in<br />
2011. Jim developed the ABFM regimen; played a major role in<br />
developing productive, fruitful interactions between the CCG/<br />
COG and BFM groups; and shaped the design <strong>of</strong> most COG ALL<br />
trials <strong>of</strong> the past 25 years. He was widely recognized as an<br />
international leader in pediatric oncology clinical research and<br />
as a mentor to many investigators. The author and the rest <strong>of</strong><br />
the pediatric oncology community miss his advice, his infectious<br />
enthusiasm, his laugh and politically incorrect sense <strong>of</strong> humor,<br />
and most <strong>of</strong> all his friendship.
ABFM THERAPY FOR PEDIATRIC ALL<br />
Author’s Disclosures <strong>of</strong> Potential Conflicts <strong>of</strong> Interest<br />
Author<br />
Employment or<br />
Leadership<br />
Positions<br />
Consultant or<br />
Advisory Role<br />
Stephen P. Hunger Bristol-Myers<br />
Squibb (U);<br />
Genzyme (U)<br />
1. Ries LA, Smith MA, Gurney JG, et al (eds). Cancer Incidence and<br />
Survival among Children and Adolescents: United States SEER Program<br />
1975-1995. Bethesda, MD: National Cancer Institute; 1999.<br />
2. Hunger SP, Lu X, Devidas M, et al. Improved survival for children and<br />
adolescents from 1990-2005: a report from the Children’s <strong>Oncology</strong> Group.<br />
J Clin Oncol. In press.<br />
3. Schultz KR, Pullen DJ, Sather HN, et al. Risk- and response-based<br />
classification <strong>of</strong> childhood B-precursor acute lymphoblastic leukemia: a combined<br />
analysis <strong>of</strong> prognostic markers from the Pediatric <strong>Oncology</strong> Group<br />
(POG) and Children’s Cancer Group (CCG). Blood. 2007;109:926-935.<br />
4. Smith M, Arthur D, Camitta B, et al. Uniform approach to risk<br />
classification and treatment assignment for children with acute lymphoblastic<br />
leukemia. J Clin Oncol. 1996;14:18-24.<br />
5. Borowitz MJ, Devidas M, Hunger SP, et al. <strong>Clinical</strong> significance <strong>of</strong><br />
minimal residual disease in childhood acute lymphoblastic leukemia and its<br />
relationship to other prognostic factors: a Children’s <strong>Oncology</strong> Group study.<br />
Blood. 2008;111:5477-5485.<br />
6. Conter V, Bartram CR, Valsecchi MG, et al. Molecular response to<br />
treatment redefines all prognostic factors in children and adolescents with<br />
B-cell precursor acute lymphoblastic leukemia: results in 3184 patients <strong>of</strong> the<br />
AIEOP-BFM ALL 2000 study. Blood. 2010;115:3206-3214.<br />
7. Nachman JB, Sather HN, Sensel MG, et al. Augmented post-induction<br />
therapy for children with high-risk acute lymphoblastic leukemia and a slow<br />
response to initial therapy. N Engl J Med. 1998;338:1663-1671.<br />
8. Riehm H, Gadner H, Welte K. The West-Berlin therapy study <strong>of</strong> acute<br />
lymphoblastic leukemia in childhood—report after 6 years (author’s transl).<br />
Klin Padiatr. 1977;189:89-102.<br />
9. Schultz KR, Bowman WP, Aledo A, et al. Improved early event-free<br />
survival with imatinib in Philadelphia chromosome-positive acute lymphoblastic<br />
leukemia: a Children’s <strong>Oncology</strong> Group study. J Clin Oncol. 2009;27:<br />
5175-5181.<br />
10. Henze G, Langermann HJ, Bramswig J, et al. The BFM 76/79 acute<br />
lymphoblastic leukemia therapy study (author’s transl). Klin Padiatr. 1981;<br />
193:145-154.<br />
11. Conter V, Valsecchi MG, Silvestri D, et al. Pulses <strong>of</strong> vincristine and<br />
dexamethasone in addition to intensive chemotherapy for children with<br />
intermediate-risk acute lymphoblastic leukaemia: a multicentre randomised<br />
trial. Lancet. 2007;369:123-131.<br />
12. Stanulla M, Schrappe M. Treatment <strong>of</strong> childhood acute lymphoblastic<br />
leukemia. Semin Hematol. 2009;46:52-63.<br />
13. Tubergen DG, Gilchrist GS, O’Brien RT, et al. Improved outcome with<br />
delayed intensification for children with acute lymphoblastic leukemia and<br />
intermediate presenting features: a Childrens Cancer Group phase III trial.<br />
J Clin Oncol. 1993;11:527-537.<br />
14. Hunger SP, Winick NJ, Sather HN, et al. Therapy <strong>of</strong> low-risk subsets <strong>of</strong><br />
childhood acute lymphoblastic leukemia: when do we say enough? Pediatr<br />
Blood Cancer. 2005;45:876-880.<br />
15. Bostrom BC, Sensel MR, Sather HN, et al. Dexamethasone versus<br />
prednisone and daily oral versus weekly intravenous mercaptopurine for<br />
patients with standard-risk acute lymphoblastic leukemia: a report from the<br />
Children’s Cancer Group. Blood. 2003;101:3809-3817.<br />
16. Matloub Y, Bostrom BC, Hunger SP, et al. Escalating intravenous<br />
methotrexate improves event-free survival in children with standard-risk<br />
acute lymphoblastic leukemia: a report for the Children’s <strong>Oncology</strong> Group.<br />
Blood. 2011;118:243-251.<br />
Stock<br />
Ownership Honoraria<br />
Amgen (B);<br />
Bristol-Myers<br />
Squibb (I);<br />
Merck (B);<br />
Pfizer (B)<br />
REFERENCES<br />
Research<br />
Funding<br />
Expert<br />
Testimony<br />
Bristol-Myers<br />
Squibb; Genzyme<br />
(U)<br />
Other<br />
Remuneration<br />
17. Gaynon PS, Steinherz PG, Bleyer WA, et al. Improved therapy for<br />
children with acute lymphoblastic leukemia and unfavorable presenting<br />
features: a follow-up report <strong>of</strong> the Childrens Cancer Group Study CCG-106.<br />
J Clin Oncol. 1993;11:2234-2242.<br />
18. Gaynon PS, Desai AA, Bostrom BC, et al. Early response to therapy and<br />
outcome in childhood acute lymphoblastic leukemia: a review. Cancer. 1997;<br />
80:1717-1726.<br />
19. Seibel NL, Steinherz PG, Sather HN, et al. Early postinduction intensification<br />
therapy improves survival for children and adolescents with highrisk<br />
acute lymphoblastic leukemia: a report from the Children’s <strong>Oncology</strong><br />
Group. Blood. 2008;111:2548-2555.<br />
20. Larsen EC, Salzer WL, Devidas M, et al. Comparison <strong>of</strong> high-dose<br />
methotrexate (HD-MTX) with Capizzi methotrexate plus asparaginase (C-<br />
MTX/ASNase) in children and young adults with high-risk acute lymphoblastic<br />
leukemia (HR-ALL): a report from the Children’s <strong>Oncology</strong> Group study<br />
AALL0232. J Clin Oncol. 2011;29:6s (suppl; abstr 3).<br />
21. Winick NJ, Salzer WL, Devidas M, et al. Dexamethasone (DEX) versus<br />
prednisone (PRED) during induction for children with high risk acute<br />
lymphoblastic leukemia (HR-ALL): a report from the Children’s <strong>Oncology</strong><br />
Group study AALL0232. J Clin Oncol 2011;29:586s (suppl; abstr 9504).<br />
22. Vora A, Wade R, Mitchell CD, et al. Improved outcome for children and<br />
young adults with T-cell acute lymphoblastic leukaemia (ALL): results <strong>of</strong> the<br />
United Kingdome Medical Research Council (MRC) Trial UKALL 2003. Blood<br />
(ASH Annual Meeting Abstracts). 2008;112:908.<br />
23. Kurtzberg J, Ernst TJ, Keating MJ, et al. Phase I study <strong>of</strong> 506U78<br />
administered on a consecutive 5-day schedule in children and adults with<br />
refractory hematologic malignancies. J Clin Oncol. 2005;23:3396-3403.<br />
24. Berg SL, Blaney SM, Devidas M, et al. Phase II study <strong>of</strong> nelarabine<br />
(compound 506U78) in children and young adults with refractory T-cell<br />
malignancies: a report from the Children’s <strong>Oncology</strong> Group. J Clin Oncol.<br />
2005;23:3376-3382.<br />
25. Dunsmore K, Devidas M, Borowitz MJ, et al. Nelarabine in combination<br />
with intensive modified BFM AALL00P2: a pilot study for the treatment<br />
<strong>of</strong> high-risk T-ALL a report from the Children’s <strong>Oncology</strong> Group. J Clin Oncol.<br />
2008;26:539s (suppl; abstr 10002).<br />
26. Winter SS, Devidas M, Wood BL, et al. Nelerabine may be safely<br />
incorporated into a phase III study for newly diagnosed T-lineage acute<br />
lymphoblastic leukemia: a report from the Children’s <strong>Oncology</strong> Group. Blood<br />
(ASH Annual Meeting Abstracts). 2010;116:865a.<br />
27. Schultz KR, Bowman WP, Slayton WB, et al. Philadelphia chromosome<br />
negative (Ph-) very high risk (VHR) acute lymphoblastic leukemia (ALL) in<br />
children and adolescents: the impact <strong>of</strong> intensified chemotherapy on early<br />
event free survival (EFS) in Children’s <strong>Oncology</strong> Group study AALL0031.<br />
Blood (ASH Annual Meeting Abstracts). 2008;112:911.<br />
28. Schultz KR, Bowman WP, Aledo A, et al. Continuous dosing imatinib<br />
with intensive chemotherapy gives equivalent outcomes to allogeneic BMT for<br />
Philadelphia chromosome-positive (Ph�) acute lymphoblastic leukemia<br />
(ALL) with longer term follow up: Updated results <strong>of</strong> Children’s <strong>Oncology</strong><br />
Group (COG) AALL0031. Pediatr Blood Cancer. 2010;54:788.<br />
29. Aricò M, Schrappe M, Hunger SP, et al. <strong>Clinical</strong> outcome <strong>of</strong> children<br />
with newly diagnosed Philadelphia chromosome-positive acute lymphoblastic<br />
leukemia treated between 1995 and 2005. J Clin Oncol. 2010;28:4755-4761.<br />
30. Hunger SP, Raetz EA, Loh ML, et al. Improving outcomes for high-risk<br />
ALL: translating new discoveries into clinical care. Pediatr Blood Cancer.<br />
2001;56:984-993.<br />
615
Refining the Role <strong>of</strong> Radiation Therapy in<br />
Pediatric Hodgkin Lymphoma<br />
By Melissa M. Hudson, MD, and Louis S. Constine, MD<br />
Overview: The role <strong>of</strong> radiation therapy in the treatment <strong>of</strong><br />
pediatric Hodgkin lymphoma has continued to be refined,<br />
motivated by the desire to avoid disruption to normal tissue<br />
development and function and secondary carcinogenesis.<br />
Such progress has occurred in tandem with modifications <strong>of</strong><br />
the multiagent chemotherapy regimens that have been used in<br />
place <strong>of</strong> or in combination with low-dose involved-field radiation<br />
that are also associated with dose-related risks <strong>of</strong><br />
cardiopulmonary and gonadal dysfunction and leukemogenesis.<br />
Consequently, treatment strategies for young patients,<br />
who have an excellent prognosis <strong>of</strong> long-term survival, utilizes<br />
RADIATION THERAPY (RT) has played a seminal role<br />
in improving outcomes for children and adolescents<br />
with Hodgkin lymphoma (HL). Following delineation <strong>of</strong> a<br />
tumoricidal dose, RT became the first modality to produce<br />
prolonged disease-free survival when delivered to consistent<br />
treatment fields <strong>of</strong> contiguous lymph nodes. Although curative<br />
for a large number <strong>of</strong> patients with localized disease, its<br />
combination with multiagent chemotherapy was needed to<br />
improve long-term survival in individuals with advancedstage<br />
and/or bulky node disease. Early RT approaches for<br />
children and adults prescribed high doses (35–44 Gy) to<br />
treatment volumes routinely extended to encompass adjacent<br />
uninvolved nodal regions. Recognition <strong>of</strong> the adverse<br />
effects <strong>of</strong> high-dose RT on musculoskeletal development in<br />
children motivated investigations <strong>of</strong> multiagent chemotherapy<br />
alone or in combination with lower radiation doses<br />
(15–25.5 Gy) to reduce treatment volumes (involvedfields).<br />
1,2 Increasing numbers <strong>of</strong> aging pediatric HL survivors<br />
consequently permitted recognition <strong>of</strong> the excess risk <strong>of</strong><br />
cardiovascular disease and secondary carcinogenesis associated<br />
with RT. 3,4 This knowledge led to the abandonment <strong>of</strong><br />
the use <strong>of</strong> RT as a single modality and its restricted use in<br />
contemporary trials.<br />
Research elucidating unique pr<strong>of</strong>iles <strong>of</strong> chemotherapyrelated<br />
toxicity in children subsequently guided the development<br />
<strong>of</strong> multiagent chemotherapy regimens that balanced<br />
the dose-related toxicity <strong>of</strong> anthracyclines, alkylating agents,<br />
and bleomycin, <strong>of</strong>ten by adding the modality <strong>of</strong> low-dose<br />
involved- field radiation therapy (IFRT). Results from these<br />
studies provide strong support for the responsiveness <strong>of</strong> HL<br />
to a variety <strong>of</strong> multiagent chemotherapy combinations that<br />
are largely derived from the original MOPP (mechlorethamine,<br />
vincristine, procarbazine, prednisone) 5 and ABVD<br />
(doxorubicin, bleomycin, vinblastine, dacarbazine) 6 combinations.<br />
The desire to maximize treatment efficacy and<br />
minimize its related long-term morbidity has increasingly<br />
focused attention on the role <strong>of</strong> RT in the treatment <strong>of</strong><br />
pediatric HL. This manuscript aims to review published<br />
data to define the optimal treatment strategies for children<br />
and adolescents with HL in regards to the use <strong>of</strong> RT.<br />
Prognostic Factors Used in Risk Designation <strong>of</strong><br />
Pediatric HL<br />
As therapy for pediatric HL becomes more effective, factors<br />
associated with outcome have become more difficult to<br />
616<br />
a risk-adapted approach that provides optimal efficacy for<br />
disease control whereas limiting toxicity associated with both<br />
radiation and chemotherapy. Because <strong>of</strong> the differences in<br />
age-related developmental status and gender-related sensitivity<br />
to chemotherapy and radiation toxicity, no single treatment<br />
approach is ideal for all pediatric patients. This<br />
manuscript summarizes results from published clinical trials<br />
with the goal <strong>of</strong> defining optimal treatment strategies for<br />
children and adolescents with Hodgkin lymphoma in regards<br />
to the use <strong>of</strong> radiation therapy.<br />
identify. Regardless, contemporary treatment for pediatric<br />
HL uses a risk-adapted and response-based paradigm that<br />
assigns the length and intensity <strong>of</strong> therapy based on diseaserelated<br />
factors such as stage, number <strong>of</strong> involved nodal<br />
regions, tumor bulk, the presence <strong>of</strong> B symptoms, and early<br />
response to chemotherapy by functional imaging. In addition<br />
to consideration <strong>of</strong> cancer-related factors that may<br />
permit therapy reduction or require dose intensification, the<br />
treatment approach may also consider unique host factors,<br />
such as age and gender, that may enhance the risk for<br />
specific treatment toxicities. Most protocols stratify groups<br />
according to low-, intermediate-, and high-risk designations.<br />
Low-risk clinical features typically include localized nodal<br />
involvement in the absence <strong>of</strong> B symptoms and bulky disease.<br />
Risk factors considered in other studies include the<br />
number <strong>of</strong> involved nodal regions, the presence <strong>of</strong> hilar<br />
adenopathy, the size <strong>of</strong> peripheral lymphadenopathy, and<br />
extranodal extension. High-risk clinical features include the<br />
presence <strong>of</strong> B symptoms, bulky mediastinal or peripheral<br />
lymphadenopathy, extranodal extension <strong>of</strong> disease, and advanced<br />
(stage IIIB- IV) disease. Bulky mediastinal lymphadenopathy<br />
is designated when the ratio <strong>of</strong> the maximum<br />
measurement <strong>of</strong> mediastinal lymphadenopathy to intrathoracic<br />
cavity on an upright chest radiograph equals or exceeds<br />
33%. Intermediate-risk features include localized<br />
disease (stages I, II, and IIIA) with unfavorable features;<br />
these cases may be treated similarly to advanced-stage<br />
disease in some treatment protocols or treated with therapy<br />
<strong>of</strong> intermediate intensity. Unfortunately, inconsistency in<br />
risk categorization across cooperative group studies <strong>of</strong>ten<br />
makes comparison <strong>of</strong> study outcomes challenging.<br />
Radiation Therapy for Low-Risk HL<br />
Considering the excellent survival rates achieved in children<br />
and adolescents with low-risk presentations <strong>of</strong> HL with<br />
From the Department <strong>of</strong> <strong>Oncology</strong>, Division <strong>of</strong> Cancer Survivorship, St. Jude’s Children’s<br />
Research Hospital, Memphis, TN; Departments <strong>of</strong> Radiation <strong>Oncology</strong> and Pediatrics,<br />
Philip Rubin Center for Cancer Survivorship, James P. Wilmot Cancer Center at University<br />
<strong>of</strong> Rochester Medical Center, Rochester, NY.<br />
Authors’ disclosures <strong>of</strong> potential conflicts <strong>of</strong> interest are found at the end <strong>of</strong> this article.<br />
Address reprint requests to Melissa M. Hudson, MD, St. Jude Children’s Research<br />
Hospital, Department <strong>of</strong> <strong>Oncology</strong>, Division <strong>of</strong> Cancer Survivorship, 262 Danny Thomas<br />
Place, Mailstop 735, Memphis, TN 38105; email: melissa.hudson@stjude.org.<br />
© <strong>2012</strong> by <strong>American</strong> <strong>Society</strong> <strong>of</strong> <strong>Clinical</strong> <strong>Oncology</strong>.<br />
1092-9118/10/1-10
RADIOTHERAPY FOR PEDIATRIC HODGKIN LYMPHOMA<br />
chemotherapy, RT, and combined-modality therapy, clinical<br />
trials have focused on reducing or eliminating agents and<br />
modalities associated with a high risk for morbidity. Numerous<br />
studies have established the effectiveness <strong>of</strong> non-crossresistant<br />
chemotherapy alone in these patients, which <strong>of</strong>fers<br />
advantages for children managed in centers lacking diagnostic<br />
and therapeutic radiation facilities and avoids the<br />
potential long-term growth inhibition, organ dysfunction,<br />
and carcinogenesis associated with RT. However,<br />
chemotherapy-alone treatment protocols typically prescribe<br />
KEY POINTS<br />
Table 1. Selected Risk-Adapted Treatment Approaches for Low-Risk Pediatric Hodgkin Lymphoma<br />
Chemotherapy (No. Cycles) Radiation (Gy) Stage No. Patients Event-Free Survival (years) Survival (years)<br />
VAMP (4) 25 15–25.5, IF CS I/II† 110 89 (10) 96 (10)<br />
COPP/ABV (4) 11 21, IF/None CS IA/B, IIA* 294 97/91 (3) 100/100 (3)<br />
OEPA/OPPA (2) 8 20–35, IF/None I, IIA 281/113 94/97 (5) NA<br />
DBVE (4) 26 25.5, IF IA, IIA, IIIA1 51 91 (6) 98 (6)<br />
Abbreviations: VAMP, vinblastine, doxorubicin, methotrexate, prednisone; ABV, doxorubicin, bleomycin, vinblastine; COPP, cyclophosphamide, vincristine,<br />
procarbazine, prednisone; OEPA, vincristine etoposide, prednisone, doxorubicin; OPPA, vincristine, procarbazine, prednisone, doxorubicin; DBVE, doxorubicin,<br />
bleomycin, vincristine, etoposide; E, extralymphatic; IF, involved-field radiation therapy.<br />
† Without bulky mediastinal (defined as one-third or more <strong>of</strong> intrathoracic ratio measured on an upright posteroanterior chest radiograph) or peripheral<br />
lymphadenopathy (defined 6 cm or more) or B symptoms.<br />
* Without adverse features defined as one or more <strong>of</strong> the following: hilar adenopathy, involvement <strong>of</strong> more than four nodal regions; mediastinal tumor with diameter<br />
greater than or equal to one-third <strong>of</strong> the chest diameter, and node or nodal aggregate with a diameter greater than 10 cm.<br />
● Pediatric Hodgkin lymphoma trials focus on maximizing<br />
treatment efficacy and minimizing risks for<br />
late toxicity associated with both radiation therapy<br />
and chemotherapy.<br />
● The use <strong>of</strong> low-dose involved-field radiation in pediatric<br />
Hodgkin lymphoma permits reduction in duration<br />
or intensity <strong>of</strong> chemotherapy and thus doserelated<br />
toxicity <strong>of</strong> anthracyclines, alkylating agents,<br />
and bleomycin that may preserve cardiopulmonary<br />
and gonadal function and reduce the risk <strong>of</strong> secondary<br />
leukemia.<br />
● Radiation has been used as an adjunct to multiagent<br />
chemotherapy in clinical trials for intermediate/highrisk<br />
pediatric Hodgkin lymphoma with the goal <strong>of</strong><br />
reducing risk <strong>of</strong> relapse in initially involved sites and<br />
preventing toxicity associated with retrieval therapy.<br />
● Compared with treatment with chemotherapy alone,<br />
adjuvant radiation produces a superior event-free<br />
survival for intermediate/high-risk children with<br />
Hodgkin lymphoma who achieve a complete response<br />
to multiagent chemotherapy, but does not affect overall<br />
survival because <strong>of</strong> the success <strong>of</strong> salvage therapy.<br />
● Radiation consolidation may facilitate local disease<br />
control in individuals with refractory/recurrent disease,<br />
especially in those who have limited or bulky<br />
sites <strong>of</strong> disease progression/recurrence, or persistent<br />
disease that does not completely respond to chemotherapy.<br />
● Future directions in the use <strong>of</strong> radiation therapy<br />
include reducing the targeted volume to include only<br />
the initially involved nodes rather than the lymph<br />
node regions that harbored those nodes; this may<br />
further reduce radiation-associated toxicities.<br />
higher cumulative doses <strong>of</strong> alkylating agents, anthracyclines,<br />
and bleomycin, which may produce late treatment<br />
morbidity from cardiopulmonary and gonadal injury and<br />
secondary leukemia. In early trials, high-dose RT provided<br />
an alternative therapeutic option for skeletally mature patients<br />
that avoided MOPP chemotherapy-related infertility<br />
and leukemogenesis. 7 Later trials combined low-dose (15–<br />
25.5 Gy) IFRT with multiagent chemotherapy and sequentially<br />
reduced the number <strong>of</strong> chemotherapy cycles, especially<br />
those including alkylating agents (Table 1).<br />
Once the effectiveness <strong>of</strong> combined modality regimens<br />
with fewer cycles <strong>of</strong> multiagent chemotherapy was established,<br />
contemporary studies employed a response-based<br />
paradigm to guide further reduction <strong>of</strong> chemotherapy and<br />
RT exposure. Regimens utilizing low-dose IFRT with chemotherapy<br />
combinations delivered at maximal dose intensity<br />
such as OPPA (vincristine, procarbazine, prednisone, doxorubicin)<br />
8 and DBVE (doxorubicin, bleomycin, vincristine,<br />
etoposide) 9 produced excellent outcomes after treatment<br />
with only two cycles <strong>of</strong> chemotherapy. A combined modality<br />
approach using nonalkylator-based chemotherapy and<br />
response-based low-dose IFRT also proved to be successful.<br />
For example, consortium investigators from St. Jude, Stanford,<br />
and Dana Farber demonstrated that local control was<br />
not compromised by reducing IFRT dose to 15 Gy in low-risk<br />
patients who achieved an early complete response to VAMP<br />
(vinblastine, doxorubicin, methotrexate, prednisone) chemotherapy.<br />
10<br />
Other groups undertook clinical trials aiming to eliminate<br />
RT for low-risk patients who achieved a complete response<br />
to chemotherapy. 8,11 A randomized controlled trial implemented<br />
by the Children’s Cancer Group (CCG) reported a<br />
significantly (stratified log-rank test; p � 0.057) higher<br />
3-year event-free survival (EFS) in patients who received 21<br />
Gy IFRT consolidation (97%; standard error [SE] � 1.7%),<br />
compared with those treated with four cycles <strong>of</strong> COPP/ABV<br />
(cyclophosphamide, vincristine, procarbazine, prednisone/<br />
doxorubicin, bleomycin, vinblastine) chemotherapy alone<br />
(91%; SE 2.8%). 11 German investigators subsequently observed<br />
no difference in disease-free survival among nonirradiated<br />
(97%; SE � 2%) and irradiated (94%; SE � 2%)<br />
low-risk girls treated with OPPA (vincristine, prednisone,<br />
procarbazine, doxorubicin) and boys treated with OEPA<br />
(substitution <strong>of</strong> etoposide for procarbazine in the OPPA<br />
combination). 8 Likewise, the EFS for children and adolescents<br />
treated with four cycles <strong>of</strong> VAMP chemotherapy after<br />
achieving an early complete response (89%; SE � 5.7%) did<br />
not differ from those treated with four cycles <strong>of</strong> VAMP and<br />
response-based IFRT (87%; SE � 6.4%) for those who did not<br />
achieve an early complete remission. 12 Common to all these<br />
617
Table 2. Selected Risk-Adapted Treatment Approaches for Intermediate-Risk Pediatric Hodgkin Lymphoma<br />
Chemotherapy (Number <strong>of</strong> Cycles) Radiation (Gy) Stage No. Patients Event-Free Survival (years) Survival (years)<br />
COPP/ABV (6) 11 21, IF CS I/II*, CS IIB, CS III 394 84 (3) 100 (3)<br />
OEPA/OPPA (2) � COPP (2) 8 20–35, IF II EA, IIB, IIIA 212 92 (5) N/A<br />
OEPA/OPPA (2) � COPDAC (2) 15 20–35, IF II EB, III EA/B, IIIB, IVA/B 139 88.3 (5) 98.5 (5)<br />
ABVE-PC (3–5) 18 21, IF IB, IIA, IIIA 53 84 (5) 95 (5)<br />
Abbreviations: COPP, cyclophosphamide, vincristine, procarbazine, prednisone; ABV, doxorubicin, bleomycin, vinblastine; OEPA, vincristine etoposide, prednisone,<br />
doxorubicin; OPPA, vincristine, procarbazine, prednisone, doxorubicin; COPDAC, cyclophosphamide, vincristine, prednisone, dacarbazine; ABVE-PC, doxorubicin,<br />
bleomycin, vincristine, etoposide-prednisone, cyclophosphamide; E, extralymphatic; IF, involved-field radiation therapy.<br />
* With adverse disease features defined as one or more <strong>of</strong> the following: hilar adenopathy, involvement <strong>of</strong> more than four nodal regions; mediastinal tumor with<br />
diameter greater than or equal to one-third <strong>of</strong> the chest diameter, and node or nodal aggregate with a diameter greater than 10 cm.<br />
studies is that because <strong>of</strong> effective retrieval therapy among<br />
relapsed patients, overall survival did not differ among the<br />
irradiated and nonirradiated groups.<br />
Radiation Therapy for Intermediate/High-Risk HL<br />
Concerns regarding long-term chemotherapy and RTrelated<br />
toxicities have led to the development <strong>of</strong> divergent<br />
therapeutic approaches for children and adolescents compared<br />
with that <strong>of</strong> adults. In contrast to adult trials, clinical<br />
trials for intermediate/high-risk pediatric HL typically use<br />
RT as an adjunct to multiagent chemotherapy, with the goal<br />
<strong>of</strong> reducing risk <strong>of</strong> relapse and preventing toxicity associated<br />
with retrieval therapy. Combination chemotherapy including<br />
vinca alkaloids, alkylating agents, anthracyclines, and<br />
<strong>of</strong>ten etoposide provides the cornerstone <strong>of</strong> therapy (Tables 2<br />
and 3). Gender-based regimens consider that male patients<br />
are more vulnerable to gonadal toxicity from alkylating<br />
agent chemotherapy and that female patients have a substantial<br />
risk <strong>of</strong> breast cancer after chest RT. In this regard,<br />
German Multi-Center investigators have undertaken a series<br />
<strong>of</strong> gender-based risk-adapted trials aiming to reduce<br />
gonadal toxicity in male patients while maintaining the<br />
excellent disease-free survival accomplished with the OPPA/<br />
COPP regimen. The DAL-HD-90 study established that<br />
substitution <strong>of</strong> etoposide for procarbazine in the OPPA<br />
combination (OEPA) in boys produces comparable EFS to<br />
that <strong>of</strong> girls treated with OPPA and is associated with<br />
hormonal parameters suggesting a lower risk <strong>of</strong> gonadal<br />
toxicity. 13,14 In the GPOH-HD 2002 trial, substitution <strong>of</strong><br />
dacarbazine for procarbazine (OEPA-COPDAC) in boys produced<br />
comparable results to standard OPPA-COPP in girls<br />
when used in combination with IFRT. 15 Long-term follow-up<br />
is needed to determine if restriction <strong>of</strong> alkylating agent<br />
cumulative dose translates into improved rates <strong>of</strong> fertility<br />
preservation.<br />
Risk-adapted multiagent chemotherapy for intermediate/<br />
high-risk HL is typically followed by consolidative RT to<br />
involved sites <strong>of</strong> disease. Effective systemic therapy coupled<br />
with advancements in diagnostic imaging has led to increasingly<br />
restricted treatment fields that generally encompass<br />
lymph node regions initially involved at the time <strong>of</strong> diagnosis;<br />
field refinement is then routinely made to account for<br />
tumor regression with chemotherapy. 16 IFRT is the most<br />
common approach used in pediatric HL trials, although<br />
some groups are now evaluating involved-nodal and limited<br />
volume conformal (tailored field), intensity modulated, and<br />
proton RT as approaches that further reduce potential<br />
injury to normal tissues.<br />
In the past decade, several trials have investigated the<br />
benefit that adjuvant RT contributes to survival outcomes<br />
among intermediate/high-risk children with HL who achieve<br />
a complete response to multiagent chemotherapy. In the<br />
CCG’s randomized controlled trial using COPP/ABV hybrid<br />
chemotherapy, the projected 3-year EFS among patients<br />
who achieved a complete response to initial therapy, was<br />
92% (SE � 1.9%) for those randomized to receive low-dose<br />
IFRT and 87% (SE � 2.2%) for those randomized to receive<br />
no further therapy. 11 The difference in 3-year EFS was most<br />
marked for stage IV patients randomized to receive<br />
combined-modality therapy with IFRT (90%, SE � 5.5%)<br />
compared with those randomized to receive chemotherapy<br />
alone (81%, SE � 6.9%). Likewise, omission <strong>of</strong> RT for<br />
patients completely responding to risk- and gender-based<br />
OEPA/COPP or OPPA/COPP chemotherapy resulted in significantly<br />
lower EFS in intermediate/high-risk patients compared<br />
with irradiated patients (79% vs. 91%, p � 0.0008). 8<br />
Notably, this study also demonstrated a survival benefit <strong>of</strong><br />
providing a 5–10 Gy RT boost to lymph node regions with<br />
an “insufficient remission” following chemotherapy, thereby<br />
overcoming the adverse prognostic implications <strong>of</strong> bulky<br />
mediastinal lymphadenopathy. 17 For both studies, estimates<br />
for overall survival did not differ between the irradiated<br />
and nonirradiated groups because <strong>of</strong> successful salvage<br />
therapy after relapse. 8,11<br />
Contemporary trials have investigated if chemotherapy<br />
and RT can be limited in patients who achieve a rapid early<br />
Table 3. Selected Risk-Adapted Treatment Approaches for High-Risk Pediatric Hodgkin Lymphoma<br />
HUDSON AND CONSTINE<br />
Chemotherapy (Number <strong>of</strong> Cycles) Radiation (Gy) Stage No. Patients Event-Free Survival (years) Survival (years)<br />
OEPA/OPPA (2) � COPP (4) 8 20–35, IF II EB, III EA/B, IIIB, IVA/B 265 91 (5) N/A<br />
OEPA/OPPA (2) � COPDAC (4) 15 20–35, IF II EB, III EA/B, IIIB, IVA/B 239 86.9 (5) 94.9 (5)<br />
ABVE-PC (3–5) 18 21, IF IB, IIA, IIIA 163 85 (5) 95 (5)<br />
BEACOPP (4); COPP/ABV (4) (RER; girls) 19 None IIB, IIIB, IV 38 94 (5) 97 (5)<br />
BEACOPP (4); ABVD (2) (RER; boys) 19 21, IF IIB, IIIB, IV 34<br />
BEACOPP (8) (SER) 19 21, IF IIB, IIIB, IV 25<br />
Abbreviations: OEPA, vincristine etoposide, prednisone, doxorubicin; OPPA, vincristine, procarbazine, prednisone, doxorubicin; COPP, cyclophosphamide,<br />
vincristine, procarbazine, prednisone; COPDAC, cyclophosphamide, vincristine, prednisone, dacarbazine; ABVE-PC, doxorubicin, bleomycin, vincristine, etoposideprednisone,<br />
cyclophosphamide; BEACOPP, bleomycin, etoposide, doxorubicin, cyclophosphamide, vincristine, procarbazine, prednisone; ABV, doxorubicin, bleomycin,<br />
vinblastine; ABVD, doxorubicin, bleomycin, vinblastine, dacarbazine; E, extralymphatic; IF, involved-field radiation therapy; RER, rapid early response; SER, slow early<br />
response.<br />
618
RADIOTHERAPY FOR PEDIATRIC HODGKIN LYMPHOMA<br />
response to dose-intensive chemotherapy regimens. These<br />
studies have been facilitated by assessment <strong>of</strong> interim or end<br />
<strong>of</strong> chemotherapy response based on anatomic or functional<br />
changes on computed tomography or functional imaging like<br />
positron emission tomography. The Pediatric <strong>Oncology</strong><br />
Group utilized a response-based therapy utilizing dosedense<br />
ABVE-PC (doxorubicin, bleomycin, vincristine,<br />
etoposide-prednisone, cyclophosphamide) for patients with<br />
unfavorable advanced-stage disease in combination with 21<br />
Gy IFRT. 18 The dose-dense approach permitted reduction in<br />
chemotherapy exposure in 63% <strong>of</strong> patients who achieved a<br />
rapid early response to three ABVE-PC cycles. Five-year<br />
EFS was comparable for rapid early responders (86%) and<br />
slow early responders (83%) treated with three and five<br />
cycles <strong>of</strong> ABVE-PC, respectively, followed by 21 Gy IFRT.<br />
The CCG (CCG-59704) evaluated response-adapted therapy<br />
featuring four cycles <strong>of</strong> the dose-intensive BEACOPP<br />
(bleomycin, etoposide, doxorubicin, cyclophosphamide, vincristine,<br />
procarbazine, prednisone) regimen followed by a<br />
gender-tailored consolidation for pediatric patients with<br />
stage IIB, IIIB with bulk disease, and IV HL. 19 For rapid<br />
early responding girls, an additional four courses <strong>of</strong> COPP/<br />
ABV (without IFRT) were given in an effort to reduce breast<br />
cancer risk. Rapid early responding boys received two cycles<br />
<strong>of</strong> ABVD followed by IFRT. Slow early responders received<br />
four additional courses <strong>of</strong> BEACOPP and IFRT. Rapid early<br />
response (defined by resolution <strong>of</strong> B symptoms and greater<br />
than 70% reduction in tumor volume) was achieved by 74%<br />
<strong>of</strong> patients after four BEACOPP cycles and 5-year EFS<br />
among the cohort is 94% (median follow-up 6.3 years). 19<br />
Results from this study support that early intensification<br />
followed by less intense response-based therapy results in<br />
high EFS. However, the potential for late treatment effects<br />
on cardiopulmonary and gonadal function have not been<br />
evaluated.<br />
Radiation Therapy in Refractory/Relapsed HL<br />
The generally excellent outcome <strong>of</strong> children and adolescents<br />
with HL limits opportunities to evaluate retrieval<br />
therapy. This is particularly true in regard to the benefits <strong>of</strong><br />
using RT in the setting <strong>of</strong> refractory/relapsed disease. Uniformly,<br />
chemotherapy is the recommended retrieval therapy,<br />
with the choice <strong>of</strong> specific agents, dose-intensity, and<br />
number <strong>of</strong> cycles determined by the initial therapy, disease<br />
characteristics at progression/relapse, and response to retrieval<br />
therapy. In children with localized favorable (relapse<br />
after 12 months after completing therapy) disease recurrences<br />
whose original therapy involved reduced cycles <strong>of</strong><br />
risk-adapted therapy, IRFT consolidation may be <strong>of</strong>fered<br />
following treatment with more intensive conventional chemotherapy.<br />
Autologous hematopoietic cell transplantation<br />
(autoHCT) is the recommended approach for patients who<br />
develop refractory/relapsed disease during or within 1 year<br />
after completing therapy. 20 Results <strong>of</strong> investigations<br />
(largely comprised <strong>of</strong> adult patients) evaluating the use <strong>of</strong><br />
IFRT immediately before or after transplant have been<br />
conflicting in their support <strong>of</strong> a potential benefit conferred by<br />
RT. 21-23 However, most studies are limited by their retrospective<br />
nature, nonrandom treatment assignments, and<br />
small patient numbers. Nevertheless, IFRT is <strong>of</strong>ten used to<br />
consolidate local control in individuals with refractory/recurrent<br />
disease, especially in those who have limited or bulky<br />
sites <strong>of</strong> disease recurrence or persistent disease that does<br />
not completely respond to salvage chemotherapy. In the<br />
latter, local radiotherapy can be considered to promote local<br />
disease resolution before autoHCT rather than as consolidation<br />
following autoHCT. However, concerns regarding toxicity<br />
have prevented the use <strong>of</strong> IFRT as a standard adjuvant to<br />
high-dose chemotherapy and autoHCT in HL. Prospective<br />
trials are needed to definitely establish the long-term risks<br />
and survival benefits provided by inclusion <strong>of</strong> RT in salvage<br />
therapy approaches.<br />
Advances in the Delivery <strong>of</strong> Radiation<br />
As previously stated, contemporary protocols are testing<br />
field reductions from involved lymph node regions to involved<br />
nodal sites. In addition, the standard technique for<br />
radiation delivery in pediatric HL is a two-dimensional<br />
CT-based approach. However, three-dimensional conformal<br />
radiation therapy (3DCRT), intensity-modulated RT<br />
(IMRT), or proton therapy may be <strong>of</strong>fer the ability to reduce<br />
exposure to critical normal tissues such as heart, lungs, and<br />
developing breast tissue. Uncertainty exists about the potential<br />
for increased late effects from IMRT, particularly<br />
secondary malignancy, bacause IMRT results in a lower<br />
dose to a larger volume compared with conventional techniques.<br />
Unfortunately, any benefits <strong>of</strong> lower doses to critical<br />
organs, or changes (either an increase or decrease) in the<br />
potential for a subsequent malignancy, will not be testable<br />
for many years.<br />
Conclusion<br />
The role <strong>of</strong> RT in the treatment <strong>of</strong> pediatric HL has<br />
evolved considerably over the last 50 years, primarily motivated<br />
by the desire to avoid its adverse long-term effects.<br />
Although elimination <strong>of</strong> RT has been a key focus <strong>of</strong> contemporary<br />
trials, this modality continues to play an important<br />
role in optimizing survival outcomes for many children and<br />
adolescents with the disease. For those with low-risk HL,<br />
the addition <strong>of</strong> low-dose IFRT permits reduction in chemotherapy<br />
duration or intensity and thus the potential doserelated<br />
toxicity <strong>of</strong> anthracyclines, alkylating agents, and<br />
bleomycin that may preserve cardiopulmonary and gonadal<br />
function. For those with intermediate/high-risk HL, the<br />
superior disease-free survival demonstrated by some trials<br />
may ultimately translate into a survival benefit by avoiding<br />
the need for more toxic salvage therapy. In the setting <strong>of</strong><br />
disease with suboptimal chemotherapy response, RT consolidation<br />
provides an effective alternative modality to enhance<br />
tumor control. Until the availability <strong>of</strong> validated biologically<br />
based prognostic factors, the inclusion <strong>of</strong> RT consolidation in<br />
pediatric HL treatment regimens will continued to be guided<br />
by prognostic factors at diagnosis correlated with disease<br />
burden and chemotherapy responsiveness as assessed by<br />
functional imaging. For patients who require RT to optimize<br />
disease control, advances in radiation technology and reduced<br />
volume radiation strategies are anticipated to greatly<br />
reduce exposure to normal tissues and the subsequent risk<br />
<strong>of</strong> late RT-associated toxicity. 24<br />
Acknowledgements<br />
Research grant support: Dr. Hudson is supported in part by<br />
the Cancer Center Support (CORE) grant CA 21765 from the<br />
National Cancer Institute and by the <strong>American</strong> Lebanese Syrian<br />
Associated Charities (ALSAC).<br />
619
Authors’ Disclosures <strong>of</strong> Potential Conflicts <strong>of</strong> Interest<br />
Author<br />
Melissa M. Hudson*<br />
Louis S. Constine*<br />
*No relevant relationships to disclose.<br />
Employment or<br />
Leadership<br />
Positions<br />
Consultant or<br />
Advisory Role<br />
1. Donaldson SS, Kaplan HS. Complications <strong>of</strong> treatment <strong>of</strong> Hodgkin’s<br />
disease in children. Cancer Treat Rep. 1982;66:977-989.<br />
2. Donaldson SS, Link MP. Combined modality treatment with low-dose<br />
radiation and MOPP chemotherapy for children with Hodgkin’s disease.<br />
J Clin Oncol. 1987;5:742-749.<br />
3. Bhatia S, Robison LL, Oberlin O, et al. Breast cancer and other second<br />
neoplasms after childhood Hodgkin’s disease. N Engl J Med. 1996;334:745-<br />
751.<br />
4. Hancock SL, Tucker MA, Hoppe RT. Factors affecting late mortality<br />
from heart disease after treatment <strong>of</strong> Hodgkin’s disease. JAMA. 1993;270:<br />
1949-1955.<br />
5. Devita VT, Jr., Serpick AA, Carbone PP. Combination chemotherapy in<br />
the treatment <strong>of</strong> advanced Hodgkin’s disease. Ann Intern Med. 1970;73:881-<br />
895.<br />
6. Bonadonna G, Santoro A. ABVD chemotherapy in the treatment <strong>of</strong><br />
Hodgkin’s disease. Cancer Treat Rev. 1982;9:21-35.<br />
7. Donaldson SS, Whitaker SJ, Plowman PN, et al. Stage I-II pediatric<br />
Hodgkin’s disease: Long-term follow-up demonstrates equivalent survival<br />
rates following different management schemes. J Clin Oncol. 1990;8:1128-<br />
1137.<br />
8. Dorffel W, Luders H, Ruhl U, et al. Preliminary results <strong>of</strong> the multicenter<br />
trial GPOH-HD 95 for the treatment <strong>of</strong> Hodgkin’s disease in children<br />
and adolescents: Analysis and outlook. Klin Padiatr. 2003;215:139-145.<br />
9. Schwartz CL. Special issues in pediatric Hodgkin’s disease. Eur J<br />
Haematol. Suppl 2005;55-62.<br />
10. Donaldson SS, Hudson MM, Lamborn KR, et al. VAMP and low-dose,<br />
involved-field radiation for children and adolescents with favorable, earlystage<br />
Hodgkin’s disease: Results <strong>of</strong> a prospective clinical trial. J Clin Oncol.<br />
2002;20:3081-3087.<br />
11. Nachman JB, Sposto R, Herzog P, et al. Randomized comparison <strong>of</strong><br />
low-dose involved-field radiotherapy and no radiotherapy for children with<br />
Hodgkin’s disease who achieve a complete response to chemotherapy. J Clin<br />
Oncol. 2002;20:3765-3771.<br />
12. Metzger ML, Weinstein HJ, Hudson MM, et al. Results <strong>of</strong> a prospective<br />
clinical trial <strong>of</strong> VAMP alone without irradiation for pediatric favorable,<br />
early-stage Hodgkin lymphoma patients who achieve an early complete<br />
response. J Clin Oncol Proc Am Soc Clin Oncol. 2011;29:585s.<br />
13. Gerres L, Bramswig JH, Schlegel W, et al. The effects <strong>of</strong> etoposide on<br />
testicular function in boys treated for Hodgkin’s disease. Cancer. 1998;83:<br />
2217-2222.<br />
14. Schellong G, Potter R, Bramswig J, et al. High cure rates and reduced<br />
long-term toxicity in pediatric Hodgkin’s disease: The German-Austrian<br />
multicenter trial DAL-HD-90. The German-Austrian Pediatric Hodgkin’s<br />
Disease Study Group. J Clin Oncol. 1999;17:3736-3744.<br />
620<br />
Stock<br />
Ownership Honoraria<br />
REFERENCES<br />
Research<br />
Funding<br />
HUDSON AND CONSTINE<br />
Expert<br />
Testimony<br />
Other<br />
Remuneration<br />
15. Mauz-Korholz C, Hasenclever D, Dorffel W, et al. Procarbazine-free<br />
OEPA-COPDAC chemotherapy in boys and standard OPPA-COPP in girls<br />
have comparable effectiveness in pediatric Hodgkin’s lymphoma: The GPOH-<br />
HD-2002 study. J Clin Oncol. 2010;28:3680-3686.<br />
16. Yahalom J, Mauch P. The involved field is back: Issues in delineating<br />
the radiation field in Hodgkin’s disease. Ann Oncol. 2002;13 Suppl 1:79-83.<br />
17. Ruhl U, Albrecht M, Dieckmann K, et al. Response-adapted radiotherapy<br />
in the treatment <strong>of</strong> pediatric Hodgkin’s disease: An interim report at 5<br />
years <strong>of</strong> the German GPOH-HD 95 trial. Int J Radiat Oncol Biol Phys.<br />
2001;51:1209-1218.<br />
18. Schwartz CL, Constine LS, Villaluna D, et al. A risk-adapted, responsebased<br />
approach using ABVE-PC for children and adolescents with<br />
intermediate- and high-risk Hodgkin lymphoma: The results <strong>of</strong> P9425. Blood.<br />
2009;114:2051-2059.<br />
19. Kelly KM, Sposto R, Hutchinson R, et al. BEACOPP chemotherapy is a<br />
highly effective regimen in children and adolescents with high-risk Hodgkin<br />
lymphoma: A report from the Children’s <strong>Oncology</strong> Group. Blood. 2011;117:<br />
2596-2603.<br />
20. Colpo A, Hochberg E, Chen YB. Current Status <strong>of</strong> Autologous Stem Cell<br />
Transplantation in Relapsed and Refractory Hodgkin’s Lymphoma. Oncologist.<br />
<strong>2012</strong>;17:80-90.<br />
21. Majhail NS, Bajorunaite R, Lazarus HM, et al. Long-term survival and<br />
late relapse in 2-year survivors <strong>of</strong> autologous haematopoietic cell transplantation<br />
for Hodgkin and non-Hodgkin lymphoma. Br J Haematol. 2009;147:<br />
129-139.<br />
22. Moskowitz AJ, Perales MA, Kewalramani T, et al. Outcomes for<br />
patients who fail high dose chemoradiotherapy and autologous stem cell<br />
rescue for relapsed and primary refractory Hodgkin lymphoma. Br J Haematol.<br />
2009;146:158-163.<br />
23. Wendland MM, Asch JD, Pulsipher MA, et al. The impact <strong>of</strong> involved<br />
field radiation therapy for patients receiving high-dose chemotherapy followed<br />
by hematopoietic progenitor cell transplant for the treatment <strong>of</strong><br />
relapsed or refractory Hodgkin disease. Am J Clin Oncol. 2006;29:189-195.<br />
24. Koh ES, Tran TH, Heydarian M, et al. A comparison <strong>of</strong> mantle versus<br />
involved-field radiotherapy for Hodgkin’s lymphoma: Reduction in normal<br />
tissue dose and second cancer risk. Radiat Oncol. 2007;2:13.<br />
25. Donaldson SS, Link MP, Weinstein HJ, et al. Final results <strong>of</strong> a<br />
prospective clinical trial with VAMP and low-dose involved-field radiation for<br />
children with low-risk Hodgkin’s disease. J Clin Oncol. 2007;25:332-337.<br />
26. Tebbi CK, Mendenhall N, London WB, et al. Treatment <strong>of</strong> stage I, IIA,<br />
IIIA1 pediatric Hodgkin disease with doxorubicin, bleomycin, vincristine and<br />
etoposide (DBVE) and radiation: A Pediatric <strong>Oncology</strong> Group (POG) study.<br />
Pediatr Blood Cancer. 2006;46:198-202.
Role <strong>of</strong> Doxorubicin in Rhabdomyosarcoma:<br />
Is the Answer Knowable?<br />
Overview: The role <strong>of</strong> doxorubicin in treatment <strong>of</strong> rhabdomyosarcoma<br />
(RMS) has been controversial for 30 years. Despite<br />
its known activity in RMS, because <strong>of</strong> its risk <strong>of</strong><br />
cardiotoxicity, its use is not justified in low-risk patients who<br />
have an excellent chance <strong>of</strong> cure with vincristine, actinomycin<br />
with or without cyclophosphamide, and primary tumor treatment.<br />
For patients with intermediate and high risks, the<br />
risk/benefit ratio must be carefully considered. In addition, the<br />
peak incidence <strong>of</strong> RMS is in toddlers, with whom the risk <strong>of</strong><br />
THE ROLE <strong>of</strong> doxorubicin in the treatment <strong>of</strong> rhabdomyosarcoma<br />
has been controversial for over three decades.<br />
The first report <strong>of</strong> its activity in rhabdomyosarcoma<br />
was by Massimo and colleagues in 1969, 1 followed by an<br />
early phase II study in children by Tan, which showed that<br />
doxorubicin was active against both newly diagnosed and<br />
previously treated rhabdomyosarcoma. 2<br />
However, one <strong>of</strong> the major disadvantages <strong>of</strong> doxorubicin<br />
is its cardiotoxicity, which was recognized early in its use. 3<br />
Risk factors for cardiotoxicity include young age at<br />
administration, higher cumulative dose, and radiation fields<br />
including the heart. With the peak incidence <strong>of</strong> rhabdomyosarcoma<br />
being in toddlers, and the majority <strong>of</strong> patients being<br />
under age 10 at diagnosis, cardiotoxicity is a factor in the<br />
risk/benefit calculation. Low-risk patients (low stage and<br />
clinical group, favorable histology) have excellent outcomes—<br />
with survival rates <strong>of</strong> 80% to 90% with VAC (vincristine,<br />
actinomycin, cyclophosphamide) or VA (vincristine, actinomycin)<br />
and primary tumor treatment alone; in them, the use<br />
<strong>of</strong> doxorubicin with its potential for long-term cardiotoxicity<br />
cannot be justified.<br />
The goal <strong>of</strong> this discussion is to review the historic data<br />
on use <strong>of</strong> doxorubicin in rhabdomyosarcoma, as well as to<br />
discuss ongoing clinical studies utilizing it.<br />
Historical Perspectives<br />
North <strong>American</strong> Experience. Patients with group III and<br />
group IV rhabdomyosarcoma were randomly assigned in Intergroup<br />
Rhabdomyosarcoma Study (IRS) to receive therapy<br />
with VAC and radiation, with or without doxorubicin. The<br />
dose <strong>of</strong> doxorubicin was 60 mg/m 2 , given at 10 to 12 week<br />
intervals (alternating with actinomycin) either alone or with<br />
continuous daily oral cyclophosphamide. There was no difference<br />
in outcome between regimens and the conclusion was<br />
that doxorubicin did not improve the outcome <strong>of</strong> patients<br />
with group III or IV disease. 4 It should be noted that the<br />
intervals between treatment cycles were much longer in the<br />
early RMS studies than treatment intervals are in the current<br />
era, and much <strong>of</strong> the cyclophosphamide was given orally.<br />
IRS II, conducted between 1978 and 1984, again evaluated<br />
the role <strong>of</strong> doxorubicin in patients with group III and IV<br />
rhabdomyosarcoma, but this time given along with vincristine<br />
and intravenous cyclophosphamide as “pulsed” therapy.<br />
Patients were randomly assigned between repetitive pulse<br />
VAC or repetitive pulse VAC/VDC (D � doxorubicin). 5 Once<br />
again, there was no difference in outcomes between the two<br />
regimens.<br />
By Carola A. S. Arndt, MD<br />
cardiotoxicity <strong>of</strong> anthracyclines is higher. A number <strong>of</strong> trials<br />
both in North America and Europe, which are reviewed in this<br />
article, have investigated the role <strong>of</strong> doxorubicin in RMS, with<br />
no conclusive outcomes. In addition, differences in riskgroup<br />
assignment on two sides <strong>of</strong> the Atlantic further complicate<br />
comparisons and analyses. The current European EpSSG<br />
2005 study for high-risk RMS (by the European definition) may<br />
come closest to giving an answer to the role <strong>of</strong> doxorubicin in<br />
RMS.<br />
IRS III, conducted between 1984 and 1991, had a very<br />
complex study design. 5 For patients with group II tumors,<br />
the major objective was to see whether the addition <strong>of</strong><br />
doxorubicin to VA and radiation therapy (RT) improved the<br />
outcome <strong>of</strong> favorable histology tumors. Patients were randomly<br />
assigned to treatment with VA or VA plus doxorubicin<br />
(with RT in both cases). The outcome was better for the<br />
patients treated with VA plus doxorubicin (89% vs. 54%<br />
5 years survival, p � 0.03). However, when historic controls<br />
from IRS II treated with VA and RT were included, the<br />
statistical significance disappeared. Moreover, the VA regimen<br />
had a worse outcome on IRS III than the identical<br />
therapy on IRS II, further confounding the results. Nevertheless,<br />
the paper stated, “In this randomized comparison,<br />
there was suggestive statistical evidence that the addition <strong>of</strong><br />
doxorubicin (30 mg/m 2 /d IV X 2 during weeks 3, 6, 12, 15, 21,<br />
and 24) to a basic VA regimen improved clinical outcome.”<br />
Patients with group III and group IV tumors were treated<br />
with doxorubicin containing regimens that also included<br />
other agents in addition to VAC, making the individual<br />
contribution <strong>of</strong> doxorubicin impossible to determine. Patients<br />
with clinical group I and II unfavorable histology<br />
tumors received pulsed VDC and VAC plus cisplatin, and<br />
showed significant improvement in 5-year progression free<br />
survival and survival rates compared with similar patients<br />
treated less intensively on IRS II (71% � 6% and 80% � 6%<br />
versus 59% � 5% and 71% � 5%, respectively; p � 0.002 and<br />
0.01, respectively). However, the regimens were quite complex,<br />
making the individual contribution <strong>of</strong> doxorubicin<br />
impossible to determine.<br />
In a series <strong>of</strong> “phase II window” studies for high-risk<br />
patients with metastatic disease conducted by the Children’s<br />
<strong>Oncology</strong> Group (COG) between 1988 and 2000,<br />
the combination <strong>of</strong> doxorubicin/ifosfamide and etoposide/<br />
ifosfamide had the highest response rates, although there<br />
was no survival difference. 6 So once again, the role <strong>of</strong><br />
doxorubicin was not able to be determined conclusively.<br />
A pilot study utilizing VDC alternating with ifosfamide/<br />
From the Department <strong>of</strong> Pediatric and Adolescent Medicine, Mayo Clinic Children’s<br />
Center, Rochester, MN.<br />
Author’s disclosure <strong>of</strong> potential conflicts <strong>of</strong> interest are found at the end <strong>of</strong> this article.<br />
Address reprint requests to Carola A. S. Arndt, MD, Mayo Clinic, Mayo Clinic Children’s<br />
Center, Department <strong>of</strong> Pediatric and Adolescent Medicine, 200 First Street SW, Rochester,<br />
MN; email: carndt@mayo.edu.<br />
© <strong>2012</strong> by <strong>American</strong> <strong>Society</strong> <strong>of</strong> <strong>Clinical</strong> <strong>Oncology</strong>.<br />
1092-9118/10/1-10<br />
621
etoposide for intermediate risk rhabdomyosarcoma reported<br />
excellent results, 7 but when compared to similar patients<br />
treated on IRS IV, a difference in outcome could not be<br />
demonstrated. 8 The study was a small pilot study, which<br />
included other agents (ifosfamide and etoposide in addition<br />
to doxorubicin).<br />
European Experience<br />
European studies have incorporated anthracyclines for<br />
rhabdomyosarcoma for many years, using doxorubicin or<br />
epirubicin. The German studies CWS-81, 86, and 91 utilized<br />
doxorubicin in combination with other drugs (ifosfamide,<br />
cyclophosphamide, etoposide in regimens VACA, VAIA, and<br />
EVAIA) but again there were no randomizations in the use<br />
<strong>of</strong> doxorubicin. 9-11 The Malignant Mesenchymal Tumor<br />
(MMT) 84 study used doxorubicin only for patients with poor<br />
response to upfront chemotherapy, again without randomization.<br />
12 In the MMT 95 study, high-risk nonmetastatic<br />
patients were randomly assigned to receive IVA (ifosfamide,<br />
vincristine, actinomycin) or IVA plus carboplatin, epirubicin,<br />
and etoposide—and found no difference in outcome. 13<br />
KEY POINTS<br />
● The role <strong>of</strong> doxorubicin in treatment <strong>of</strong> rhabdomyosarcoma<br />
(RMS) remains controversial.<br />
● The use <strong>of</strong> doxorubicin in patients with low-risk RMS<br />
is not justified.<br />
● Several studies have failed to show benefit in the use<br />
<strong>of</strong> anthracycline, but the studies were complex, confounding<br />
the issue.<br />
● Although differences in risk group assignment between<br />
Europe and North America make comparisons<br />
difficult, the current EpSSG 2005 study may shed<br />
light on the role <strong>of</strong> anthracycline in high-risk (by<br />
European definition) RMS.<br />
● EpSSG will study the addition <strong>of</strong> doxorubicin to<br />
ifosfamide, actinomycin, vincristine, not a substitution<br />
study <strong>of</strong> doxorubicin for actinomycin.<br />
Author’s Disclosure <strong>of</strong> Potential Conflicts <strong>of</strong> Interest<br />
Author<br />
Carola A. S. Arndt*<br />
*No relevant relationships to disclose.<br />
Employment or<br />
Leadership<br />
Positions<br />
Consultant or<br />
Advisory Role<br />
1. Massimo L, Cottafava F, Mori PG, et al. [Preliminary clinical trial <strong>of</strong><br />
adriamycin in solid malignant tumors in infants]. Minerva Pediatr. 1969;21:<br />
2182-2186.<br />
2. Tan C, Etcubanas E, Wollner N, et al. Adriamycin—an antitumor<br />
antibiotic in the treatment <strong>of</strong> neoplastic diseases. Cancer. 1973;32:9-17.<br />
3. Gilladoga AC, Manuel C, Tan CT, et al. The cardiotoxicity<br />
<strong>of</strong> adriamycin and daunomycin in children. Cancer. 1976;37:1070-<br />
1078.<br />
4. Maurer HM, Beltangady M, Gehan EA, et al. The Intergroup Rhabdomyosarcoma<br />
Study-I. A final report. Cancer. 1988;61:209-220.<br />
622<br />
More recently, a phase II window study <strong>of</strong> single-agent<br />
doxorubicin in patients with high risk metastatic disease<br />
showed a 65% overall response rate after two courses <strong>of</strong><br />
60 mg/m 2 at 21 day intervals. 14 The current high-risk<br />
European pediatric s<strong>of</strong>t tissue sarcoma study (EpSSG 2005)<br />
is randomly assigning patients with high-risk RMS to IVA<br />
versus IVA with doxorubicin. High-risk patients will also be<br />
randomly assigned to discontinue therapy or continue with<br />
maintenance chemotherapy, which may complicate the analysis.<br />
This study will be the purest in evaluating the addition<br />
<strong>of</strong> doxorubicin (as opposed to the substitution <strong>of</strong> doxorubicin<br />
for actinomycin) and may come the closest to addressing the<br />
value <strong>of</strong> doxorubicin in patients with high-risk RMS. Adding<br />
doxorubicin to IVA will likely make that arm <strong>of</strong> the study<br />
more toxic. Significant differences in patient classification<br />
may make trans-Atlantic comparisons difficult, however. 15<br />
Conclusion<br />
Despite over 30 years <strong>of</strong> clinical trials in rhabdomyosarcoma<br />
on both sides <strong>of</strong> the Atlantic, we have yet to define the<br />
appropriate use <strong>of</strong> doxorubicin in RMS. Clearly, it is not<br />
appropriate to investigate further the use <strong>of</strong> this cardiotoxic<br />
drug in patients with low-risk disease who have an outstanding<br />
prognosis. What about the intermediate-risk patients<br />
(by North <strong>American</strong> definition)? Currently, COG has<br />
chosen other new agents to investigate, such as topoisomerase<br />
inhibitors, and in the future, multityrosine kinase<br />
inhibitors. Is it justifiable to expose toddlers with<br />
intermediate-risk RMS to a cardiotoxic agent that has never<br />
been proven to improve outcome when there are newer,<br />
perhaps more interesting drugs available? At this time, most<br />
(but not all) investigators would say no. What about patients<br />
with node-positive alveolar histology or other patients at<br />
high risk <strong>of</strong> recurrence? They are considered “high risk” by<br />
the European definition. The outcome is poor for these<br />
patients, who are the subjects <strong>of</strong> the IVA versus IVADo<br />
study in EpSSg 2005. We hope that this study will provide<br />
some clarity for us in a subset <strong>of</strong> patients, as long as it is not<br />
confounded by the randomization between maintenance<br />
chemotherapy and stopping chemotherapy, or differences in<br />
toxicity by the addition <strong>of</strong> doxorubicin to IVA, which may<br />
compromise dose intensity. For patients with metastatic<br />
disease, adding doxorubicin to regimens that incorporate<br />
other agents has also not improved outcome.<br />
Stock<br />
Ownership Honoraria<br />
REFERENCES<br />
Research<br />
Funding<br />
Expert<br />
Testimony<br />
CAROLA A. S. ARNDT<br />
Other<br />
Remuneration<br />
5. Maurer HM, Gehan EA, Beltangady M, et al. The Intergroup Rhabdomyosarcoma<br />
Study-II. Cancer. 1993;71:1904-1922.<br />
6. Lager JJ, Lyden ER, Anderson JR, et al. Pooled analysis <strong>of</strong> phase II<br />
window studies in children with contemporary high-risk metastatic rhabdomyosarcoma:<br />
a report from the S<strong>of</strong>t Tissue Sarcoma Committee <strong>of</strong> the<br />
Children’s <strong>Oncology</strong> Group. J Clin Oncol. 2006;24:3415-3422.<br />
7. Arndt CA, Nascimento AG, Schroeder G, et al. Treatment <strong>of</strong> intermediate<br />
risk rhabdomyosarcoma and undifferentiated sarcoma with alternating<br />
cycles <strong>of</strong> vincristine/doxorubicin/cyclophosphamide and etoposide/<br />
ifosfamide. Eur J Cancer. 1998;34:1224-1229.
DOXORUBICIN IN RHABDOMYOSARCOMA<br />
8. Arndt CA, Hawkins DS, Meyer WH, et al. Comparison <strong>of</strong> results <strong>of</strong> a<br />
pilot study <strong>of</strong> alternating vincristine/doxorubicin/cyclophosphamide and etoposide/ifosfamide<br />
with IRS-IV in intermediate risk rhabdomyosarcoma: A report<br />
from the Children’s <strong>Oncology</strong> Group. Pediatr Blood Cancer. 2008;50:33-36.<br />
9. Dantonello TM, Int-Veen C, Harms D, et al. Cooperative trial CWS-91 for<br />
localized s<strong>of</strong>t tissue sarcoma in children, adolescents, and young adults. J Clin<br />
Oncol. 2009;27:1446-1455.<br />
10. Koscielniak E, Jürgens H, Winkler K, et al. Treatment <strong>of</strong> s<strong>of</strong>t tissue<br />
sarcoma in childhood and adolescence. A report <strong>of</strong> the German Cooperative S<strong>of</strong>t<br />
Tissue Sarcoma Study. Cancer. 1992;70:2557-2567.<br />
11. Koscielniak E, Harms D, Henze G, et al. Results <strong>of</strong> treatment for s<strong>of</strong>t<br />
tissue sarcoma in childhood and adolescence: a final report <strong>of</strong> the German Cooperative<br />
S<strong>of</strong>t Tissue Sarcoma Study CWS-86. J Clin Oncol. 1999;17:3706-3719.<br />
12. Flamant F, Rodary C, Rey A, et al. Treatment <strong>of</strong> non-metastatic rhabdomyosarcomas<br />
in childhood and adolescence. Results <strong>of</strong> the second study <strong>of</strong> the<br />
International <strong>Society</strong> <strong>of</strong> Paediatric <strong>Oncology</strong>: MMT84. Eur J Cancer. 1998;34:<br />
1050-1062.<br />
13. Stevens M, Rey A, et al. SIOP MMT 95: Intensified (6 drug) versus<br />
standard (IVA) chemotherapy for high risk non metastatic rhabdomyosarcoma<br />
(RMS). J Clin Oncol. 2004;22:14s (suppl; abstr 8515).<br />
14. Bergeron C, Thiesse P, Rey A, et al. Revisiting the role <strong>of</strong> doxorubicin in<br />
the treatment <strong>of</strong> rhabdomyosarcoma: an up-front window study in newly<br />
diagnosed children with high-risk metastatic disease. Eur J Cancer. 2008;44:<br />
427-431.<br />
15. Sultan I, Ferrari A. Selecting multimodal therapy for rhabdomyosarcoma.<br />
Expert Rev Anticancer Ther. 2010;10:1285-1301.<br />
623
PONTINE GLIOMAS IN CHILDREN:<br />
TO BIOPSY OR NOT TO BIOPSY<br />
CHAIR<br />
Mark W. Kieran, MD, PhD<br />
Dana-Farber Cancer Institute<br />
Boston, MA<br />
SPEAKERS<br />
Stephanie Puget, MD, PhD<br />
Necker Hospital, Universite Paris-Descartes<br />
Paris, France<br />
Nicholas K. Foreman, MD<br />
The Children’s Hospital<br />
Aurora, CO
Identification <strong>of</strong> Novel Biologic Targets in the<br />
Treatment <strong>of</strong> Newly Diagnosed Diffuse<br />
Intrinsic Pontine Glioma<br />
By Nathan J. Robison, MD, and Mark W. Kieran, MD, PhD<br />
Overview: Diffuse intrinsic pontine gliomas (DIPGs) carry an<br />
extremely poor prognosis. Standard practice has been to base<br />
the diagnosis on classic imaging and clinical characteristics<br />
and to treat with focal radiation therapy, usually accompanied<br />
with experimental therapy. As a result <strong>of</strong> the desire to avoid<br />
upfront biopsy, little has been learned regarding the molecular<br />
features <strong>of</strong> this disease. Findings from several autopsy series<br />
have included loss <strong>of</strong> p53 and PTEN, and amplification <strong>of</strong><br />
DIFFUSE INTRINSIC pontine glioma is a disease <strong>of</strong><br />
childhood that carries an abysmal prognosis. Marked<br />
advances in oncology over the last five decades have not been<br />
paralleled in this disease. DIPG is now the main cause <strong>of</strong><br />
brain tumor death in children. 1 The standard <strong>of</strong> care for<br />
DIPG remains focal radiation therapy alone, which alleviates<br />
symptoms in over 75% <strong>of</strong> patients; however rapid<br />
disease progression is almost universal. 2 Median overall<br />
survival is less than 1 year, and the 2-year survival rate<br />
is less than 10%. 1,3 Concerted experimental clinical trials<br />
over several decades have yielded no improvement in<br />
progression-free or overall survival. 4<br />
DIPG biology is not well understood. In the magnetic<br />
resonance imaging era, DIPG has become a radiologic diagnosis<br />
with no clear indication to biopsy, except in rare cases<br />
with atypical features. Paucity <strong>of</strong> upfront tissue available for<br />
analysis has been one <strong>of</strong> the main reasons that knowledge<br />
<strong>of</strong> this disease has failed to keep pace with that <strong>of</strong> other<br />
cancers. In recent years, an exponential increase in the<br />
technology for molecular pr<strong>of</strong>iling has begun to fundamentally<br />
change approaches to cancer treatment. There is a<br />
growing awareness that traditional diagnostic classifications<br />
based on microscopic appearance <strong>of</strong> the tumor must be<br />
reconsidered in light <strong>of</strong> our greater understanding <strong>of</strong> cancer<br />
biology and the relevant signal transduction pathways that<br />
lead to new therapeutic options. Tumors previously thought<br />
to represent a single diagnosis have been found, by molecular<br />
characterization, to in fact represent multiple distinct<br />
diseases. Conversely, recurring mutations such as BRAF<br />
and EGFR have been identified across seemingly unrelated<br />
cancers. With the advent <strong>of</strong> molecularly targeted therapy,<br />
these features take on new therapeutic importance. Identification<br />
<strong>of</strong> targetable oncogenic molecules has become a<br />
major focus <strong>of</strong> translational oncology research. However,<br />
identification <strong>of</strong> drugable targets in DIPG has lagged significantly<br />
behind other cancers. Initial attempts to use targeted<br />
modalities in DIPG treatment, relying by necessity on<br />
targets identified in adult glioblastoma rather than those<br />
specifically identified in DIPG, have shown no efficacy. 5-7<br />
Clearly, a better understanding <strong>of</strong> DIPG biology is needed.<br />
Initial Molecular Studies <strong>of</strong> DIPG<br />
In one <strong>of</strong> the earliest molecular genetic studies <strong>of</strong> DIPG,<br />
gene sequencing showed p53 mutation in five <strong>of</strong> seven DIPG<br />
(71%) samples analyzed. Portions <strong>of</strong> 17p that included the<br />
p53 gene were lost in four; none showed EGFR amplification.<br />
The same small series noted loss <strong>of</strong> the long arm <strong>of</strong><br />
PDGFR. Based on these and other findings, murine models<br />
have been generated and provide a new tool for preclinical<br />
testing. DIPG biopsy at diagnosis has increasingly become<br />
incorporated into national protocols at several centers, bringing<br />
the prospect <strong>of</strong> a better understanding <strong>of</strong> DIPG biology in<br />
the future. Initial analyses <strong>of</strong> pretreatment tumors cast valuable<br />
new light and establish the importance <strong>of</strong> p53 inactivation<br />
and the RTK-PI3K pathway in this disease.<br />
chromosome 10 in four cases. 8 Similarly, p53 mutations<br />
occurring in 8 <strong>of</strong> 13 (61%) high-grade pontine glioma specimens<br />
obtained at autopsy have also been reported. 9 An<br />
examination <strong>of</strong> 28 malignant brainstem gliomas, including<br />
18 pretreatment and 10 postmortem specimens identified<br />
TP53 mutation in six tumors (21%) while EGFR amplification<br />
was seen in three (11%). EGFRvIII mutations were not<br />
seen, and no correlation was seen between TP53 and EGFR<br />
abnormalities. 10<br />
Postmortem Findings<br />
Several more recent studies have evaluated series <strong>of</strong> DIPG<br />
samples collected at autopsy. 11-13 Tissue samples collected<br />
at autopsy have been shown to yield DNA and RNA <strong>of</strong><br />
quality sufficient for genomic analysis in the majority <strong>of</strong><br />
cases. 16 However, autopsy findings must be interpreted with<br />
caution. Significant alterations may occur in tumor biology<br />
between the time <strong>of</strong> initial presentation and post-treatment<br />
recurrence or progression. 13 Radiation and chemotherapy<br />
can select for new genomic aberrations at the time <strong>of</strong><br />
recurrence and loss <strong>of</strong> targets present at diagnosis. 17 Discovered<br />
targets from autopsy specimens, though arguably<br />
relevant to the treatment <strong>of</strong> progressive advanced-stage<br />
disease, may be less so to new-onset disease. Some <strong>of</strong> the<br />
studies described here do include a small number <strong>of</strong> upfront<br />
biopsy specimens, 13 although these were cases in which<br />
atypical features were present and were the reason these<br />
tumors underwent biopsy. The extent to which these molecular<br />
findings are representative <strong>of</strong> the biology <strong>of</strong> classic<br />
newly diagnosed DIPG is therefore uncertain.<br />
A genome-wide analysis <strong>of</strong> 11 DIPGs, including 9 postmortem<br />
and 2 pretreatment samples, were analyzed by hybridized<br />
single-nucleotide polymorphism (SNP) array, with PCR<br />
confirmation. 11 PDGFR amplification was seen in four tumors<br />
(36%). Chromosomes commonly showing large areas <strong>of</strong><br />
gain included 1q, 9q, 17q, and 10p. Loss <strong>of</strong> chromosomes<br />
11p, 17p, 14q, 18p, and 22q was also common. TP53, on<br />
chromosome 17p, demonstrated hemizygous deletion in<br />
seven (64%) <strong>of</strong> 11 DIPGs. PTEN deletion was seen in three<br />
From the Dana-Farber Children’s Hospital Cancer Center, Boston, MA.<br />
Authors’ disclosures <strong>of</strong> potential conflicts <strong>of</strong> interest are found at the end <strong>of</strong> this article.<br />
Address reprint requests to Mark W. Kieran, MD, PhD, Pediatric Medical Neuro-<br />
<strong>Oncology</strong>, Dana-Farber Children’s Hospital Cancer Center, 450 Brookline Avenue, Rm<br />
SW331, Boston, MA 02215; email: mark_kieran@dfci.harvard.edu.<br />
© <strong>2012</strong> by <strong>American</strong> <strong>Society</strong> <strong>of</strong> <strong>Clinical</strong> <strong>Oncology</strong>.<br />
1092-9118/10/1-10<br />
625
(27%), including one <strong>of</strong> two pretreatment specimens. Loss <strong>of</strong><br />
heterozygosity was seen <strong>of</strong> genes involved in DNA repair<br />
pathways, including RPA1 and MYH1 on 17p, MNAT1 and<br />
RAD51L1 on 14q, and MSH4 on 1p, among others. Low-level<br />
amplification <strong>of</strong> PARP-1, involved in DNA repair, was seen<br />
in three (27%) cases.<br />
DIPG specimens obtained at autopsy from 11 patients<br />
showed gain <strong>of</strong> MDM4, an inhibitor <strong>of</strong> p53 activity, in seven<br />
cases (64%). Immunohistochemistry for p53 was positive in<br />
four (36%). Abnormalities <strong>of</strong> genes involved in the RTK-<br />
PI3K pathway were common, including amplification <strong>of</strong><br />
EGFR in three (27%), PDGFRA in two (18%), and IRS2 in<br />
one, and loss <strong>of</strong> PTEN (associated with loss <strong>of</strong> chromosome<br />
10q) in six (54%). Gains <strong>of</strong> 1q, 7q, and 7p were common. 12<br />
A larger recent study similarly revealed focal amplifications<br />
in the PI3K pathway, most commonly involving PDG-<br />
FRA (30%) and MET (26%), in 20 <strong>of</strong> 43 (47%) DIPGs. 13<br />
Thirty percent <strong>of</strong> tumors had focal amplifications <strong>of</strong> genes<br />
regulating retinoblastoma (RB) protein phosphorylation,<br />
most commonly CDK4, CDK6, and CCND2. Notably, 21%<br />
had concurrent amplification <strong>of</strong> both PI3K and RB signaling<br />
pathways. In one case, for which autopsy specimen both<br />
from primary tumor and from an area <strong>of</strong> cerebellar extension<br />
were analyzed, significant heterogeneity was observed<br />
in copy number alterations within the same tumor. It is<br />
unclear whether this same heterogeneity would be seen in a<br />
tumor prior to treatment. Intratumoral biologic heterogeneity<br />
is one <strong>of</strong> the limitations <strong>of</strong> a personalized medicine<br />
approach. On the other hand, this may simply highlight the<br />
limitations <strong>of</strong> molecular data from post-treatment autopsy<br />
specimens.<br />
Very recently, whole-genome sequencing <strong>of</strong> DNA from<br />
seven DIPGs with matched normal tissue showed mutations<br />
in histone H3, is<strong>of</strong>orms H3.3 or H3.1, resulting in substitution<br />
to methionine at lysine 27, in five <strong>of</strong> seven patients. 14 In<br />
a validation cohort, similar mutations were found in 78% <strong>of</strong><br />
DIPGs and 22% <strong>of</strong> non-brainstem pediatric glioblastomas.<br />
Similar mutations have not been seen in adult glioblastoma.<br />
It is hypothesized that the mutation results in a gain-<strong>of</strong>function<br />
phenotype, affecting epigenetic gene expression<br />
regulation. Of note, H3F3A, the gene encoding the more<br />
KEY POINTS<br />
● DIPG diagnosis is based on radiographic and clinical<br />
criteria rather than biopsy because <strong>of</strong> concerns related<br />
to the morbidity <strong>of</strong> biopsy in the pons.<br />
● Current treatment approaches include focal radiation<br />
therapy and experimental agents based on pathways<br />
identified in adult glioblastoma multiforme.<br />
● No progress has been made in the event-free or<br />
overall survival <strong>of</strong> DIPG over the last four decades.<br />
● Advances in neurosurgical techniques and molecular<br />
pr<strong>of</strong>iling <strong>of</strong> small samples have resulted in the development<br />
<strong>of</strong> national protocols that include pretherapy<br />
biopsy and molecular pr<strong>of</strong>iling <strong>of</strong> DIPGs.<br />
● Specific targets in pretherapy DIPGs have been identified,<br />
and in conjunction with new preclinical models<br />
<strong>of</strong> DIPG, can now be tested with direct translation to<br />
the clinic.<br />
626<br />
commonly affected H3.3 is<strong>of</strong>orm, is located on chromosome<br />
1q, a region <strong>of</strong> large-scale chromosomal gain in more than<br />
20% <strong>of</strong> DIPGs, as well as <strong>of</strong> other pediatric high-grade<br />
gliomas.<br />
An important area <strong>of</strong> target discovery is the identification<br />
<strong>of</strong> tumor-specific cell surface markers, which may not drive<br />
malignant behavior, but which can be used to localize<br />
delivery <strong>of</strong> cytotoxic agents or induce an immune response.<br />
In one such study, increased staining <strong>of</strong> interleukin-13<br />
receptor alpha2 chain (IL-13Ralpha2), which is not expressed<br />
at significant levels in normal brain, was seen in 17<br />
<strong>of</strong> 28 (61%) brainstem gliomas. 15 Increased expression <strong>of</strong><br />
IL-13Ralpha2 is also seen in adult high-grade gliomas.<br />
Pretreatment Biopsy Analysis Series<br />
ROBISON AND KIERAN<br />
Developments in neurosurgical technique, as well as developments<br />
in tissue processing and extraction techniques,<br />
which have decreased the amount <strong>of</strong> tissue necessary for<br />
analysis, have significantly decreased the potential risks <strong>of</strong><br />
DIPG biopsy while increasing the information obtained.<br />
Experience from several European countries where biopsies<br />
are performed as part <strong>of</strong> a formal clinical trial has shown<br />
biopsy <strong>of</strong> DIPG to be safe and feasible. 18 This has important<br />
implications for our prospects <strong>of</strong> gaining a better understanding<br />
<strong>of</strong> this disease.<br />
In a high-grade glioma study that included seven pretreatment<br />
DIPGs, high-resolution analysis <strong>of</strong> genomic imbalances<br />
showed PDGR amplification in two tumors (29%). 19 In<br />
another small study <strong>of</strong> formalin-fixed paraffin-embedded<br />
tissue from 13 DIPGs, 10 <strong>of</strong> which were pretreatment<br />
samples, high-resolution 244 K oligo array comparative<br />
genomic hybridization identified PDGFRA amplification<br />
(4q11–13), confirmed by qPCR, in 2 DIPGs (15%). One tumor<br />
had amplification <strong>of</strong> the cell cycle regulator cyclin D1<br />
(11q13). CDKN2A/CDKN2B deletions were notably lacking,<br />
as was EGFR amplification. One <strong>of</strong> 12 scorable tumors<br />
showed MYCN amplification. 1q gain, which is seen a<br />
diverse array <strong>of</strong> both the central nervous system and other<br />
pediatric malignancies, was seen in 3 tumors (23%); in 2,<br />
this was associated with 1p loss. Loss <strong>of</strong> 17p, the site <strong>of</strong> the<br />
well-characterized tumor suppressor gene p53, was seen in<br />
four tumors (31%); loss <strong>of</strong> 14q was likewise seen in four. 17<br />
Included on each <strong>of</strong> these chromosomal arms (1p, 17p, 14q)<br />
are a number <strong>of</strong> genes involved in DNA repair pathways,<br />
previously noted to be lost in DIPG. 11 Loss <strong>of</strong> 10q, which is<br />
the most common aberration in adult GBM, and relatively<br />
common in pediatric high-grade glioma (HGG), was only<br />
seen in one DIPG. One tumor had a balanced genome, with<br />
no aberrations identified by aCGH. 17<br />
Very recently, we have reported on a mutational analysis<br />
<strong>of</strong> 20 classic DIPG tumors all biopsied at diagnosis. 20 Using<br />
OncoMap, a mass spectrometric method <strong>of</strong> allele detection<br />
that analyzes for the presence <strong>of</strong> 983 different mutations in<br />
115 oncogenes, nine <strong>of</strong> 20 (45%) tumors were found to have<br />
no identifiable mutations. In keeping with findings <strong>of</strong> earlier<br />
studies, eight (40%) tumors were found to harbor TP53<br />
mutations. As OncoMap assays for only the seven most<br />
frequent TP53 mutations, this may underestimate the TP53<br />
mutation prevalence. Interestingly, five <strong>of</strong> eight identified<br />
TP53 mutations were at the 273 arginine locus. PI3KCA<br />
mutations were seen in three tumors, which is the first<br />
report <strong>of</strong> a mutated oncogene in newly diagnosed DIPG<br />
samples. Additional PI3K pathway-related abnormalities
NOVEL TARGETS IN THE TREATMENT OF DIPG<br />
included ATM and MPL mutations, both in the same tumor,<br />
and PDGFR amplification in an additional three (15%)<br />
TP53-mutated tumors. The presence <strong>of</strong> detectable PI3Krelated<br />
abnormalities in seven (35%) tumors suggests that<br />
the PI3K pathway may play an important role in the genesis<br />
<strong>of</strong> this disease. PTEN deletions (on chromosome 10q) were<br />
seen in two tumors. Notably absent were mutations in many<br />
genes commonly implicated in malignant gliomas and other<br />
pediatric tumors, including RB1, EGFR, MET, CTNNB1, N-,<br />
H-, or K-RAS, MLH1, EPHA genes and the tyrosine-kinase<br />
domain <strong>of</strong> KIT or PDGFRA. 20 Expression analysis by immunohistochemistry<br />
in an overlapping cohort showed EGFR<br />
positivity without EGFR amplification in eight <strong>of</strong> 20 patients<br />
(40%) and absent PTEN expression in the majority. This<br />
reinforces the emerging importance <strong>of</strong> the RTK-PI3K pathway<br />
in this disease. 21<br />
The potential <strong>of</strong> identifying targetable mutations with<br />
high-throughput mutational pr<strong>of</strong>iling suggests a new clinical<br />
utility to biopsy at diagnosis. In the emerging era <strong>of</strong><br />
personalized medicine, tissue-less diagnosis may rapidly<br />
cease to be a viable option. In the United States, a recently<br />
opened multi-institutional trial for patients with DIPG mandates<br />
tumor biopsy in all cases at the time <strong>of</strong> diagnosis, and<br />
involves a treatment program determined by specific molecular<br />
findings. The purpose <strong>of</strong> this study is fundamentally<br />
two-fold: to enable detailed molecular analysis <strong>of</strong> a larger<br />
number <strong>of</strong> primary specimens than has ever previously been<br />
possible, and to evaluate a specific molecularly guided treatment<br />
approach (<strong>Clinical</strong>Trial.gov; NCT01182350).<br />
Comparative Genomics<br />
Molecular analysis has shown significant differences between<br />
pediatric and adult HGG, 19,22,23 as well as between<br />
DIPG and other pediatric HGG. 11,13,17 EGFR amplification,<br />
for instance, the most common focal abnormality in adult<br />
high-grade glioma, appears to be relatively rare in DIPG, as<br />
in other pediatric high-grade gliomas. However, underlying<br />
similarities are also seen. 17 Proneural, proliferative, and<br />
mesenchymal expression subgroups described in adult GBM<br />
have also been described in pediatric GBM regardless <strong>of</strong><br />
site. 19 Interestingly, PDFRA amplification, which is consistently<br />
seen in a minority <strong>of</strong> DIPGs, is associated with<br />
secondary GBM in both adults and children. 17,19<br />
Preclinical Models<br />
Until very recently, DIPG research has been limited by<br />
the lack <strong>of</strong> faithful animal models. The development <strong>of</strong> a<br />
genetically engineered mouse model <strong>of</strong> brainstem glioma<br />
using the RCAS/tv-a system has recently been reported. 24<br />
Upregulation <strong>of</strong> PDGF signaling in the Nestin� cells <strong>of</strong> the<br />
subventricular zone in the fourth ventricle gives rise to<br />
dorsal pontine glioblastoma. Similarly, a mouse model <strong>of</strong><br />
DIPG generated by retroviral vector-induced upregulation <strong>of</strong><br />
PDGF signal nonspecifically in cells <strong>of</strong> the dorsolateral pons<br />
has also been reported. 25 These models provide a potential<br />
tool for preclinical testing in DIPG, which had previously not<br />
been possible.<br />
A recent preclinical study highlights the potential role <strong>of</strong> a<br />
unique pontine postnatal neural progenitor cell population<br />
in the genesis <strong>of</strong> DIPG. A human Nestin� and Olig2�<br />
neural progenitor cell population that is unique to the<br />
ventral pons and immunophenotypically similar to DIPG<br />
has generated significant interest in the cell type that gives<br />
rise to DIPG. 26 This cell population wanes after infancy,<br />
but then peaks again at 6 years, the age at which DIPG<br />
incidence is highest. A similar Olig2� postnatal cell progenitor<br />
cell population was noted in the mouse pons, and was<br />
characterized by Hedgehog (Hh) pathway activation. Selective<br />
further upregulation <strong>of</strong> the Hh pathway in these Olig-2<br />
positive cells resulted in postnatal DIPG-like pontine hypertrophy,<br />
with proliferation <strong>of</strong> PDGFR-alpha� oligodendrocyte<br />
precursor cells, although without dysplastic<br />
transformation. Upregulation <strong>of</strong> Hh pathway in neurospheres<br />
from a DIPG autopsy-derived cell line cells resulted<br />
in increased self-renewal. These findings suggest that DIPG<br />
may represent a deregulation <strong>of</strong> Hh-activated neuronal<br />
progenitor cells, with additional molecular events causing<br />
malignant transformation.<br />
Conclusion<br />
Any conclusion regarding the biology <strong>of</strong> DIPG must, at<br />
this point, be considered preliminary. Existing series are<br />
small; with the one exception, all rely at least partially if<br />
not exclusively on autopsy specimens. Nonetheless, there<br />
are certain consistent findings that provide an important<br />
window into the biology <strong>of</strong> this disease. P53 loss or mutation<br />
is clearly present in a substantial portion <strong>of</strong> tumors. The<br />
RTK-Ras-PI3K-Akt pathway is affected in a substantial<br />
portion <strong>of</strong> tumors as well. The prevalence <strong>of</strong> PTEN loss<br />
suggests that mTOR may be a valid target for further<br />
study. 27 The identification <strong>of</strong> a possible Hh-dependent cancer<br />
stem cell provides an intriguing additional potential<br />
target, and raises the possibility <strong>of</strong> rational multitarget<br />
combinations. Significant differences between DIPG and<br />
both pediatric and adult HGG confirm that DIPG treatment<br />
cannot simply be based on the biology or treatment <strong>of</strong><br />
nonbrainstem HGG. 28 Many important questions remain<br />
unanswered. Ongoing collection and analysis <strong>of</strong> pretreatment<br />
tumors using advanced molecular techniques such as<br />
next-generation sequence and epigenetic pr<strong>of</strong>iling will be<br />
vital. 20<br />
The question may rightly be asked whether identification<br />
<strong>of</strong> drugable targets will be sufficient to alter outcome <strong>of</strong><br />
DIPG. Adult high-grade glioma remains a poor-prognosis<br />
disease despite the abundance <strong>of</strong> biologic data already<br />
available. Indeed, for most cancers, with some notable exceptions,<br />
the initial use <strong>of</strong> targeted strategies has not resulted<br />
in the immediate dramatic changes in outcome that<br />
some might have hoped. However, molecular targeting as a<br />
therapeutic strategy is still in its infancy. Identification <strong>of</strong><br />
genomic aberrations and overexpressed genes is necessary<br />
but not sufficient; a deeper understanding <strong>of</strong> tumor biology<br />
is required for truly informed target selection. Other important<br />
problems, such as ensuring adequate drug delivery and<br />
overcoming drug resistance, must be dealt with. Nonetheless,<br />
historic experience in such malignancies as pediatric<br />
acute lymphoblastic leukemia, which over the span <strong>of</strong> a few<br />
decades was transformed from an incurable to a largely<br />
curable disease, teaches us that innovative approaches and<br />
persistent concerted investigational efforts can ultimately<br />
result in real and lasting change. We are optimistic that new<br />
molecular information in DIPG will prove the crucial first<br />
step toward a better understanding <strong>of</strong> the biology <strong>of</strong> this<br />
disease, which, in turn, will be the crucial first step toward<br />
better treatments and a better outcome.<br />
627
Authors’ Disclosures <strong>of</strong> Potential Conflicts <strong>of</strong> Interest<br />
Employment or<br />
Leadership<br />
Positions<br />
Consultant or<br />
Advisory Role<br />
Author<br />
Nathan J. Robison*<br />
Mark W. Kieran AstraZeneca;<br />
Celgene; Infinity;<br />
Merck; Novartis<br />
*No relevant relationships to disclose.<br />
1. Hargrave D, Bartels U, Bouffet E. Diffuse brainstem glioma in children:<br />
critical review <strong>of</strong> clinical trials. Lancet Oncol. 2006;7:241-248.<br />
2. Fangusaro J. Pediatric high-grade gliomas and diffuse intrinsic pontine<br />
gliomas. J Child Neurol. 2009;24:1409-1417.<br />
3. Khatua S, Moore KR, Vats TS, et al. Diffuse intrinsic pontine gliomacurrent<br />
status and future strategies. Childs Nerv Syst. 2011;27:1391-1397.<br />
4. Cohen KJ, Heideman RL, Zhou T, et al. Temozolomide in the treatment<br />
<strong>of</strong> children with newly diagnosed diffuse intrinsic pontine gliomas: a report<br />
from the Children’s <strong>Oncology</strong> Group. Neuro-oncology. 2011;13:410-416.<br />
5. Haas-Kogan DA, Banerjee A, Poussaint TY, et al. Phase II trial <strong>of</strong><br />
tipifarnib and radiation in children with newly diagnosed diffuse intrinsic<br />
pontine gliomas. Neuro-oncology. 2011;13:298-306.<br />
6. Pollack IF, Stewart CF, Kocak M, et al. A phase II study <strong>of</strong> gefitinib and<br />
irradiation in children with newly diagnosed brainstem gliomas: a report from<br />
the Pediatric Brain Tumor Consortium. Neuro-oncology. 2011;13:290-297.<br />
7. Pollack IF, Jakacki RI, Blaney SM, et al. Phase I trial <strong>of</strong> imatinib in<br />
children with newly diagnosed brainstem and recurrent malignant gliomas: a<br />
Pediatric Brain Tumor Consortium report. Neuro-oncology. 2007;9:145-160.<br />
8. Louis DN, Rubio MP, Correa KM, et al. Molecular genetics <strong>of</strong> pediatric<br />
brain stem gliomas. Application <strong>of</strong> PCR techniques to small and archival<br />
brain tumor specimens. J Neuropathol Exp Neurol. 1993;52:507-515.<br />
9. Zhang S, Feng X, Koga H, et al. p53 gene mutations in pontine gliomas<br />
<strong>of</strong> juvenile onset. Biochem Biophys Res Commun. 1993;196:851-857.<br />
10. Gilbertson RJ, Hill DA, Hernan R, et al. ERBB1 is amplified and<br />
overexpressed in high-grade diffusely infiltrative pediatric brain stem glioma.<br />
Clin Cancer Res. 2003;9:3620-3624.<br />
11. Zarghooni M, Bartels U, Lee E, et al. Whole-genome pr<strong>of</strong>iling <strong>of</strong><br />
pediatric diffuse intrinsic pontine gliomas highlights platelet-derived growth<br />
factor receptor alpha and poly (ADP-ribose) polymerase as potential therapeutic<br />
targets. J Clin Oncol. 2010;28:1337-1344.<br />
12. Warren KE, Killian K, Suuriniemi M, et al. Genomic aberrations in<br />
pediatric diffuse intrinsic pontine gliomas. Neuro-oncology. 2011. doi:10.1093/<br />
neuonc/nor190.<br />
13. Paugh BS, Broniscer A, Qu C, et al. Genome-wide analyses identify<br />
recurrent amplifications <strong>of</strong> receptor tyrosine kinases and cell-cycle regulatory<br />
genes in diffuse intrinsic pontine glioma. J Clin Oncol. 2011;29:3999-4006.<br />
14. Wu G, Broniscer A, McEachron TA, et al. Somatic histone H3 alterations<br />
in pediatric diffuse intrinsic pontine gliomas and non-brainstem glioblastomas.<br />
Nat Genet. <strong>2012</strong>. doi:10.1038/ng. 1102.<br />
15. Joshi BH, Puri RA, Leland P, et al. Identification <strong>of</strong> interleukin-13<br />
628<br />
Stock<br />
Ownership Honoraria<br />
REFERENCES<br />
Research<br />
Funding<br />
Advantagene;<br />
Amersham;<br />
AstraZeneca;<br />
Celgene; Merck<br />
KGaA; Novartis;<br />
Schering-Plough;<br />
Transmolecular<br />
Expert<br />
Testimony<br />
ROBISON AND KIERAN<br />
Other<br />
Remuneration<br />
receptor alpha2 chain overexpression in situ in high-grade diffusely infiltrative<br />
pediatric brainstem glioma. Neuro-oncology. 2008;10:265-274.<br />
16. Broniscer A, Baker JN, Baker SJ, et al. Prospective collection <strong>of</strong> tissue<br />
samples at autopsy in children with diffuse intrinsic pontine glioma. Cancer.<br />
2010;116:4632-4637.<br />
17. Barrow J, Adamowicz-Brice M, Cartmill M, et al. Homozygous loss <strong>of</strong><br />
ADAM3A revealed by genome-wide analysis <strong>of</strong> pediatric high-grade glioma<br />
and diffuse intrinsic pontine gliomas. Neuro-oncology. 2011;13:212-222.<br />
18. Roujeau T, Machado G, Garnett MR, et al. Stereotactic biopsy <strong>of</strong> diffuse<br />
pontine lesions in children. J Neurosurg. 2007;107:1-4.<br />
19. Paugh BS, Qu C, Jones C, et al. Integrated molecular genetic pr<strong>of</strong>iling<br />
<strong>of</strong> pediatric high-grade gliomas reveals key differences with the adult disease.<br />
J Clin Oncol. 2010;28:3061-3068.<br />
20. Grill J, Puget S, Andreiuolo F, et al. Critical oncogenic mutations in<br />
newly diagnosed pediatric diffuse intrinsic pontine glioma. Pediatr Blood<br />
Cancer. 2011. doi:10.1002/pbc. 24060.<br />
21. Geoerger B, Hargrave D, Thomas F, et al. Innovative therapies for<br />
children with cancer pediatric phase I study <strong>of</strong> erlotinib in brainstem glioma<br />
and relapsing/refractory brain tumors. Neuro-oncology. 2011;13:109-118.<br />
22. Faury D, Nantel A, Dunn SE, et al. Molecular pr<strong>of</strong>iling identifies<br />
prognostic subgroups <strong>of</strong> pediatric glioblastoma and shows increased YB-1<br />
expression in tumors. J Clin Oncol. 2007;25:1196-1208.<br />
23. Bax DA, Mackay A, Little SE, et al. A distinct spectrum <strong>of</strong> copy number<br />
aberrations in pediatric high-grade gliomas. Clin Cancer Res. 2010;16:3368-<br />
3377.<br />
24. Becher OJ, Hambardzumyan D, Walker TR, et al. Preclinical evaluation<br />
<strong>of</strong> radiation and perifosine in a genetically and histologically accurate<br />
model <strong>of</strong> brainstem glioma. Cancer Res. 2010;70:2548-2557.<br />
25. Masui K, Suzuki SO, Torisu R, et al. Glial progenitors in the brainstem<br />
give rise to malignant gliomas by platelet-derived growth factor stimulation.<br />
Glia. 2010;58:1050-1065.<br />
26. Monje M, Mitra SS, Freret ME, et al. Hedgehog-responsive candidate<br />
cell <strong>of</strong> origin for diffuse intrinsic pontine glioma. Proc Natl Acad Sci USA.<br />
2011;108:4453-4458.<br />
27. Geoerger B, Kieran MW, Grupp S, et al. Phase II trial <strong>of</strong> temsirolimus<br />
in children with high-grade glioma, neuroblastoma and rhabdomyosarcoma.<br />
Eur J Cancer. <strong>2012</strong>;48:253-262.<br />
28. Gururangan S, Chi SN, Young Poussaint T, et al. Lack <strong>of</strong> efficacy <strong>of</strong><br />
bevacizumab plus irinotecan in children with recurrent malignant glioma and<br />
diffuse brainstem glioma: a Pediatric Brain Tumor Consortium study. J Clin<br />
Oncol. 2010;28:3069-3075.
Is Biopsy Safe in Children with Newly<br />
Diagnosed Diffuse Intrinsic Pontine Glioma?<br />
By Stephanie Puget, MD, PhD, Thomas Blauwblomme, MD, and Jacques Grill, MD, PhD<br />
Overview: Diffuse intrinsic pontine gliomas (DIPGs), with a<br />
median survival <strong>of</strong> 9 months, represent the biggest therapeutic<br />
challenge in pediatric neuro-oncology. Despite many clinical<br />
trials, no major improvements in treatment have been made<br />
over the past 30 years. In most cases, biopsy is not needed for<br />
diagnosis because DIPG diagnosis is based on a typical<br />
clinical picture with radiologic evidence on magnetic resonance<br />
imaging. Therefore, little data on newly diagnosed DIPG<br />
have been published and are confounded by including autopsy<br />
(i.e., postradiation therapy) cases. In most cancers, advancing<br />
to cure has been linked to the discovery <strong>of</strong> relevant biomarkers,<br />
only found by access to tissue. Therefore, to further<br />
understand the biology <strong>of</strong> DIPG, fresh tissue samples must be<br />
DIFFUSE INTRINSIC pontine gliomas (DIPGs), which<br />
represent 15% <strong>of</strong> all childhood brain tumors, are inoperable<br />
neoplasms. Response to radiation therapy is only<br />
transient and chemotherapy has not improved long-term<br />
survival. The development <strong>of</strong> targeted therapies for these<br />
tumors has been hampered by the lack <strong>of</strong> knowledge <strong>of</strong> their<br />
biology and many trials to date have been carried out based<br />
on the misconception that DIPG biology is similar either to<br />
its adult counterparts 1-3 or to other pediatric supratentorial<br />
malignant gliomas. 1,4 Diagnosis is usually based on the<br />
association <strong>of</strong> a short history <strong>of</strong> less than 2 months, cranial<br />
nerves palsies, long-tract signs, and ataxia with typical<br />
imaging findings. DIPG is usually described as an infiltrating<br />
tumor mass <strong>of</strong> the pons, hypointense on T1 and hyperintense<br />
on T2 and fluid-attenuated inversion recovery; by<br />
definition, at least 50% <strong>of</strong> the pons should be involved.<br />
Contrast enhancement, if any, is usually limited and annular.<br />
Grading <strong>of</strong> these lesions has been difficult based on<br />
small biopsies and could therefore not be linked to outcome.<br />
Biopsy <strong>of</strong> these tumors has been controversial, and most<br />
neurosurgical teams limit the use biopsy to patients with<br />
lesions that have unusual presentation. However, the urgent<br />
need to improve the prognosis for patients with these<br />
devastating tumors has led to the reconsideration <strong>of</strong> the role<br />
<strong>of</strong> stereotactic biopsy for patients with DIPG. This article<br />
will address the feasibility and safety <strong>of</strong> stereotactic biopsy<br />
for patients with DIPG, its diagnostic yield, and its role in<br />
redefining this tumor by its molecular signature and pr<strong>of</strong>iling<br />
targeted therapy.<br />
Is It Safe to Perform a Biopsy for Patients Newly<br />
Diagnosed with DIPG?<br />
Stereotactic biopsies are now completely integrated in the<br />
diagnosis and management <strong>of</strong> several intracranial lesions.<br />
The role <strong>of</strong> stereotactic biopsy for patients with newly<br />
diagnosed DIPG remains controversial, and currently the<br />
general attitude is not to biopsy the tumors <strong>of</strong> these patients.<br />
Stereotactic biopsy <strong>of</strong> brainstem tumors is an old procedure;<br />
it became popular after the first report <strong>of</strong> this procedure<br />
in 1978. 5 Ten years later, arguments against brainstem<br />
biopsy were strong because it was believed to have no use, to<br />
be dangerous, and to <strong>of</strong>fer poor yield. 6-8 The 1993 manuscript<br />
published by Albright and colleagues changed the<br />
course <strong>of</strong> pediatric DIPG management by claiming that<br />
obtained at diagnosis. However, most neurosurgical teams are<br />
reluctant to perform biopsy in pediatric patients, citing potential<br />
risks and lack <strong>of</strong> direct benefit. Yet, in reviewing 90<br />
patients with and the published data on brainstem biopsy,<br />
these procedures have a diagnostic yield and morbidity and<br />
mortality rates similar to those reported for other brain<br />
locations. In addition, the quality and quantity <strong>of</strong> the material<br />
obtained confirm the diagnosis and inform an extended molecular<br />
screen, including biomarker study—information important<br />
to designing next-generation trials with targeted agents.<br />
Stereotactic biopsies can be considered a safe procedure in<br />
well-trained neurosurgical teams and could be incorporated in<br />
well-defined protocols for patients with DIPG.<br />
magnetic resonance imaging (MRI) “. . . scans provide images<br />
that are virtually diagnostic <strong>of</strong> brainstem gliomas and<br />
yield prognostic information equivalent to that obtainable<br />
from biopsies ...”. 8 Since then, the neurosurgical world was<br />
divided into “for” and “against” the brainstem biopsy. Despite<br />
the reluctance <strong>of</strong> some neurosurgical teams, others<br />
chose to perform biopsies <strong>of</strong> brainstem lesions in children<br />
and adults for both unusual lesions and typical ones as part<br />
<strong>of</strong> a clinical trial. 9-16 Papers on stereotactic biopsies in the<br />
brainstem published in the last 20 years represent a substantial<br />
amount <strong>of</strong> knowledge, yet unfortunately <strong>of</strong>ten report<br />
mixed series <strong>of</strong> adults and children with a wide range <strong>of</strong><br />
diagnoses. These mixed series cite morbidity rates between<br />
0% and 10% and mortality rates between 0% to 3% (Table 1).<br />
However, when biopsy data on pediatric patients with DIPG<br />
are extracted, the diagnostic yield ranges from 96% to 100%,<br />
with no mortality and morbidity less than 5% for the largest<br />
series.<br />
Samadani and Judy performed a meta-analysis <strong>of</strong> 13<br />
studies <strong>of</strong> stereotactic biopsy <strong>of</strong> brainstem lesions in 381<br />
children and adults. 17 With a diagnostic yield <strong>of</strong> 96%, this<br />
study reported one death attributable to a biopsy <strong>of</strong> a<br />
vascular lesion in an adult, with rates <strong>of</strong> permanent and<br />
transient neurologic deficits <strong>of</strong> 4% and 1%, respectively. A<br />
few years later, a second meta-analysis on brainstem lesions<br />
in pediatric patients was published by Pincus and colleagues.<br />
13 This review <strong>of</strong> 192 children revealed a diagnostic<br />
yield <strong>of</strong> 94.9%, and mortality and morbidity rates <strong>of</strong> 0.7%<br />
and 4.9%, respectively. Recently, Rajshekhar and Moorthy<br />
reported a series <strong>of</strong> stereotactic biopsies in 106 children with<br />
brainstem masses. With no mortality or permanent morbidity<br />
reported, the authors highlighted that “. . . this procedure<br />
is safe in children and the benefits outweigh the risks<br />
in patients who are appropriately selected to undergo this<br />
procedure. . . ” 11 A few years ago, our group started to use<br />
From the Necker Enfants Malades Hospital, Université Paris Descartes, Sorbonne Paris<br />
Cité, France; Gustave Roussy Cancer Institute, Universite Paris Sud, Villejuif, France.<br />
Authors’ disclosures <strong>of</strong> potential conflicts <strong>of</strong> interest are found at the end <strong>of</strong> this article.<br />
Address reprint requests to Stephanie Puget, MD, PhD, Department <strong>of</strong> Pediatric Neurosurgery,<br />
Necker Enfants Malades Hospital, 149 rue de Sèvres, 75015 Paris, Université Paris<br />
Descartes, Sorbonne Paris Cité, France; email: stephanie.puget@gmail.com<br />
© <strong>2012</strong> by <strong>American</strong> <strong>Society</strong> <strong>of</strong> <strong>Clinical</strong> <strong>Oncology</strong>.<br />
1092-9118/10/1-10<br />
629
Lead Author Year<br />
stereotactic biopsies for patients with DIPG to obtain both<br />
pathologic confirmation and immunohistochemical assessment<br />
<strong>of</strong> specific biomarkers before including patients in<br />
trials <strong>of</strong> targeted agents. 10,18,19 During the 10-year period,<br />
90 children with pontine lesions resembling DIPG received<br />
this technique at Necker Enfants Malades Hospital in Paris.<br />
All patients received corticosteroids at least 3 days before<br />
the procedure. Using the Leksell stereotactic system, a<br />
transcerebellar approach was used for all patients. The<br />
procedure was carried out with the patient in a prone<br />
position under general anesthesia and the stereotactic coordinates<br />
were determined by computed tomography or MRI.<br />
The contrast enhancement was targeted when possible (i.e.,<br />
when it was not too anteriorly close to the pyramidal tract or<br />
KEY POINTS<br />
Table 1. Brainstem Stereotactic Biopsies in Children and Mixed Patient Series<br />
No. <strong>of</strong><br />
Patients (Type) Technique (Frame and Route)<br />
● Every attempt to treat children with diffuse intrinsic<br />
pontine glioma has failed in the last 30 years.<br />
● In most cases, biopsy is not needed to diagnose a<br />
diffuse intrinsic pontine glioma because diagnosis is<br />
based on the association <strong>of</strong> a typical clinical picture<br />
and radiologic evidence on magnetic resonance imaging.<br />
● Tissue samples are required to further understand<br />
the biology <strong>of</strong> these lesions and research new therapeutic<br />
targets.<br />
● Stereotactic biopsy for patients with diffuse intrinsic<br />
pontine glioma is as safe as biopsy for patients with<br />
supratentorial lesions and have a diagnostic yield<br />
above 90%.<br />
● The quality and quantity <strong>of</strong> the material obtained<br />
allow for the confirmation <strong>of</strong> the diagnosis and for the<br />
performance <strong>of</strong> an extended molecular screen, including<br />
biomarker study, permitting children with newly<br />
diagnosed diffuse intrinsic pontine glioma to enroll<br />
in next-generation clinical trials with targeted<br />
therapies.<br />
Diagnostic<br />
Yield, % Histopathologic Diagnosis, %, Type<br />
Morbidity,<br />
n (%) Mortality, %<br />
Puget <strong>2012</strong> 90 (C) Leksell, all TC 100 100 DIPG 4 T (4.4) 0<br />
Dellaretti 2011 44 (C) Talairach, 42 TF, 2 TC 93 36 HGG<br />
34.6 LGG, miscellaneous<br />
13 (9.8) 0<br />
Rajshekhar 2010 106 (C) BRW or CRW, 77 TF, 29 TC 100 90 glioma<br />
10 inflammatory<br />
3 T (2.8) 0<br />
Pirotte 2007 20 (C) PET ( 18FDG), 3 TF, 17 TC 100 75 glioma<br />
1 T (5) 0<br />
25 others (PNET, teratoma, germinoma) 1 P (5)<br />
Pincus (meta-analysis) 2006 192 (C) CRW, TF and TC 94.9a —PNET, neurocytoma, ependymoma,<br />
vasculitis, germinoma. . .<br />
— (4.9) b 0.7c Samadani (meta-analysis) 2003 381 (A CRW BRW, Todd-Weillis, 96 31 HGG<br />
—T (4) 0.3<br />
and C) Leksell, Richert, 292 TF<br />
23 LGG<br />
10 metatasis<br />
16 hematomas and miscellaneous<br />
—P (1)<br />
d<br />
Abbreviations: —, not reported; 18 FDG, 18F-fluorodeoxyglucose; A, adults; BRW, Brown-Roberts-Wells; C, children; CRW, Cosman-Robert-Wells; HGG, high-grade<br />
glioma; LGG, low-grade glioma; P, permanent; PET, positron emission tomography; T, transient; TC, transcerebellar; TF, transfrontal.<br />
a<br />
Range, 75% to 100%<br />
b<br />
Range, 0% to 16%<br />
c<br />
0% to 3.3%<br />
d<br />
One adult patient.<br />
630<br />
PUGET, BLAUWBLOMME, AND GRILL<br />
near the nuclei <strong>of</strong> the cranial nerves). In patients with no<br />
contrast enhancement, we targeted the infiltrative part,<br />
close to the middle cerebellar peduncle (Fig. 1). Using a<br />
single trajectory, we used to take two samples at the beginning<br />
<strong>of</strong> our experience. This number was increased up to<br />
eight samples in the past 5 years without additional risk,<br />
allowing for DNA and RNA extraction. The diagnostic yield<br />
was 100% in our series, with no mortality or permanent<br />
deficit, and we observed transient worsening <strong>of</strong> neurologic<br />
deficit in four patients.<br />
From the literature and our own data, stereotactic biopsy<br />
for children with DIPG can be considered a safe procedure<br />
with high diagnostic yield (Table 1). In a large series <strong>of</strong><br />
stereotactic brain biopsy in 270 adults, it has been shown<br />
that increasing numbers <strong>of</strong> specimens obtained per trajectory<br />
and brainstem lesions were not significant risk factors<br />
for morbidity. 20<br />
Two routes have been described for brainstem biopsies:<br />
the transcerebellar approach and the transfrontal approach<br />
(Table 1, Figs. 1 and 2). The transfrontal route is longer and<br />
allows sampling <strong>of</strong> masses located in all the segments <strong>of</strong> the<br />
brainstem. The positioning <strong>of</strong> the entry site has to be chosen<br />
carefully to avoid the ventricles, the vascular structures,<br />
and the tentorium. The use <strong>of</strong> s<strong>of</strong>tware is recommended for<br />
planning this trajectory. The transcerebellar approach is<br />
shorter, through the middle cerebellar peduncle, and has<br />
less eloquent structures in its trajectory. 11,21 It can be used<br />
only for pontine and upper medullary masses, and because<br />
<strong>of</strong> this, the transcerebellar route for DIPG is preferred by the<br />
authors. In a series comparing both routes, no significant<br />
differences were reported regarding the rates <strong>of</strong> complication<br />
and diagnostic yield. 22 The use <strong>of</strong> functional MRI and<br />
nuclear medicine techniques to define biopsy targets might<br />
increase the diagnostic and prognostic value <strong>of</strong> the tissue<br />
samples obtained, but the studies comparing the two routes<br />
are mainly concentrated in adults. However, Pirotte and<br />
colleagues published a series about 20 children with infiltrative<br />
brainstem tumors in which it was suggested that the<br />
integration <strong>of</strong> metabolic information from positron emission<br />
tomography might improve the diagnostic yield <strong>of</strong> the biopsy<br />
sampling. 23
SAFETY OF DIPG BIOPSY IN CHILDREN<br />
Fig. 1. Plan neuroimaging that shows the trajectory<br />
from a trancerebellar approach. E � Entry; T � Target.<br />
(A) Three-dimensional neuroimaging plan. (B) Axial<br />
gadolinium-enhanced T1-weighted magnetic resonance<br />
image. (C) Sagittal gadolinium-enhanced T1-weighted<br />
magnetic resonance image. (D) Coronal gadoliniumenhanced<br />
T1-weighted magnetic resonance image.<br />
In addition, to increase direct application <strong>of</strong> chemotherapeutic<br />
agents to the DIPG, attempts have been pursued to<br />
circumvent the blood-brain barrier using the convectionenhanced<br />
delivery technique. Theoretically, the safety <strong>of</strong><br />
this procedure is similar to that <strong>of</strong> biopsy. Preclinical models<br />
and clinical trials have demonstrated the feasibility, efficacy,<br />
and safety <strong>of</strong> this technique. 24 Moreover, two children<br />
with brainstem lesions demonstrated no neurologic deficits<br />
after infusion. 25 Therapeutic protocols are currently under<br />
way, but the choice <strong>of</strong> the appropriate agent to deliver with<br />
this technique remains in question.<br />
Fig. 2. Plan neuroimaging showing the trajectory<br />
with a transfrontal approach. E � Entry; T � Target.<br />
(A) Three-dimensional neuroimaging plan. (B) Axial<br />
gadolinium-enhanced T1-weighted magnetic resonance<br />
image. (C) Sagittal gadolinium-enhanced T1-weighted<br />
magnetic resonance image. (D) Coronal Gadoliniumenhanced<br />
T1-weighted magnetic resonance image.<br />
Why Perform a Biopsy for Patients with Newly<br />
Diagnosed DIPG?<br />
The lack <strong>of</strong> samples from patients with newly diagnosed<br />
DIPG has limited our understanding <strong>of</strong> tumor biology, which<br />
hinders the development <strong>of</strong> newer therapies for patients<br />
with these devastating tumors. As targeted therapy development<br />
undoubtedly requires tissue, it could be argued that<br />
such advances will only be optimized with the knowledge<br />
that biopsies provide in terms <strong>of</strong> tumor biology and the<br />
identification <strong>of</strong> new targets. In recent papers, authors<br />
defend the idea that a biopsy <strong>of</strong> patients with newly diag-<br />
631
nosed DIPG should be performed to increase our knowledge<br />
<strong>of</strong> tumor biology to provide clues to improve their prognosis.<br />
9,24,26 However, for patients with typical DIPG evidence<br />
on MRI, biopsies should not be performed only to confirm the<br />
diagnosis, because the risk <strong>of</strong> the procedure, even though<br />
minimal, is not null. Biopsy should be part <strong>of</strong> a wellconducted<br />
clinical trial or a research program approved by<br />
an ethics committee. 9,26 The authors agree with the conclusions<br />
<strong>of</strong> Wilkinson and Harris, which stated that “. . . Once<br />
emotional and social interests are taken into account there<br />
seems little doubt that brainstem biopsy could be lawful<br />
even if there was no benefit to the child’s medical interests<br />
. . .”. 27<br />
Even if the amount <strong>of</strong> tumor tissue obtained by stereotactic<br />
biopsy is limited by the size <strong>of</strong> the needle, it has been<br />
shown that it could provide enough tissue for histopathologic<br />
diagnosis and immunohistochemical staining. Moreover, the<br />
authors have recently shown that this surgical technique<br />
could allow multiple biopsies samples (up to eight) to provide<br />
enough tissue for further genomic analyses and stem-cell<br />
culture. 26,28,29 In the author series, one or two biopsies were<br />
used for histologic diagnosis and immunohistochemistry.<br />
The remaining biopsies were snap-frozen with cytologic<br />
control smears within minutes <strong>of</strong> surgical removal, and<br />
nucleic acids were extracted later. A median <strong>of</strong> 3.325 microgram<br />
(microg) <strong>of</strong> DNA (range, 0.805 microg to 21.5 microg)<br />
and 2.332 microg <strong>of</strong> RNA (range, 0.048 microg to 15.84<br />
microg) could be extracted from the biopsies. One to three<br />
samples were needed to obtain enough nucleic acids, depending<br />
on the infiltrative rate <strong>of</strong> the tumor cells. An<br />
integrated molecular pr<strong>of</strong>ile was created to identify two<br />
distinct subgroups <strong>of</strong> DIPG with specific abnormalities: one<br />
showing a mesenchymal gene expression pr<strong>of</strong>ile and the<br />
other showing a more proliferative gene expression signature.<br />
The former group <strong>of</strong> tumors showed a better survival<br />
than the latter. The poor-prognosis group defined by gene<br />
expression pr<strong>of</strong>iling showed an oligodendroglial differentiation<br />
that could be correlated with an adverse prognosis in<br />
another validation cohort. In addition, this group <strong>of</strong> tumors<br />
showed amplification in platelet-derived growth factor<br />
Authors’ Disclosures <strong>of</strong> Potential Conflicts <strong>of</strong> Interest<br />
Author<br />
Stephanie Puget*<br />
Thomas Blauwblomme*<br />
Jacques Grill*<br />
*No relevant relationships to disclose.<br />
Employment or<br />
Leadership<br />
Positions<br />
Consultant or<br />
Advisory Role<br />
1. Bax DA, Mackay A, Little SE, et al. A distinct spectrum <strong>of</strong> copy number<br />
aberrations in pediatric high-grade gliomas. Clin Cancer Res. 16:3368-3377.<br />
2. Qu HQ, Jacob K, Fatet S, et al. Genome-wide pr<strong>of</strong>iling using singlenucleotide<br />
polymorphism arrays identifies novel chromosomal imbalances in<br />
pediatric glioblastomas. Neuro Oncol. 12:153-163.<br />
3. Paugh BS, Qu C, Jones C, et al. Integrated molecular genetic pr<strong>of</strong>iling <strong>of</strong><br />
pediatric high-grade gliomas reveals key differences with the adult disease.<br />
J Clin Oncol. 28:3061-3068.<br />
4. Zarghooni M, Bartels U, Lee E, et al. Whole-genome pr<strong>of</strong>iling <strong>of</strong> pediatric<br />
diffuse intrinsic pontine gliomas highlights platelet-derived growth factor<br />
receptor alpha and poly (ADP-ribose) polymerase as potential therapeutic<br />
targets. J Clin Oncol. 28:1337-1344.<br />
5. Gleason CA, Wise BL, Feinstein B. Stereotactic localization (with com-<br />
632<br />
receptor-alpha with or without mutation in the external<br />
domain and MET gain that could represent relevant therapeutic<br />
targets. This study is the first to comprehensively<br />
define the biologic alterations <strong>of</strong> DIPG at diagnosis, allowing<br />
the discovery <strong>of</strong> novel therapeutic targets. 29 The material<br />
obtained also afforded screening for oncogenic mutations<br />
with currently available targeted drugs. Our group was able<br />
to identify PI3KCA mutations in 15% <strong>of</strong> DIPG, which <strong>of</strong>fers<br />
another relevant therapeutic target. 26 Finally, cultures <strong>of</strong><br />
DIPG could be derived as cell lines or as neurospheres. 28<br />
They represent irreplaceable tools both for preclinical studies<br />
<strong>of</strong> new therapeutic agents and for understanding the<br />
oncogenesis <strong>of</strong> DIPG.<br />
Recently, there has been a worldwide resurgence <strong>of</strong> interest<br />
in pediatric brainstem biopsy in the hopes that molecular<br />
pr<strong>of</strong>iling could help to find new therapeutic targets. To this<br />
end, several international consensus conferences on DIPG<br />
have been organized in North America and Europe in the<br />
past 3 years. 30,31 One can feel that there is a growing body<br />
<strong>of</strong> evidence that biopsy for most patients with newly diagnosed<br />
DIPG is now considered rational because the procedure<br />
might alter treatment with targeted therapy and might<br />
help in correlative biology with appropriate biomarker response.<br />
Biopsy might also help to guide therapies for patients<br />
with relapsed disease, to look for active treatments,<br />
and to develop relevant biologic models. 20<br />
Conclusion<br />
DIPG remains a leading cause <strong>of</strong> death for children with<br />
brain tumors. The role <strong>of</strong> diagnostic biopsy for patients with<br />
these tumors has been controversial because <strong>of</strong> the high<br />
eloquence <strong>of</strong> the brainstem and the lack <strong>of</strong> direct benefit for<br />
the patient. Based on the literature and our own data,<br />
stereotactic biopsy for patients with DIPG is approximately<br />
as safe and diagnostic as supratentorial biopsy, and the<br />
amount <strong>of</strong> tissue obtained allows for molecular analysis.<br />
This technique should be <strong>of</strong>fered to these patients and opens<br />
new perspectives for the characterization <strong>of</strong> biomarkers that<br />
permit children with newly DIPG to enroll in nextgeneration<br />
clinical trials with targeted therapies. 26<br />
Stock<br />
Ownership Honoraria<br />
REFERENCES<br />
Research<br />
Funding<br />
PUGET, BLAUWBLOMME, AND GRILL<br />
Expert<br />
Testimony<br />
Other<br />
Remuneration<br />
puterized tomographic scanning), biopsy, and radi<strong>of</strong>requency treatment <strong>of</strong><br />
deep brain lesions. Neurosurgery. 1978;2:217-222.<br />
6. Stroink AR, H<strong>of</strong>fman HJ, Hendrick EB, et al. Diagnosis and management<br />
<strong>of</strong> pediatric brain-stem gliomas. J Neurosurg. 1986;65:745-750.<br />
7. Epstein F, McCleary EL. Intrinsic brain-stem tumors <strong>of</strong> childhood:<br />
surgical indications. J Neurosurg. 1986;64:11-15.<br />
8. Albright AL, Packer RJ, Zimmerman R, et al. Magnetic resonance scans<br />
should replace biopsies for the diagnosis <strong>of</strong> diffuse brain stem gliomas: a<br />
report from the Children’s Cancer Group. Neurosurgery. 1993;33:1026-1029.<br />
9. Leach PA, Estlin EJ, Coope DJ, et al. Diffuse brainstem gliomas in<br />
children: should we or shouldn’t we biopsy? Br J Neurosurg. 2008;22:619-624.<br />
10. Roujeau T, Machado G, Garnett MR, et al. Stereotactic biopsy <strong>of</strong> diffuse<br />
pontine lesions in children. J Neurosurg. 2007;107:1-4 (suppl).
SAFETY OF DIPG BIOPSY IN CHILDREN<br />
11. Rajshekhar V, Moorthy RK: Status <strong>of</strong> stereotactic biopsy in children<br />
with brain stem masses: insights from a series <strong>of</strong> 106 patients. Stereotact<br />
Funct Neurosurg. 2010;88:360-366.<br />
12. Chico-Ponce de Léon F, Perezpeña-Diazconti M, Castro-Sierra E, et al.<br />
Stereotactically-guided biopsies <strong>of</strong> brainstem tumors. Childs Nerv Syst.<br />
2003;19:305-310.<br />
13. Pincus DW, Richter EO, Yachnis AT, et al. Brainstem stereotactic<br />
biopsy sampling in children. J Neurosurg. 2006;104:108-114.<br />
14. Valdes-Gorcia J, Espinoza-Diaz DM, Paredes-Diaz E. Stereotactic biopsy<br />
<strong>of</strong> brain stem and posterior fossa lesions in children. Acta Neurochir<br />
(Wien). 1998;140:899-903.<br />
15. St George EJ, Walsh AR, Sgouros S. Stereotactic biopsy <strong>of</strong> brain<br />
tumours in the paediatric population. Childs Nerv Syst. 2004;20:163-167.<br />
16. Dellaretti M, Touzet G, Reyns N, et al. Correlation among magnetic<br />
resonance imaging findings, prognostic factors for survival, and histological<br />
diagnosis <strong>of</strong> intrinsic brainstem lesions in children. J Neurosurg Pediatr.<br />
8:539-543.<br />
17. Samadani U, Judy KD. Stereotactic brainstem biopsy is indicated for<br />
the diagnosis <strong>of</strong> a vast array <strong>of</strong> brainstem pathology. Stereotact Funct<br />
Neurosurg. 2003;81:5-9.<br />
18. Geoerger B, Hargrave D, Thomas F, et al. Innovative therapies for<br />
children with cancer pediatric phase I study <strong>of</strong> erlotinib in brainstem glioma<br />
and relapsing/refractory brain tumors. Neuro Oncol. 2011;13:109-118.<br />
19. Geoerger B, Morland B, Ndiaye A, et al. Target-driven exploratory<br />
study <strong>of</strong> imatinib mesylate in children with solid malignancies by the<br />
Innovative Therapies for Children with Cancer (ITCC) European Consortium.<br />
Eur J Cancer. 2009;45:2342-2351.<br />
20. McGirt MJ, Woodworth GF, Coon AL, et al. Independent predictors <strong>of</strong><br />
morbidity after image-guided stereotactic brain biopsy: a risk assessment <strong>of</strong><br />
270 cases. J Neurosurg. 2005;102:897-901.<br />
21. Backlund EO. A new instrument for stereotaxic brain tumour biopsy.<br />
Technical note. Acta Chir Scand. 1971;137:825-827.<br />
22. Dellaretti M, Reyns N, Touzet G, et al. Stereotactic biopsy for brainstem<br />
tumors: comparison <strong>of</strong> transcerebellar with transfrontal approach. Stereotact<br />
Funct Neurosurg. <strong>2012</strong>;90:79-83.<br />
23. Pirotte BJ, Lubansu A, Massager N, et al. Results <strong>of</strong> positron emission<br />
tomography guidance and reassessment <strong>of</strong> the utility <strong>of</strong> and indications for<br />
stereotactic biopsy in children with infiltrative brainstem tumors. J Neurosurg.<br />
2007;107:392-399.<br />
24. Khatua S, Moore KR, Vats TS, et al. Diffuse intrinsic pontine gliomacurrent<br />
status and future strategies. Childs Nerv Syst. 27:1391-1397.<br />
25. Lonser RR, Warren KE, Butman JA, et al. Real-time image-guided<br />
direct convective perfusion <strong>of</strong> intrinsic brainstem lesions. Technical note.<br />
J Neurosurg. 2007;107:190-197.<br />
26. Grill J, Puget S, Andreiuolo F, et al. Critical oncogenic mutations in<br />
newly diagnosed pediatric diffuse intrinsic pontine glioma. Pediatr Blood<br />
Cancer. <strong>2012</strong>;58:489-491.<br />
27. Wilkinson R, Harris J. Moral and legal reasons for altruism in the case<br />
<strong>of</strong> brainstem biopsy in diffuse glioma. Br J Neurosurg. 2008;22:617-618.<br />
28. Thirant C, Bessette B, Varlet P, et al. <strong>Clinical</strong> relevance <strong>of</strong> tumor cells<br />
with stem-like properties in pediatric brain tumors. PLoS One. 2011;6:e16375.<br />
29. Puget S, Philippe C, Bax DA, et al. Mesenchymal transition and<br />
PDGFRA amplification/mutation are key distinct oncogenic events in pediatric<br />
diffuse intrinsic pontine gliomas. PLoS One. <strong>2012</strong>;7:e30313.<br />
30. Jansen MH, van Vuurden DG, Vandertop WP, et al. Diffuse intrinsic<br />
pontine gliomas: a systematic update on clinical trials and biology. Cancer<br />
Treat Rev. <strong>2012</strong>;38:27-35.<br />
31. Bartels U, Hawkins C, Vézina G, et al. Proceedings <strong>of</strong> the diffuse<br />
intrinsic pontine glioma (DIPG) Toronto Think Tank: advancing basic and<br />
translational research and cooperation in DIPG. J Neurooncol. 2011;105:119-<br />
125.<br />
633
Ethics <strong>of</strong> Biopsy in Children with Newly<br />
Diagnosed Diffuse Intrinsic Pontine Glioma<br />
Overview: To understand the ethical dilemmas that beset<br />
this issue <strong>of</strong> biopsy in children with newly diagnosed diffuse<br />
intrinsic pontine glioma, one must understand both the history<br />
behind it and the current dominant interpretation <strong>of</strong> ethics<br />
through the medium <strong>of</strong> the institutional review boards. It is<br />
IN DIFFUSE intrinsic pontine glioma (DIPG), the stage<br />
for the controversy over biopsy was set in the early 1990s.<br />
The early optimism engendered by the <strong>of</strong>ten dramatic responses<br />
to radiation had faded and the fatal nature <strong>of</strong> DIPGs<br />
had become all too obvious. The morbidity from biopsy in<br />
that era was low, but there was no apparent benefit in<br />
looking at the tissue over diagnosis by radiology. Indeed,<br />
histology seemed more confusing than radiology given that<br />
tumor grade did not influence survival. 1 The standard diagnosis<br />
then became the “typicality” <strong>of</strong> the scan. If the scan<br />
was typical there was no need to biopsy. Oddly enough, this<br />
standard method <strong>of</strong> making a diagnosis was never standardized<br />
and there remained a huge variation in what was<br />
considered typical or atypical. 2<br />
DIPGs were then diagnosed by scan. Radiation was administered<br />
as a standard therapy and the tumors would<br />
shrink. The quality <strong>of</strong> life for many would improve for a few<br />
months. However, all children with “typical” (and, indeed,<br />
most with atypical scans) would die. Time to death was<br />
stunningly uniform across all studies, with a median <strong>of</strong> 9 to<br />
10 months.<br />
Given the lack <strong>of</strong> curative therapy, the children with this<br />
terrible disease became the subject <strong>of</strong> literally dozens <strong>of</strong><br />
phase I and II trials. These were passed easily by institutional<br />
review boards (IRBs) because there was a “possibility”<br />
<strong>of</strong> direct benefit. No IRB seemed to have reviewed the<br />
increasing evidence, added by each consecutive trial, that<br />
there was no benefit from these non–biology-directed phase<br />
I and II trials. 3 Children with this terrible disease suffered<br />
again and again from vomiting and diarrhea from yet<br />
another fruitless trial. Perhaps more than 1,000 children<br />
worldwide were entered onto these trials with no benefit and<br />
no questioning from the IRBs.<br />
By the turn <strong>of</strong> this century, DIPG deaths were making up<br />
to 30% to 40% <strong>of</strong> the deaths in the pediatric neuro-oncology<br />
program as prognosis improved for other brain tumors.<br />
There had been an explosion <strong>of</strong> biologic information about<br />
most pediatric brain tumors, but virtually nothing was<br />
known about the biology <strong>of</strong> DIPGs. There were discussions<br />
with parents and parental lead advocacy groups. A proposal<br />
was made by me at the Children’s Hospital Colorado in 2008<br />
to biopsy a limited number <strong>of</strong> children with all the costs<br />
being borne by institutional money. In the proposal it was<br />
stated that there was no direct benefit to the children<br />
concerned, instead <strong>of</strong> saying the benefit was in the order <strong>of</strong><br />
phase I or II trials. The local IRB declined to approve it<br />
citing federal regulations that procedures done on children<br />
should <strong>of</strong>fer the possibility <strong>of</strong> benefit if they were not<br />
minimal risk. 4 The IRB asked for a panel <strong>of</strong> experts to be<br />
assembled and this led to a national panel being assembled<br />
by the U.S. Food and Drug Administration (FDA). Although<br />
634<br />
By Nicholas K. Foreman, MD<br />
also important to understand that this article represents the<br />
author’s personal viewpoint. At a consensus meeting to discuss<br />
the issue <strong>of</strong> biopsies in Diffuse Intrinsic Pontine Glioma<br />
(DIPG) at the National institutes <strong>of</strong> Health held in the fall <strong>of</strong><br />
2011, there were a variety <strong>of</strong> opinions expressed.<br />
the proposal was supported by most members <strong>of</strong> the panel,<br />
the ethicists on the panel were opposed and the local IRB<br />
rejected the proposal. 5<br />
This brought up some interesting ethical issues by various<br />
panel members and participants. I am going to give some<br />
personal comments on these in the following sections<br />
First, That There Has to Be the Possibility <strong>of</strong> Some<br />
Direct Benefit to the Child to Allow a Procedure or<br />
a Therapy That May Cause More Than Minimal Harm<br />
This does not have to be quantified and may in fact be<br />
immeasurably small (such as the continuing proposals for<br />
non–biology-based phase I and II trials in this tumor).<br />
Indeed, it would appear it could even be speculative. IRBs<br />
(and apparently ethicists) considered surgical procedures to<br />
be in a different class from chemotherapy regimens, although<br />
the latter may inflict weeks <strong>of</strong> misery on a child. The<br />
French data showing a low risk to the procedure was<br />
insufficient in the view <strong>of</strong> the ethicists. 6<br />
So, if one said that the biopsy would include looking for<br />
v600e mutation <strong>of</strong> BRAF or a H3F3A mutation and that the<br />
child could at progression be considered a candidate for<br />
therapy targeting these mutations, this might pass muster—even<br />
though there is no evidence these agents are<br />
effective in childhood brain tumors.<br />
Second, That Societal or Family Benefit Should Not<br />
Be Regarded in a Discussion <strong>of</strong> the Ethics <strong>of</strong><br />
Performing Biopsy on a Child<br />
One <strong>of</strong> the ethicists said that consideration <strong>of</strong> societal or<br />
family benefit was more in line with European ethics than<br />
<strong>American</strong>. The basis for this statement is difficult to find and<br />
perhaps lies more in the interpretation <strong>of</strong> ethics. Family<br />
members at the meeting said the death <strong>of</strong> their child seemed<br />
in vain, that nothing was learned and more children would<br />
go on to die as a result <strong>of</strong> this terrible disease. They felt that<br />
both they and the child (when older) would have had comfort<br />
from knowing that others would be less likely to suffer. One<br />
<strong>of</strong> the panel members implicitly stated that the family had<br />
no rights in this matter and could not consent to their child’s<br />
undergoing a procedure without direct benefit even when<br />
the child had a fatal disease. Some <strong>of</strong> the parents were<br />
From the University <strong>of</strong> Colorado, Aurora, CO.<br />
Author’s disclosures <strong>of</strong> potential conflicts <strong>of</strong> interest are found at the end <strong>of</strong> this article.<br />
Address reprint requests to Nicholas K. Foreman, MD, Pr<strong>of</strong>essor <strong>of</strong> Pediatrics, Seebaum-<br />
Tschetter Chair <strong>of</strong> Neuro-<strong>Oncology</strong>, University <strong>of</strong> Colorado, 13123 E. 16 th Ave., B115,<br />
Aurora, CO 80045; email: nicholas.foreman@childrenscolorado.org.<br />
© <strong>2012</strong> by <strong>American</strong> <strong>Society</strong> <strong>of</strong> <strong>Clinical</strong> <strong>Oncology</strong>.<br />
1092-9118/10/1-10
CHILDREN WITH DIFFUSE INTRINSIC PONTINE GLIOMA<br />
outraged by this and considered this a paternalistic void <strong>of</strong><br />
their rights.<br />
Third, That All Alternatives to Performing Biopsies on<br />
Children Should Be Considered<br />
Basically what was being asked was whether there some<br />
way <strong>of</strong> advancing knowledge and therapy without performing<br />
biopsies, which would seem absolutely uncontroversial.<br />
This statement and the failure <strong>of</strong> this proposal put many <strong>of</strong><br />
the parental advocacy groups firmly behind an effort to<br />
obtain autopsy samples. This effort was strikingly successful<br />
and has resulted in great advances in the knowledge base for<br />
this disease. However it should be noted even in the recent<br />
Nature Genetics article showing a potential driving muta-<br />
KEY POINTS<br />
● Diffuse intrinsic pontine gliomas are almost uniformly<br />
fatal.<br />
● There is considerable controversy over the ethics <strong>of</strong><br />
biopsying the tumor.<br />
● The issue <strong>of</strong> direct benefit is central to this controversy.<br />
● The issue <strong>of</strong> whether familial or societal benefit<br />
should be considered when a biopsy on a child is<br />
contemplated is controversial.<br />
● The issue <strong>of</strong> who declares a procedure standard is<br />
uncertain.<br />
Author’s Disclosures <strong>of</strong> Potential Conflicts <strong>of</strong> Interest<br />
Author<br />
Nicholas K. Foreman*<br />
*No relevant relationships to disclose.<br />
Employment or<br />
Leadership<br />
Positions<br />
Consultant or<br />
Advisory Role<br />
1. Albright AL, Packer RJ, Zimmerman R, et al. Magnetic resonance<br />
scans should replace biopsies for the diagnosis <strong>of</strong> diffuse brain stem<br />
gliomas: a report from the Children’s Cancer Group. Neurosurgery. 1993:<br />
33:1026-1030.<br />
2. Hankinson TC, Campagna EJ, Foreman NK, et al. Interpretation <strong>of</strong><br />
magnetic resonance images in diffuse intrinsic pontine glioma: a survey <strong>of</strong><br />
pediatric neurosurgeons. J Neurosurg Pediatr. 2011; 8:97-102.<br />
3. Frazier JL, Lee J, Thomale UW, et al. Treatment <strong>of</strong> diffuse intrinsic<br />
brainstem gliomas: Failed approaches and future strategies. J Neurosurg<br />
Pediatr. 2009;3:259-269.<br />
4. Anderson BD, Adamson PC, Weiner SL et al. Tissue collection for<br />
tion in a specific histone gene in many autopsy samples, a<br />
small number <strong>of</strong> upfront biopsies were used to suggest that<br />
this was a driving mutation not a passenger mutation. 7 The<br />
article did not say how these specimens were obtained.<br />
Fourth, Declaration that a Procedure Is Standard<br />
A participant who serves on an IRB pointed out that most<br />
IRBs relied on local expertise to declare what was standard.<br />
If an institution’s neurosurgeons declared that biopsy was<br />
standard, then it was not in fact an IRB issue. This brings up<br />
the whole question <strong>of</strong> what is a standard and whether its<br />
determination at a local level should be influenced by national<br />
consensus. Is there an obligation on the part <strong>of</strong> local<br />
investigators when considering a procedure standard to<br />
discuss with their IRB whether there is national controversy<br />
about that standard? Do IRBs have an ethical obligation to<br />
investigate what is “standard” or whether a therapy has a<br />
“realistic” possibility <strong>of</strong> benefit?<br />
As a Thought-Provoking Exercise<br />
Let’s say that there had never been a statement in an<br />
influential journal indicating that biopsies <strong>of</strong> typical DIPG<br />
were not needed for diagnosis and should not be done.<br />
Let’s say there were biopsies performed throughout the<br />
1990s and an H3F3A mutation was identified in 1999<br />
(purely hypothetical given technical considerations).<br />
Let’s say that targeted therapy doubled survival time and<br />
cured a small number (say 10% to 20%) by 2005.<br />
Is it possible that fear <strong>of</strong> a biopsy (with a known very small<br />
morbidity and mortality risk) resulted in shortening the<br />
lives <strong>of</strong> many and even the loss <strong>of</strong> life?<br />
Stock<br />
Ownership Honoraria<br />
REFERENCES<br />
Research<br />
Funding<br />
Expert<br />
Testimony<br />
Other<br />
Remuneration<br />
correlative studies in childhood cancer clinical trials: ethical considerations<br />
and special imperatives. J Clin Oncol. 2004;1:4846-4850.<br />
5. U.S Food and Drug Administration. Minutes <strong>of</strong> the Joint Meeting <strong>of</strong><br />
the Pediatric and <strong>Oncology</strong>/Advisory Committee, April 27, 2009. http://www.fda.<br />
gov/downloads/AdvisoryCommittees/CommitteesMeetingMaterials/Pediatric-<br />
AdvisoryCommittee/UCM171523.pdfBrainstemgliomas. Accessed March 2, <strong>2012</strong>.<br />
6. Roujeau T, Machado G, Garnett MR, et al. Stereotactic biopsy <strong>of</strong> diffuse<br />
pontine lesions in children. J Neurosurg. 2007;107(1 Suppl):1-4.<br />
7. Wu G, Broniscer A, McEachron TA, et al. Somatic histone H3 alterations<br />
in pediatric diffuse intrinsic pontine gliomas and non-brainstem glioblastomas.<br />
Nat Genet. <strong>2012</strong>; 44:251-253.<br />
635
TRANSITION AND DECISION MAKING FOR<br />
PALLIATIVE CARE OF PATIENTS WITH<br />
PEDIATRIC CANCER<br />
CHAIR<br />
Jennifer W. Mack, MD, MPH<br />
Dana-Farber Cancer Institute<br />
Boston, MA<br />
SPEAKERS<br />
Chris Feudtner, MD, PhD, MPH<br />
Children’s Hospital <strong>of</strong> Philadelphia<br />
Philadelphia, PA<br />
Pamela S. Hinds, PhD, RN<br />
Children’s National Medical Center<br />
Washington, DC
Communication and Decision Support for<br />
Children with Advanced Cancer and<br />
Their Families<br />
By Jennifer W. Mack, MD, MPH, Chris Feudtner, MD, PhD, MPH,<br />
and Pamela S. Hinds, PhD, RN<br />
Overview: Clinician communication related to treatment decision<br />
making is a fundamentally important health care intervention<br />
and is <strong>of</strong>ten reported by parents <strong>of</strong> seriously ill<br />
children to be the most valued <strong>of</strong> clinician skills. Since<br />
different children and families have different communication<br />
styles and expectations, and since these may change over the<br />
course <strong>of</strong> the illness experience, one <strong>of</strong> the early and recurring<br />
tasks is to clarify and work with these diverse styles and<br />
expectations. Adopting a stance <strong>of</strong> compassionate desire to<br />
know more about patients and families, in addition to imparting<br />
information, is vital, and can be facilitated by following a<br />
general strategy <strong>of</strong> “ask, tell, ask.” In addition to the exchange<br />
<strong>of</strong> information, communication between clinician and patient<br />
PALLIATIVE CARE is centered on children with lifethreatening<br />
illnesses and their families. As such, the<br />
values and goals held by individual children and families<br />
guide plans for care, and the child’s quality <strong>of</strong> life and<br />
symptoms—physical, psychologic, spiritual—are <strong>of</strong> primary<br />
importance. These principles affirm that the child is valued<br />
and guide care from diagnosis through all phases <strong>of</strong> illness.<br />
For patients with progressive cancer or serious complications,<br />
communication practices founded on these principles<br />
set the stage for palliative care when it is needed. In this<br />
article, we <strong>of</strong>fer practical guidance for clinicians who care<br />
for children with advanced cancer, addressing various aspects<br />
<strong>of</strong> communication and supportive decision making<br />
across the illness trajectory (see Fig. 1).<br />
Communication across the Illness Experience<br />
Communication, central to the work <strong>of</strong> caring for children<br />
with advanced cancer, has three primary purposes. First,<br />
communication allows for the exchange <strong>of</strong> information and<br />
the development <strong>of</strong> shared knowledge. Communication goes<br />
both ways: from the family and child as well as to them, and<br />
sometimes wordlessly, in shared silence. Second, communication<br />
serves as the foundation for a relationship between<br />
the child, parent, and clinician, in which the parents and<br />
child can feel known and understood. Third, communication<br />
provides a forum for decision making. The type <strong>of</strong> decision<br />
may change over time, but even when parents and children<br />
approach painful decisions about end-<strong>of</strong>-life care, they can<br />
use familiar processes and relationships as an anchor.<br />
Sharing Information<br />
The first communication goal that spans the illness trajectory<br />
is the sharing <strong>of</strong> information. Clinicians must both<br />
seek information from and impart information to children<br />
and their families, and what we choose to discuss sets a tone<br />
for the central issues <strong>of</strong> care.<br />
A useful general communication strategy is “ask, tell,<br />
ask,” in which communication from the clinician is framed<br />
by the child and family. For example, when discussing a<br />
cancer diagnosis, one might start by asking the child or<br />
family, “What is your understanding <strong>of</strong> the diagnosis so far?”<br />
and family also involves the signaling and exchange <strong>of</strong> emotions,<br />
in which the pace, verbal inflection, and body language<br />
<strong>of</strong> the conversation are fundamental. Discussions about prognosis<br />
and goals <strong>of</strong> care, while needing to be handled in a<br />
gentle manner, should start early in the illness experience and<br />
be revised whenever there is a relapse or major complication.<br />
Children <strong>of</strong>ten want to participate in these conversations to a<br />
degree <strong>of</strong> their own choosing, which they themselves can<br />
clarify. Effective and empathetic clinician communication can<br />
greatly facilitate decision making and care for children with<br />
advanced cancer and their families, and provide a substantial<br />
source <strong>of</strong> comfort.<br />
Their answer helps the clinician to know where to start, and<br />
correct any misconceptions. After communicating information,<br />
the clinician can use a final question to check understanding<br />
(“I want to make sure that I’ve explained things.<br />
Can you tell me what you are taking away from this<br />
conversation?”) and assess the emotional effect (“How are<br />
you feeling now?”) <strong>of</strong> the news.<br />
Even at diagnosis, conversations may be incomplete without<br />
honest discussion about the child’s future, including the<br />
nature <strong>of</strong> the illness and the prognosis. These conversations<br />
are hard—cognitively, emotionally, socially, and spiritually—and<br />
clinicians may understandably avoid them. 1,2 Yet<br />
most parents worry about their child’s future from the<br />
day they hear the word “cancer,” and want prognostic<br />
information, even if they find it upsetting. 3 In addition,<br />
starting such conversations at diagnosis helps with transition<br />
to palliative care if treatment fails.<br />
We recommend the following steps when talking about<br />
prognosis, 4,5 in line with the general strategy <strong>of</strong> “ask, tell,<br />
ask”:<br />
● Consider what the child or family needs to know about<br />
prognosis, and when they need to know it. In some situations,<br />
if the family does not want prognostic information,<br />
then the clinician need not discuss it. In other situations,<br />
such as a very poor prognosis or acute deterioration, true<br />
informed consent may not be possible without it.<br />
● Ask the family and/or child what they understand about<br />
what lies ahead.<br />
● Unless you have already decided that the family needs<br />
From the Department <strong>of</strong> Pediatric <strong>Oncology</strong> and the Center for Outcomes and Policy<br />
Research, Dana-Farber Cancer Institute, Boston, MA; Harvard Medical School, Boston,<br />
MA; Pediatric Advanced Care Team, Department <strong>of</strong> Medical Ethics, PolicyLab, The<br />
Children’s Hospital <strong>of</strong> Philadelphia, Philadelphia, PA; Department <strong>of</strong> Pediatrics, The<br />
Perelman School <strong>of</strong> Medicine at the University <strong>of</strong> Pennsylvania, Philadelphia, PA; Children’s<br />
National Medical Center, Washington, DC; The George Washington University,<br />
Washington, DC.<br />
Authors’ disclosures <strong>of</strong> potential conflicts <strong>of</strong> interest are found at the end <strong>of</strong> this article.<br />
Address reprint requests to Jennifer Mack, MD MPH, 450 Brookline Ave., Boston, MA<br />
02215; email: jennifer_mack@dfci.harvard.edu.<br />
© <strong>2012</strong> by <strong>American</strong> <strong>Society</strong> <strong>of</strong> <strong>Clinical</strong> <strong>Oncology</strong>.<br />
1092-9118/10/1-10<br />
637
detailed prognostic information, ask them what they want to<br />
know. Some patients and families will wait for the physician<br />
to <strong>of</strong>fer such information. Clinicians who do not routinely<br />
<strong>of</strong>fer prognostic information may leave some children and<br />
families without information they want.<br />
● For families who want information, provide an honest<br />
estimate, without euphemisms. Adult patients with cancer<br />
are more likely to understand the prognosis with one negative<br />
fact; although providing reasons to be more hopeful<br />
than the mortality or morbidity data suggest may make<br />
physicians feel better, it can cause families to be overly<br />
optimistic. 6<br />
● Ask the parents and the child how they are feeling, and<br />
respond appropriately.<br />
● Remind the child and/or family that you will be see them<br />
through. Although we cannot ensure that the course will be<br />
easy, we can help them to feel less alone.<br />
When treatment fails or results in a life-threatening<br />
complication, this information must also be shared. As<br />
outlined above, we recommend assessing the child’s and<br />
family’s understanding <strong>of</strong> the situation, providing information,<br />
then checking understanding and asking how they<br />
are doing (ask, tell, ask). Importantly, families value clear<br />
language, without euphemisms. Clinicians sometimes focus<br />
on treatment rather than prognosis, 7 perhaps because this<br />
is something we can <strong>of</strong>fer in a difficult situation. Conveying<br />
a poor prognosis with honesty, however, can help parents<br />
make the best decisions for their children. Many families<br />
also appreciate honest expressions <strong>of</strong> empathy or sadness.<br />
While there are no “right” words for this situation, one might<br />
say, “I wish the news was different. I am afraid that the<br />
scans show the cancer has returned. Unfortunately we know<br />
that when the cancer comes back, it is not curable. This is<br />
such sad news.” Often the precise words used are less<br />
important than the relationship between the clinician and<br />
the patient and family, since communication works best in a<br />
context <strong>of</strong> caring and trust.<br />
Developing Relationships<br />
The second communication goal, the development <strong>of</strong> a<br />
relationship between the parents, child, and clinician, also<br />
begins at diagnosis. Children and families need to know that<br />
their caregiver cares about them, can be trusted, respects<br />
KEY POINTS<br />
● “Ask, tell, ask” is a useful general communication<br />
strategy.<br />
● Each child and family has a different style and set <strong>of</strong><br />
expectations regarding communication and making<br />
decisions, and clinicians need to clarify and work with<br />
these styles and expectations.<br />
● Emotions play a vital role in communication—and<br />
are <strong>of</strong>ten the most important substance being communicated.<br />
● Asking a child “What makes you happiest?” can be a<br />
gateway to many important conversations.<br />
● Asking a parent “What do you need to do to feel like<br />
you are being the best parent you can be for your<br />
child?” can likewise open up deep lines <strong>of</strong> discussion.<br />
638<br />
MACK, FEUDTNER, AND HINDS<br />
and listens to them, and will continue to be there as the<br />
illness evolves. These relationships develop organically and<br />
in diverse directions, rather than following a prescribed<br />
path. Nevertheless, clinicians can foster such relationship<br />
with questions and behaviors that engage the child and<br />
parents in teaching the clinician about who they are.<br />
Most importantly, this work is fostered by listening.<br />
Children especially may not always wish to talk about issues<br />
related to their illness, but sometimes, they may leave the<br />
conversational door ajar. Clinicians who are patiently and<br />
consistently listening are more likely to detect and use that<br />
opening. Furthermore, by listening, clinicians reassure children<br />
and parents that their words have meaning.<br />
Questions can help build relationships, and <strong>of</strong>ten have<br />
dual roles: they allow the child and family to express<br />
themselves, and give clinicians information that helps them<br />
to provide the best possible care. Posing these questions<br />
and listening to the answers does not necessarily require<br />
much time; sitting by the bedside for a few minutes at the<br />
end <strong>of</strong> the day may suffice to learn about one or two <strong>of</strong> these<br />
issues, and reinforces the message that those in the room<br />
are valued.<br />
What kind <strong>of</strong> a person are you? How would you describe<br />
yourself? These questions unfold across multiple conversations,<br />
with many possible prompts: “What are you proudest<br />
<strong>of</strong> ?” “How do you like to be thought <strong>of</strong> ?” “Where do you find<br />
strength and support?” The goal is to learn about this child,<br />
and this family, as unique individuals. Early conversations<br />
(about family members, school, and interests, for example)<br />
may evolve over time into deeper ones about spirituality<br />
or the meaning <strong>of</strong> illness (“Do you ever wonder why this<br />
happened to you?”).<br />
Tell me about a time when you were the happiest. This<br />
question has particular value in getting to know the child<br />
as a person. Children with cancer share a lot <strong>of</strong> tough times<br />
with their medical team. This question allows them to share<br />
what is good and meaningful to them, which is good in itself,<br />
for the child, family, and clinician. Additionally, this can<br />
aid clinicians in conversations about care when treatment<br />
options have failed—“Remember when you told me about<br />
...?Iamwondering if we can find our way to more times<br />
like that.”<br />
How does communication work in your family? How do<br />
you want it to work? Asking this question shows respect for<br />
the family system, which may be influenced by cultural<br />
values, past experience, and the unique ways that a family<br />
comes together around a child’s illness. Asking questions<br />
and listening are the best tools to understand these differences.<br />
Some families have very different preferences from<br />
those <strong>of</strong> the clinician; for example, a family may wish to<br />
exclude an adolescent patient from conversations about the<br />
diagnosis. By asking at the beginning, the clinician can<br />
learn what matters to the family and start a conversation<br />
about how to proceed. A plan that works well for everyone<br />
requires ongoing conversation, but this question reminds<br />
families and patients that their preferences are paramount.<br />
As you think about the future, what do you worry about the<br />
most? Worries about the future are nearly universal at the<br />
time <strong>of</strong> diagnosis, yet <strong>of</strong>ten go unspoken. This question<br />
allows the clinician to acknowledge these difficult emotions.<br />
Unrealistic worries can be corrected, and reasonable worries<br />
can be recognized, understood, and met with empathy. This
COMMUNICATION AND DECISION SUPPORT FOR PEDIATRIC ADVANCED CANCER<br />
Fig. 1. Overview <strong>of</strong> communication issues and strategies.<br />
639
question also sets the stage for formulating <strong>of</strong> goals <strong>of</strong> care,<br />
which may address fears about outcomes.<br />
After a relapse or serious complication, communication<br />
can continue to affirm and deepen the relationship between<br />
child, family, and clinician. Useful questions focus on the<br />
emotional and existential experience <strong>of</strong> illness and hopes<br />
and fears for the future, and allow children and families to<br />
reflect on the child’s life and its meaning. Some examples<br />
are below, although not every question is appropriate for<br />
every child; as always, careful listening may help identify<br />
good times for these conversations.<br />
● Do you ever think about what all <strong>of</strong> this means?<br />
● As you think about the future, is there anything that<br />
you are especially afraid <strong>of</strong>? Is there anything that you are<br />
especially hoping for?<br />
● I know you have pain, but is there anything that is even<br />
worse than the pain? 8<br />
● Is there anything you’d especially like for your family to<br />
know about you?<br />
● How do you like to be thought <strong>of</strong> ?<br />
Making Decisions<br />
The third goal <strong>of</strong> communication, shared decision making,<br />
is also relevant throughout illness. Starting treatment with<br />
language based on goals <strong>of</strong> care can help to make the<br />
transition to palliative care feel more natural. For most<br />
children, the initial goal <strong>of</strong> treatment is cure. Yet evidence<br />
suggests that parents consider palliation a priority even<br />
during initial, cure-focused care. 9 Thus, we recommend making<br />
goals an explicit part <strong>of</strong> all conversations about cancer<br />
treatment: “As you think about your (your child’s) illness,<br />
what is most important to you?” and “Aside from this most<br />
important goal, what else are you hoping may be possible?”<br />
Once goals have been identified, clinicians can respond by<br />
framing care, including treatment options, in the context <strong>of</strong><br />
these goals. Rather than relying on parents to decide on<br />
their own, clinicians may make recommendations about care<br />
consistent with parents’ goals and values. For example, one<br />
might say, “You just told me how important it is for her to<br />
have a good quality <strong>of</strong> life. Given that, I wouldn’t recommend<br />
intensive chemotherapy, because I am worried that she will<br />
spend a lot <strong>of</strong> time in the hospital and the clinic, instead <strong>of</strong><br />
doing things she enjoys. I’d like to recommend that we <strong>of</strong>fer<br />
her any treatments we think may make her feel better,<br />
but stay away from treatment that is likely to be difficult or<br />
make her feel unwell.” In doing so, clinicians can join<br />
parents in what may be very difficult decisions. Individual<br />
parents have a wide range <strong>of</strong> preferences for involvement in<br />
decision making, and preferred roles can change over time<br />
as parental experience and the nature <strong>of</strong> decisions change.<br />
Thus clinicians may wish to ask parents how they want<br />
decision making to look, and work to support their preferences.<br />
Starting Conversations Early<br />
Conversations about prognosis and goals <strong>of</strong> care are appropriate<br />
for all children with cancer and their families.<br />
They address fears about the future, help clinicians learn<br />
what the child and family consider important, and allow the<br />
child, family, and clinician to know one another as people.<br />
Clinicians can return to these discussions and, in doing so,<br />
reassure children and their families that they matter, that<br />
640<br />
they are known, and that their care team will do everything<br />
possible to uphold their wishes.<br />
Emotions and Communication<br />
MACK, FEUDTNER, AND HINDS<br />
In all conversations, emotions play several roles in shaping<br />
the interaction. 10 First, emotions affect the way people<br />
think and behave. Emotions like fear, sadness, anger, joy,<br />
happiness, surprise, relief, guilt, shame, disgust, or contempt<br />
(which constitute a core set <strong>of</strong> emotions that many<br />
psychologists identify as primary emotional responses) orient<br />
a person toward different aspects <strong>of</strong> a situation and color<br />
their interpretation <strong>of</strong> it. Second, whether in the background<br />
or foreground <strong>of</strong> any particular interaction, emotions are<br />
part <strong>of</strong> what individuals communicate to each other, wittingly<br />
or not: by a combination <strong>of</strong> word choice, vocal inflection,<br />
body language, and other cues, people “show” how they<br />
feel. Third, what people “show” each other may or may not<br />
accurately reflect how they truly feel, and resultant misunderstandings<br />
can pr<strong>of</strong>oundly alter the tone and outcome <strong>of</strong><br />
particular conversations and future interactions. Fourth,<br />
when children are seriously ill, parents <strong>of</strong>ten have both<br />
strong negative feelings (their child is so ill they are afraid<br />
or angry) and strong positive feelings (they love their child<br />
with boundless affection and pride), further complicating<br />
the handling <strong>of</strong> emotions, communication, and decision making.<br />
11<br />
Because <strong>of</strong> these and other important effects <strong>of</strong> emotions<br />
on communication, clinicians who care for children with<br />
advanced cancer and their families must become aware <strong>of</strong><br />
and seek to improve their own emotional communication<br />
skills. As yet, there are no well-established evidenced-based<br />
“best practices” regarding the emotional side <strong>of</strong> communication;<br />
12 acknowledging this, we recommend the following<br />
techniques:<br />
Keep tabs on your own emotions. Even before a conversation<br />
begins, and periodically during the encounter, clinicians<br />
should spend a moment focusing on their own feelings, and<br />
name the emotions they have; labeling how one is feeling<br />
helps in self-regulating the effect that emotions have on<br />
one’s own thinking and behavior.<br />
Engage in a common purpose. At the outset <strong>of</strong> the conversation,<br />
after <strong>of</strong>fering a personal greeting, work to quickly<br />
establish with the patient and family what you are all trying<br />
to get out <strong>of</strong> the discussion: “Thanks for meeting with me.<br />
There are a few things I know I want to discuss with you.<br />
What do you want to talk about, what would be helpful?”<br />
Slow down. Often clinicians, feeling either time pressure<br />
or the need to discuss lots <strong>of</strong> information, move the conversation<br />
forward with at their own quickened pace. Slowing<br />
down is a cardinal way to show empathy and caring. Even<br />
if fewer facts are covered in a given period <strong>of</strong> time, the<br />
emotional quality <strong>of</strong> the conversation is enhanced.<br />
Listen and summarize. Observational studies <strong>of</strong> physicians<br />
talking “with” patients show that physicians do most<br />
<strong>of</strong> the talking, but that patients leave encounters far more<br />
satisfied when physicians do less talking and more listening.<br />
A particularly effective technique is for the clinician to ask<br />
a question, listen for 30 seconds to a minute or two, then<br />
summarize what has been said: “Okay, let me see if I heard<br />
you correctly, you are most concerned about . . .”<br />
Solicit permission. Before addressing topics that may be<br />
difficult to discuss, solicit permission to do so: “There is
COMMUNICATION AND DECISION SUPPORT FOR PEDIATRIC ADVANCED CANCER<br />
something else, something important, that may be hard to<br />
talk about but I think we need to talk about, okay?” 13<br />
Ask about what they are hoping for. When confronting<br />
serious illness, people <strong>of</strong>ten become focused on making it<br />
go away. When asked, “In trying to take the best care <strong>of</strong> you<br />
that I can, it will help me to know what you are hoping<br />
for—can you share your thoughts with me?”, the first<br />
answer will <strong>of</strong>ten be a cure or miracle. Be patient and<br />
supportive. “I also hope that can happen. What else are you<br />
hoping for?” The subsequent hopes are likely to be very<br />
helpful in understanding distressing symptoms or previously<br />
unarticulated fears or goals that can be addressed by<br />
good comprehensive care. 14<br />
Provide a “warning shot” when bad news is coming. If you<br />
are aware that you about to share some bad news, provide<br />
the patient or family with a warning shot: “I wish the news<br />
was different. The test shows that the cancer has come<br />
back.”<br />
When people cry or shout, stop. People experience another<br />
person as being empathic if this other person seems to “get”<br />
how they are feeling and respond in an appropriate manner.<br />
Continuing to talk when someone cries, shouts, or becomes<br />
visibly distraught, is a recipe for being perceived as lacking<br />
empathy.<br />
Observe and name. People <strong>of</strong>ten show their feelings but<br />
may not verbalize how they are feeling. In many situations,<br />
simply stating that you are aware that they have strong<br />
feelings can be helpful: “I see how much this news has upset<br />
you.”<br />
Before ending, be clear about next steps. Do not end an<br />
emotionally charged conversation without a clear sense <strong>of</strong><br />
what will happen next and when you will talk again. Map<br />
out a plan. Make a commitment about the time and place <strong>of</strong><br />
your next conversation.<br />
The only way to incorporate these skills into routine<br />
clinical practice is to practice. Faculty as well as trainees, in<br />
our experience, benefit from role-playing scenarios. And<br />
although most clinicians dread being videotaped during a<br />
scenario, no other educational method matches direct observation<br />
<strong>of</strong> one’s own conversational style and habits to motivate<br />
and guide positive change.<br />
Specific Communication Issues Often Encountered<br />
during Advanced Illness<br />
When a child’s cancer is no longer curable, the child and<br />
family are likely to experience additional significant treatment<br />
or care decisions such as whether to participate in a<br />
phase I trial, implement a do-not-attempt resuscitation<br />
order, or end all treatment efforts and focus on end-<strong>of</strong>-life<br />
options. 15 The child and family members come to end-<strong>of</strong>-life<br />
treatment decision making affected by exposure to other ill<br />
children and their families who preceded them in reaching<br />
such a decision point, and by their own personal experience<br />
with treatment. Both <strong>of</strong> these sources <strong>of</strong> experience produce<br />
pr<strong>of</strong>ound, lingering images for the child and family and<br />
affect how they engage in end-<strong>of</strong>-life decision making. It is<br />
therefore helpful to clinicians to directly ask the child and<br />
parents about the treatment experiences they find themselves<br />
reflecting on when making end-<strong>of</strong>-life treatment decisions.<br />
Engagement in end-<strong>of</strong>-life discussions. A proportion <strong>of</strong><br />
these families will initiate the end-<strong>of</strong>-life discussion because<br />
they have been carefully (and typically, privately) considering<br />
what they would do if their ill child’s illness became<br />
terminal. A separate proportion <strong>of</strong> families will not be able to<br />
engage in this kind <strong>of</strong> discussion because doing so is culturally<br />
unacceptable. To do so means that the families must<br />
face the risk <strong>of</strong> both losing their child and <strong>of</strong> being excommunicated<br />
from their larger community <strong>of</strong> relatives, friends,<br />
or fellow worshippers for “giving up” on their child. It is too<br />
much to expect these families to engage in end-<strong>of</strong>-life decision<br />
making. But the majority <strong>of</strong> families <strong>of</strong> children whose<br />
disease cannot be cured, particularly those comfortable in<br />
the culture <strong>of</strong> North America, will expect to be part <strong>of</strong><br />
discussions about their ill child’s end-<strong>of</strong>-life care options. 16,17<br />
This majority <strong>of</strong> parents also anticipates having their preferences<br />
for end-<strong>of</strong>-life care honored by the health care<br />
system because they are the “child’s parent until the child’s<br />
last breath.” Even though the majority <strong>of</strong> parents indicate<br />
a preference for involvement, they also readily identify<br />
end-<strong>of</strong>-life decision making to be the most difficult <strong>of</strong> all their<br />
care decisions. 15<br />
Change over time. One <strong>of</strong> the striking findings from studies<br />
involving parents <strong>of</strong> children with relapsed, advanced,<br />
and incurable cancer is their self-observation <strong>of</strong> not being<br />
the same individuals the care team knew at the point <strong>of</strong><br />
diagnosis or during cure-oriented treatment. Instead, the<br />
parents describe being more watchful and wary, or being<br />
less inclined to know all <strong>of</strong> the treatment details and<br />
preferring to focus on only the one most pressing concern at<br />
any given time. 18,19 The first implication <strong>of</strong> this parent<br />
self-observation is that clinicians should anticipate being<br />
surprised at some point during end-<strong>of</strong>-life decision making<br />
or care by parental responses to the discussion. The second<br />
implication for clinicians is that developing a strategy for<br />
keeping aligned with the parents’ evolving care preferences<br />
and their expressions <strong>of</strong> those preferences is beneficial. Such<br />
a strategy can include occasionally and directly asking<br />
parents to describe their self-observed changes and their<br />
current care preferences.<br />
Observing signs <strong>of</strong> dying. Parents indicate that to be<br />
ready to participate in end-<strong>of</strong>-life discussions, they must<br />
first come to believe that their child will not recover. Factors<br />
that help parents to believe that their child is going to die<br />
include the information and care recommendations received<br />
from clinicians, and statements that their ill child has made<br />
about not continuing treatment if it became pointless. Parents’<br />
beliefs are also influenced by their child’s physical<br />
changes such as swelling, bleeding, or changes in breathing<br />
patterns, and by concerns about adverse events that could<br />
result from further anticancer treatment. 15,20 For these<br />
parents, clinicians can facilitate their participation in decision<br />
making by mentioning the child’s actual changes, including<br />
those physically apparent and those apparent only<br />
on scans, and by interpreting these changes for the parents,<br />
making it clear that all that can be done has been done and<br />
done well.<br />
Parents’ personal sense <strong>of</strong> “being a good parent.” Parent<br />
participation in end-<strong>of</strong>-life discussions is also influenced by<br />
the parent-named factor <strong>of</strong> “being a good parent.” 21,22 This<br />
factor represents an internal definition held by parents that<br />
is their personal benchmark <strong>of</strong> how well they did by their<br />
child before and during end-<strong>of</strong>-life care. Although there are<br />
aspects <strong>of</strong> this definition that are unique to each parent,<br />
641
there are commonalities as well. Clinicians may anticipate<br />
these commonalities to include the parents’ desire to be<br />
certain that their ill child feels loved by the parent, and that<br />
the parent has been a positive role model, and has continued<br />
to provide care and to meet the child’s needs, including<br />
protection from suffering and maintaining vestiges <strong>of</strong> health<br />
for as long as is possible. Finally, parents want to be certain<br />
that they have made prudent decisions on behalf <strong>of</strong> their<br />
children. Clinician recognition <strong>of</strong> the existence <strong>of</strong> this internal<br />
definition and inviting parents to share their own<br />
definition <strong>of</strong> being a “good parent” to their seriously ill child<br />
will help parents to trust that the end-<strong>of</strong>-life decisionmaking<br />
process includes thoughtful recognition <strong>of</strong> their<br />
emotional well-being as parents.<br />
Child participation in conversations. We find that not all<br />
parents are comfortable with their child being included in<br />
end-<strong>of</strong>-life decision making. Limited research supports the<br />
ability <strong>of</strong> children treated for cancer to participate in such<br />
discussions and, indeed, a preference for inclusion. A careful<br />
exploration by clinicians with the parents regarding their<br />
preferences for including their child is warranted before<br />
initiating actual discussion with the child. In this clinicianparent<br />
preparatory discussion, parents can identify the<br />
terms they are comfortable using with their ill child so that<br />
clinicians, parents, and the ill child can use the same<br />
language to facilitate shared understanding <strong>of</strong> the seriousness<br />
<strong>of</strong> the clinical situation. The process <strong>of</strong> “ask, tell, ask” is<br />
very fitting for this preparatory phase <strong>of</strong> end-<strong>of</strong>-life discussions.<br />
Although the research evidence is limited, the findings<br />
that have been reported coupled with reported clinical<br />
experience support the ability <strong>of</strong> children (age 6 and older)<br />
to indicate their preferences for continuing or discontinuing<br />
cure-oriented treatment. 23 Additionally, children age 10 and<br />
older have been studied in terms <strong>of</strong> their stated preferences<br />
for enrolling on a phase I trial, not implementing a resuscitation<br />
order, or initiating terminal care. 24,25 These children<br />
were able to state their reasons for their end-<strong>of</strong>-life prefer-<br />
Authors’ Disclosures <strong>of</strong> Potential Conflicts <strong>of</strong> Interest<br />
Author<br />
Jennifer W. Mack*<br />
Chris Feudtner*<br />
Pamela S. Hinds*<br />
*No relevant relationships to disclose.<br />
Employment or<br />
Leadership<br />
Positions<br />
Consultant or<br />
Advisory Role<br />
1. Miyaji NT. The power <strong>of</strong> compassion: truth-telling among <strong>American</strong><br />
doctors in the care <strong>of</strong> dying patients. Soc Sci Med. 1993;36:249-264.<br />
2. Christakis NA, Iwashyna TJ. Attitude and self-reported practice regarding<br />
prognostication in a national sample <strong>of</strong> internists. Arch Intern Med.<br />
1998;158:2389-2395.<br />
3. Mack JW, Wolfe J, Grier HE, et al. Communication about prognosis<br />
between parents and physicians <strong>of</strong> children with cancer: parent preferences<br />
and the impact <strong>of</strong> prognostic information. J Clin Oncol. 2006;24:5265-5270.<br />
4. Back AL, Arnold RM. Discussing prognosis: “how much do you want to<br />
know?” talking to patients who do not want information or who are ambivalent.<br />
J Clin Oncol. 2006;24:4214-4217.<br />
5. Back AL, Arnold RM. Discussing prognosis: “how much do you want to<br />
know?” talking to patients who are prepared for explicit information. J Clin<br />
Oncol. 2006;24:4209-4213.<br />
6. Robinson TM, Alexander SC, Hays M, et al. Patient-oncologist commu-<br />
642<br />
ences, with the most commonly reported factor being concern<br />
for loved ones such as family members and favorite<br />
clinicians. Documented fears <strong>of</strong> children regarding end <strong>of</strong> life<br />
include having pain and being alone. Assurances from parents<br />
and clinicians about being available and determined to<br />
help are <strong>of</strong> comfort to a seriously ill child.<br />
Treatment team communication and dynamics. It is also<br />
important to involve the child’s health care team in end-<strong>of</strong>life<br />
discussions and treatment decision making. Team tension<br />
can emerge when members <strong>of</strong> the team do not feel<br />
included in decision making. Patients and parents report<br />
being able to sense team tension. Teams that establish time<br />
to discuss end-<strong>of</strong>-life care options among themselves and to<br />
review child and family preferences related to end-<strong>of</strong>-life<br />
options may be able to prevent or diminish this kind <strong>of</strong><br />
tension.<br />
Conclusion<br />
Clinician communication related to treatment decision<br />
making is a fundamentally important health care intervention<br />
and is <strong>of</strong>ten reported by parents <strong>of</strong> seriously ill children<br />
to be the most valued <strong>of</strong> clinician skills. Words and the<br />
manner in which clinicians convey their concerns about the<br />
ill child and about the family during the periods <strong>of</strong> diagnosis,<br />
treatment, and transition, devoting themselves to supporting<br />
patient and parental decision making, become sources<br />
<strong>of</strong> substantial influence on child and parent responses to<br />
challenging times. Clinician attention to this powerful skill<br />
set will likely be a source <strong>of</strong> special and well-remembered<br />
comfort for the child and family.<br />
Acknowledgments<br />
This study was supported in part by Grants No. R21<br />
NR008634, R21 NR010026, and RO1 NR012026 from the National<br />
Institute <strong>of</strong> Nursing Research, Cancer Center Support<br />
Grant No. P30 CA21765 from the National Cancer Institute,<br />
and an <strong>American</strong> Cancer <strong>Society</strong> Mentored Research Scholar<br />
Grant.<br />
Stock<br />
Ownership Honoraria<br />
REFERENCES<br />
Research<br />
Funding<br />
MACK, FEUDTNER, AND HINDS<br />
Expert<br />
Testimony<br />
Other<br />
Remuneration<br />
nication in advanced cancer: Predictors <strong>of</strong> patient perception <strong>of</strong> prognosis.<br />
Support Care Cancer. 2008;16:1049-1057.<br />
7. The AM, Hak T, Koeter G, et al. Collusion in doctor-patient communication<br />
about imminent death: an ethnographic study. BMJ. 2000;321:1376-1381.<br />
8. Cassel EJ. The nature <strong>of</strong> suffering and the goals <strong>of</strong> medicine. N Engl<br />
J Med. 1982;306:639-645.<br />
9. Wolfe J, Klar N, Grier HE, et al. Understanding <strong>of</strong> prognosis among<br />
parents <strong>of</strong> children who died <strong>of</strong> cancer: impact on treatment goals and<br />
integration <strong>of</strong> palliative care. JAMA. 2000;284:2469-2475.<br />
10. Feudtner C. Collaborative communication in pediatric palliative care: a<br />
foundation for problem-solving and decision-making. Pediatr Clin North Am.<br />
2007;54:583-607, ix.<br />
11. Feudtner C, Carroll KW, Hexem KR, et al. Parental hopeful patterns <strong>of</strong><br />
thinking, emotions, and pediatric palliative care decision making: a prospective<br />
cohort study. Arch Pediatr Adolesc Med. 2010;164:831-839.
COMMUNICATION AND DECISION SUPPORT FOR PEDIATRIC ADVANCED CANCER<br />
12. Stone D, Patton B, Heen S. Difficult Conversations: How to Discuss<br />
What Matters Most. New York: Viking; 1999.<br />
13. Feudtner C. Partnering leadership. Arch Pediatr Adolesc Med. 2011;<br />
165:487-488.<br />
14. Feudtner C. The breadth <strong>of</strong> hopes. N Engl J Med. 2009;361:2306-<br />
2307.<br />
15. Hinds PS, Oakes L, Furman W, et al. Decision making by parents and<br />
healthcare pr<strong>of</strong>essionals when considering continued care for pediatric patients<br />
with cancer. Oncol Nurs Forum. 1997;24:1523-1528.<br />
16. Hinds PS, Oakes L, Quargnenti A, et al. An international feasibility<br />
study <strong>of</strong> parental decision making in pediatric oncology. Oncol Nurs Forum.<br />
2000;27:1233-1243.<br />
17. Michelson KN, Koogler TK, Skipton K, et al. Parents’ reactions to<br />
participating in interviews about end-<strong>of</strong>-life decision making. J Palliat Med.<br />
2006;9:1329-1338.<br />
18. Hinds PS, Birenbaum LK, Clarke-Steffen L, et al. Coming to terms:<br />
parents’ response to a first cancer recurrence in their child. Nurs Res.<br />
1996;45:148-153.<br />
19. Hinds PS, Birenbaum LK, Pedrosa AM, et al. Guidelines for the<br />
recurrence <strong>of</strong> pediatric cancer. Semin Oncol Nurs. 2002;18:50-59.<br />
20. Hinds PS OL, Furman W. End <strong>of</strong> life decision-making in pediatric<br />
oncology. In: Ferrell B CN (ed). Oxford Textbook <strong>of</strong> Palliative Nursing Care. 2d<br />
ed. New York: Oxford University Press, 2006;945-958.<br />
21. Hinds PS, Oakes LL, Hicks J, et al. “Trying to be a good parent” as<br />
defined by interviews with parents who made phase I, terminal care, and<br />
resuscitation decisions for their children. J Clin Oncol. 2009;27:5979-5985.<br />
22. Maurer SH, Hinds PS, Spunt SL, et al. Decision making by parents<br />
<strong>of</strong> children with incurable cancer who opt for enrollment on a phase I trial<br />
compared with choosing a do not resuscitate/terminal care option. J Clin<br />
Oncol. 2010;28:3292-3298.<br />
23. Nitschke R, Humphrey GB, Sexauer CL, et al. Therapeutic choices<br />
made by patients with end-stage cancer. J Pediatr. 1982;101:471-476.<br />
24. Hinds PS, Drew D, Oakes LL, et al. End-<strong>of</strong>-life care preferences <strong>of</strong><br />
pediatric patients with cancer. J Clin Oncol. 2005;23:9146-9154.<br />
25. Freyer DR. Care <strong>of</strong> the dying adolescent: special considerations. Pediatrics.<br />
2004;113:381-388.<br />
643
ADDRESSING THE IMBALANCE OF SUPPLY AND<br />
DEMAND: INTEGRATING ADVANCED PRACTICE<br />
PROVIDERS INTO SURVIVORSHIP CARE<br />
CHAIR<br />
Michael Goldstein, MD<br />
Beth Israel Deaconess Medical Center and Harvard Medical School<br />
Boston, MA<br />
SPEAKERS<br />
James Ross Waisman, MD<br />
Breastlink Medical Group, Inc.<br />
Todd Alan Pickard, PA-C, MMSc<br />
University <strong>of</strong> Texas M. D. Anderson Medical Center<br />
Houston, TX<br />
Mary S. McCabe, RN, MA<br />
Memorial Sloan-Kettering Cancer Center<br />
New York, NY
Planning for the Future: The Role <strong>of</strong> Nurse<br />
Practitioners and Physician Assistants in<br />
Survivorship Care<br />
By Mary S. McCabe, RN, MA, and Todd Alan Pickard, PA-C, MMSc<br />
Overview: The number <strong>of</strong> cancer survivors in the United<br />
States now approaches 12 million individuals, with an estimated<br />
7.2% <strong>of</strong> the general population aged 18 years or older<br />
reporting a previous cancer diagnosis. These figures highlight<br />
a number <strong>of</strong> questions about the care <strong>of</strong> survivors—how<br />
patients at risk for a known set <strong>of</strong> health problems should be<br />
followed, by whom, and for how long. At the same time that<br />
oncologists are developing strategies to provide services to<br />
this growing population, there are economic and systems<br />
challenges that have relevance to the previous questions,<br />
including a predicted national shortage <strong>of</strong> physicians to provide<br />
oncology services. Nurse practitioners (NPs) and physician<br />
assistants (PAs) have been identified as members <strong>of</strong> the<br />
AS A RESULT <strong>of</strong> the ongoing advances in early detection<br />
and treatment, the number <strong>of</strong> cancer survivors in<br />
the United States now approaches 12 million individuals,<br />
with an estimated 7.2% <strong>of</strong> the general population aged 18<br />
years or older reporting a previous cancer diagnosis. 1,2 The<br />
5-year relative survival rate for adult cancer survivors has<br />
reached 67%, with the largest number <strong>of</strong> survivors having<br />
been treated for breast, prostate, and colorectal cancers. 2<br />
These optimistic figures also highlight the challenges and<br />
opportunities facing the oncology community as we work to<br />
make survivorship a formal period <strong>of</strong> care. Such a focus<br />
requires not only appropriate surveillance for recurrence,<br />
but also the comprehensive rehabilitation <strong>of</strong> the posttreatment<br />
patient: 1) follow-up medical care tailored to the<br />
problems <strong>of</strong> specific populations, 2) psychosocial support,<br />
and 3) health promotion education that includes diet, exercise,<br />
and screening for new primary cancers. These core<br />
services are coupled together with the imperative that we<br />
share the care <strong>of</strong> these individuals with their community<br />
primary care physician (PCP), thus assuring a coordination<br />
<strong>of</strong> care that leads to effective communication and improved<br />
quality <strong>of</strong> life for cancer survivors (Sidebar 1).<br />
In 2005, the Institute <strong>of</strong> Medicine issued a seminal report,<br />
From Cancer Patient to Cancer Survivor: Lost in Transition,<br />
which served as an early guide to the development <strong>of</strong><br />
survivorship-specific services and models <strong>of</strong> care. 3 It was the<br />
first comprehensive proposal detailing the follow-up care<br />
needs <strong>of</strong> survivors and the novel provider arrangements that<br />
could be implemented to provide these services. This report,<br />
and others that followed, refocused our thinking about how<br />
survivorship care can and should be delivered. Initially,<br />
survivorship research focused on the long-term consequences<br />
<strong>of</strong> cancer therapy in the pediatric cancer survivor,<br />
in particular the serious morbidity and premature mortality<br />
resulting from exposure to radiation and chemotherapeutic<br />
agents in developing organs and normal tissue. And now,<br />
over the past 10 years, research has increasingly focused on<br />
the long-term and late effects experienced by individuals<br />
treated for adult-onset cancers as well. 4,5,6 Traditionally,<br />
the follow-up care <strong>of</strong> the survivor <strong>of</strong> an adult-onset cancer<br />
was primarily focused on an evaluation <strong>of</strong> recurrence, but<br />
this single focus is insufficient. We now know that survivors<br />
e56<br />
health care team who can help reduce the oncology supply and<br />
demand gap in a number <strong>of</strong> ways. The ASCO Study <strong>of</strong> Collaborative<br />
Practice Arrangements (SCPA) in 2011 concluded that<br />
oncology patients were aware and satisfied when their care<br />
was provided by NPs and PAs; there was an increase in<br />
productivity in practices that utilized NPs and PAs; utilizing<br />
the full scope <strong>of</strong> practice <strong>of</strong> NPs and PAs was financially<br />
advantageous; and, physicians, NPs, and PAs are highly satisfied<br />
with their collaborative practices. Increasingly, the<br />
oncology and health policy literature contains evidence supporting<br />
innovative provider models. There is still much work to<br />
be done to move beyond pilot data to establish the true value<br />
<strong>of</strong> these models.<br />
face a variety <strong>of</strong> health risks that are dependent on treatment<br />
exposures, genetic predisposition, comorbid health<br />
conditions, and lifestyle behaviors and that the identified<br />
issues cross multiple domains, including the medical, psychological,<br />
and social. 7,8,9,10 Because breast cancer survivors<br />
are the most widely studied group to date, they serve as an<br />
excellent example <strong>of</strong> the range <strong>of</strong> late effects a survivor can<br />
face. Medical late effects include anthracycline-related cardiomyopathy,<br />
osteoporosis, cognitive dysfunction, and infertility;<br />
psychological effects include fear <strong>of</strong> recurrence, sexual<br />
dysfunction, and fatigue; and, social issues relate to return<br />
to work and changes in role functioning. These examples<br />
raise the question <strong>of</strong> how patients at risk for a known set <strong>of</strong><br />
health problems should be followed, by whom, and for how<br />
long. At the same time that oncologists are developing<br />
strategies to provide services to this growing population—<br />
formally extending oncology care through the survivorship<br />
period—there are other economic and systems challenges<br />
that have relevance to the previous questions.<br />
Drivers <strong>of</strong> Workforce Change<br />
There are a number <strong>of</strong> factors in the U.S. health care<br />
system today that will affect the ability for oncology physicians<br />
to provide care for patients with cancer. The growing<br />
population <strong>of</strong> individuals over the age <strong>of</strong> 65 with an increasing<br />
incidence <strong>of</strong> cancer is a significant driver for an increasing<br />
demand for oncology care. 11 The improvement in cancer<br />
therapies and the resulting reduction in mortality rates<br />
mean that there will be an increase in the number <strong>of</strong> years<br />
that a patient will live with cancer through all phases <strong>of</strong><br />
the disease. Essentially, there will be a larger population <strong>of</strong><br />
individuals diagnosed with cancer, and they are likely to live<br />
longer with the disease. This drives the increased demand<br />
From the Memorial Sloan-Kettering Cancer Center and University <strong>of</strong> Texas M. D.<br />
Anderson Cancer Center.<br />
Authors’ disclosures <strong>of</strong> potential conflicts <strong>of</strong> interest are found at the end <strong>of</strong> this article.<br />
Address reprint requests Mary S. McCabe, 1275 York Avenue, New York, NY 10065;<br />
email: mccabem@mskcc.org.<br />
© <strong>2012</strong> by <strong>American</strong> <strong>Society</strong> <strong>of</strong> <strong>Clinical</strong> <strong>Oncology</strong>.<br />
1092-9118/10/1-10
NPS AND PAS IN SURVIVORSHIP CARE<br />
Sidebar 1. Essential Components <strong>of</strong> Survivorship Care<br />
● Surveillance for recurrence<br />
● Screening for new cancers<br />
● Identification and management <strong>of</strong> the consequences<br />
<strong>of</strong> the cancer and its treatment<br />
● Health promotion strategies<br />
● Coordination/communication between oncology specialists<br />
and primary care providers<br />
for oncology services up and will continue to do so for the<br />
foreseeable future. 12<br />
The number <strong>of</strong> physicians entering the field <strong>of</strong> oncology<br />
has decreased, and the supply <strong>of</strong> oncologists will not be able<br />
to meet the demand for oncology services by 2020. In fact,<br />
there is a predicted shortage <strong>of</strong> more than 4,000 physicians<br />
to provide oncology services. Several factors for this gap<br />
have been identified and include fewer residents entering<br />
oncology training programs, a growing population <strong>of</strong><br />
retirement-age oncologists, a shift in the health care system<br />
to a focus on primary care, and decreasing reimbursement<br />
for chemotherapy infusion and other services that oncology<br />
practices require to remain fiscally sound. The <strong>American</strong><br />
<strong>Society</strong> <strong>of</strong> <strong>Clinical</strong> <strong>Oncology</strong> (ASCO) Workforce Study completed<br />
in 2007 identified the gap between supply <strong>of</strong> oncology<br />
physicians and demand for oncology services. There will be<br />
an estimated demand <strong>of</strong> 57.2 to 60.7 million visits per year<br />
by 2020 and only an estimated supply <strong>of</strong> 45.6 to 47.8 million<br />
visits per year. Thus, there is a gap <strong>of</strong> 11.6 to 12.9 million<br />
visits per year by 2020 (Fig. 1). This study highlighted the<br />
fact that the increasing gap in supply and demand requires<br />
new ways <strong>of</strong> providing oncology services such as increasing<br />
oncology training programs to produce more oncologists and<br />
shifting services traditionally provided by oncologists, such<br />
as survivorship, to the primary care setting. 13<br />
Nurse practitioners (NPs) and physician assistants (PAs)<br />
have been identified as members <strong>of</strong> the health care team<br />
who can help reduce the oncology supply and demand gap in<br />
a number <strong>of</strong> ways. 14 Fortunately, there is already a significant<br />
number <strong>of</strong> NPs and PAs practicing in oncology settings<br />
working with physicians to provide diagnostic evaluation,<br />
patient education, infusion services, treatment monitoring,<br />
KEY POINTS<br />
● There is an increasing number <strong>of</strong> survivors who have<br />
been successfully treated for cancer.<br />
● The number <strong>of</strong> oncologists will not be able to meet the<br />
demand for oncology services by 2020.<br />
● Nurse practitioners (NP) and physician assistants<br />
(PA) are pr<strong>of</strong>essionally prepared to help reduce the<br />
gap in care.<br />
● Innovative care models are being implemented internationally<br />
using NPs and PAs as important colleagues<br />
in the care <strong>of</strong> cancer survivors.<br />
● Additional research is needed to evaluate the most<br />
efficient, highest quality care models in a variety <strong>of</strong><br />
settings and health systems.<br />
surveillance, and symptom management. 15-20 The ASCO<br />
Workforce Study identified that NPs and PAs can extend the<br />
services that oncology physicians provide and allow physicians<br />
to focus on evaluating new patients, creating treatment<br />
plans, and addressing changes in patient condition.<br />
NPs and PAs can free up oncologists from the more routine<br />
needs <strong>of</strong> patients with cancer so that they can focus on more<br />
complex care. 21 The Workforce Study also identified transitioning<br />
survivorship care from oncology physicians to the<br />
primary care setting as a means to close the gap in supply<br />
and demand. It was estimated that a reduction <strong>of</strong> 10% to<br />
20% in the number <strong>of</strong> patients being seen by an oncologist<br />
during the monitoring phase could be achieved. 13 Although<br />
the Workforce Study focused on transitioning survivorship<br />
care from oncologists to primary care physicians, it can be<br />
argued that NPs and PAs can also provide this care. Many<br />
cancer centers and oncology practices have established survivorship<br />
programs that utilize NPs and PAs. These programs<br />
have enabled oncologists to shift post-treatment care<br />
to other members <strong>of</strong> the team while keeping the patients<br />
within the institution or practice. 22<br />
Drivers <strong>of</strong> Collaborative Practice<br />
After the ASCO Workforce Study was completed, it was<br />
apparent that NPs and PAs would be an important part <strong>of</strong><br />
the solution to close the gap in supply and demand. However,<br />
there has been a limited understanding <strong>of</strong> the NP and<br />
PA workforce in oncology as well as the scope <strong>of</strong> services they<br />
provide, their effect on productivity, patient satisfaction,<br />
and physician satisfaction in the utilization <strong>of</strong> NPs and PAs.<br />
To address these issues, ASCO released the results for a<br />
Study <strong>of</strong> Collaborative Practice Arrangements (SCPA) in<br />
2011. The study had five main conclusions: oncology patients<br />
were aware and satisfied when the care they received was<br />
provided by NPs and PAs; there was an increase in productivity<br />
in practices that utilized NPs and PAs; utilizing the<br />
full scope <strong>of</strong> practice <strong>of</strong> NPs and PAs was financially advantageous<br />
and was a main driver <strong>of</strong> the model <strong>of</strong> collaborative<br />
practice chosen; perceptions <strong>of</strong> workload for oncologists,<br />
NPs, and PAs did not correlate to objective measures <strong>of</strong> work<br />
production; and physicians, NPs, and PAs are highly satisfied<br />
with their collaborative practices.<br />
Another interesting result <strong>of</strong> the SCPA was the list <strong>of</strong><br />
services provided by NPs and PAs in oncology practices. The<br />
top three services as identified by more than 80% <strong>of</strong> the<br />
practices surveyed were assessing patients during treatment<br />
visits, pain and symptom management, and follow-up<br />
care for patients in remission. The bottom three services<br />
with 20% or less <strong>of</strong> practices indicating these services were<br />
NPs and PAs providing night or weekend call, survivorship<br />
clinics, and “other” services. What is striking about these<br />
results is that NPs and PAs provided a large portion <strong>of</strong><br />
follow-up care for patients in remission, but this was not<br />
within a formal survivorship program. This seems to indicate<br />
that NPs and PAs have the skills needed to take care <strong>of</strong><br />
cancer survivors but that they are not yet doing so within<br />
structured programs.<br />
There have been three main collaborative practice models<br />
described between oncology physicians, NPs, and PAs. Each<br />
model has a different level <strong>of</strong> physician interaction with the<br />
patient and has a different effect on productivity and patient<br />
volume. The first model, incident-to-practice, is defined as<br />
autonomous practice <strong>of</strong> an NP/PA while the physician is<br />
e57
present in the <strong>of</strong>fice suite but does not routinely see patients.<br />
This practice model allows for the physician and NP/PA to<br />
work in parallel, thus increasing productivity and patient<br />
volume while allowing the physician to be available for<br />
patient care as needed. This model was the most prevalent<br />
in the SCPA. The second model, shared practice, is defined<br />
as physician and NP/PA care that is given together. The<br />
physician always sees the patient during the clinical visit.<br />
This model has a modest effect on productivity and patient<br />
volume. This model was the second most prevalent in the<br />
SCPA. The third model, independent practice, is defined as<br />
NP/PA care that is provided to a separate panel <strong>of</strong> patients<br />
without physician participation or presence. This model<br />
increased productivity and patient volume, but it is generally<br />
reimbursed at 85% <strong>of</strong> physician rates. This model was<br />
the least prevalent in the SCPA.<br />
NPs and PAs: Who Are They?<br />
Acknowledging that oncology has always been a multidisciplinary<br />
specialty, it is not difficult to ask the question <strong>of</strong><br />
who is the appropriate provider for various groups <strong>of</strong> survivors,<br />
based on a risk model <strong>of</strong> recurrence and late effects. As<br />
we frequently see, NPs and PAs can and are increasingly<br />
filling this role.<br />
NPs are licensed advanced practice nurses who have<br />
training as registered nurses before taking postgraduate<br />
training. Their education programs include degree-granting<br />
and postgraduate education programs (master’s or doctoral<br />
degree) and are accredited. To become licensed/certified as<br />
an NP, individuals must pass a certification examination<br />
that assesses national competencies <strong>of</strong> the overall role and<br />
at least one specialty area <strong>of</strong> practice, such as adultgerontology<br />
or pediatrics. NPs may also specialize in oncology,<br />
and the education for this expertise is supplemental<br />
to the core curriculum. Overall, the NP curriculum is broadbased<br />
and includes three separate level graduate-level<br />
courses in advanced physiology/pathophysiology, heath assessment,<br />
and pharmacology, as well as a variety <strong>of</strong> clinical<br />
experiences. Extensive clinical training provides the graduate<br />
with strong physical assessment skills and focuses on a<br />
comprehensive approach to patient care with an emphasis<br />
e58<br />
MCCABE AND PICKARD<br />
Fig. 1. Growth <strong>of</strong> gap between supply<br />
and demand for oncologist visits, 2005 to<br />
2020. 13<br />
on health promotion. This expertise is a strong complement<br />
to the skills <strong>of</strong> both the physician oncologist and the primary<br />
care physician. In support <strong>of</strong> the NP role as a key member <strong>of</strong><br />
the survivorship care team, the 2011 Institute <strong>of</strong> Medicine<br />
Report, The Future <strong>of</strong> Nursing: Leading Change, Advancing<br />
Health, highlights that nurses should practice to the full<br />
extent <strong>of</strong> their education and training—something the survivorship<br />
provider role <strong>of</strong>fers. 23 A second key message in<br />
this report is that nurses should be full partners with<br />
physicians and other health pr<strong>of</strong>essionals in redesigning<br />
health care in the United States. This is in synch with the<br />
ASCO SCPA that calls for the greater inclusion <strong>of</strong> NPs in<br />
oncology practice models and reports that such models<br />
result in increasing productivity for the practice and provide<br />
high physician and NP satisfaction, as well as satisfied<br />
patients.<br />
First proposed in the mid 1960s, PAs are pr<strong>of</strong>essional<br />
members <strong>of</strong> the health care team who are trained in the<br />
medical model. They practice in physician-PA teams<br />
throughout the United States, Canada, the United Kingdom,<br />
Australia, and the Netherlands. The PA educational<br />
model includes a general medical education for a period <strong>of</strong><br />
24–28 months (there are a handful <strong>of</strong> programs lasting<br />
36 months). The first year is spent in didactic training in<br />
gross anatomy, pathophysiology, medical terminology, microbiology,<br />
pharmacology, principals and practices <strong>of</strong> medicine,<br />
history and physical examination skills, bedside<br />
procedures, and major diseases <strong>of</strong> all body systems. The<br />
second year is spent in clinical rotations in pediatrics,<br />
internal medicine, family practice, emergency medicine,<br />
general surgery, obstetrics and gynecology, and a variety <strong>of</strong><br />
electives. The remaining coursework in the last semester <strong>of</strong><br />
training includes community health, epidemiology, biostatistics,<br />
pr<strong>of</strong>essional practice, and medical research. The<br />
master’s degree is currently the terminal degree for the PA<br />
pr<strong>of</strong>ession with a few programs <strong>of</strong>fering a clinical doctorate.<br />
There are a significant number <strong>of</strong> PA programs housed<br />
within medical schools where PA students and medical<br />
students take basic science courses together. Most PA students<br />
perform clinical rotations side by side with medical<br />
students.
NPS AND PAS IN SURVIVORSHIP CARE<br />
The educational curriculum and certification for the PA<br />
pr<strong>of</strong>ession is based on a single national standard. Education<br />
accreditation for all PA programs is through the Accreditation<br />
Review Commission on Education for the Physician<br />
Assistant (ARC-PA). The ARC-PA protects the interests <strong>of</strong><br />
the public and PA pr<strong>of</strong>ession by defining the standards for<br />
PA education and evaluating PA educational programs<br />
within the territorial United States to ensure their compliance<br />
with those standards. Graduation from an ARC-PA<br />
accredited PA program is a requirement for licensure in<br />
every state. There is a single national certification examination<br />
through the National Commission on the Certification<br />
<strong>of</strong> Physician Assistants (NCCPA) and is a requirement for<br />
licensure in every state.<br />
Practice as a PA is collaborative by definition. The pr<strong>of</strong>ession<br />
has defined itself and is based on a commitment to<br />
working with physicians in collaborative teams. 24 PAs do<br />
not practice medicine independent <strong>of</strong> their relationship with<br />
physicians. However, autonomous practice is very common,<br />
as physicians are not required to be present when services<br />
are provided by PAs. Prescriptive authority is granted to<br />
PAs in all 50 states.<br />
Integrating NP and PA Providers into<br />
Survivorship Programs<br />
The pr<strong>of</strong>essional training <strong>of</strong> both NPs and PAs lends itself<br />
very well to the comprehensive, holistic care proposed for<br />
survivors. This skill set allows them to provide the multiple<br />
services needed in survivorship care including age-specific<br />
cancer screening, general wellness, disease prevention, patient<br />
and caregiver counseling, nutrition counseling, and<br />
psychosocial assessment. Because many NPs and PAs have<br />
demonstrated the ability to develop expertise in oncology<br />
and are involved in the care <strong>of</strong> patients with cancer in active<br />
treatment, it will be important to assess the value <strong>of</strong> the<br />
inclusion <strong>of</strong> NPs and PAs as providers <strong>of</strong> survivorship care<br />
and determine whether such care can be transitioned immediately<br />
post-treatment to primary care providers as some<br />
have suggested. 25,26,27 And although NPs and PAs are<br />
increasingly taking on central roles in survivorship care, the<br />
question remains whether oncologists, patients, and payors<br />
will embrace this model nationally.<br />
An important issue related to the transition <strong>of</strong> survivorship<br />
care to other providers is the establishment <strong>of</strong> survivorship<br />
eligibility criteria and surveillance guidelines. Such<br />
guidance is a foundational step toward determining which<br />
patients are appropriate for survivorship care and when<br />
they can be transitioned from the oncologist to an NP or PA.<br />
Having clearly defined eligibility criteria based on stage <strong>of</strong><br />
disease and risk <strong>of</strong> recurrence <strong>of</strong>fers a consistent approach<br />
and important direction for those pr<strong>of</strong>essionals providing<br />
survivorship care, as well as cancer survivors themselves. In<br />
addition, guidance about the identification and management<br />
<strong>of</strong> long-term toxicities and late effects <strong>of</strong> cancer treatments<br />
are critical to the effective transition to other providers.<br />
Although limited guidance currently exists for patients with<br />
adult onset cancers, groups such as ASCO, the <strong>American</strong><br />
Cancer <strong>Society</strong>, and the National Comprehensive Cancer<br />
Network have all convened national groups <strong>of</strong> experts to<br />
address this important need.<br />
Although NPs and PAs have the training and experience<br />
to provide quality care to cancer survivors, it is important to<br />
realize that there a number <strong>of</strong> factors that should be ad-<br />
dressed before proceeding. Beyond the need for direction<br />
from oncologists through survivorship eligibility criteria and<br />
survivorship algorithms, there are basic regulatory considerations<br />
and logistics to consider. Unlike physicians, who<br />
enjoy a relatively stable scope <strong>of</strong> practice from state to state,<br />
the legal scope <strong>of</strong> practice for NPs and PAs can have<br />
significant variation. Some states allow broad prescriptive<br />
authority and other states tightly regulate this authority<br />
such that NPs and PAs cannot prescribe certain medications.<br />
The definition <strong>of</strong> physician supervision, collaboration,<br />
and delegation can also vary from state to state. This is<br />
important when determining if physicians have to be physically<br />
present or review a certain percentage <strong>of</strong> charts.<br />
Understanding the practice act for NPs and PAs in each<br />
state is crucial to ensure that they have the support from<br />
physicians they need as well as a defined operational structure<br />
in any survivorship program. 28 These logistical issues<br />
are not insurmountable nor are they prohibitive to incorporating<br />
NPs and PAs into survivorship programs. They are<br />
simply issues that need to be addressed to create an effective<br />
and efficient survivorship program. NPs and PAs have the<br />
training, medical knowledge, and skills to provide survivorship<br />
care; they are committed to collaborative relationships<br />
with physicians; and they have a demonstrated role in<br />
oncology with a commitment to lifelong learning. 29,21<br />
Models <strong>of</strong> Survivorship Care<br />
Internationally, as oncology physicians in hospitals, clinics,<br />
and private practices begin to develop more efficient<br />
approaches to the follow up <strong>of</strong> cancer survivors, they increasingly<br />
look to the incorporation <strong>of</strong> NPs and PAs as partners.<br />
There are well-established practice arrangements based on<br />
the Wagner Chronic Care Model from which to choose, as<br />
well as the multidisciplinary care model that has long been<br />
used in the care <strong>of</strong> pediatric cancer survivors. 30 The choice <strong>of</strong><br />
model is most <strong>of</strong>ten based on the complexity <strong>of</strong> the patient<br />
(risk for long-term and late effects) and the optimal financial<br />
reimbursement arrangement for the medical group. The<br />
Shared Survivorship Visit is common and has evolved into<br />
two types <strong>of</strong> survivorship care based on the diagnosis <strong>of</strong> the<br />
survivors being seen. First, is the multidisciplinary model,<br />
where survivors with a variety <strong>of</strong> diagnoses are cared for,<br />
and the survivor is seen by both the physician and NP or PA,<br />
with each provider focusing on particular aspects <strong>of</strong> the visit.<br />
This model is very useful for complicated patients, and in<br />
the general medicine experience, studies have shown that<br />
NP/PAs spend more time counseling the patient than if the<br />
patient were seen only by the physician. The second type <strong>of</strong><br />
shared visit is the disease-specific model in which the<br />
physician-NP/PA team focuses on one disease, such as<br />
breast cancer survivors. Many groups start with this type <strong>of</strong><br />
care model when setting up formal survivorship services as<br />
a way to pilot the sharing <strong>of</strong> care among providers and to<br />
assess the financial returns. Although the billing specifics<br />
may differ by state and payor, both types <strong>of</strong> shared visit<br />
models <strong>of</strong>fer the opportunity for joint billing in the physician’s<br />
name. The second category <strong>of</strong> visit is the Independent<br />
Survivorship Visit, which includes two types <strong>of</strong> care models<br />
where the NP or PA sees the survivors independently <strong>of</strong> the<br />
physician. The first type <strong>of</strong> independent visit is the consultative<br />
model; this common approach is easily established as<br />
a one-time or annual visit with an NP or PA during which<br />
general survivorship issues are discussed and a Treatment<br />
e59
Summary and Care Plan are provided. When specific issues<br />
are identified, the NP or PA makes referrals to appropriate<br />
specialists, such as medical subspecialists, physical therapists,<br />
or psychiatry. Although one advantage <strong>of</strong> this model is<br />
that the provider can serve an unrestricted survivor population,<br />
it can be difficult to be expert in the long-term<br />
follow-up issues <strong>of</strong> many types <strong>of</strong> survivors. Another independent<br />
visit type that is becoming increasingly adopted is<br />
the integrated model, where the NP or PA is imbedded with<br />
the treatment team, but sees the patient independently.<br />
The strength <strong>of</strong> this model is that it frees up the treating<br />
physician and allows the patient to continue to be seen in a<br />
familiar location with familiar staff, providing continuity <strong>of</strong><br />
care because <strong>of</strong> the proximity and relationship among providers.<br />
Both types <strong>of</strong> independent care models require a<br />
cultural change for some oncology physician groups, but as<br />
discussed in the ASCO workforce report, there are considerable<br />
pressures to encourage change in how care is delivered<br />
(Sidebar 2). 31<br />
Future Directions in Caring for Survivors<br />
Increasingly, the oncology and health policy literature<br />
contains papers providing support for innovative provider<br />
Authors’ Disclosures <strong>of</strong> Potential Conflicts <strong>of</strong> Interest<br />
models for the follow-up care <strong>of</strong> cancer survivors. Coupled<br />
with these early study results that show equivalent outcomes<br />
and excellent patient satisfaction when survivors<br />
receive care from NPs and PAs, are the economic and quality<br />
care incentives for change. <strong>Oncology</strong> has long been a multidisciplinary<br />
specialty, and reconfiguring the responsibilities<br />
<strong>of</strong> the team members can and should result in better, more<br />
efficient care at a time when the number <strong>of</strong> cancer survivors<br />
is rapidly increasing. Such new models focus on the key<br />
elements <strong>of</strong> a tailored visit utilizing a risk-based approach to<br />
care that allows the specific needs <strong>of</strong> the survivor to be<br />
addressed. In addition, such models all incorporate a communication<br />
link with the community primary care provider.<br />
If we are to overcome the barriers to shared-care between<br />
specialists and generalists, we need to incorporate a bidirectional<br />
communication plan that designates which provider is<br />
responsible for what specific medical services. This sharing<br />
<strong>of</strong> care has long been used for other disease groups such as<br />
patients with cardiac disease and diabetes. It is time for us<br />
to move out <strong>of</strong> our subspecialty isolation and assure coordinated<br />
care for our survivors. In addition, using NPs and PAs<br />
as a communication link can assist with the implementation<br />
<strong>of</strong> risk-based care and the transition <strong>of</strong> survivors back to the<br />
PCP when they are at low risk <strong>of</strong> recurrence and low risk <strong>of</strong><br />
late effects, not needing cancer follow-up. There is still much<br />
work to be done to move beyond pilot data and clinical<br />
hunches about the value <strong>of</strong> these models. We need to<br />
evaluate the patient health outcomes and the financial<br />
efficiency <strong>of</strong> these new models as well. Just as oncology has<br />
been a leader in many aspects <strong>of</strong> health care, it is an<br />
opportunity and necessity to do so in this important field <strong>of</strong><br />
cancer survivorship. We owe it to our patients to assure<br />
them the highest quality <strong>of</strong> life possible after cancer.<br />
Employment or<br />
Leadership Consultant or Stock<br />
Research<br />
Expert<br />
Other<br />
Author<br />
Mary S. McCabe*<br />
Positions Advisory Role Ownership Honoraria Funding Testimony Remuneration<br />
Todd Alan Pickard<br />
*No relevant relationships to disclose.<br />
CEMINES CEMINES (I)<br />
1. Rowland JH. Cancer Survivorship: Rethinking the cancer control continuum.<br />
Semin Oncol Nurs. 2008;24:145-152.<br />
2. Underwood JM, Townsend JS, Stewart ST, et al. Surveillance <strong>of</strong> demographic<br />
characteristics and health behaviors among adult cancer survivors -<br />
behavioral risk factor surveillance system, United States, 2009 CDC Surveillance<br />
Summaries. <strong>2012</strong>;61(SS01):1-23.<br />
3. Hewitt J, Greenfield S, Stovall E. (eds). From cancer patient to cancer<br />
survivor: Lost in transition. Washington, DC: National Academies Press; 2005.<br />
4. Bhatia S. Role <strong>of</strong> genetic susceptibility in development <strong>of</strong> treatmentrelated<br />
adverse outcomes in cancer survivors. Cancer Epidemiol Biomarkers<br />
Prev. 2011;20:2048-2067.<br />
5. Carmack CL, Basen-Engquist K, Gritz ER. Survivors at higher risk for<br />
adverse late outcomes due to psychosocial and behavioral risk factors. Cancer<br />
Epidemiol Biomarkers Prev. 2011;20:2068-2077.<br />
6. Oeffinger KC, Tonorezos ES. The cancer is over, now what? Cancer.<br />
2011;117(10 suppl):2250-2257.<br />
7. Miller KD, Triano LR. Medical issues in cancer survivors - a review.<br />
Cancer. 2008;14:375-387.<br />
8. Oeffinger KC, Mertens AC, Sklar CA, et al. Chronic health conditions in<br />
adult survivors <strong>of</strong> pediatric cancer. N Engl J Med. 2006;355:1572-1582.<br />
9. Rao AO, Demark-Wahnefried W. The older cancer survivor. Crit Rev Onc<br />
Hematol. 2006;60:131-143.<br />
e60<br />
Sidebar 2. Models <strong>of</strong> Survivorship Care: Physician–<br />
Nurse Practitioner/Physician Assistant<br />
Shared-Visit Model<br />
Multidisciplinary clinic<br />
Disease-specific clinic<br />
Independent Visit Model<br />
Consultative clinic<br />
Integrated clinic<br />
REFERENCES<br />
MCCABE AND PICKARD<br />
10. Holland J, Weiss T. The new standard <strong>of</strong> quality cancer care: Integrating<br />
the psychosocial aspects in routine cancer from diagnosis through survivorship.<br />
Cancer. 2008;14:425-428.<br />
11. Parry C, Kent EE, Mariotto AE, et al. Cancer survivors: A booming<br />
population. Cancer Epidemiol Biomarkers Prev. 2011;20:1996-2005.<br />
12. Bunnell CA, Shulman LN. Will we be able to care for cancer patients in<br />
the future? <strong>Oncology</strong>. 2010;24:1343-1352.<br />
13. Erikson C, Salsberg E, Forte G, et al. Future supply and demand for<br />
oncologists. J Oncol Pract. 2007;3:79-86.<br />
14. Street D, Cossman JS. Does familiarity breed respect? physician<br />
attitudes toward nurse practitioners in a medically underserved state. JAm<br />
Acad Nurse Prac. 2010;22:431-439.<br />
15. <strong>American</strong> Academy <strong>of</strong> Physician Assistants. Specialty Practice: PAs in<br />
<strong>Oncology</strong>. March 2011:1-4.<br />
16. Buswell LA, Ponte PR, Shulman LN. Provider practice models in<br />
ambulatory oncology practice: Analysis <strong>of</strong> productivity, revenue, and provider<br />
and patient satisfaction. J Oncol Pract. 2009;5:188-192.<br />
17. Friese CR, Hawley ST, Griggs JJ, et al. Employment <strong>of</strong> nurse practitioners<br />
and physician assistants in breast cancer care. J Oncol Pract.<br />
2010;6:312-316.<br />
18. Hinkel JM, Vandergrift JL, Perkel SJ, et al. Practice and productivity<br />
<strong>of</strong> physician assistants and nurse practitioners in outpatient oncology clinics
NPS AND PAS IN SURVIVORSHIP CARE<br />
at national comprehensive cancer network institutions. J Oncol Pract. 2010;<br />
6:182-187.<br />
19. Hylton H. <strong>Clinical</strong> partnership: The role <strong>of</strong> physician assistants in<br />
oncology practice. ASCO Daily News, June 4, 2011:21B.<br />
20. Ross AC, Polansky MN, Parker PA, et al. Understanding the role <strong>of</strong><br />
physician assistants in oncology. J Oncol Pract. 2010;6:26-30.<br />
21. Polansky M. Strategies for workplace learning used by entry-level<br />
physician assistants. J Physician Assist Educ. 2011;22:43-50.<br />
22. Towle EL, Barr TR, Goldstein M, et al. Results <strong>of</strong> the ASCO study <strong>of</strong><br />
collaborative practice arrangements. J Oncol Pract. 2011;7:278-282.<br />
23. IOM (Institutes <strong>of</strong> Medicine). The future <strong>of</strong> nursing: Leading change,<br />
advancing health. Washington, DC: The National Academies Press; 2011.<br />
24. <strong>American</strong> Academy <strong>of</strong> Physician Assistants. Pr<strong>of</strong>essional Issues: The<br />
Physician-PA Team. October 2011; 1-4.<br />
25. Watts SA, Gee J, O’Day ME, et al. Nurse practitioner-led multidisciplinary<br />
teams to improve chronic illness care: The unique strengths <strong>of</strong> nurse<br />
practitioners applied to shared medical appointments/group visits. J Am Acad<br />
Nurse Pract. 2009;21:167-172.<br />
26. Lewis R, Neal RD, Williams NH, et al. Nurse-led vs. conventional<br />
physician-led follow-up for patients with cancer: Systematic review. J Adv<br />
Nurs. 2009;65:706-723.<br />
27. Coniglio D, Pickard T, Wei S. Commentary: Physician Assistant Perspective<br />
on the Results <strong>of</strong> the ASCO Study <strong>of</strong> Collaborative Practice Arrangements.<br />
J Oncol Pract. 2011;7:283-284.<br />
28. Gage EA, Pailler M, Zevon MA, et al. Structuring survivorship care:<br />
Discipline-specific clinician perspectives. J Cancer Surv. 2011;5:217-225.<br />
29. Korber K. How do NPs and PAs get information and learn? Adv NPs<br />
PAs. 2011;2:14.<br />
30. Wagner E. Chronic disease management; what will it take to improve<br />
care for chronic illness? Effective <strong>Clinical</strong> Practice. 1998;1:2-4.<br />
31. Gilbert SM, Miller DC, Hollenbeck BK, et al. Cancer survivorship:<br />
Challenges and changing paradigms. J Urol. 2008;179:431-438.<br />
e61
DOING IT RIGHT, AND FOR LESS: IMPLEMENTING<br />
PRACTICE CHANGES TO MANAGE THE GROWING<br />
COMPLEXITIES, INEFFICIENCIES, AND<br />
COSTS OF CANCER CARE<br />
CHAIR<br />
Adam Brufsky, MD, PhD<br />
University <strong>of</strong> Pittsburgh School <strong>of</strong> Medicine<br />
Pittsburgh, PA<br />
SPEAKERS<br />
Bruce E. Hillner, MD<br />
Virginia Commonwealth University<br />
Richmond, VA<br />
Henry Orlando Otero, MD<br />
Virginia Mason Medical Center<br />
Seattle, WA
Driving Evidence-Based Standardization <strong>of</strong><br />
Care within a Framework <strong>of</strong> Personalized<br />
Medicine<br />
By Adam Brufsky, MD, PhD, Kathleen Lokay, BBA, and Melissa McDonald, MS<br />
Overview: Cancer care in the United States faces a number <strong>of</strong><br />
key challenges today that are causing payers, referring physicians,<br />
and patients alike to question the value <strong>of</strong> the care, in<br />
terms <strong>of</strong> both outcomes and costs. New technologies in the<br />
form <strong>of</strong> pharmaceuticals and biologics, prognostic tests, and<br />
new radiation therapy tools and techniques <strong>of</strong>fer the promise<br />
<strong>of</strong> improved outcomes, but their cost-effectiveness is <strong>of</strong>ten<br />
unclear. Oncologists themselves are caught in the middle, as<br />
they are the prescribers <strong>of</strong> such technologies and the entity<br />
billing for such services but have only limited ability to<br />
influence the pricing models for these services. Finally, as<br />
BY 2004, the University <strong>of</strong> Pittsburgh Medical Center<br />
(UPMC) Cancer Centers had expanded to approximately<br />
40 sites <strong>of</strong> services in a 100-mile radius in Western<br />
Pennsylvania. Concerns over the consistency and quality <strong>of</strong><br />
care across such a diverse network were validated through<br />
the results <strong>of</strong> internal surveys that demonstrated wide variability<br />
in approaches to cancer care. Pressures and concerns<br />
from the large payers in the region were also driving an<br />
imperative to collaborate around a solution to containing the<br />
rising costs. In addition, with the University <strong>of</strong> Pittsburgh<br />
Cancer Institute (UPCI) as its NCI-designated cancer center,<br />
UPMC needed tools for increasing awareness <strong>of</strong>, and<br />
accrual to, clinical trials. The solution for all <strong>of</strong> these needs<br />
was the development and implementation <strong>of</strong> the Via <strong>Oncology</strong><br />
Pathways program to improve quality, ensure consistency<br />
<strong>of</strong> care, reduce hospital admissions, and reduce total<br />
cost <strong>of</strong> care. To date, the program has served UPMC well for<br />
more than 7 years and provides a firm foundation for UPMC’s<br />
overall health care reform strategy for accountable care.<br />
Differences between Via <strong>Oncology</strong> Pathways and<br />
<strong>Oncology</strong> Guidelines<br />
Although excellent sources <strong>of</strong> oncology guidelines exist<br />
today, adherence to these guidelines is more difficult to<br />
assess, especially in community-based oncology practices,<br />
where more than 80% <strong>of</strong> cancer care is delivered. A recent<br />
analysis by IntrinsiQ <strong>of</strong> its data set <strong>of</strong> more than 17,000<br />
patients per month suggests that, for example, for non–<br />
small cell lung cancer, adherence to guidelines is 100% for<br />
first-line therapy but only 60% for second- and third-line<br />
therapies (Ed Kissell, IntrinsiQ, email communication, September<br />
2009). Even within guidelines, case studies routinely<br />
collected through physician surveys by reputable third parties<br />
suggest that a substantial amount <strong>of</strong> unexplained variability<br />
exists, in large part due to the inclusive nature <strong>of</strong><br />
multiple options as standards <strong>of</strong> care. Case study surveys<br />
conducted in 2005 within the UPMC Cancer Centers<br />
(UPMC-CC) revealed a high amount <strong>of</strong> variability within<br />
guidelines that could not be easily explained. No one will<br />
dispute that cancer is a very complex disease or that the<br />
nature <strong>of</strong> decision making is highly dependent on physician<br />
judgment for each unique patient. However, experiences in<br />
other fields <strong>of</strong> business suggest that oncology care will<br />
e62<br />
oncology care becomes more complex because <strong>of</strong> increased<br />
understanding <strong>of</strong> the pathogenesis <strong>of</strong> the many subtypes <strong>of</strong><br />
cancer, the community-based oncologist who cares for patients<br />
with all cancer subtypes is confronted with maintaining an<br />
up-to-date knowledge base that is expanding rapidly. Although<br />
no single solution for these issues exists in oncology today,<br />
the experience at the University <strong>of</strong> Pittsburgh Medical Center<br />
has demonstrated that a clinical pathways program can reduce<br />
unwarranted variability in both treatment and outcomes,<br />
drive adherence to evidence-based medicine, and, in the process,<br />
reduce the growth rate in the total costs <strong>of</strong> cancer care.<br />
benefit from a certain level <strong>of</strong> standardization for the majority<br />
<strong>of</strong> clinical presentations.<br />
Development and Maintenance <strong>of</strong> Via<br />
<strong>Oncology</strong> Pathways<br />
Starting in 2005, UPMC developed and maintained algorithms<br />
(Via <strong>Oncology</strong> Pathways) for oncology clinical decision<br />
making that UPMC physicians use to determine the<br />
best management for any given state and stage <strong>of</strong> cancer.<br />
This effort has been painstaking and has involved the<br />
cooperation <strong>of</strong> the majority <strong>of</strong> the academic and clinical<br />
experts at UPMC as well as numerous physicians from other<br />
practices. A committee exists for each major disease category<br />
(eg, colorectal) and consists <strong>of</strong> two chairpersons: one<br />
academically based oncologist specializing in that disease<br />
and one community-based oncologist with a background and<br />
patient concentration in that disease. The disease committees<br />
are composed <strong>of</strong> practicing oncologists from UPMC as<br />
well as all other practices that use the Via <strong>Oncology</strong> Pathways<br />
program. The committees convene quarterly to review<br />
new clinical literature, pathway results, and the appropriateness<br />
<strong>of</strong> the granularity <strong>of</strong> the algorithms (eg, defining<br />
the states and stages <strong>of</strong> disease and unique patient scenarios<br />
at which decisions should be made). Through their collaborative<br />
work, a single “best” treatment for the majority <strong>of</strong><br />
clinical scenarios in oncology care is defined. These best<br />
treatments are based on reviewing the literature in a consistent<br />
decision hierarchy. First, the committees look for the<br />
most effective treatment based on the current literature. In<br />
cases where there is a single “best” (i.e., most effective)<br />
treatment, that becomes the Via <strong>Oncology</strong> Pathway for that<br />
case. However, if there is more than one treatment with<br />
comparable efficacy, then the committees look for the least<br />
toxic therapy with the goal <strong>of</strong> maximizing patient quality <strong>of</strong><br />
life and outcomes and minimizing unnecessary costs <strong>of</strong><br />
From the University <strong>of</strong> Pittsburgh, Pittsburgh, PA; D3 <strong>Oncology</strong> Solutions (Via <strong>Oncology</strong><br />
Pathways), an affiliate <strong>of</strong> the University <strong>of</strong> Pittsburgh Medical Center, Pittsburgh, PA.<br />
Authors’ disclosures <strong>of</strong> potential conflicts <strong>of</strong> interest are found at the end <strong>of</strong> this article.<br />
Address reprint requests Adam Brufsky, MD, PhD, Division <strong>of</strong> Hematology/<strong>Oncology</strong>,<br />
University <strong>of</strong> Pittsburgh Cancer Center, 300 Halket St, Pittsburgh, PA 15213; email:<br />
brufskyam@msx.upmc.edu.<br />
© <strong>2012</strong> by <strong>American</strong> <strong>Society</strong> <strong>of</strong> <strong>Clinical</strong> <strong>Oncology</strong>.<br />
1092-9118/10/1-10
CLINICAL PATHWAYS STANDARDIZING PERSONALIZED MEDICINE<br />
toxicity management, such as hospitalizations. Finally, if there<br />
is more than one treatment with comparable efficacy and<br />
toxicity, the committees look for the least costly alternative to<br />
reduce unnecessary health care expenditures without compromising<br />
clinical benefit. In addition to a single best therapy<br />
as the primary pathway, the committees include options for<br />
common scenarios, such as neuropathy, poor performance<br />
status, drug shortages, etc. However, such options are only<br />
presented and counted as “on pathway” when the physician<br />
notes the specific scenarios. For those patient scenarios not<br />
addressed by the pathway, it is anticipated and expected that<br />
physicians will choose a treatment “<strong>of</strong>f pathway.” There is<br />
no penalty for such decisions and, in fact, adherence rates<br />
approaching 100% would cause concern. The expectation <strong>of</strong><br />
the Via <strong>Oncology</strong> Pathways program’s disease committees is<br />
adherence rates in the 70% to 90% range overall.<br />
Imbedding Data-Driven Personalized Medicine within<br />
the Via <strong>Oncology</strong> Pathways<br />
As part <strong>of</strong> the quarterly process for the Via <strong>Oncology</strong><br />
Pathways, the disease committees review not only data<br />
regarding alternative treatments but also biomarkers and<br />
other prognostic testing to drive to “personalized” medicine<br />
where the data are appropriate. The same level <strong>of</strong> rigor is<br />
applied to all alternatives with the additional requirement<br />
placed on biomarkers and prognostic tests <strong>of</strong> whether the<br />
results actually drive care decisions. If the data are robust<br />
and show a positive effect on care decisions, the pathways<br />
are updated to include these tests as recommendations (eg,<br />
anaplastic lymphoma kinase translocation–positive drives<br />
to crizotinib, epidermal growth factor receptor mutation<br />
drives to erlotinib, etc). For those tests without adequate<br />
data or firm clinical relevance, the committees develop<br />
literature-based explanations to be included in the pathways<br />
to discourage the ordering <strong>of</strong> these high-cost tests. As<br />
a result, the Via <strong>Oncology</strong> Pathways drive evidence-based<br />
personalized medicine only when it has been shown to make<br />
a difference in patient care.<br />
Physician Engagement Strategies<br />
A number <strong>of</strong> strategies have been successful in engaging<br />
oncologists both at UPMC and in other markets in the<br />
KEY POINTS<br />
● Issues in oncology today <strong>of</strong> quality, variability, and<br />
cost are driving the need for physician-led solutions.<br />
● A rigorous clinical pathways program includes physician<br />
experts who meet routinely to evaluate evidence<br />
and develop care algorithms that standardize<br />
personalized medicine.<br />
● Strategies to engage physicians must be employed,<br />
including involvement in the development and maintenance<br />
<strong>of</strong> the pathways content.<br />
● Decision support tools are critical components to<br />
achieving adherence and measuring results.<br />
● Early results demonstrate the success <strong>of</strong> clinical<br />
pathways programs in terms <strong>of</strong> adoption, adherence,<br />
reducing variability in care, patient outcomes, and<br />
cost containment.<br />
adoption <strong>of</strong> and adherence to the Via <strong>Oncology</strong> Pathways.<br />
Key to their support is an understanding <strong>of</strong> the realities <strong>of</strong><br />
the rising costs <strong>of</strong> cancer care and the likely changes that<br />
will be imposed by payers if no alternative solutions are<br />
proposed. Beyond these compelling reasons, other factors<br />
that drive physician acceptance include (1) creating an open<br />
and transparent disease committee process that allows<br />
participation by all physicians, (2) allowing unique patient<br />
scenarios to be addressed in the pathways, and (3) emphasizing<br />
that treating <strong>of</strong>f-pathway is not a negative outcome<br />
but an expected one because <strong>of</strong> the nature <strong>of</strong> the unique<br />
patient scenarios encountered daily.<br />
Decision Support Tools Are Critical for the Use and<br />
Measurement <strong>of</strong> <strong>Clinical</strong> Pathways<br />
The development and maintenance <strong>of</strong> the clinical content<br />
<strong>of</strong> the Via <strong>Oncology</strong> Pathways are certainly the most critical<br />
components in reducing unwarranted variability and adhering<br />
to evidence-based medicine. However, without tools to<br />
either deliver such content to the oncologists or measure<br />
their adherence to the pathways, the clinical content alone is<br />
no more valuable than other online resources or reference<br />
textbooks. Within UPMC, the need for such tools became<br />
evident quickly after the development <strong>of</strong> the clinical pathways<br />
and the implementation <strong>of</strong> a cumbersome paper-based<br />
process. Substantial investment was made by UPMC in<br />
developing a Web-based s<strong>of</strong>tware application for the delivery<br />
<strong>of</strong> the Via <strong>Oncology</strong> Pathways via a decision-support tool<br />
to the physicians. The delivery format is patient-specific and<br />
provided in real time. The Via <strong>Oncology</strong> Pathways are<br />
navigated through a question-and-answer format where the<br />
critical questions that drive that disease-specific pathway<br />
are presented and the physician is navigated to the appropriate<br />
branch <strong>of</strong> the decision tree based on his or her<br />
response. At the end <strong>of</strong> each node <strong>of</strong> the decision tree, local<br />
clinical trial options are presented, followed by the pathway<br />
treatment option including the full details <strong>of</strong> the order (ie,<br />
drugs, doses, schedule, other medications, etc). An easy-touse<br />
process is also available for selecting an <strong>of</strong>f-pathway<br />
treatment. Physicians are never prevented from going <strong>of</strong>f<br />
pathway but, rather, are asked to describe the reason for<br />
going <strong>of</strong>f pathway (from a predefined list).<br />
Results <strong>of</strong> Via <strong>Oncology</strong> Pathways Use<br />
Currently at UPMC, approximately 90 medical oncologists<br />
and 30 radiation oncologists use the Via <strong>Oncology</strong><br />
Pathways in their daily patient care. These oncologists<br />
practice at 40 sites <strong>of</strong> service, including the flagship academic<br />
center in the heart <strong>of</strong> Pittsburgh and communitybased<br />
sites over a 100-mile radius throughout Western<br />
Pennsylvania. In 2011, the UPMC physicians confirmed a<br />
pathways status for 94% <strong>of</strong> their patient visits (195,000<br />
visits) and achieved an on-pathway rate <strong>of</strong> 82% 2 for their<br />
treatment decisions (18,000 treatment decisions). The original<br />
premise <strong>of</strong> Via <strong>Oncology</strong> Pathways was to find the<br />
minimum number <strong>of</strong> therapies to meet the needs <strong>of</strong> the<br />
majority <strong>of</strong> patient scenarios. This result seems to reflect<br />
that the goals <strong>of</strong> reducing unwarranted variability, adhering<br />
to evidence-based medicine, and ensuring that each patient’s<br />
care is personalized, have been achieved. Other<br />
practices using the Via <strong>Oncology</strong> Pathways have generated<br />
comparable results.<br />
e63
A key question in the implementation <strong>of</strong> a pathways<br />
program is whether such reductions in variability also<br />
reduce or slow the rate <strong>of</strong> growth in cancer care costs. UPMC<br />
has collaborated with two external parties to explore these<br />
questions in three separate studies. First, through a multiyear<br />
effort with Highmark Blue Cross Blue Shield <strong>of</strong> Pennsylvania<br />
(Highmark), the largest commercial payer for<br />
UPMC, two studies were conducted to compare the effects<br />
<strong>of</strong> implementing the Via <strong>Oncology</strong> Pathways (then called<br />
the UPMC Pathways) program on growth rate in the total<br />
costs <strong>of</strong> cancer care with providers without such a program<br />
(Robert Wanovich, BOCP, Highmark Blue Cross Blue<br />
Shield, email communication, July 2007 and August 2009).<br />
To determine if a growth-rate reduction occurred, it was<br />
necessary to also study the rate <strong>of</strong> growth <strong>of</strong> cancer costs in<br />
a control arm <strong>of</strong> patients being treated by non-UPMC<br />
physicians who do not use any type <strong>of</strong> formal pathways<br />
program. The two studies examined the effects <strong>of</strong> implementing<br />
the pathways for non–small cell lung cancer and<br />
breast cancer. In both studies, two periods were measured:<br />
the 12-month period before the implementation <strong>of</strong> the pathway<br />
within UPMC and the 12-month period after the implementation.<br />
Total costs <strong>of</strong> care were calculated by Highmark<br />
for the control arm and the experimental arm, using data<br />
sets <strong>of</strong> actual claims payments for inpatient, outpatient, and<br />
pharmacy benefit claims. The results in both studies demonstrated<br />
a positive difference in the rate <strong>of</strong> growth favoring<br />
the experimental (i.e., UPMC) arm.<br />
In the study <strong>of</strong> the breast cancer pathway, preliminary<br />
results showed a 16% growth rate <strong>of</strong> costs for the nonpathways<br />
arm between the two 12-month periods compared with<br />
7% for the UPMC arm. In addition, hospital admissions per<br />
100 patients decreased by 15% in the UPMC arm compared<br />
with a 2% increase in the nonpathways arm. This difference<br />
suggests that reductions in unwarranted variability combined<br />
with a prioritization <strong>of</strong> treatments with lower toxicities<br />
can affect the rate <strong>of</strong> hospitalizations for patients with<br />
cancer. Such a reduction is not only a source <strong>of</strong> cost savings<br />
but also a likely improvement in patient quality <strong>of</strong> life and<br />
overall outcomes.<br />
In the study <strong>of</strong> the non–small cell lung cancer pathway,<br />
actual growth rates between the two 12-month periods were<br />
6% for the control arm and 1% for the UPMC arm. Included<br />
in this growth rate difference were reduced hospitalization<br />
costs <strong>of</strong> 12% in the UPMC arm, compared with a 4% increase<br />
in the control arm.<br />
Although the Highmark studies focused on the effect <strong>of</strong><br />
the Via <strong>Oncology</strong> Pathways program on the rate <strong>of</strong> growth<br />
in cancer costs in Western Pennsylvania, a third study with<br />
IntrinsiQ examined the possible implications <strong>of</strong> the Via<br />
<strong>Oncology</strong> Pathways program on the care provided elsewhere<br />
in the United States (Ed Kissell, IntrinsiQ, email communication,<br />
September 2009). IntrinsiQ is a national leader in<br />
the field <strong>of</strong> oncology information management and their<br />
proprietary chemotherapy ordering s<strong>of</strong>tware, Intellidose, is<br />
used by approximately 700 oncologists in more than 100<br />
practices. Their data set is widely used by key information<br />
consumers such as pharmaceutical companies because <strong>of</strong> its<br />
comprehensive nature, both in terms <strong>of</strong> clinical granularity<br />
and its ability to accurately project patterns <strong>of</strong> care for the<br />
entire United States. By reviewing the actual treatment<br />
decisions made by its oncologist customers and comparing<br />
those decisions to the Via <strong>Oncology</strong> Pathways recommenda-<br />
e64<br />
Table 1. Adjuvant Chemotherapy Usage and <strong>Clinical</strong> Trial<br />
Participation Rates for 174 Patients Node-Negative,<br />
HER2/Neu-Negative/Unknown, ER-Positive Breast Cancer<br />
(On-Pathway Rate: 90.8%)<br />
Treatment Selected, %<br />
<strong>Clinical</strong><br />
Oncotype Risk n TC AC Tamoxifen Anastrozole Trial<br />
High 34 76.5 11.8 — — 5.9<br />
Intermediate 73 80.8 2.7 6.8 6.8 1.4<br />
Low 40 — — 55.0 30.0 7.5<br />
Not Ranked 27 59.3 7.4 14.8 3.7 7.4<br />
Abbreviations: AC, doxorubicin/cyclophosphamide; ER, estrogen receptor;<br />
HER2/neu, human epidermal growth factor receptor-2; TC, docetaxel/cyclophosphamide.<br />
tions, IntrinsiQ was able to calculate the financial effects on<br />
the costs <strong>of</strong> drugs to payers if actual care had instead<br />
complied with the Via <strong>Oncology</strong> Pathways. By its calculations,<br />
payers could save approximately 40% on oncology<br />
drug costs if the Via <strong>Oncology</strong> Pathways had been followed<br />
in all retrospective prescribing decisions. Assuming that<br />
actual Via <strong>Oncology</strong> Pathways are followed in only 80% to<br />
85% <strong>of</strong> patient scenarios, such savings would likely be in the<br />
24% to 32% range.<br />
Finally, a pilot study <strong>of</strong> Via <strong>Oncology</strong> Pathways and<br />
Horizon Blue Cross Blue Shield <strong>of</strong> New Jersey (Horizon) has<br />
generated encouraging results in an early analysis (Richard<br />
Popiel, MD, Horizon Healthcare Innovations, and Richard<br />
Weininger, MD, CareCore National, email communication,<br />
February <strong>2012</strong>). Two large practices in northern and southern<br />
New Jersey implemented the Via <strong>Oncology</strong> Pathways in<br />
the third quarter <strong>of</strong> 2010. The study compares total cost <strong>of</strong><br />
care (excluding radiation oncology, because the participating<br />
practices provided only medical oncology services) between<br />
these Via <strong>Oncology</strong> Pathways practices (experimental arm)<br />
and the remaining practices in New Jersey (control arm)<br />
over the same periods. The initial results are being validated<br />
by Horizon with possible publication in the second half <strong>of</strong><br />
<strong>2012</strong>.<br />
Breast Cancer Treatment Patterns at UPMC-CC<br />
An analysis <strong>of</strong> the patterns <strong>of</strong> care for patients with breast<br />
cancer within UPMC-CC for the 12 months that ended May<br />
31, 2011, was performed to describe utilization patterns and<br />
concordance with the Via <strong>Oncology</strong> Pathways. This period<br />
was selected to avoid the confounding changes in care<br />
patterns in the second half <strong>of</strong> 2011 because <strong>of</strong> concerns from<br />
the United States Food and Drug Administration over the<br />
use <strong>of</strong> bevacuzimab to treat patients with breast cancer. The<br />
analysis included new chemotherapy treatment decisions for<br />
Table 2. Frequency and Description <strong>of</strong> Treatment Decisions<br />
According to the Via <strong>Oncology</strong> Pathways and Accrual to <strong>Clinical</strong><br />
Trials for 104 Patients with HER2/Neu-Negative/Unknown,<br />
ER-Negative, PR-Negative Breast Cancer Receiving Adjuvant<br />
Chemotherapy (On-Pathway Rate: 79.8%)<br />
Node Statis<br />
No. <strong>of</strong><br />
Patients<br />
BRUFSKY, LOKAY, AND MCDONALD<br />
Treatment Selected, %<br />
TC AC TAC Other<br />
<strong>Clinical</strong><br />
Trial<br />
Negative 82 62.2 17.1 — 11 9.8<br />
Positive (1–3 nodes) 22 18.2 59.1 4.5 4.5 13.6<br />
Abbreviations: AC, doxorubicin/cyclophosphamide; ER, estrogen receptor;<br />
HER2/neu, human epidermal growth factor receptor-2; PR, progesterone receptor;<br />
TAC, docetaxel/doxorubicin/cyclophosphamide; TC, docetaxel/cyclophosphamide.
CLINICAL PATHWAYS STANDARDIZING PERSONALIZED MEDICINE<br />
Table 3. Frequency and Description <strong>of</strong> Treatment Decisions<br />
According to the Via <strong>Oncology</strong> Pathways and Accrual to <strong>Clinical</strong><br />
Trials for 109 Patients with HER2/Neu-Positive, ER-Positive<br />
Breast Cancer Receiving Adjuvant Chemotherapy<br />
(On Pathway Rate: 82.6%)<br />
Node Status<br />
patients with metastatic breast cancer receiving adjuvant<br />
and first-line treatment as documented by the UPMC-CC<br />
oncologists in the Via <strong>Oncology</strong> Pathways system. Excluded<br />
from the analysis were hormone therapy decisions (except<br />
for patients with node-negative or HER2-negative disease<br />
receiving adjuvant treatment) and biologic drugs prescribed<br />
as monotherapy (these are assumed to be maintenance<br />
therapy following combination therapy).<br />
For this period and this population <strong>of</strong> decisions, the<br />
on-pathway rate (ie, treatment decisions per the Via <strong>Oncology</strong><br />
Pathways recommendation divided by all treatment<br />
decisions) was 85.7% (505 treatment decisions). Tables 1<br />
through 4 describe the frequency and description <strong>of</strong> treatment<br />
decisions according to the Via <strong>Oncology</strong> Pathways and<br />
Authors’ Disclosures <strong>of</strong> Potential Conflicts <strong>of</strong> Interest<br />
Author<br />
No. <strong>of</strong><br />
Patients<br />
Treatment Selected, %<br />
TCH AC Other<br />
Employment or<br />
Leadership<br />
Positions<br />
Consultant or<br />
Advisory Role<br />
Adam Brufsky Celgene;<br />
Genentech;<br />
Novartis; Roche<br />
Kathleen Lokay D3 <strong>Oncology</strong><br />
Solutions<br />
Melissa McDonald D3 <strong>Oncology</strong><br />
Solutions<br />
<strong>Clinical</strong><br />
Trial<br />
Negative 70 78.6 2.9 14.3 4.3<br />
Positive 39 84.6 — 7.7 7.7<br />
Abbreviations: AC, doxorubicin/cyclophosphamide; ER, estrogen receptor;<br />
HER2/neu, human epidermal growth factor receptor-2; TCH, docetaxel/carboplatin<br />
with concurrent trastuzumab, followed by trastuzumab.<br />
Table 4. Frequency and Description <strong>of</strong> Treatment Decisions<br />
According to the Via <strong>Oncology</strong> Pathways and Accrual to <strong>Clinical</strong><br />
Trials for 118 Patients with HER2/Neu-Negative/Unknown,<br />
ER-Positive, PR-Positive Metastatic Breast Cancer Receiving<br />
First-Line Chemotherapy (On-Pathway Rate: 86.4%)<br />
Treatment Selected (%)<br />
Capecitabine 27.1<br />
Paclitaxel/Bevacizumab 22.9<br />
Paclitaxel (protein-bound)/Bevacizumab 19.5<br />
Paclitaxel (protein-bound) 8.5<br />
Paclitaxel 5.9<br />
Docetaxel/Gemcitabine 2.5<br />
Other 8.5<br />
<strong>Clinical</strong> Trial 5.1<br />
Abbreviations: ER, estrogen receptor; HER2/neu, human epidermal growth<br />
factor receptor-2; PR, progesterone receptor.<br />
1. Paik S, Tang G, Shak S, et al. Gene expression and benefit <strong>of</strong> chemotherapy<br />
in women with node-negative, estrogen receptor-positive breast<br />
cancer. J Clin Oncol. 2006;24:3726-3734.<br />
accrual to clinical trials. All <strong>of</strong>f-pathways decisions are<br />
aggregated and described as “Other.” These data are reported<br />
“as is” based on physician answers to questions<br />
within the Via <strong>Oncology</strong> Pathways s<strong>of</strong>tware application and<br />
have not been validated against original medical records.<br />
In the adjuvant setting, the pathway for breast cancer<br />
provides guidance for the use <strong>of</strong> Oncotype DX recurrence<br />
scores (RS) in both node-positive and node-negative settings.<br />
Table 1 shows a breakdown <strong>of</strong> chemotherapy usage for<br />
patients with node-negative, HER2/Neu-negative disease.<br />
Chemotherapy usage is high (� 88%) in the high-risk group<br />
and was not reported in the low-risk group. This usage is<br />
consistent with retrospective data that suggest patients<br />
with node-negative disease with high RS scores benefit from<br />
chemotherapy and those with low RS scores (� 18) do not. 1<br />
Intermediate-risk patients showed more variability, with<br />
84% receiving chemotherapy and 14% receiving hormonal<br />
treatment alone.<br />
In the first-line HER2/Neu-negative metastatic setting<br />
(Table 4), usage reflects an 86% concordance with the Via<br />
<strong>Oncology</strong> Pathways. This presentation provides an example<br />
<strong>of</strong> how the disease committees provide options for numerous<br />
patient scenarios, each with different associated treatments.<br />
Conclusion<br />
The results from the UPMC experience with the Via<br />
<strong>Oncology</strong> Pathways program and those reported by other<br />
clinical pathways programs 2 suggest that these are effective<br />
models for improving quality, reducing unwarranted variability<br />
in care, and reducing the rate <strong>of</strong> growth in the cost <strong>of</strong><br />
cancer care. A robust pathways program that reduces unwarranted<br />
variability can serve as the vehicle to improve the<br />
value <strong>of</strong> cancer care to patients, payers, and providers<br />
through increasing quality and decreasing costs. Oncologists<br />
must take an active role in defining and implementing<br />
cost-effective care for patients, payers, and referring physicians,<br />
or suffer the alternatives that potentially compromise<br />
quality, access to care, and practice viability. If appropriately<br />
implemented, clinical practice guidelines and pathways<br />
are possible solutions to improving the quality and<br />
cost-effectiveness <strong>of</strong> cancer care that preserves physician<br />
decision-making, ensures access to evidence-based personalized<br />
medicine, and eliminates nonvalue-added administrative<br />
hurdles such as prior authorizations.<br />
Stock<br />
Ownership Honoraria<br />
Celgene;<br />
Genentech; Lilly;<br />
Novartis; Roche<br />
REFERENCES<br />
Research<br />
Funding<br />
Celgene;<br />
Genentech; Lilly;<br />
Novartis; Roche<br />
Expert<br />
Testimony<br />
Other<br />
Remuneration<br />
2. Neubauer MA, Hoverman JR, Kolodziej M, et al. Cost effectiveness <strong>of</strong><br />
evidence-based treatment guidelines for the treatment <strong>of</strong> non-small-cell lung<br />
cancer in the community setting. JOP. 2010;6(1):12-18.<br />
e65
PATIENT PORTALS IN ONCOLOGY<br />
CHAIR<br />
Elizabeth S. Rodriguez, DNP, RN, OCN<br />
Memorial Sloan-Kettering Cancer Center<br />
New York, NY<br />
SPEAKERS<br />
Henry Feldman, MD<br />
Division <strong>of</strong> <strong>Clinical</strong> Informatics<br />
Brookline, MA
The Future <strong>of</strong> <strong>Oncology</strong> Care with Personal<br />
Health Records<br />
By Henry Feldman, MD, and Elizabeth S. Rodriguez, DNP, RN, OCN<br />
Overview: Personal health records (PHRs) and patients’ access<br />
to their own clinical information through a patient portal<br />
are changing the patient-physician relationship. Historically,<br />
health care providers have been gatekeepers <strong>of</strong> patients’<br />
medical records. Now, these portals provide patients access<br />
to clinical information, electronic messaging with the clinical<br />
team, and appointment and billing information. This type <strong>of</strong><br />
access supports patient empowerment by engaging patients<br />
in their own care.<br />
Patients desire online access to information. The health<br />
care industry, like any other, must respond to the needs <strong>of</strong> its<br />
consumers. <strong>Oncology</strong> practices face unique challenges to<br />
meeting this need because <strong>of</strong> the complex nature <strong>of</strong> medical<br />
records <strong>of</strong> patients with cancer . Health care providers worry<br />
about the consequences <strong>of</strong> patients receiving “bad news”<br />
online, thereby increasing patient anxiety. This anxiety may, in<br />
PERSONAL HRs, also referred to as patient portals,<br />
began in 2000 with Beth Israel Deaconess Medical<br />
Center launching PatientSite online, 1 which allowed patients<br />
to view laboratory results, medication lists, imaging<br />
reports, pathology reports, and financial reports, and to send<br />
messages to their health care providers. Early during the<br />
implementation, there was much controversy about whether<br />
patients would be frightened or overwhelmed by the technical<br />
aspects <strong>of</strong> data and whether they would inundate their<br />
providers with questions about the data in their record.<br />
These concerns were most strongly raised by two groups <strong>of</strong><br />
providers: oncologists and psychiatrists. These fears were<br />
not borne out over the 12 years since deployment <strong>of</strong> that<br />
system, and in the intervening years, many large medical<br />
centers and health care systems have developed and deployed<br />
large-scale PHRs. As <strong>of</strong> this writing, tens <strong>of</strong> millions<br />
<strong>of</strong> patients have access to PHRs sharing select clinical data<br />
over the Internet in a secure manner.<br />
As part <strong>of</strong> The <strong>American</strong> Recovery and Reinvestment Act<br />
<strong>of</strong> 2009, the federal government, under the auspices <strong>of</strong><br />
the Department <strong>of</strong> Health and Human Services decided<br />
that the US health care system would benefit from the<br />
increased use <strong>of</strong> electronic health records (EHRs). To ensure<br />
that government funds were allocated for useful progress,<br />
a set <strong>of</strong> phased “meaningful use” guidelines were<br />
enacted by the Centers for Medicare & Medicaid Services<br />
that vendors and health care providers would have to meet,<br />
under the auspices <strong>of</strong> the Office <strong>of</strong> the National Coordinator<br />
for Health Information Technology (ONC-HIT). One <strong>of</strong> these<br />
requirements in phase II is the use <strong>of</strong> PHRs. The required<br />
features <strong>of</strong> PHRs include inpatient and outpatient data that<br />
must be shared. As such, practices must start addressing<br />
these issues in relation to their practice areas, including<br />
oncology.<br />
PHRs and patient portals are rapidly being viewed as a<br />
way to encourage patient empowerment and engage patients<br />
in their own care. Multiple studies have shown poor<br />
retention <strong>of</strong> complex technical information by patients during<br />
<strong>of</strong>fice visits. 2,3 In subspecialty fields, such as oncology,<br />
extremely complex care plans are presented, with many<br />
options and complications, and it is unreasonable to expect<br />
e66<br />
turn, increase providers’ workload by creating additional calls<br />
or visits to the <strong>of</strong>fice.<br />
These valid concerns require careful consideration when<br />
implementing a PHR or patient portal into a practice. Providers<br />
will benefit from a clear understanding <strong>of</strong> actual compared<br />
with potential risks and benefits. Much <strong>of</strong> the concerns about<br />
the negative effect on providers’ workload and the potential<br />
increase in patients’ anxiety have not been borne out. On the<br />
other hand, the implementation strategy, governance structure,<br />
and end-user education are crucial components to ensuring<br />
success.<br />
Successful implementation <strong>of</strong> a PHR or patient portal affords<br />
the opportunity to improve patient satisfaction and<br />
increase efficiency in provider workflow. The possibility exists<br />
to improve patient outcomes by engaging the patient in<br />
decision making and follow through.<br />
any patient to recall everything from an <strong>of</strong>fice or inpatient<br />
visit. These tools can help leverage your time and improve<br />
patient-physician communication. 4,5 However, like any intervention,<br />
these tools have a risk-benefit ratio and side<br />
effects and must be managed carefully.<br />
Governance<br />
One <strong>of</strong> the most difficult issues practices face in implementing<br />
this type <strong>of</strong> project, and particularly one that<br />
crosses many stakeholder areas and needs, is governance.<br />
PHRs involve a particularly complex area <strong>of</strong> governance, as<br />
much <strong>of</strong> the data are regulated by various government<br />
agencies, the data serve multiple purposes, the information<br />
has not traditionally been intended for lay readers, and data<br />
are being shared between content experts (providers) and<br />
lay readers (patients and surrogates) who have different<br />
ideas <strong>of</strong> immediacy, privacy, and explanatory material. All <strong>of</strong><br />
these factors need to be brought to the table in planning and<br />
monitoring the implementation and use <strong>of</strong> a PHR. 6,7 Clearly,<br />
representation by the facility’s stakeholders is crucial; these<br />
stakeholders include the technical staff <strong>of</strong> the practice;<br />
clinical staff, including physicians and nurses; informatics<br />
staff; business operations; and, most importantly, patients.<br />
Practices may also consider addinglegal staff, as PHRs are<br />
becoming more regulated.<br />
This governance structure has several key tasks to initiate<br />
and monitor the project:<br />
● Ensure that all regulatory issues are addressed<br />
● Work with patients and caregivers to determine what<br />
data need to be shared, when, and in what formats<br />
From Beth Israel Deaconess Medical Center, Harvard Medical School, Boston, MA;<br />
Ambulatory Care Services, Department <strong>of</strong> Nursing, Memorial Sloan-Kettering Cancer<br />
Center, New York, NY.<br />
Authors’ disclosures <strong>of</strong> potential conflicts <strong>of</strong> interest are found at the end <strong>of</strong> this article.<br />
Address reprint requests to Henry Feldman, MD, Division <strong>of</strong> <strong>Clinical</strong> Informatics, Beth<br />
Israel Deaconess Medical Center, 1330 Beacon St., Suite 400, Brookline, MA 02446; email:<br />
hfeldman@bidmc.harvard.edu.<br />
© <strong>2012</strong> by <strong>American</strong> <strong>Society</strong> <strong>of</strong> <strong>Clinical</strong> <strong>Oncology</strong>.<br />
1092-9118/10/1-10
PERSONAL HEALTH RECORDS<br />
● Determine what resources are needed at all phases <strong>of</strong><br />
the project<br />
● Set policies that all providers can work with to support<br />
clinical care and patient empowerment<br />
● Decide what data are shared immediately, embargoed,<br />
or blocked<br />
● Provide end-user support and education<br />
● Develop communication policies to prevent messaging<br />
overload or under-alerting for providers and patients<br />
These governing bodies must be given authority to make<br />
changes to the applications as needs arise and to enact<br />
policies for both staff and patients. If a practice has multiple<br />
specialties, it is recommended that representatives from<br />
several specialty areas be consulted, at least initially about<br />
the design and needs analysis.<br />
Embargos Compared with Blocking<br />
One <strong>of</strong> the commonly voiced concerns about displaying<br />
clinical data directly to patients is that a patient may see a<br />
report with ominous findings before his or her clinician has<br />
seen it. The most common phrase is similar to “I do not want<br />
my patient finding out their cancer has recurred from a<br />
website!” This is a reasonable concern and, although such a<br />
situation may be rare, most clinicians understand that<br />
pathology, radiology, and some laboratory results come back<br />
with a differential rather than a simple diagnosis, which<br />
may be quite concerning. This problem is similar to notes by<br />
medical students, which <strong>of</strong>ten have academic leanings<br />
rather than concise clinical pictures (“that purpuric rash<br />
could be Ebola but more likely is idiopathic thrombocytopenic<br />
purpura”). Such notes can produce anxiety in patients,<br />
whereas clinicians have the expertise to know the likelihoods<br />
<strong>of</strong> specific outcomes and can filter out much <strong>of</strong> the<br />
noise to summarize for the patient.<br />
As such, the governance structure needs to closely look at<br />
the evidence for how to treat a given piece <strong>of</strong> clinical<br />
information. The basic three choices are to share it immediately,<br />
embargo it for a specified time period, or block it<br />
(never show it). Most medical centers have chosen to show<br />
KEY POINTS<br />
● Personal health records (PHRs) and patient portals<br />
provide patients with increased access to their own<br />
medical information.<br />
● Important considerations exist for the use <strong>of</strong> PHRs<br />
and portals related to the types <strong>of</strong> data to display, the<br />
timeliness <strong>of</strong> data release, electronic communication<br />
with patients, and end-user education.<br />
● A governance structure is essential for successful<br />
system implementation and sustainability while<br />
keeping providers engaged and protecting patient<br />
safety and privacy.<br />
● PHRs and patient portals can serve as a platform for<br />
online patient communities and as a way to share<br />
clinical trial information.<br />
● Evaluation <strong>of</strong> how patients and providers use these<br />
sites can be easily accomplished through analysis on<br />
the backend <strong>of</strong> the system.<br />
routine labs (such as basic chemistry pr<strong>of</strong>iles) immediately,<br />
embargo pathology and radiology results, and to block such<br />
results as HIV antibody testing. Most embargos last 1 to 6<br />
days, and there is a safety net in knowing that if a clinician<br />
fails to see a report for some reason, the patient will see it<br />
after 6 days, and an important finding can be addressed at<br />
that time. We address the special topic <strong>of</strong> clinical notes in<br />
the section on Types <strong>of</strong> Data to Share.<br />
Communication Strategies<br />
Patient-physician communication is hard enough in person,<br />
as any clinician can tell you, and it becomes infinitely<br />
more complex when delivered through electronic media.<br />
Furthermore, because health care, and oncology in particular,<br />
is a “team sport,” it is <strong>of</strong>ten overwhelming to a patient to<br />
figure out who to message about a given need. Front-line<br />
providers are also overwhelmed by information overload and<br />
do not welcome additional channels <strong>of</strong> communication. Interestingly,<br />
patients are also surprisingly cautious about<br />
adding to clinicians’ workloads, and so messaging strategies<br />
need to work with both the practice style and patients’<br />
information needs. 8 Whatever strategy is adopted, it is<br />
crucial that it be tightly integrated into practice workflows,<br />
or it will not be advantageous to the clinician.<br />
Messages typically can be divided into administrative<br />
(medication refills, scheduling, referrals, financial matters)<br />
and clinical (new medication request, educational needs,<br />
and symptom/follow-up inquiries). Dividing basic tasks between<br />
clinical and nonclinical messaging makes sense and is<br />
a good start. However, many lines are blurred and because<br />
<strong>of</strong> this, some practices have adopted a system in which a<br />
central person is responsible for triaging all messages to a<br />
practice, which can be very successful. One subspecialty<br />
clinic on PatientSite uses a total communication strategy in<br />
this manner, where all communication to and from the clinic<br />
is through the PHR and this single point <strong>of</strong> communication,<br />
with a guaranteed response time. This system has been<br />
quite successful for this clinic, but it required a large<br />
investment (a full-time equivalent staff) by the clinic and<br />
policy makers. A corollary to this is that every person who<br />
receives a message is a single point <strong>of</strong> failure if he or she<br />
either does not check messages or is unavailable to do so,<br />
and so coverage schemes are required. A strategy that most<br />
clinics adopt for this is a two-level process, where most<br />
clinical messages are received by all clinical providers on<br />
that patient’s team (physicians, nurse practitioners, registered<br />
nurses, and administrative staff) and whoever responds<br />
to the patient removes the message from the other<br />
users’ queues. As a backup, each individual can denote<br />
another user to provide coverage <strong>of</strong> his or her inbox.<br />
Whatever your strategy is, it is vital to manage expectations;<br />
that is, to explain to patients what the turnaround<br />
time is for messaging and what can be handled through<br />
messaging and what should be handled through call-in or<br />
other mechanisms. Closed-loop communication also helps<br />
eliminate patients’ anxiety about the status <strong>of</strong> messages.<br />
Typically, this type <strong>of</strong> communication can be performed<br />
by the system automatically (for example, sending readreceipts<br />
or notification once a prescription has been sent<br />
to the pharmacy). Patients understand that clinicians are<br />
busy, and do not mind waiting for most responses for a<br />
reasonable period <strong>of</strong> time, as long as that time is a known<br />
e67
quantity. Determining the appropriate time period is an<br />
important task for the governance committee.<br />
One common complaint has been that time spent messaging<br />
with patients is nonreimbursed. However, in the upcoming<br />
capitated Accountable Care Organization (ACO) era,<br />
time communicating electronically is time well spent if it can<br />
save an <strong>of</strong>fice visit, admission, or test. Unlike telephone or<br />
face-to-face communication, the asynchronous nature <strong>of</strong><br />
electronic communication allows the provider to prioritize<br />
the communication, increasing efficiency. 9 The economics <strong>of</strong><br />
PHRs will change over the next few years, and funding <strong>of</strong><br />
these projects need to take this into effect.<br />
Types <strong>of</strong> Data to Share<br />
The most controversial area <strong>of</strong> PHRs is what the patient<br />
should be allowed to see and do. Most PHR providers agree<br />
that financial information is noncontroversial to share, as is<br />
medication and scheduling information; in fact, sharing this<br />
information through a PHR can <strong>of</strong>ten reduce phone calls to<br />
the clinic. Radiology and pathology reports are much more<br />
controversial, but many practices now have several years <strong>of</strong><br />
experience with little untoward events, with use <strong>of</strong> embargos.<br />
One area <strong>of</strong> concern is the question <strong>of</strong> releasing clinician<br />
notes to patients. There are two large-scale examples <strong>of</strong><br />
sharing notes have demonstrated little downside and have<br />
been well received. The OpenNotes trial in 2011 included<br />
20,000 patients across three medical centers, and opened<br />
all primary care notes to the patients. 10,11 (You can view<br />
videos about the experience at myopennotes.org.) The<br />
oncology-specific example is from the University <strong>of</strong> Texas<br />
M. D. Anderson Cancer Center in Texas, which opened all<br />
clinician notes to patients, and approximately 60,000 patients<br />
have viewed their notes through their PHRs at<br />
myMDAnderson.org. The standard concern is that patients<br />
will be scared by what they see in the chart; however, hiding<br />
medical issues from patients stands on shaky ethical<br />
grounds. Furthermore, under federal Health Insurance Portability<br />
and Accountability Act (HIPAA) laws and some state<br />
laws, patients must be able to view their medical records.<br />
Your governance committee should work closely with<br />
patients, providers, and other stakeholders to decide what<br />
will and will not be shown and whether there will be<br />
embargos, and have clear policies with transparent reasoning<br />
for each decision. These policies should be posted clearly<br />
on the PHR for patients to see. You should also be prepared<br />
to readdress policies as evidence and technology advance.<br />
Provider Education<br />
Like any clinical tool, full PHR use requires education for<br />
the team that will be using it. <strong>Clinical</strong> champions will<br />
increase your success <strong>of</strong> adoption, as providers are more<br />
willing to accept advice from “one <strong>of</strong> their own” than from<br />
dedicated training staff. If your PHR is so disruptive to<br />
clinical workflows that a major training effort is required,<br />
you may wish to reexamine the system design. Education is<br />
needed about how to manage expectations <strong>of</strong> patients and<br />
about the use <strong>of</strong> inappropriate phrases in documentation.<br />
For instance, clinicians should use dyspnea rather than<br />
“SOB” (or shortness <strong>of</strong> breath), as patients may not understand<br />
the abbreviation. Although patients have the right to<br />
see their records, few do so in reality, and clinician documentation<br />
<strong>of</strong>ten reflects this.<br />
e68<br />
Patient Education<br />
Patient education is needed in several areas, some technologic<br />
and some personal. Low technical literacy may be a<br />
problem if your patient population is primarily older and low<br />
health or English literacy may be a problem if you have a<br />
large immigrant population. How you will address specific<br />
groups <strong>of</strong> users and help them utilize your systems to their<br />
best advantage should be worked out during the design<br />
phase <strong>of</strong> the project. You also must develop a process for<br />
educating patients about the rollout <strong>of</strong> the PHR itself, and<br />
about communication expectations <strong>of</strong> the implementation. If<br />
the patient education needs become too cumbersome, it<br />
would be wise to reexamine the design.<br />
Once patients are using the PHR, your practice needs to<br />
decide how much patient education materials you will include<br />
for them to be self-sufficient. Patients are very selfreliant,<br />
and will use Google and other search engines to try<br />
and understand what they read in their PHR and will<br />
overcome the jargon. 8 Your practice may decide on either<br />
pre-made or new materials for patients. You should also<br />
decide if you will send materials to patients (based on<br />
diagnoses or on clinician selection) or simply provide libraries<br />
for them to browse/search. For instance, when you<br />
display laboratory results, do you show the internal name<br />
for the tests (such as “Na” for sodium) or do you show the full<br />
names with links to basic explanations <strong>of</strong> each lab type?<br />
Dana-Farber Cancer Center and Memorial Sloan-Kettering<br />
Cancer Center have opted to use www.labtestsonline.org<br />
to provide patients a reliable source <strong>of</strong> information in understandable<br />
terms. 12,13 This public website is for patients<br />
and caregivers and <strong>of</strong>fers a large library <strong>of</strong> definitions and<br />
explanations <strong>of</strong> common lab tests as a resource.<br />
Communities<br />
Many medical practices have considered implementing a<br />
patient community function in their PHRs. Communities<br />
are similar to social networking chat areas, where patients<br />
discuss symptoms, therapies, and other topics <strong>of</strong> mutual<br />
concern. The use <strong>of</strong> a community function is legally complex,<br />
and needs to be carefully considered by any practice. Although<br />
many <strong>of</strong> these chat areas already exist on the<br />
Internet, they are not hosted at medical practices and<br />
HIPAA-covered entities.<br />
The challenge with implementing communities is that<br />
patients could breach each other’s confidential health information.<br />
Because the communities are hosted and provided<br />
by a practice, patient breaches could be viewed as privacy<br />
breaches, which, under current HIPAA and other privacy<br />
regulations, can carry dire consequences. Also, patients<br />
could give each other incorrect medical advice; again, because<br />
the community is a service provided by the practice,<br />
there is potential for liability for the practice. Some institutions<br />
have dedicated health care providers to scan or moderate<br />
chat functions, which is clearly expensive, and it is<br />
unclear that practice-hosted communities <strong>of</strong>fer benefit over<br />
similar functions already on the Internet outside <strong>of</strong> a practice.<br />
State laws can vary around liability and should be<br />
carefully watched.<br />
Protocols and Studies<br />
FELDMAN AND RODRIGUEZ<br />
PHRs also provide a powerful tool for monitoring patients<br />
who are already on protocols and for recruiting potential
PERSONAL HEALTH RECORDS<br />
study participants. One key problem in smaller practices is<br />
that both patients and clinicians may be unaware <strong>of</strong> ongoing<br />
studies in a particular disease; technology can effectively<br />
match up clinicians, patients, and protocols. The PHR is also<br />
an efficient way to securely gather information from patients<br />
on protocols, such as symptoms, quality-<strong>of</strong> life scores, and<br />
other research information. These uses have to be implemented<br />
in an ethical way that meets health services research<br />
guidelines but can be a powerful tool when done well.<br />
Evaluating the Intervention<br />
Most PHRs include some logging function to record what a<br />
user did and saw while using the system. These data are rich<br />
for mining to determine what patients are doing with the<br />
system, what functions are used most frequently, and what<br />
aspects are never used. Regular reporting and analysis <strong>of</strong><br />
usage patterns are crucial, as is periodic re-engagement<br />
with stakeholders to address shifting needs and technologies.<br />
For example, in 2006, few designers thought about the<br />
need to access patient portals from cell phones other than<br />
Authors’ Disclosures <strong>of</strong> Potential Conflicts <strong>of</strong> Interest<br />
Author<br />
Henry Feldman*<br />
Elizabeth S. Rodriguez*<br />
*No relevant relationships to disclose.<br />
Employment or<br />
Leadership<br />
Positions<br />
Consultant or<br />
Advisory Role<br />
1. Sands DZ, Halamka JD, Pellaton D. PatientSite: a Web-based clinical<br />
communication and health education tool. HIMSS Proc. 2001;3:1-6.<br />
2. Tarn DM, Flocke A. New prescriptions: how well do patients remember<br />
important information? Fam Med. 2011;43:254-259.<br />
3. Nightingale SL. Do physicians tell patients enough about prescription<br />
drugs? Do patients think so? Postgrad Med. 1983;74:169-175.<br />
4. Siteman E, Businger A, Gandhi T, et al. Patient review <strong>of</strong> selected<br />
electronic health record data improves visit experience. AMIA Annu Symp<br />
Proc. 2006:1101.<br />
5. Siteman E, Businger A, Gandhi T, et al. Clinicians recognize value <strong>of</strong><br />
patient review <strong>of</strong> their electronic health record data, AMIA Annu Symp Proc.<br />
2006:1101.<br />
6. Reti SR, Feldman HJ, Ross SE, et al. Governance for personal health<br />
records. J Am Med Inform Assoc. 2009;16:14-17.<br />
7. Halamka J, Aranow M, Ascenzo C, et al. Health care IT collaboration in<br />
Massachusetts: the experience <strong>of</strong> creating regional connectivity. JAmMed<br />
Inform Assoc. 2005;12:596-601.<br />
through text messaging, but after the introduction <strong>of</strong> the<br />
iPhone and other smartphones, this access became an important<br />
area for PHR designers to consider, as patients have<br />
became much more sophisticated in their demands for mobile<br />
interaction.<br />
Conclusion<br />
PHRs are coming to practices around the United States,<br />
and oncology practices large and small will have to adopt<br />
these technologies. Practices will confront many <strong>of</strong> the<br />
issues described in this chapter, and will need careful<br />
governance, stakeholder engagement, and constant reassessment.<br />
PHRs serve many masters across many domains<br />
and needs. Patients are already demanding access to their<br />
records, and government regulations are requiring these in<br />
upcoming years. Thus, instead <strong>of</strong> thinking <strong>of</strong> PHRs as a<br />
burden, oncologists should view them as an opportunity to<br />
promote their practice, engage and educate their patients,<br />
and, possibly, in the ACO era, lower the utilization in their<br />
practice.<br />
Stock<br />
Ownership Honoraria<br />
REFERENCES<br />
Research<br />
Funding<br />
Expert<br />
Testimony<br />
Other<br />
Remuneration<br />
8. Earnest MA, Ross SE, Wittevrongel L, et al Use <strong>of</strong> a patient-accessible<br />
electronic medical record in a practice for congestive heart failure: patient and<br />
physician experiences. J Am Med Inform Assoc. 2004;11:410-417.<br />
9. Rodriguez ES. Using a patient portal for electronic communication with<br />
oncology patients: implications for nurses. Oncol Nurs Forum. 2010;37:667-<br />
671.<br />
10. Walker J, et al. Inviting patients to read their doctors’ notes: patients<br />
and doctors look ahead - patient and physician surveys. Ann Intern Med.<br />
2011;155:811-819, 2011.<br />
11. Delbanco T, Walker J, Darer JD, et al. Open notes: doctors and patients<br />
signing on. Ann Intern Med. 2010;153:121-125.<br />
12. Wald JS, Burk K, Gardner K, et al. Sharing electronic laboratory<br />
results in a patient portal-a feasibility pilot. Stud Health Technol Inform.<br />
2007;129(Pt 1):18-22.<br />
13. Rodriguez ES, Thom B, Schneider SM. Nurse and physician perspectives<br />
<strong>of</strong> cancer patients having online access to their lab results. Oncol Nurs<br />
Forum. 2011;38:476-482.<br />
e69
THE ASCO QUALITY ONCOLOGY PRACTICE<br />
INITIATIVE AND BEYOND<br />
CHAIR<br />
Joseph O. Jacobson, MD<br />
North Shore Medical Center<br />
Salem, MA<br />
SPEAKERS<br />
Michael N. Neuss, MD<br />
Vanderbilt-Ingram Cancer Center<br />
Nashville, TN<br />
Robert Hauser, PhD, PharmD<br />
<strong>American</strong> <strong>Society</strong> <strong>of</strong> <strong>Clinical</strong> <strong>Oncology</strong><br />
Alexandria, VA
Measuring and Improving Value <strong>of</strong> Care in<br />
<strong>Oncology</strong> Practices: ASCO Programs from<br />
Quality <strong>Oncology</strong> Practice Initiative to the<br />
Rapid Learning System<br />
By Joseph O. Jacobson, MD, MSc, Michael N. Neuss, MD, and<br />
Robert Hauser, PharmD, PhD<br />
Overview: Rising cancer care costs are no longer sustainable.<br />
Medical oncologists must focus on providing the<br />
maximum value to their patients; improving short-term, intermediate<br />
and long-term outcomes; and managing overall costs.<br />
Accurate measurement <strong>of</strong> outcomes and overall cost is essential<br />
to informing providers and institutions and in the quest for<br />
continuous improvement in value. The ASCO Quality <strong>Oncology</strong><br />
Practice Initiative (QOPI) is an excellent tool for sampling<br />
WHEN PUBLISHED in 2001, Crossing the Quality<br />
Chasm: A New Health System for the Twenty-first<br />
Century defined quality <strong>of</strong> care as “the degree to which<br />
health services for individuals and populations increase the<br />
likelihood <strong>of</strong> desired health outcomes and are consistent<br />
with current pr<strong>of</strong>essional knowledge.” 1 Six core components<br />
<strong>of</strong> quality were identified: safety, effectiveness, patientcenteredness,<br />
timeliness, efficiency, and equitability. The<br />
publication received widespread attention and served as a<br />
clarion call to action to providers and health care organizations<br />
to begin a relentless focus on quality <strong>of</strong> care.<br />
The Need to Focus on the Value <strong>of</strong> Cancer Care<br />
At the time <strong>of</strong> publication <strong>of</strong> Crossing the Quality Chasm,<br />
health care costs as a percentage <strong>of</strong> gross domestic product<br />
had exceeded 14%. In 2010, the percentage had increased to<br />
17.9%, and total health care expenditures had reached $2.6<br />
trillion, according to the Center for Medicare and Medicaid<br />
Services. Further rises in health care costs are now recognized<br />
as unsustainable. The Patient Protection and Affordable<br />
Care Act was signed into law by President Obama in<br />
2010 at least partly in response to the recognition that<br />
spiraling health care costs threatened the economic health <strong>of</strong><br />
the nation.<br />
In the United States, the cost <strong>of</strong> cancer care is increasing<br />
at a faster rate than nonmalignant conditions. If annual<br />
direct cancer care costs cannot be contained, they are projected<br />
to reach $173 billion by 2020, representing a 39%<br />
increase compared with 2010. 2 Increases are caused by rises<br />
in both the cost <strong>of</strong> therapy and the extent <strong>of</strong> care. 3 Rises in<br />
cost <strong>of</strong> therapy are partly justified by dramatic advances in<br />
the management <strong>of</strong> cancer that have occurred in the past<br />
decade. For example, a new generation <strong>of</strong> targeted chemotherapeutic<br />
agents has emerged with unparalleled activity<br />
and with reduced toxicity compared with standard chemotherapy.<br />
4 However, the benefit <strong>of</strong> other new and expensive<br />
technologies such as robotic surgery for early-stage prostate<br />
cancer is still largely unproven. 5<br />
Porter and Teisberg argue that quality alone is insufficient<br />
to justify the incorporation <strong>of</strong> a new technology or<br />
agent into routine use. 6 They argue persuasively that value<br />
is a far better means to assess the effect <strong>of</strong> a change in care,<br />
defining value simply as “outcomes achieved per dollar<br />
spent.” “Dollars spent” is intended to include the cost <strong>of</strong> care<br />
e70<br />
processes <strong>of</strong> care in medical oncology practices. To achieve<br />
the larger goal <strong>of</strong> improving the value <strong>of</strong> cancer care, ASCO is<br />
investing in the development <strong>of</strong> a Rapid Learning System,<br />
which will leverage emerging information technologies to<br />
more accurately measure outcomes (including those reported<br />
by the patient) and costs, resulting in highly efficient, effective,<br />
and safe cancer care.<br />
over a full set <strong>of</strong> interventions needed to manage a specific<br />
medical condition. 7 In this paradigm, Porter describes three<br />
tiers <strong>of</strong> outcome. Tier 1 includes measures <strong>of</strong> success familiar<br />
to oncologists, including survival and response to treatment.<br />
Tier 2 assesses the process <strong>of</strong> recovery and focuses on<br />
disutility <strong>of</strong> care including treatment delays, toxicities, adverse<br />
events, and errors. The long-term outcome and late<br />
treatment effects constitute Tier 3 outcomes.<br />
Defining the value <strong>of</strong> the care that we provide to our<br />
patients is vital for medical oncologists as we prepare for<br />
new reimbursement models. As Porter notes, “value—neither<br />
an abstract ideal nor a code word for cost reduction—<br />
should define the framework for performance improvement<br />
in health care.” 7 For medical oncologists, this requires new<br />
attention to delivery <strong>of</strong> cancer care with a focus on the full<br />
spectrum <strong>of</strong> services and on providing care in which value<br />
can be quantified. Various models for providing value-based<br />
cancer care have been described. 8-10<br />
Providing High-Value Care Requires an Organized<br />
Health Care Delivery System<br />
Bohmer has observed that high-value health care organizations<br />
have four common habits: (1) specification and<br />
planning; (2) intentional infrastructure design; (3) measurement<br />
and oversight and; (4) self-study. 11 He observes that<br />
many health care organizations succeed at accomplishing<br />
some <strong>of</strong> these goals, but only a handful manage them all; it<br />
is these few that have succeeded in delivering high-value<br />
cancer care. Examples include Intermountain Healthcare<br />
and Mayo Clinic.<br />
Specification and Planning<br />
Bohmer defines specification as separating heterogeneous<br />
patient populations into clinically meaningful subsets. As he<br />
notes, “Many hospitals and clinicians do not plan care<br />
From the Dana-Farber Cancer Institute, Boston, MA; Vanderbilt-Ingram Cancer Center,<br />
Nashville, TN; Department <strong>of</strong> Quality and Guidelines, <strong>American</strong> <strong>Society</strong> <strong>of</strong> <strong>Clinical</strong><br />
<strong>Oncology</strong>, Alexandria, VA.<br />
Authors’ disclosures <strong>of</strong> potential conflicts <strong>of</strong> interest are found at the end <strong>of</strong> this article.<br />
Address reprint requests Joseph Jacobson, MD, MSc, 450 Brookline Ave., Dana-Farber<br />
Cancer Institute, Boston, MA; email: joseph_jacobson@dcfi.harvard.edu.<br />
© <strong>2012</strong> by <strong>American</strong> <strong>Society</strong> <strong>of</strong> <strong>Clinical</strong> <strong>Oncology</strong>.<br />
1092-9118/10/1-10
IMPROVING VALUE OF CARE IN ONCOLOGY<br />
processes in advance in such detail; instead, they treat each<br />
new patient or problem as a random draw from a heterogeneous<br />
population and must, therefore, reinvent the strategy<br />
for solving it.” For much <strong>of</strong> its history, medical oncology has<br />
approached care this way. Fortunately, advances in genomics<br />
have begun to identify clinically meaningful disease<br />
subsets. Patients with adenocarcinoma <strong>of</strong> the lung, for<br />
example, are now routinely tested for the presence <strong>of</strong> activating<br />
mutations <strong>of</strong> epidermal growth factor or translocations<br />
<strong>of</strong> anaplastic lymphoma kinase (ALK). 12,13 Patients<br />
with lung cancer with these genotypes now have access to<br />
targeted therapies that have transformed the natural history<br />
<strong>of</strong> their disease. By “planning,” Bohmer argues for the<br />
development <strong>of</strong> pre-emptive treatment approaches for patients<br />
using clinical practice guidelines or clinical pathways.<br />
Intentional Infrastructure Design<br />
The gradual shift <strong>of</strong> medicine from “guild” to pr<strong>of</strong>ession is<br />
embodied by this habit. 14 This entails designing care to<br />
maximize the use <strong>of</strong> each team member’s skills and expertise<br />
and requires engagement <strong>of</strong> the patient and family. The<br />
medical home model now being implemented widely in<br />
primary care is an example <strong>of</strong> effective infrastructure design,<br />
and a pilot in one medical oncology practice suggests<br />
that the model can be extended. 9 Multidisciplinary cancer<br />
care—when well designed—is a powerful tool to leverage<br />
personnel and limited resources. 15<br />
Measurement and Oversight<br />
High-value organizations go far beyond measurement<br />
required for external reporting to the development <strong>of</strong> a set <strong>of</strong><br />
internal measurements that become integral to accountability<br />
and organizational management.<br />
Self-Study<br />
Bohmer notes that “high-value organizations treat clinical<br />
knowledge as an organizational as well as individual property...these<br />
organizations deliberately nurture a culture<br />
that supports learning. ...” Inthis environment, in which<br />
KEY POINTS<br />
● Medical oncologists must strive for high value <strong>of</strong><br />
cancer care—defined as outcomes achieved per dollars<br />
spent.<br />
● Reliable, accurate means to assess the quality and<br />
cost across the spectrum <strong>of</strong> cancer services must be<br />
created to drive high-value care.<br />
● The ASCO Quality <strong>Oncology</strong> Practice Initiative<br />
(QOPI) is available to all U.S. medical oncologists to<br />
sample care processes and to compare their performance<br />
over time and against national benchmarks.<br />
● The ASCO QOPI Certification Program (QCP) allows<br />
practices to demonstrate high quality <strong>of</strong> care and safe<br />
chemotherapy administration practices.<br />
● ASCO is developing a Rapid Learning System that<br />
leverages new information technologies with the goal<br />
<strong>of</strong> improving the value <strong>of</strong> cancer care by focusing on<br />
effectiveness, safety, efficiency, and quality.<br />
specification and planning are built into systems, variations<br />
from guidelines and pathways are studied to understand<br />
and learn from variation. This is the clinical equivalent <strong>of</strong><br />
the managerial approach <strong>of</strong> “managing by exception.”<br />
Managing What We Measure<br />
At the 1962 Yale University Commencement Address,<br />
John F. Kennedy noted that “for the great enemy <strong>of</strong> truth is<br />
very <strong>of</strong>ten not the lie—deliberate, contrived, and dishonest—but<br />
the myth—persistent, persuasive, and unrealistic.”<br />
This admonition is as true in medicine as it is in politics.<br />
Without data to inform our delivery <strong>of</strong> care, we are dependent<br />
on subjectivity and biases and at risk <strong>of</strong> squandering<br />
precious resources and mistreating our patients. 16 The collection<br />
<strong>of</strong> accurate, granular, and timely data is vital for<br />
organizational and provider growth and in the quest for<br />
excellence. Berwick, James, and Coye describe two overlapping<br />
types <strong>of</strong> performance assessment, “measurement for<br />
selection” and “measurement for improvement.” 17<br />
Measurement for selection. This approach to improving<br />
the quality <strong>of</strong> care is based on the premise that health care<br />
outcomes studied in any setting (e.g., individual practitioner<br />
or health care organization) will have a distribution <strong>of</strong><br />
performance levels. In principle, this consumer-based approach<br />
to value simply requires the selection <strong>of</strong> the desired<br />
outcome, review <strong>of</strong> the provider or organizational performance,<br />
and selection based on identifying the best<br />
performer—“get better care by choosing better care.” 17 A<br />
major shortcoming <strong>of</strong> this approach is that it does not<br />
improve the overall quality <strong>of</strong> care; care is simply shifted<br />
toward high performing practitioners or organizations. 18 In<br />
addition, measurement for selection remains an inexact<br />
science in medicine. Outcome variation across providers and<br />
organizations is <strong>of</strong>ten because <strong>of</strong> the complex interactions <strong>of</strong><br />
patient-specific factors (which are <strong>of</strong>ten imperfectly measured)<br />
with the health care system. In addition, measurement<br />
for selection <strong>of</strong>ten relies on a series <strong>of</strong> summary<br />
statistics lacking sufficient specificity or granularity to draw<br />
precise conclusions for most medical conditions or procedures.<br />
Measurement for improvement. Using this approach, focus<br />
is shifted toward the processes <strong>of</strong> care rather than the<br />
providers, and, when successful, the entire distribution <strong>of</strong><br />
performance levels is shifted to a higher level. Institutions<br />
such as Intermountain Healthcare (IHC) that have experienced<br />
major strides in improving quality over the past<br />
decade have consciously shifted from measuring for selection<br />
to measuring for improvement. 18 IHC success has<br />
depended on building a comprehensive clinical information<br />
system, engaging clinicians to develop standardized approaches<br />
(care process models [CPMs]) to the management<br />
<strong>of</strong> common conditions, encouraging clinicians to vary from<br />
these approaches when justified by circumstances, and continuously<br />
modifying the CPMs based on analysis <strong>of</strong> variations<br />
and on emerging literature.<br />
The Emergence <strong>of</strong> the Quality <strong>Oncology</strong> Practice<br />
Initiative (QOPI)<br />
In response to the Institute <strong>of</strong> Medicine’s National Cancer<br />
Policy Board recommendations published in 1999, 19 ASCO<br />
commissioned researchers at the Harvard School <strong>of</strong> Public<br />
Health and at the RAND Corporation to undertake a study<br />
e71
<strong>of</strong> oncology care in the United States. The National Initiative<br />
for Cancer Care Quality (NICCQ) surveyed breast and<br />
colorectal cancer care in five U.S. metropolitan areas. 20 This<br />
cross-sectional analysis <strong>of</strong> processes <strong>of</strong> care identified widely<br />
varying rates <strong>of</strong> adherence to recommended care and identified<br />
significant opportunities for improvement. 21<br />
In 2002, Joseph Simone, MD, co-chair <strong>of</strong> the National<br />
Cancer Policy Board, recruited seven volunteer medical<br />
oncologists from community practices to consider alternative<br />
methods for measuring cancer quality in ambulatory<br />
practices. This “alpha group” <strong>of</strong> clinicians created a measurement<br />
system, now known as the Quality <strong>Oncology</strong><br />
Practice Initiative (QOPI), based on the goal <strong>of</strong> “promoting<br />
excellence in cancer care by helping oncology practices<br />
create a culture <strong>of</strong> self-examination and improvement.” 22<br />
Through a series <strong>of</strong> pilot projects, a collection <strong>of</strong> consensusbased<br />
and evidence-based performance indicators was selected,<br />
a chart abstraction methodology was created, and a<br />
system for rapid feedback <strong>of</strong> results to practices was created.<br />
In 2006, based on the success <strong>of</strong> a series <strong>of</strong> pilot projects,<br />
ASCO <strong>of</strong>fered all U.S. members the opportunity to participate<br />
in QOPI. QOPI has grown rapidly since the initial<br />
national rollout. Nearly 100 performance indicators have<br />
been created, spanning multiple domains and diseases.<br />
More than 700 practices have enrolled in QOPI, and semiannual<br />
data collections now routinely include 250 to 300<br />
practices (Fig. 1). A rigorous QOPI Certification Program<br />
was introduced in 2010. 23 At the end <strong>of</strong> 2011, 105 practices<br />
had achieved certification.<br />
ASCO remains at the forefront <strong>of</strong> physician specialty<br />
societies in its quest to continuously improve the value <strong>of</strong><br />
care provided to patients with cancer. This manuscript<br />
describes the current state <strong>of</strong> the QOPI program and preliminary<br />
planning for an entirely new approach to measuring<br />
and improving cancer care based on principles recently<br />
developed by the Institute <strong>of</strong> Medicine for rapid learning. 24<br />
How QOPI Works—A Consideration <strong>of</strong> Effort,<br />
Expense, and Rewards<br />
The <strong>American</strong> <strong>Society</strong> <strong>of</strong> <strong>Clinical</strong> <strong>Oncology</strong> (ASCO) Quality<br />
<strong>Oncology</strong> Practice Initiative (QOPI) has two major com-<br />
e72<br />
JACOBSON, NEUSS, AND HAUSER<br />
Fig. 1. Growth <strong>of</strong> the Quality <strong>Oncology</strong> Practice Initiative<br />
(QOPI) Since 2006.<br />
ponents. The self-reported QOPI practice survey provides a<br />
structured method and benchmark comparison designed to<br />
help practices understand their areas <strong>of</strong> strength and weakness<br />
in patient care, both in absolute and comparative<br />
terms. By giving practices a way to measure their performance<br />
on a variety <strong>of</strong> process measures designed to look at<br />
specific domains <strong>of</strong> care, practitioners can identify areas <strong>of</strong><br />
potential opportunity. By providing comparative data, they<br />
allow practices a way to understand if their weaknesses are<br />
similar to or different from comparable groups. 25,26 Through<br />
the QOPI certification process, they can obtain external<br />
verification <strong>of</strong> their performance as well as compliance with<br />
policies and procedures consistent with the ASCO/ <strong>Oncology</strong><br />
Nursing <strong>Society</strong> (ONS) chemotherapy safety standards. 23<br />
QOPI participation is open to all U.S. ASCO members<br />
twice yearly. An estimated 10% to 20% <strong>of</strong> members have<br />
submitted records. Most <strong>of</strong> the practices that have participated<br />
have participated repeatedly. The chart abstraction<br />
and data submission process seems daunting at first, but<br />
becomes easier over time. Identifying charts for analysis is<br />
easier with electronic billing or charting systems, which<br />
allow sequential lists <strong>of</strong> patients with specific diagnoses to<br />
be arranged and facilitate chart review without physically<br />
finding and handling a paper chart, though both systems are<br />
equally acceptable. Interestingly, some practices have found<br />
that review in paper records is faster than electronic formats.<br />
Our experience, having reviewed charts in both, is<br />
that the electronic format allows much quicker abstraction<br />
as reviewers become more familiar with the data elements,<br />
but it still takes 20–40 minutes per chart, with the time<br />
dependent on the number and complexity <strong>of</strong> modules. At the<br />
Vanderbilt-Ingram Cancer Center, physicians and nurses<br />
perform the chart reviews. In private practices, abstraction<br />
is <strong>of</strong>ten done by both nurses and/or research data technicians.<br />
Assuming 40–100 charts are reviewed at an average<br />
employee expense <strong>of</strong> from $40 to $200 (the latter for physicians)<br />
per hour, the cost for a round <strong>of</strong> data submission<br />
ranges from $1,600 to $20,000. Based on reports <strong>of</strong> participants,<br />
it seems reasonable to estimate $4,000 per practice<br />
per round; the expense <strong>of</strong> the effort is dependent on the pay<br />
scale <strong>of</strong> those performing abstraction.
IMPROVING VALUE OF CARE IN ONCOLOGY<br />
Table 1. QOPI Participation and QOPI Certification Compared<br />
Program QOPI Participation QOPI Certification<br />
Cost to ASCO Members Free $3000–15,000 (Depending<br />
on practice size and<br />
number <strong>of</strong> locations)<br />
Measures 89 questions in 7<br />
modules [1]<br />
Passing score on 5 modules<br />
(Currently, score to pass is<br />
72% on general measures,<br />
80% on appropriate<br />
adjuvant treatment)<br />
Standards Not applicable Documented policies<br />
consistent with 17 <strong>of</strong><br />
ASCO/ONS chemotherapy<br />
administration [i]standards<br />
External Audit No Yes<br />
Duration <strong>of</strong> Certification Not applicable 3 years<br />
Comments Comparative data are<br />
available to help a<br />
practice measure its<br />
achievement<br />
Pricing is variable for larger<br />
groups<br />
Abbreviations: QOPI, Quality <strong>Oncology</strong> Practice Initiative; ONS, <strong>Oncology</strong><br />
Nursing <strong>Society</strong>.<br />
ASCO requests that each QOPI practice identify a practice<br />
steward who will take responsibility to educate practitioners<br />
about their achievement in QOPI. Our experience in<br />
distributing these results in both community and academic<br />
settings has been that physicians accept the good results<br />
with a tremendous amount <strong>of</strong> pride, and simultaneously<br />
doubt the reliability or importance <strong>of</strong> measures where their<br />
performance lags. Nevertheless, they almost always improve<br />
their performance in these areas, although it is very<br />
difficult to determine whether true performance or documentation<br />
<strong>of</strong> performance is really changing. From personal<br />
experience, one <strong>of</strong> the authors (MNN) notes that he has<br />
learned much and changed how he interacts with patients.<br />
For example, he is much more attentive to infertility counseling<br />
and interventions before starting chemotherapy and<br />
trying to prevent constipation in patients on narcotic analgesics<br />
than previously.<br />
QOPI certification <strong>of</strong> a large academic practice such as<br />
Vanderbilt-Ingram Cancer Center poses challenges. Just<br />
like preparing for any serious certification exercise, such as<br />
the Commission on Cancer, Joint Commission, or Magnet<br />
Nursing Certification, it is a significant team effort that<br />
requires commitment <strong>of</strong> time and resources. Achieving certification<br />
requires that a practice surpass a threshold<br />
achievement level during a round <strong>of</strong> data submission and<br />
demonstrate compliance with a subset <strong>of</strong> the chemotherapy<br />
administration safety standards developed by ASCO and<br />
ONS. 27 The number <strong>of</strong> charts needed for review is dependent<br />
on the practice size (Table 1).<br />
The benefits <strong>of</strong> participation and certification vary by<br />
institution. The intrinsic reward—assessing practice or<br />
practitioner performance—is great. Other benefits are possible.<br />
Listing <strong>of</strong> certification has likely attracted some patients.<br />
More tangible advantages have been realized by some<br />
practices. For example, some practices have been provided<br />
payment bonuses from payers. Some exclusive quality programs<br />
recognize participation or certification as part <strong>of</strong><br />
inclusion standards. Finally, some practices have negotiated<br />
rate premiums from payers based on the achievement <strong>of</strong><br />
certification (though antitrust concerns preclude any specific<br />
detailing <strong>of</strong> these arrangements).<br />
In summary, QOPI participation and certification require<br />
significant effort associated with nontrivial expenses for<br />
practices. There are intrinsic rewards and the opportunity<br />
for self-examination and improvement with even a partial<br />
data submission, although full and broad participation and<br />
benchmark comparisons <strong>of</strong>fer more.<br />
The ASCO Rapid Learning System (RLS)<br />
The field <strong>of</strong> oncology is entering an unprecedented time<br />
where advances in technology, drug development, treatment<br />
techniques, genomics, and research are driving health care<br />
costs to unsustainable levels. 28 As a result, there is an<br />
increasing need for higher quality care, at lower costs, with<br />
better outcomes. To meet these needs and demonstrate<br />
success, each component (quality, cost, and outcomes) must<br />
be accurately defined, measured, and reported. Fortunately,<br />
we have defined quality measures in oncology (QOPI), costs<br />
can be captured via claims and other means, and outcomes<br />
can be defined and reported though electronic medical record<br />
(EMR) systems. Unfortunately, all <strong>of</strong> these data reside<br />
in many different electronic (sometimes paper) systems that<br />
do not interconnect. If these important data continue to<br />
reside in many disparate systems, reporting high-value<br />
cancer care will be burdensome or impossible. However,<br />
there is a solution on the horizon: a learning health care<br />
system. As defined by the Institute <strong>of</strong> Medicine, a learning<br />
health care system is grounded in the principles <strong>of</strong> improving<br />
effectiveness, safety, efficiency, and quality, paired with<br />
evaluating the processes shared by engineering and medicine.<br />
29 ASCO is taking the lead to develop a rapid learning<br />
health system in oncology to accomplish the goal <strong>of</strong> improving<br />
effectiveness, safety, efficiency, and quality in oncology.<br />
The ASCO RLS will have the ability to utilize evidencebased<br />
medicine in the form <strong>of</strong> publications, guidelines, and<br />
measures, pair it with patient-level data from multiple<br />
Health Information Technology systems, aggregate it, analyze<br />
it, and turn the data into usable knowledge.<br />
QOPI provides a uniquely strong foundation, but ASCO<br />
recognizes that members and the oncology community will<br />
benefit from an initiative with even greater sophistication.<br />
In 2011, the ASCO Board directed that immediate work<br />
begin on construction <strong>of</strong> a cutting-edge rapid learning system<br />
for oncology.<br />
What Is a Rapid Learning System?<br />
In the most basic form, an RLS is a technology platform<br />
that allows for the collection <strong>of</strong> data from different health<br />
information technology systems to be collected, aggregated,<br />
analyzed, and turned into usable information to improve<br />
treatment and drive scientific learning and discovery from<br />
every patient at every encounter (Fig. 2). 30 From a culture<br />
standpoint, an RLS speeds and improves how we learn<br />
(Fig. 3).<br />
What Can We Do With an RLS?<br />
The output from an RLS is potentially unlimited. The<br />
community will have the ability to obtain real-time clinical<br />
decision support, identify patients for clinical trials, conduct<br />
population health research, conduct health outcomes research,<br />
conduct comparative effectiveness research, provide<br />
quality benchmarking, possibly use for REMS reporting,<br />
and identify rare adverse events by detecting early warning<br />
signals.<br />
e73
One example <strong>of</strong> the power <strong>of</strong> an oncology RLS can be<br />
illustrated with the example <strong>of</strong> erythropoietin stimulating<br />
agents (ESAs). Based on studies demonstrating reduced<br />
need for blood transfusion in chemotherapy patients, the<br />
first <strong>of</strong> these agents was approved for oncology use in 1993.<br />
ESAs subsequently became a commonly used component<br />
<strong>of</strong> cancer supportive care during active treatment. Beginning<br />
in 2004, emerging safety concerns led to product<br />
label warnings. In 2007, additional studies suggesting<br />
negative effects on survival and disease progression caused<br />
the United States Food and Drug Administration (FDA) and<br />
CMS to issue further product label warnings and to restrict<br />
coverage (Product Package Insert). However, if we had<br />
access to data from an RLS at the time, with real-time<br />
capture <strong>of</strong> millions <strong>of</strong> clinical data points and patient outcomes,<br />
it is reasonable to conclude that we could have<br />
identified these safety signals before 2004. If the use <strong>of</strong><br />
ESAs from 2004 -2007 had been at levels seen between<br />
2008 and 2011, the quality <strong>of</strong> health care provided to our<br />
patients would have been improved, and CMS could have<br />
saved more than $3 billion (assuming a conservative estimate<br />
<strong>of</strong> 6,000 practicing oncologists prescribing ESAs during<br />
these years).<br />
e74<br />
Fig. 2. Proposed Model for an <strong>Oncology</strong> Rapid Learning System (RLS).<br />
Why Did ASCO Undertake an RLS?<br />
JACOBSON, NEUSS, AND HAUSER<br />
ASCO has been committed to quality <strong>of</strong> cancer care since<br />
the original founders met in 1964 to share ideas and research<br />
on how they were improving patient care and quality<br />
outcomes in the field. ASCO continues to meet its<br />
quality driven mission with the QOPI program, and an<br />
RLS is just the next iteration in how we learn, share<br />
knowledge, and improve quality outcomes. ASCO has a<br />
patient-driven mission, which also focuses on education and<br />
quality. ASCO has built trust with its membership and will<br />
always keep members’ best interest and outcomes front and<br />
center.<br />
How Is ASCO Proceeding? What Will It Take to Build<br />
an RLS?<br />
ASCO has developed a plan, led by the Board <strong>of</strong> Directors<br />
and an RLS Advisory Group, in which a solid foundation is<br />
built on trust and the following seven guiding principles:<br />
1. Rigorous–the hallmarks <strong>of</strong> RLS will be methodological<br />
rigor and comprehensiveness<br />
2. Patient-focused–RLS will reduce morbidity and mortality<br />
and incorporate the humane values <strong>of</strong> the pr<strong>of</strong>ession<br />
Fig. 3. Comparison between the Current Model <strong>of</strong><br />
Care and a Proposed Model Based on a Fully Functional<br />
Rapid Learning System (RLS).
IMPROVING VALUE OF CARE IN ONCOLOGY<br />
3. Transparent–the methodology <strong>of</strong> RLS and the incentives<br />
<strong>of</strong> every collaborator and contributor will be<br />
visible<br />
4. Independent–the credibility <strong>of</strong> RLS depends on oncology<br />
peer decision-making and collaboration without<br />
compromise <strong>of</strong> principles. There will be a firewall<br />
between industry and guideline and standards development<br />
5. Inclusive–ASCO is open to collaborating with any organization,<br />
public or private, that accepts its collaboration<br />
principles. RLS will be accessible and usable<br />
from non-ASCO platforms<br />
6. Streamlined–RLS data collection will be efficient and<br />
not burdensome to providers<br />
7. Sustainable–ASCO will make a substantial investment<br />
in RLS and needs a sustainable economic model. Any<br />
commercialization <strong>of</strong> RLS will be supported by ASCO’s<br />
educational and quality endeavors<br />
The plan for building the RLS is to begin with small pilot<br />
projects to build the foundation <strong>of</strong> the system. The foundation<br />
<strong>of</strong> the RLS is comprised <strong>of</strong> three key components:<br />
1. Data Governance<br />
2. Data Standards<br />
3. Flexible Technology Platform.<br />
ASCO will convene experts in the area <strong>of</strong> data governance<br />
to develop a defined set <strong>of</strong> policies and procedures on how<br />
data are collected, secured, accessed, aggregated, and reported.<br />
The <strong>Society</strong> will engage experts from government<br />
agencies, law firms, health systems, academic centers, community<br />
practices, and other industries to ensure the RLS is<br />
built with the strongest and most transparent data governance<br />
polices possible. These policies will be maintained by<br />
a board <strong>of</strong> experts who will be responsible for stewarding<br />
and updating the documents.<br />
Authors’ Disclosures <strong>of</strong> Potential Conflicts <strong>of</strong> Interest<br />
Author<br />
Employment or<br />
Leadership<br />
Positions<br />
Consultant or<br />
Advisory Role<br />
Joseph O. Jacobson*<br />
Michael N. Neuss EMeRged Systems<br />
(U), (L)<br />
Robert Hauser Amgen; Novartis<br />
*No relevant relationships to disclose.<br />
1. Institute <strong>of</strong> Medicine: Crossing the Quality Chasm: A New Health<br />
System for the Twenty-first Century. Washington DC: National Academies<br />
Press, 2001.<br />
2. Mariotto AB, Robin Yabr<strong>of</strong>f K, Shao Y, et al. Projections <strong>of</strong> the Cost <strong>of</strong><br />
Cancer Care in the United States: 2010. J Natl Cancer Inst. 2011;103:117-<br />
128.<br />
3. Smith TJ, Hillner BE. Bending the Cost Curve in Cancer Care. N Engl<br />
J Med. 2011;364:2060-2065.<br />
4. Stegmeier F, Warmuth M, Sellers WR, et al. Targeted Cancer Therapies<br />
in the Twenty-First Century: Lessons From Imatinib. Clin Pharmacol Ther.<br />
2010;87:543-552.<br />
5. Barbash GI, Glied SA. New technology and health care costs-the case <strong>of</strong><br />
robot-assisted surgery. N Engl J Med. 2010;363:701-704.<br />
6. Porter ME, Teisberg EO. Redefining health care: Creating value-based<br />
competition on results. Boston, MA: Harvard Business School Press, 2006.<br />
7. Porter ME. What is value in health care? N Engl J Med. 2010;363:2477-<br />
2481.<br />
8. Pollock RE. Value-based health care: The MD Anderson experience. Ann<br />
Surg. 2008;248:510-518.<br />
A second key project in the process is to define data<br />
standards. ASCO will work with and convene panels <strong>of</strong><br />
experts to begin to build on work in medical data standards<br />
and enhance the oncology data standards already developed.<br />
For data to be easily aggregated, common definitions are<br />
needed in which all systems report. Developing this data<br />
“ontology” in oncology, will be paramount to the success <strong>of</strong><br />
the system.<br />
ASCO will soon be selecting a technology platform on<br />
which to build the RLS. Choosing the correct technology<br />
platform is critical as ASCO develops the RLS. The platform<br />
will need to have the ability to work within the constraints<br />
<strong>of</strong> data and technology today, but be flexible enough to<br />
adjust to the health information technology systems and<br />
s<strong>of</strong>tware <strong>of</strong> the future. Again, ASCO will engage experts in<br />
health information technology, oncology, and developers to<br />
help guide its decisions in this arena. Once the foundation <strong>of</strong><br />
the RLS is built, then portals and applications will be built.<br />
ASCO currently envisions having physician portals, clinical<br />
decision support portals, patient portals, and numerous<br />
applications to enhance the quality <strong>of</strong> oncology care.<br />
Conclusion<br />
ASCO has developed and earned the trust <strong>of</strong> its members<br />
since 1964 and has a dedicated mission to improve quality<br />
and a history <strong>of</strong> successful projects and programs. QOPI and<br />
the QOPI Certification Program allow ASCO participants to<br />
measure and improve practice performance, to compare<br />
themselves to their peers, and to guarantee that they adhere<br />
to safe chemotherapy administration practices. To continue<br />
to lead the field, and to respond to the urgent need to<br />
improve the value <strong>of</strong> cancer care, ASCO is investing in the<br />
development <strong>of</strong> a comprehensive information technologybased<br />
Rapid Learning System.<br />
Stock<br />
Ownership Honoraria<br />
REFERENCES<br />
Research<br />
Funding<br />
Expert<br />
Testimony<br />
Other<br />
Remuneration<br />
9. Sprandio J. <strong>Oncology</strong> patient-centered medical home and accountable<br />
cancer care. Community <strong>Oncology</strong>. 2010;7:565-572.<br />
10. Bach PB, Mirkin JN, Luke JJ. Episode-based payment for cancer care:<br />
A proposed pilot for Medicare. Health Aff (Millwood). 2011;30:500-509.<br />
11. Bohmer RM. The four habits <strong>of</strong> high-value health care organizations.<br />
N Engl J Med. 2011;365:2045-2047.<br />
12. Landi L, Cappuzzo F. Targeted therapies: Front-line therapy in lung<br />
cancer with mutations in EGFR. Nat Rev Clin Oncol. 2011;8:571-573.<br />
13. Shaw AT, Yeap BY, Solomon BJ, et al. Effect <strong>of</strong> crizotinib on overall<br />
survival in patients with advanced non-small-cell lung cancer harbouring<br />
ALK gene rearrangement: A retrospective analysis. Lancet Oncol. 2011;12:<br />
1004-1012.<br />
14. Sox HC. Medical Pr<strong>of</strong>essionalism and the Parable <strong>of</strong> the Craft Guilds.<br />
Ann Intern Med. 2007;147:809-810.<br />
15. Jacobson JO. Multidisciplinary Cancer Management: A Systems-Based<br />
Approach to Deliver Complex Care. J Oncol Pract. 2010;6:274-275.<br />
16. Auerbach AD, Landefeld CS, Shojania KG. The Tension between<br />
Needing to Improve Care and Knowing How to Do It. N Engl J Med.<br />
2007;357:608-613.<br />
e75
17. Berwick DM, James B, Coye MJ. Connections between quality measurement<br />
and improvement. Medical Care. 2003;41:130-138.<br />
18. James BC, Savitz LA. How Intermountain Trimmed Health Care Costs<br />
Through Robust Quality Improvement Efforts. Health Affairs. 2011;30:1185-<br />
1191.<br />
19. Hewitt ME, Simone JV. Ensuring Quality Cancer Care. Washington<br />
DC: National Academies Press, 1999.<br />
20. Schneider EC, Malin JL, Kahn KL, et al. Developing a system to assess<br />
the quality <strong>of</strong> cancer care: ASCO’s national initiative on cancer care quality.<br />
[erratum appears in J Clin Oncol 2004 Nov 15;22:4656]. J Clin Oncol.<br />
2004;22:2985-2991.<br />
21. Malin JL, Schneider EC, Epstein AM, et al. Results <strong>of</strong> the National<br />
Initiative for Cancer Care Quality: How can we improve the quality <strong>of</strong> cancer<br />
care in the United States? [erratum appears in J Clin Oncol 2006;24:1966].<br />
J Clin Oncol. 2006;24:626-634.<br />
22. Neuss MN, Desch CE, McNiff KK, et al. A process for measuring the<br />
quality <strong>of</strong> cancer care: The Quality <strong>Oncology</strong> Practice Initiative. J Clin Oncol.<br />
2005;23:6233-6239.<br />
23. McNiff K, Bonelli KR, Jacobson JO. Quality <strong>Oncology</strong> Practice Initiative<br />
Certification Program: Overview, Measure Scoring Methodology, and Site<br />
Assessment Standards. J Oncol Pract. 2009;5:270-276.<br />
e76<br />
JACOBSON, NEUSS, AND HAUSER<br />
24. Institute <strong>of</strong> Medicine: A Foundation for Evidence-Driven Practice: A<br />
Rapid Learning System for Cancer Care: Workshop Summary. Washington<br />
DC: The National Academies Press, 2010.<br />
25. Neuss MN, Desch CE, McNiff KK, et al. A Process for Measuring the<br />
Quality <strong>of</strong> Cancer Care: The Quality <strong>Oncology</strong> Practice Initiative. J Clin<br />
Oncol. 2005;23:6233-6239.<br />
26. Neuss M, Jacobson J, McNiff K, et al. Evolution and elements <strong>of</strong> the<br />
Quality <strong>Oncology</strong> Practice Initiative measure set. Cancer Control. 2009;16:<br />
312-317.<br />
27. Jacobson JO, Polovich M, McNiff KK, et al. <strong>American</strong> <strong>Society</strong> <strong>of</strong> <strong>Clinical</strong><br />
<strong>Oncology</strong>/<strong>Oncology</strong> Nursing <strong>Society</strong> Chemotherapy Administration Safety<br />
Standards. J Clin Oncol. 2009;27:5469-5475.<br />
28. Garber AM, Skinner J. Is <strong>American</strong> health care uniquely inefficient.<br />
Journal <strong>of</strong> Economic Perspectives. 2008;22:27-50.<br />
29. Grossmann C, Goolsby WA, Olsen L, et al. Engineering a Learning<br />
Healthcare System: A Look at the Future - Workshop Summary, The<br />
Learning Health System Series Roundtable on Value and Science-Driven<br />
Healthcare. Washington DC: National Academies Press, 2011.<br />
30. Abernethy AP, Etheredge LM, Ganz PA, et al. Rapid-learning system<br />
for cancer care. J Clin Oncol. 2010;28:4268-4274.
PHYSICIAN WELLNESS: COPING WITH REPETITIVE<br />
LOSS AND WORK-RELATED STRESS<br />
CHAIR<br />
Michael J. Fisch, MD, MPH<br />
University <strong>of</strong> Texas M. D. Anderson Cancer Center<br />
Houston, TX<br />
SPEAKERS<br />
Teresa Gilewski, MD<br />
Memorial Sloan-Kettering Cancer Center<br />
New York, NY<br />
Joshua Hauser, MD<br />
Northwestern University<br />
Chicago, IL
Buoyancy: A Model for Self-Reflection about<br />
Stress and Burnout in <strong>Oncology</strong> Providers<br />
Overview: Burnout is a prevalent syndrome among oncology<br />
providers marked by exhaustion, a sense <strong>of</strong> ineffectiveness,<br />
and depersonalization. This syndrome can have enormous<br />
influence on patient care as well as the provider’s career<br />
fulfillment and personal and family well-being. A buoyancy<br />
BURNOUT IS a familiar syndrome in health care.<br />
Maslach defines it as “a prolonged response to chronic<br />
emotional and interpersonal stressors on the job, and [it] is<br />
defined by the three dimensions <strong>of</strong> exhaustion, cynicism, and<br />
inefficacy.” 1 It is well described in medical oncologists, 2,3<br />
pediatric oncologists, 4,5 surgical oncologists, 6,7 primary care<br />
physicians, 8 nurses, 9 residents, 10-12 medical students, 13 and<br />
others, including physician leaders. 14,15 Provider burnout<br />
has been associated with varied factors, including higher<br />
workloads, less time with patients, lower experience and<br />
training, young marriages, having young children, lower<br />
reimbursements, higher levels <strong>of</strong> debt, work-family conflict,<br />
poor health, unrealistic expectations, and anger issues. The<br />
adverse consequences <strong>of</strong> burnout are not difficult to predict,<br />
and include despair, lower fulfillment, low productivity,<br />
increased turnover, loss <strong>of</strong> boundaries with patients, increased<br />
medical errors, and suboptimal care. Several excellent<br />
reviews about ways to understand and combat burnout<br />
in oncology have been published. 16,17<br />
The focus <strong>of</strong> this manuscript is to describe a practical<br />
paradigm for the opposite <strong>of</strong> burnout: buoyancy. Buoyancy is<br />
a force that keeps one afloat, in contrast to sinking. Posing<br />
this concept as the dependent variable in a multivariable<br />
model, 10 parameters that might be expected to absorb most<br />
<strong>of</strong> the variance in a model <strong>of</strong> buoyancy are depicted in Figure<br />
1. They are listed in no particular order, do not have equal<br />
importance within persons or among individuals, and are<br />
expected to vary over time. These theoretical parameters<br />
are: (1) autonomy, (2) exercise <strong>of</strong> skill, (3) establishing and<br />
maintaining meaningful relationships, (4) being awake to<br />
the present reality (mindfulness), (5) gratitude, (6) courage,<br />
(7) appreciation <strong>of</strong> impermanence, (8) compassionate mind<br />
frame, (9) finding and keeping one’s sense <strong>of</strong> safety and<br />
security, and (10) answering the question, “Do I matter?”<br />
The parameters <strong>of</strong> buoyancy can be compared with one’s<br />
finances, in which the different parts <strong>of</strong> a portfolio contribute<br />
to the bottom line. For instance, if “managing fear” and<br />
“finding and keeping your safety and security” stocks take a<br />
brief hit, they may rebound splendidly, or perhaps one’s<br />
overall buoyancy will trend upward by virtue <strong>of</strong> a compensatory<br />
increase in other areas. With this in mind, one can see<br />
how buoyancy would be expected to vary day to day and year<br />
to year. However, a solid manager <strong>of</strong> the metaphoric portfolio<br />
(i.e., you) could keep the net performance in a consistently<br />
good range and find ways to hedge against extreme<br />
underperformance. Extreme underperformance in buoyancy<br />
would likely manifest as burnout.<br />
A detailed examination <strong>of</strong> some <strong>of</strong> the parameters contributing<br />
to buoyancy will reveal how they can manifest and<br />
change over time.<br />
By Michael J. Fisch, MD, MPH<br />
model is proposed as a method to take stock <strong>of</strong> key parameters<br />
that may contribute to happiness and resiliency. Selfmonitoring<br />
<strong>of</strong> personal buoyancy parameters may help<br />
oncology providers prevent burnout.<br />
Autonomy (Freedom to Choose)<br />
<strong>Oncology</strong> providers are well aware <strong>of</strong> autonomy as an<br />
ethical principle and routinely apply it to medical decision<br />
making. For clinicians, it is worth reflecting on whether one<br />
has a modicum <strong>of</strong> control over the work environment and<br />
schedule as well as how one proceeds to undertake various<br />
projects. With changes in health care and the trend toward<br />
larger organizational structures, some providers may feel<br />
that they have lost ground in this dimension. It is worthwhile<br />
to consider what has not been lost, to determine which<br />
expectations are realistic, and to have a dialogue and negotiation<br />
with colleagues on key points <strong>of</strong> autonomy when<br />
necessary.<br />
Exercise <strong>of</strong> Skill (Doing Your Thing)<br />
This parameter is focused on realizing what one’s<br />
strengths and passions are and making sure one does not<br />
compromise too much in this area. For example, people who<br />
enjoy taking care <strong>of</strong> patients with complex conditions and<br />
mastering difficult diagnoses would not fare that well in a<br />
high-volume clinic focused on breast cancer survivorship<br />
care. This extends not only to the obvious focus <strong>of</strong> care but<br />
also to the aspects <strong>of</strong> work that make it most fulfilling. For<br />
example, the nontechnical, humanistic aspects <strong>of</strong> work (especially<br />
in settings associated with medical failure) are <strong>of</strong>ten<br />
some <strong>of</strong> the most gratifying experiences. 18 Understanding<br />
that these aspects can be incorporated into one’s focus, while<br />
continuing to build on skills in communication and compassionate<br />
care, can render any patient care setting or individual<br />
encounter more enriching. 19<br />
Early in my career, I received this advice: “Do not let<br />
medical school ruin your education.” Medicine, with its<br />
potentially endless layers, is sufficiently absorbing that it<br />
can overgrow other dimensions <strong>of</strong> skill and interest if allowed.<br />
Think deeply about your other skills and interests,<br />
perhaps music, reading mystery novels, cooking, boating,<br />
helping others in the community, teaching or coaching<br />
children, martial arts, and so on. Make the opportunity to<br />
exercise skills in some <strong>of</strong> these areas also. Insights derived<br />
from other realms can help us grow as physicians. David<strong>of</strong>f,<br />
for one, described how music enriches our connection to our<br />
own emotions, sensitizes us to being more empathic, and<br />
From the Department <strong>of</strong> General <strong>Oncology</strong>, University <strong>of</strong> Texas M. D. Anderson Cancer<br />
Center, Houston, TX.<br />
Author’s disclosures <strong>of</strong> potential conflicts <strong>of</strong> interest are found at the end <strong>of</strong> this article.<br />
Address reprint requests to Michael J. Fisch, MD, MPH, Department <strong>of</strong> General<br />
<strong>Oncology</strong>, University <strong>of</strong> Texas M. D. Anderson Cancer Center, 1515 Holcombe Blvd., Unit<br />
0410, Houston, TX 77030; email: mfisch@mdanderson.org.<br />
© <strong>2012</strong> by <strong>American</strong> <strong>Society</strong> <strong>of</strong> <strong>Clinical</strong> <strong>Oncology</strong>.<br />
1092-9118/10/1-10<br />
e77
also provides insight on how clinical practice might be<br />
learned, taught, and performed more effectively. 20<br />
Establishment <strong>of</strong> Meaningful Relationships<br />
Seventeenth-century poet John Donne realized that “no<br />
man is an island entire <strong>of</strong> itself.” Recognizing one’s connections<br />
to others and reflecting on how to tap into and nurture<br />
key relationships can get lost in the course <strong>of</strong> daily routine,<br />
and we clinicians are indeed busy. Providers may face<br />
burnout when they focus too heavily on their relationship<br />
with their patients and lose sight <strong>of</strong> their colleagues, other<br />
staff, and collaborators who are critical to success and<br />
well-being. This increases vulnerability to mistakes, loss <strong>of</strong><br />
boundaries, feelings <strong>of</strong> pr<strong>of</strong>essional impotence, and being<br />
perceived as ineffective. The demands on providers are<br />
particularly high early in their careers when they are trying<br />
to establish their niche, skills, and reputation. This frequently<br />
coincides with young marriages and young children.<br />
The work-family conflict that <strong>of</strong>ten ensues is clearly a risk<br />
factor for burnout, and physicians must be aware that<br />
attention, time, and open communication are necessary<br />
KEY POINTS<br />
● Stress and burnout are prevalent problems among<br />
oncology providers.<br />
● The syndrome <strong>of</strong> burnout can adversely affect patient<br />
care outcomes and reduce the health and fulfillment<br />
<strong>of</strong> clinicians.<br />
● Buoyancy refers to the force, in opposition to burnout,<br />
that uplifts the provider.<br />
● The individual parameters that constitute overall<br />
buoyancy vary between individuals and also vary<br />
within individuals over time (and sometimes day to<br />
day).<br />
● Examining one’s own buoyancy factors can be useful<br />
to guide choices and habits in such a way that<br />
maximizes buoyancy and minimizes risk <strong>of</strong> burnout.<br />
e78<br />
outside the work setting to sustain growth and fulfillment in<br />
the workplace. 21 One way to enhance success in this dimension<br />
is to learn communication skills and leadership skills.<br />
Workshops on these topics are <strong>of</strong>ten available to early career<br />
physicians but overlooked in favor <strong>of</strong> content-focused learning.<br />
These behavior-oriented workshops help with skills<br />
such as authentic listening, giving and receiving feedback,<br />
reserving judgment, noticing one’s own emotions, negotiation,<br />
and other critical aspects <strong>of</strong> relating to others and<br />
mindful awareness <strong>of</strong> oneself. Finally, some practices and<br />
programs have created forums for self-disclosure and dialogue<br />
among physicians (see http://theheart<strong>of</strong>medicine.org<br />
for ideas and resources).<br />
Being Awake to Your Present Reality<br />
MICHAEL J. FISCH<br />
Fig 1. Dimensions <strong>of</strong> buoyancy.<br />
<strong>Oncology</strong> providers generally are good achievers and are<br />
quite adept at striving for specific goals for themselves and<br />
their patients. Being awake to your present reality amounts<br />
to checking in with yourself about basic conditions that are<br />
frequently nearly out <strong>of</strong> view—in one’s life blind spot, as it<br />
were. This might include simple matters such as “I am tired.<br />
I should rest now” or “I am frustrated. I should take a break<br />
for a while,” but an assessment could encompass larger<br />
matters such as “I am ill. I should see a doctor at this point”<br />
or even an appraisal <strong>of</strong> what one really wants out <strong>of</strong> life (not<br />
just work or home in isolation). Honest appraisal and<br />
regular assessment <strong>of</strong> your health, how you think and work,<br />
and your expectations can pay major dividends in buoyancy.<br />
Regarding health, taking note <strong>of</strong> your diet, the amount<br />
and quality <strong>of</strong> your sleep, and whether aerobic and resistance<br />
exercise are part <strong>of</strong> your routine is a good starting<br />
point. Willingness to get care from another physician is<br />
important too, as this allows preventive care and health<br />
maintenance as well as treatment for existing conditions.<br />
Neglected health issues such as depression and substance<br />
abuse are common among providers who suffer from impairment,<br />
disruptive behavior, or burnout. 22<br />
An appraisal <strong>of</strong> thinking and working habits is also<br />
worthwhile, as these can also escape our notice and undermine<br />
well-being. It is worthwhile to consider whether you<br />
work efficiently and delegate appropriately. Or do you tend
BUOYANCY TO PREVENT STRESS AND BURNOUT<br />
to get almost all <strong>of</strong> your needs met by trying to help others<br />
and neglecting yourself? Do you find yourself being cynical<br />
or intolerant <strong>of</strong> “fools,” catastrophizing, or being overly<br />
negative or pessimistic? Self-appraisal in these realms is<br />
healthy. We learn about the concept <strong>of</strong> “equipoise” in clinical<br />
trials and scientific experimentation, but it is equally critical<br />
to introduce equipoise in our own ways <strong>of</strong> looking at things.<br />
In other words, learn to resist becoming too enamored <strong>of</strong><br />
your own ideas and perspectives. Take a close look at your<br />
position about an issue and think, “Maybe so.” Next, try to<br />
explore alternate perspectives, ranging from one extreme to<br />
the opposite and then back toward your own estimate <strong>of</strong><br />
reality. You may end up in the same place—your position—<br />
but flexing your mind around an issue can help you find<br />
greater balance and enhance your buoyancy.<br />
Gratitude<br />
The idea that it is healthy to “count one’s blessings” is<br />
widely understood. However, having the discipline and skill<br />
to maintain this particular intention is not simple. One way<br />
to accomplish this is to think deeply about the concept <strong>of</strong><br />
buoyancy and notice that there are indeed things that are<br />
working in your life. As a further step, having noticed your<br />
buoyancy and its roots, take a moment to actually feel<br />
gratitude for achieving balance. Finally, act on gratitude by<br />
speaking or writing words <strong>of</strong> appreciation. In my practice,<br />
good habits are <strong>of</strong>ten easiest to internalize when I implement<br />
them at a patient’s bedside for the purpose <strong>of</strong> improving<br />
patient care. In that setting, start by noticing and<br />
appreciating elements <strong>of</strong> your patient’s behavior. Patients<br />
do amazing things all the time; it is not difficult to find<br />
behaviors to comment positively about. Learning to say<br />
goodbye to patients toward the end <strong>of</strong> life 23 and writing<br />
letters <strong>of</strong> condolence to family members 24 are examples <strong>of</strong><br />
showing gratitude in clinical settings. The next step would<br />
be extending this practice to your colleagues, staff, family,<br />
and friends by saying “thank you” more frequently. Many<br />
providers are surprisingly stingy when it comes to saying<br />
“thank you” in pr<strong>of</strong>essional and personal realms, perhaps as<br />
a result <strong>of</strong> high expectations from staff and colleagues.<br />
Expressing gratitude more <strong>of</strong>ten does not require lowering<br />
expectations, and it can help bolster buoyancy.<br />
Courage (Managing Fear)<br />
<strong>Oncology</strong> providers, in general, have a great deal <strong>of</strong><br />
courage. It takes courage to compete in school and achieve<br />
enough success to get trained in medicine and cancer care. It<br />
takes courage to deal with the sacrifices inherent in medicine<br />
and the great responsibility (and privilege) <strong>of</strong> caring for<br />
persons who are ill. But courage is not generic. One can be a<br />
war hero and be frightened <strong>of</strong> snakes or bugs. If you think <strong>of</strong><br />
courage as “fear management,” then it makes sense that one<br />
can be successful in some realms and still need work in<br />
others. To guard against burnout and build buoyancy, providers<br />
need to learn how to recognize their fears, give those<br />
fears a name, and reflect deeply on how to confront fear.<br />
Common fears include fear <strong>of</strong> making mistakes, being<br />
blamed or sued, or being judged or taken for granted by<br />
colleagues, friends, or family. There may be fear <strong>of</strong> losing<br />
income as a result <strong>of</strong> competition, <strong>of</strong> not being appropriately<br />
recognized or promoted, or any number <strong>of</strong> other concerns<br />
that are understandable in our modern environment. Other<br />
issues revolve around countertransference and one’s feelings<br />
about personal mortality and death. Becoming aware <strong>of</strong> our<br />
fears and developing habits and action plans for managing<br />
them is critical. Introspection is critical. For many providers,<br />
connecting with others is preferable to tackling these<br />
tasks alone.<br />
Appreciation <strong>of</strong> Impermanence<br />
To appreciate impermanence is to understand that all<br />
things change and that loss is part <strong>of</strong> the package in life.<br />
Being flexible and open to change is essential for personal<br />
growth and equanimity. I may not relish change in all<br />
instances (like my aging body or my changing bank account),<br />
but I can see this truth. The idea that change is the nature<br />
<strong>of</strong> all things is yet another topic that is easy to recite but<br />
difficult to integrate on a practical level. Change is frequently<br />
associated with emotions, both positive and negative.<br />
People cry at weddings and funerals and graduations,<br />
for instance—all events associated with transitions. How<br />
<strong>of</strong>ten to you notice your own emotions? Do you notice when<br />
you are getting angry? Are you aware <strong>of</strong> your own grieving?<br />
In oncology, dealing with loss and grief is particularly<br />
important. 16,25 In his classic article “Dealing with Our<br />
Losses,” Balfour Mount wrote, “To live is to experience<br />
loss.” 26 A mentor once told me, “You have got to learn how to<br />
lose.” We experience loss not only with our patients facing<br />
serious illness but also with changes, setbacks, and failures<br />
<strong>of</strong> every variety: experiments that fail, grants not funded,<br />
papers not accepted, organizational shifts that are not in our<br />
favor, and the list goes on. As oncology providers, we should<br />
strive to maintain our resilience while at the same time<br />
finding a way to notice and manage our emotions. Although<br />
people are variably resilient, there are related skills that can<br />
be learned and practiced. Mindfulness training, for example,<br />
has shown benefits for helping physicians become more<br />
awake to their present reality as well as positioned to accept<br />
the inevitability <strong>of</strong> change. 8<br />
Compassionate Mind Frame<br />
It is a bit embarrassing to have been concerned with the<br />
human problem all one’s life and find at the end that one<br />
has no more to <strong>of</strong>fer by way <strong>of</strong> advice than ‘try to be a little<br />
kinder.’ –Aldous Huxley<br />
I think about compassion as something that starts with<br />
empathy but requires buoyancy (like a coenzyme in a chemical<br />
reaction) to come to fruition in an objective way. Without<br />
buoyancy, empathy would pull one toward the same place<br />
as the subject <strong>of</strong> one’s empathy. The good news, then, is that<br />
all the things you do to stay buoyant position you for success<br />
in creating and maintaining a compassionate mind frame.<br />
The aspect <strong>of</strong> compassion that takes particular practice and<br />
attention is empathy. Once again, a good place to start is<br />
in the clinical realm—learning to act empathetically toward<br />
your patients and, better yet, to feel empathy, too. There<br />
are good reviews on this topic, 27,28 and communication<br />
workshops and time spent with master clinicians can help<br />
promulgate this skill. The next step toward maximizing<br />
the buoyancy derived from a compassionate mind frame is to<br />
extend these skills outside the clinical setting to staff,<br />
colleagues, students, family, friends, and strangers, and<br />
then stretch this powerful mind frame all the way to oneself.<br />
Being compassionate toward oneself is a beautiful way to<br />
stay resilient and guard against the forces <strong>of</strong> burnout. A<br />
e79
good resource for exploring the principles and some specific<br />
programs related to compassion in health care is the<br />
Schwartz Center (www.theschwartzcenter.org).<br />
Finding and Keeping Your Safety and Security<br />
Some issues are difficult to put into a defined category but<br />
nevertheless have a major effect on personal buoyancy, and<br />
therefore are worthy <strong>of</strong> careful examination. One example is<br />
money. Some people are more affected by money than<br />
others—depending on things like family size, level <strong>of</strong> preexisting<br />
debt, amount <strong>of</strong> wealth in the family, personal goals<br />
and preferences, etc. Another example is location. For some<br />
oncology providers, living near family (or a spouse’s family)<br />
makes an enormous difference in the overall well-being <strong>of</strong><br />
the family, and thus has influence on the family’s tolerance<br />
<strong>of</strong> long work hours and other aspects <strong>of</strong> sacrifice that may be<br />
called for within a particular job setting. In addition to<br />
money and location, another factor affecting buoyancy is<br />
religion and spirituality. The ability to integrate with a<br />
religious community and worship on certain days is paramount<br />
for some providers. Factors such as these (and many<br />
others that might fit this broad category) are highly variable<br />
between individuals and sometimes vary within individuals<br />
over time.<br />
Author’s Disclosures <strong>of</strong> Potential Conflicts <strong>of</strong> Interest<br />
Answering the Question, “Do I Matter?”<br />
<strong>Oncology</strong> providers make considerable sacrifices in the<br />
course <strong>of</strong> training and also in the course <strong>of</strong> daily practice. To<br />
maximize a sense <strong>of</strong> fulfillment, providers need to feel they<br />
are making a difference. They want to matter to their<br />
patients, to their organization or team, and in various<br />
groups and relationships <strong>of</strong> importance in work, family, and<br />
community settings. A loss <strong>of</strong> buoyancy is sometimes best<br />
understood by deep reflection and unmasking a “Do I matter?”<br />
crisis in some dimension. Such a crisis can also focus on<br />
the ultimate, existential dimension.<br />
Conclusion<br />
Employment or<br />
Leadership Consultant or Stock<br />
Research<br />
Author<br />
Positions Advisory Role Ownership Honoraria Funding<br />
Michael J. Fisch Elsevier (L) Biosite<br />
1. Maslach C, Schaufeli WB, Leiter MP. Job burnout. Annual Rev Psych.<br />
2001;52:397-422.<br />
2. Allegra CJ, Hall R, Yothers G. Prevalence <strong>of</strong> burnout in the U.S.<br />
<strong>Oncology</strong> community: results <strong>of</strong> a 2003 survey. J Oncol Pract. 2005;1:140-147.<br />
3. Whippen DA, Canellos GP. Burnout syndrome in the practice <strong>of</strong> oncology:<br />
results <strong>of</strong> a random survey <strong>of</strong> 1,000 oncologists. J Clin Oncol. 1991;9:<br />
1916-1920.<br />
4. Liakopoulou M, Panaretaki I, Papadakis V, et al. Burnout, staff support,<br />
and coping in pediatric oncology. Support Care Cancer. 2008;16:143-150.<br />
5. Mukherjee S, Beresford B, Glaser A, et al. Burnout, psychiatric morbidity,<br />
and work-related sources <strong>of</strong> stress in paediatric oncology staff: a review <strong>of</strong><br />
the literature. Psycho-oncology. 2009;18:1019-1028.<br />
6. Balch CM, Shanafelt TD, Sloan JA, et al. Distress and career satisfaction<br />
among 14 surgical specialties, comparing academic and private practice<br />
settings. Ann Surg. 2011;254:558-568.<br />
7. Kuerer HM, Eberlein TJ, Pollock RE, et al. Career satisfaction, practice<br />
patterns and burnout among surgical oncologists: report on the quality <strong>of</strong> life<br />
<strong>of</strong> members <strong>of</strong> the <strong>Society</strong> <strong>of</strong> Surgical <strong>Oncology</strong>. Ann Surg Oncol. 2007;14:<br />
3043-3053.<br />
8. Krasner MS, Epstein RM, Beckman H, et al. Association <strong>of</strong> an educational<br />
program in mindful communication with burnout, empathy, and<br />
attitudes among primary care physicians. JAMA. 2009;302:1284-1293.<br />
9. Bram PJ, Katz LF. A study <strong>of</strong> burnout in nurses working in hospice and<br />
hospital oncology settings. Oncol Nurs Forum. 1989;16:555-560.<br />
10. Blanchard P, Truchot D, Albiges-Sauvin L, et al. Prevalence and causes<br />
<strong>of</strong> burnout amongst oncology residents: a comprehensive nationwide crosssectional<br />
study. Eur J Cancer. 2010;46:2708-2715.<br />
11. Shanafelt TD. Enhancing meaning in work: a prescription for preventing<br />
physician burnout and promoting patient-centered care. JAMA. 2009;302:<br />
1338-1340.<br />
12. West CP, Shanafelt TD, Kolars JC. Quality <strong>of</strong> life, burnout, educational<br />
debt, and medical knowledge among internal medicine residents. JAMA.<br />
2011;306:952-960.<br />
13. Dyrbye LN, Massie FS, Jr., Eacker A, et al. Relationship between<br />
burnout and pr<strong>of</strong>essional conduct and attitudes among US medical students.<br />
JAMA. 2010;304:1173-1180.<br />
e80<br />
REFERENCES<br />
Stress and burnout are prevalent problems among oncology<br />
providers, and these problems have a substantial effect<br />
on patient care outcomes and provider health and fulfillment.<br />
Examining one’s own buoyancy factors can be useful<br />
to guide choices and habits in such as way that maximizes<br />
buoyancy and minimizes risk <strong>of</strong> burnout.<br />
Acknowledgments<br />
The author gratefully acknowledges Ms. Danielle Walsh for<br />
providing the graphics for the figure, and Ms. Joann Aaron and<br />
Mr. Bryan Tutt for providing editorial assistance.<br />
Expert<br />
Testimony<br />
MICHAEL J. FISCH<br />
Other<br />
Remuneration<br />
14. Br<strong>of</strong>fman G. Controlled burn! Physician executives must be ready to<br />
handle job burnout, career stress. Physician Executive. 2001;27:42-45.<br />
15. Johns MM III, Oss<strong>of</strong>f RH. Burnout in academic chairs <strong>of</strong> otolaryngology:<br />
head and neck surgery. Laryngoscope. 2005;115:2056-2061.<br />
16. Lyckholm L. Dealing with stress, burnout, and grief in the practice <strong>of</strong><br />
oncology. Lancet Oncol. 2001;2:750-755.<br />
17. Shanafelt T, Chung H, White H, et al. Shaping your career to maximize<br />
personal satisfaction in the practice <strong>of</strong> oncology. J Clin Oncol. 2006;24:4020-<br />
4026.<br />
18. Horowitz CR, Suchman AL, Branch WT Jr., et al. What do doctors find<br />
meaningful about their work? Ann Intern Med. 2003;138:772-775.<br />
19. Armstrong J, Holland J. Surviving the stresses <strong>of</strong> clinical oncology by<br />
improving communication. <strong>Oncology</strong>. 2004;18:363-375.<br />
20. David<strong>of</strong>f F. Music lessons: what musicians can teach doctors (and other<br />
health pr<strong>of</strong>essionals). Ann Intern Med. 2011;154:426-429.<br />
21. Dyrbye LN, West CP, Satele D, et al. Work/home conflict and burnout<br />
among academic internal medicine physicians. Arch Intern Med. 2011;171:<br />
1207-1209.<br />
22. Brown SD, Goske MJ, Johnson CM. Beyond substance abuse: atress,<br />
burnout, and depression as causes <strong>of</strong> physician impairment and disruptive<br />
behavior. J Am Coll Radiol. 2009;6:479-485.<br />
23. Back AL, Arnold RM, Tulsky JA, et al. On saying goodbye: acknowledging<br />
the end <strong>of</strong> the patient-physician relationship with patients who are<br />
near death. Ann Intern Med. 2005;142:682-685.<br />
24. Bedell SE, Cadenhead K, Graboys TB. The doctor’s letter <strong>of</strong> condolence.<br />
N Engl J Med. 2001;344:1162-1164.<br />
25. Meier DE, Back AL, Morrison RS. The inner life <strong>of</strong> physicians and care<br />
<strong>of</strong> the seriously ill. JAMA. 2001;286:3007-3014.<br />
26. Mount BM. Dealing with our losses. J Clin Oncol. 1986;4:1127-1134.<br />
27. Back A, Arnold RM, Tulsky JA. Mastering Communication with Seriously<br />
Ill Patients: Balancing Honesty with Empathy and Hope. New York:<br />
Cambridge University Press; 2009.<br />
28. Coulehan JL, Platt FW, Egener B, et al. “Let me see if I have this<br />
right . . .”: words that help build empathy. Ann Intern Med. 2001;135:221-<br />
227.
Encountering Grief in Patient Care<br />
Overview: Grief is essentially unavoidable and is a normal<br />
reaction to loss. Grief may be experienced by patients and<br />
their loved ones as well as by physicians and members <strong>of</strong> the<br />
health care team in response to the consequences <strong>of</strong> illness or<br />
death. Grief is typified by certain indicators that may significantly<br />
effect one’s emotional and physical well-being. Although<br />
these indicators tend to follow a general pattern, there<br />
is variability among individuals. Complicated grief may require<br />
EMILY DICKINSON wrote, “I measure every grief I<br />
meet with analytic eyes—I wonder if it weighs like<br />
mine—or has an easier size.” 1 For most humans, grief<br />
is unavoidable. Physicians encounter grief in their patients<br />
and their patient’s loved ones but may also experience grief<br />
on a personal level in response to caring for a patient. In the<br />
practice <strong>of</strong> medicine, grief is <strong>of</strong>ten associated with death but<br />
may also be experienced in response to loss from the ravages<br />
<strong>of</strong> illness. It is valuable for physicians to explore the process<br />
<strong>of</strong> grief, not only to enhance patient care but also to optimize<br />
their own well-being. Grief has the potential to connect us to<br />
our humanity.<br />
The Meaning <strong>of</strong> Grief<br />
A common definition <strong>of</strong> grief is a “deep and poignant<br />
distress caused by or as if by bereavement,” where bereavement<br />
refers to “suffering the death <strong>of</strong> a loved one.” 2 However,<br />
the meaning <strong>of</strong> grief, may also connote a broader<br />
application. In the Education in Palliative and End-<strong>of</strong>-Life<br />
Care (EPEC) Project oncology core module regarding loss,<br />
grief, and bereavement, grief is defined as the “experience <strong>of</strong><br />
a loss” and bereavement as the “state <strong>of</strong> living with a loss.” 3<br />
Although grief is most <strong>of</strong>ten associated with a loss from<br />
death, it is important to recognize that loss is present<br />
throughout the course <strong>of</strong> illness. There may be a multitude<br />
<strong>of</strong> losses that effect nearly every aspect <strong>of</strong> life for the patient<br />
and their loved ones. 3<br />
Some <strong>of</strong> these losses may be easily recognizable, such as<br />
the loss <strong>of</strong> one’s usual physical appearance and bodily<br />
functions. With illness, the usual roles <strong>of</strong> the patient and his<br />
or her loved ones <strong>of</strong>ten change. The former caregiver may<br />
now become the patient, thus threatening the equilibrium <strong>of</strong><br />
former relationships. The alterations in these social interactions<br />
between the patient and members <strong>of</strong> his or her<br />
community may lead to a sense <strong>of</strong> loss for the familiarity <strong>of</strong><br />
prior responsibilities. In addition, the financial costs <strong>of</strong><br />
illness, including expenses and lost wages, may induce a<br />
tangible loss <strong>of</strong> relative monetary stability. A study <strong>of</strong> 988<br />
patients with various terminal illnesses (51.8% with cancer)<br />
found that nearly 35% had moderate to high care needs (e.g.,<br />
required assistance with transportation, personal and nursing<br />
care, and homemaking). 4 These patients reported a<br />
greater economic burden than those with low care needs.<br />
There are also the subtle losses that may develop on an<br />
emotional or philosophical level, such as the loss <strong>of</strong> expectations<br />
for the future, the loss <strong>of</strong> fulfillment <strong>of</strong> dreams, and the<br />
loss <strong>of</strong> a common human perspective that death is somehow<br />
far away. When death becomes imminent or occurs, it tends<br />
to shake the foundation <strong>of</strong> how we view life.<br />
By Teresa Gilewski, MD<br />
psychiatric intervention. Caring for the seriously ill or dying<br />
patient may be particularly challenging from an emotional<br />
level and may increase the risk <strong>of</strong> burnout. Recognition <strong>of</strong><br />
these emotions is a critical aspect <strong>of</strong> providing compassionate<br />
care on a sustainable level. Various strategies may be beneficial<br />
in coping with grief, and the exploration <strong>of</strong> grief may<br />
provide greater insight into the humanistic basis <strong>of</strong> medicine.<br />
Physicians are privy to some <strong>of</strong> these losses in their<br />
interactions with patients and their loved ones. However,<br />
physicians may also experience their own grief in response<br />
to caring for patients with such losses. Physicians themselves<br />
may feel a sense <strong>of</strong> loss or distress regarding ineffective<br />
treatments, an inability to adequately respond to the<br />
patient’s suffering, or the death <strong>of</strong> a patient.<br />
Stages and Types <strong>of</strong> Grief<br />
Grief consists <strong>of</strong> multiple reactions, including behavioral,<br />
psychologic/emotional, social, spiritual, and physical changes<br />
(Table 1). 3,5 Various characteristics <strong>of</strong> the mourning process—proposed<br />
by Sigmund Freud, Eric Lindemann, Colin<br />
Parkes, and John Bowlby—have been essential to an understanding<br />
<strong>of</strong> grief and bereavement. 6 Elisabeth Kübler-Ross,<br />
MD, popularized five stages <strong>of</strong> grief in her classic book “On<br />
Death and Dying,” published in 1969. 7 When faced with<br />
death because <strong>of</strong> an incurable illness, she observed that<br />
patients commonly experienced denial, anger, bargaining,<br />
depression, and acceptance. However, these stages may be<br />
<strong>of</strong> various durations and intensities. Individuals may not<br />
experience all <strong>of</strong> these stages, nor may they occur in the<br />
same order. Since then, further evaluation has provided<br />
greater insight into the patterns <strong>of</strong> grief—either normal and<br />
uncomplicated or complicated.<br />
The Yale Bereavement Study queried 233 bereaved individuals<br />
whose family member or loved one died <strong>of</strong> natural<br />
causes. 8 Over a period <strong>of</strong> 24 months, five grief indicators<br />
were rated: disbelief, yearning, anger, depression, and acceptance.<br />
Yearning was the most <strong>of</strong>ten reported negative<br />
indicator with a peak value at 4 months post-loss. Disbelief<br />
peaked at 1 month post-loss, anger peaked at 5 months<br />
post-loss, and depression at 6 months post-loss. Interestingly,<br />
acceptance was the most commonly reported factor<br />
and continued to increase during the duration <strong>of</strong> the study.<br />
This trajectory is most consistent with an uncomplicated<br />
or normal grief reaction. However, significant expression<br />
<strong>of</strong> negative emotions after 6 months may indicate a more<br />
complex grief reaction that requires additional assessment.<br />
Complicated grief comprises a unique entity that signifies<br />
a greater level <strong>of</strong> dysfunction and portends a greater risk <strong>of</strong><br />
morbidity. Characteristic features may occur in anticipation<br />
From the Memorial Sloan-Kettering Cancer Center, New York, NY.<br />
Author’s disclosure <strong>of</strong> potential conflicts <strong>of</strong> interest are found at the end <strong>of</strong> this article.<br />
Address reprint requests Teresa Gilewski, MD, Memorial Sloan-Kettering Cancer Center,<br />
300 East 66 th St., New York, NY; email: gilewskt@mskcc.org.<br />
© <strong>2012</strong> by <strong>American</strong> <strong>Society</strong> <strong>of</strong> <strong>Clinical</strong> <strong>Oncology</strong>.<br />
1092-9118/10/1-10<br />
e81
<strong>of</strong> death or following death and include excessive bitterness,<br />
yearning, emptiness, numbness, and inability to cope with<br />
the loss. 9 An individual’s personality traits may contribute<br />
to the development <strong>of</strong> complicated grief reactions. For some,<br />
grief may be chronic, delayed, exaggerated, or masked and<br />
may necessitate psychiatric intervention. 3<br />
In their pr<strong>of</strong>essional role, physicians’ awareness <strong>of</strong> the<br />
emotions experienced by patients and their loved ones in the<br />
grieving process is essential. Physicians may also grieve for<br />
their patients’ losses or following their patients’ deaths, in a<br />
series <strong>of</strong> “minigrief” reactions. 10<br />
The Effect <strong>of</strong> Patient Care and Grief<br />
on the Physician<br />
General Effect<br />
As human beings, physicians instinctively develop emotions<br />
about their patients. It is the responsibility <strong>of</strong> the<br />
physician to acknowledge that these feelings are an integral<br />
KEY POINTS<br />
● In the medical field, grief is a normal response to loss<br />
from the consequences <strong>of</strong> illness or death.<br />
● Grief is characterized by emotional and physical<br />
indicators and may be uncomplicated or complicated;<br />
complicated grief causes significant dysfunction and<br />
requires further psychiatric intervention.<br />
● Grief is experienced by the patient and their loved<br />
ones as well as by the physician and other members <strong>of</strong><br />
the health care team.<br />
● The expectations and goals <strong>of</strong> the physician in regard<br />
to patient care and the repeated exposure to death<br />
may contribute to burnout.<br />
● Coping with grief requires acknowledgement <strong>of</strong> one’s<br />
emotions and a focus on strategies to optimize overall<br />
well-being.<br />
e82<br />
Table 1. Some Potential Indicators <strong>of</strong> Grief 3,5,22<br />
Psychologic/Emotional<br />
Disbelief<br />
Anxiety<br />
Sadness<br />
Numbness<br />
Anger<br />
Behavioral<br />
Searching/yearning<br />
Difficulty with concentration<br />
Physical<br />
Anorexia<br />
Fatigue<br />
Palpitations<br />
Difficulty sleeping<br />
Hair loss<br />
Weight change<br />
Gastrointestinal disturbance<br />
Social<br />
Social withdrawal<br />
Isolation<br />
Spiritual<br />
Questioning beliefs<br />
part <strong>of</strong> the doctor–patient relationship. 11 A key challenge is<br />
finding a balance where the physician “feels” for a patient<br />
but not to the extent that it becomes so overwhelming the<br />
physician cannot empathize with the next patient. 12<br />
Meier et al emphasized that these emotions, if left unexamined,<br />
may negatively affect patient care and the wellbeing<br />
<strong>of</strong> the physician. 13 They propose that recognition <strong>of</strong><br />
risk factors that significantly influence an emotional response<br />
and awareness <strong>of</strong> the resultant behaviors and emotions<br />
are important. Once the emotions are identified and<br />
accepted, further reflection on the emotion and discussion<br />
with a trustworthy colleague may enhance the ability <strong>of</strong><br />
physicians to reach an emotional balance that is sustainable<br />
and, thus, minimize the risk <strong>of</strong> burnout and stress. This is <strong>of</strong><br />
particular relevance when the physician encounters grief on<br />
a frequent basis, as in the field <strong>of</strong> oncology.<br />
Physician Response to Patient Death<br />
TERESA GILEWSKI<br />
There is a surprisingly limited amount <strong>of</strong> information on<br />
the physician’s response to patient deaths. A cross-sectional<br />
study <strong>of</strong> two academic hospitals in the United States evaluated<br />
the reactions <strong>of</strong> 188 physicians, including attendings,<br />
residents, and interns to the death <strong>of</strong> their patients. 14 There<br />
were 68 evaluable inpatient deaths that occurred on a<br />
general medicine or intensive care unit service. Only 12% <strong>of</strong><br />
the physicians had cared for the patient before this hospitalization;<br />
62% characterized their connection with the patient<br />
as “not close.”<br />
Most physicians had satisfying experiences with their<br />
patients. On average they experienced 2 <strong>of</strong> 14 indicators <strong>of</strong><br />
grief, with “feeling upset” or “numb” the most frequent,<br />
while approximately 23% described the death as “very<br />
disturbing.” Grief was associated with longer periods <strong>of</strong><br />
caring for the patient and was more common in female<br />
physicians. There was no significant distinction in emotional<br />
response based on level <strong>of</strong> training, although interns wanted<br />
more emotional support than attendings. Less than 25% <strong>of</strong><br />
the interns and residents viewed the attending as the most<br />
valuable source <strong>of</strong> assistance in coping.<br />
A study from the United Kingdom surveyed physicians<br />
regarding their last memorable patient death. This was<br />
described as one that caused reflection and happened in<br />
the few months before the study. 15 In a mix <strong>of</strong> house <strong>of</strong>ficers<br />
and attending staff, the majority reported coping satisfactorily<br />
after the patient’s death. Nearly 84% favored “talking<br />
with others” as a method <strong>of</strong> coping, and approximately 57%<br />
supported additional education on this topic. Approximately<br />
44% reported feelings <strong>of</strong> moderate to severe sadness.<br />
These studies raise several interesting topics for further<br />
evaluation. Historically, the primary physician <strong>of</strong>ten followed<br />
the patient as both inpatient and outpatient. However,<br />
with the increasing use <strong>of</strong> hospitalists, the patient’s<br />
primary physician may have limited contact with the patient<br />
during the final inpatient hospitalization. The significance<br />
<strong>of</strong> this factor as well as the location <strong>of</strong> the patient’s<br />
death (e.g., hospital, home, hospice center) on the primary<br />
physician’s emotional response is unclear. These studies<br />
were not specific to oncology, so oncologists who frequently<br />
care for patients over extended time periods may be at an<br />
increased risk <strong>of</strong> grief. In addition, the role <strong>of</strong> the attending<br />
physician in discussing grief with the house staff is likely<br />
underdeveloped.
ENCOUNTERING GRIEF IN PATIENT CARE<br />
Focus on Burnout<br />
Grief and continuous exposure to death may contribute to<br />
burnout. The key features <strong>of</strong> burnout identified by Maslach<br />
et al include exhaustion, depersonalization (cynicism), and<br />
ineffectiveness, and these symptoms are primarily job related.<br />
16 Factors that may increase the risk <strong>of</strong> burnout<br />
include younger age, single marital status, low self-esteem,<br />
less hardy character trait, avoidant coping approach, belief<br />
that others or chance are in control, and possibly job attitude.<br />
Burnout affects patient care as well as the individual<br />
physician.<br />
Whippen and Canellos surveyed 1,000 oncologists regarding<br />
burnout, 60% <strong>of</strong> whom were medical oncologists. 17 In the<br />
598 completed responses, overall 56% <strong>of</strong> oncologists endorsed<br />
burnout. The presence <strong>of</strong> burnout was similar across<br />
the subspecialties with 58% for medical oncologists, 48% for<br />
surgical oncologists, 44% for pediatric oncologists, and 52%<br />
for radiation oncologists. The most common potential causes<br />
<strong>of</strong> burnout were insufficient personal/vacation time (57%),<br />
continuous exposure to fatal diseases (53%), and frustration<br />
<strong>of</strong> ineffective therapies (45%). Interestingly, approximately<br />
80% felt that their career fulfilled the expectations they had<br />
in training.<br />
A high level <strong>of</strong> burnout was observed in another study <strong>of</strong><br />
261 house staff, nurses, and oncologists. 18 The most common<br />
stressor for burnout was unfavorable work events, such as<br />
frequent exposure to death and controversies over do not<br />
resuscitate status. Therefore, health care workers may be<br />
predisposed to stress and burnout from a variety <strong>of</strong> factors,<br />
including grief and exposure to patient suffering. 19-21<br />
Possible Benefits<br />
Health care pr<strong>of</strong>essionals can learn much from the patients<br />
who face death and their loved ones, if they are open<br />
to listening. Despite the personal toll that results from<br />
encountering grief on a regular basis, physicians and other<br />
health care workers may achieve a pr<strong>of</strong>ound fulfillment in<br />
the knowledge that their contribution helped ease suffering.<br />
This may favorably affect job satisfaction. 22,23 From a sociologic<br />
perspective, there is evidence to suggest that giving<br />
support may be more favorable to one’s health and wellbeing<br />
than receiving support. 24<br />
The bereaved may also benefit from engagement <strong>of</strong> the<br />
physician. In a study <strong>of</strong> patients with terminal illness and<br />
considerable care needs, their caregivers conveyed the importance<br />
<strong>of</strong> the interaction with the physician. 4 If the physician<br />
listened to them about the patient’s illness, they were<br />
less likely to feel depressed. Bereavement outcomes may<br />
also be affected by physician communication. 25<br />
Facing Grief: What to Do<br />
The Perspective <strong>of</strong> the Oncologist and Medical Team<br />
As Mount succinctly noted, “To practice oncology . . . is to<br />
augment our losses many-fold, for they are an integral part<br />
<strong>of</strong> our pr<strong>of</strong>essional existence.” 19 Grief, therefore, is a normal<br />
part <strong>of</strong> an oncologist’s life. The overall perspective <strong>of</strong> the<br />
oncologist toward these losses may be a crucial component <strong>of</strong><br />
the ability to cope with grief. If the oncologist views a cure or<br />
prolongation <strong>of</strong> life as the only significant goals <strong>of</strong> patient<br />
care, then he or she may miss a myriad <strong>of</strong> opportunities to<br />
alleviate suffering from loss and to achieve greater insight<br />
into the meaning <strong>of</strong> life. With this viewpoint, the physician<br />
may particularly struggle when dealing with grief.<br />
It is not easy to be reminded <strong>of</strong> one’s mortality on a<br />
frequent basis. It is not easy to grieve or to watch another<br />
grieve. Yet these experiences have the potential to ultimately<br />
inspire and lead to a broader understanding <strong>of</strong> the<br />
consequences <strong>of</strong> illness. A major challenge is to apply a<br />
realistic approach to attain this idealistic ambition.<br />
Realistic Suggestions to Cope with Grief<br />
Grief is a normal and universal reaction to loss; engaging<br />
one’s grief may be far more beneficial than avoiding it.<br />
There are numerous strategies that can be used to work<br />
through stress and grief. Several authors suggest that it<br />
is crucial for the physician to be aware <strong>of</strong> his or her<br />
emotions. 13,19 In addition, it is important to set reasonable<br />
work goals, enhance time management skills, and limit<br />
personal involvement with patients. 19,20 Integration <strong>of</strong> other<br />
medical team members (i.e., nurses, chaplains, social workers)<br />
may help to optimize care <strong>of</strong> the whole patient and<br />
alleviate some <strong>of</strong> the emotional burden for the individual<br />
physician and other medical caregivers. It may be beneficial<br />
to foster a supportive work environment that encourages<br />
discussion <strong>of</strong> these issues in a group or individual setting,<br />
such as “Death Rounds”. 26<br />
On a more personal level, a variety <strong>of</strong> suggestions for<br />
coping include habitual exercise, optimal rest, interests<br />
outside <strong>of</strong> work, writing one’s thoughts, reading other’s ideas<br />
about these issues, laughter, and a focus on inner growth,<br />
such as mindful meditation and expansion <strong>of</strong> one’s view <strong>of</strong><br />
life to reflect a more global/philosophic perspective. 20,23 In<br />
particular, the field <strong>of</strong> narrative medicine employs reflective<br />
writing to explore the physician’s relationship with self,<br />
patients, colleagues, and society. 27 Physicians may also turn<br />
to their own spiritual beliefs for support.<br />
A classic symbol <strong>of</strong> grief is tears. In response to sadness,<br />
does a physician crying indicate weakness? There may be<br />
various responses to this question. 28 However, it may be<br />
completely appropriate to cry for a patient or the loss <strong>of</strong> a<br />
patient. In fact, it may be viewed by the patient or family as<br />
a sign <strong>of</strong> compassion. 29<br />
Bereavement Care by the Physicians<br />
Bereavement care initiated by the physician may be very<br />
meaningful to patient’s families. 30 In addition, it may improve<br />
physician well-being and job satisfaction, although<br />
further research is necessary to definitively confirm an<br />
association. 22<br />
Bedell et al noted that in the 1800s, it was common for<br />
physicians to send a letter <strong>of</strong> condolence to a patient’s<br />
family. 31 However, with the changing times, this practice<br />
has diminished. Yet, for the bereaved, the words <strong>of</strong> the<br />
physician may provide comfort and exemplify a more humanistic<br />
approach to patient care. The authors suggest that<br />
the letter include an articulation <strong>of</strong> the physician’s sympathy,<br />
a personal reference to the patient, and an acknowledgment<br />
<strong>of</strong> the family’s devotion to the patient. Obviously, only<br />
sincere thoughts and emotions should be included.<br />
A recent Canadian survey <strong>of</strong> medical oncologists, radiation<br />
oncologists, and palliative care physicians focused on<br />
bereavement practices. 32 The authors evaluated how <strong>of</strong>ten<br />
the 535 physicians practiced one <strong>of</strong> the following: making a<br />
e83
telephone call to the family, sending a sympathy card, or<br />
attending a funeral. Approximately 33.3% (95% CI 29.3–<br />
37.4%) reported usually or always, 30.5% (95% CI 26.5–<br />
34.4%) reported sometimes, and 36.2% (95% CI 32.1–40.3%)<br />
reported rarely or never. The most common practice was a<br />
phone call whereas attending a funeral was rare. A small<br />
number <strong>of</strong> physicians, approximately 9% (95% CI 6.6–<br />
11.5%), felt that sending a condolence note was the responsibility<br />
<strong>of</strong> the physician.<br />
The most common identified barrier to contacting the<br />
patient’s family was lack <strong>of</strong> time and resources, although<br />
some felt uncomfortable and were not certain whom to<br />
contact. Multivariate analysis revealed that more frequent<br />
bereavement practice was associated with female sex, working<br />
in an academic center, conviction that physicians have<br />
a responsibility to write a condolence card, and absence <strong>of</strong> a<br />
Author’s Disclosure <strong>of</strong> Potential Conflicts <strong>of</strong> Interest<br />
Author<br />
Teresa Gilewski*<br />
*No relevant relationships to disclose.<br />
Employment or<br />
Leadership<br />
Positions<br />
Consultant or<br />
Advisory Role<br />
1. Dickinson, E. The Complete Poems <strong>of</strong> Emily Dickinson. Boston: Little<br />
Brown; 1924.<br />
2. Merriam-Webster. www.merriam-webster.com. Accessed February 6,<br />
<strong>2012</strong>.<br />
3. Emanuel LL, Ferris FD, von Gunten DF, Von Roenn J. Education in<br />
Palliative and End-<strong>of</strong>-Life Care - <strong>Oncology</strong> (EPEC-O): Participant’s Handbook.<br />
Module 4: Loss, Grief and Bereavement. EPEC Project. Chicago, IL:<br />
2005;M4-1-M4-24.<br />
4. Emmanuel EJ, Fairclough DL, Slutsman J, et al. Understanding economic<br />
and other burdens <strong>of</strong> terminal illness: The experience <strong>of</strong> patients and<br />
their caregivers. Ann Inter Med. 2000;132:451-459.<br />
5. Casarett D, Kutner JS, Abrahm J, et al. Life after death: A practical<br />
approach to grief and bereavement. Ann Intern Med. 2001;134:208-215.<br />
6. Wo<strong>of</strong> WR, Carter YH. The grieving adult and the general practitioner:<br />
A literature review in two parts (Part 1). Br J Gen Pract. 1997;47:443-448.<br />
7. Kübler-Ross E. On Death and Dying. New York, NY, Macmillan, 1969.<br />
8. Macicjewski PK, Zhang B, Block SD, et al. An empirical examination <strong>of</strong><br />
the stage theory <strong>of</strong> grief. JAMA. 2007;297:716-723.<br />
9. Tomarken A, Holland J, Schachter S, et al. Factors <strong>of</strong> complicated grief<br />
pre-death in caregivers <strong>of</strong> cancer patients. Psychooncology. 2008;17:105-111.<br />
10. Holland JC. Management <strong>of</strong> grief and loss: Medicine’s obligation and<br />
challenge. J Am Med Womens Assoc. 2002;57:95-96.<br />
11. Gorlin R, Zucker HD. Physicians’ reactions to patients: A key to<br />
teaching humanistic medicine. N Engl J Med. 1983;308:1059-1063.<br />
12. Novack DH, Suchman AL, Clark W, et al. Calibrating the physician:<br />
Personal awareness and effective patient care. JAMA. 1997;278:502-509.<br />
13. Meier DE, Back AL, Morrison RS. The inner life <strong>of</strong> physicians and care<br />
<strong>of</strong> the seriously ill. JAMA. 2001;286:3007-3014.<br />
14. Redinbaugh EM, Sullivan AM, Block SD, et al. Doctors’ emotional<br />
reactions to recent death <strong>of</strong> a patient: Cross sectional study <strong>of</strong> hospital<br />
doctors. BMJ. 2003;327:185-189.<br />
15. Moores TS, Castle KL, Shaw KL, et al. ‘Memorable patient deaths’:<br />
reactions <strong>of</strong> hospital doctors and their need for support. Med Edu. 2007;41:<br />
942-946.<br />
16. Maslach C, Schaufeli WB, Leiter MP. Job burnout. Annu Rev Psychol.<br />
2001;52:397-422.<br />
e84<br />
palliative care program, with the most significant being a<br />
palliative care specialty.<br />
For some patient’s families, it may also be meaningful to<br />
visit with them in person. Physicians may not recognize the<br />
importance <strong>of</strong> these bereavement practices because <strong>of</strong> limited<br />
feedback from the patient’s loved ones.<br />
Conclusion<br />
Grief is a normal response to loss and a fundamental<br />
component <strong>of</strong> the field <strong>of</strong> oncology. Health care workers in<br />
oncology face the challenge <strong>of</strong> repeated exposure to serious<br />
illness and death. Enhanced acknowledgment and recognition<br />
<strong>of</strong> the indicators <strong>of</strong> grief and reasons for grief are<br />
essential to providing sustainable compassionate care. Engaging<br />
one’s grief may result in considerable insight into the<br />
essence <strong>of</strong> patient care and the soul <strong>of</strong> medicine.<br />
Stock<br />
Ownership Honoraria<br />
REFERENCES<br />
Research<br />
Funding<br />
Expert<br />
Testimony<br />
TERESA GILEWSKI<br />
Other<br />
Remuneration<br />
17. Whippen DA, Canellos GP. Burnout syndrome in the practice <strong>of</strong><br />
oncology: Results <strong>of</strong> a random survey <strong>of</strong> 1,000 oncologists. J Clin Oncol.<br />
1991;9:1916-1921.<br />
18. Kash KM, Holland JC, Breitbart W, et al. Stress and burnout in<br />
oncology. <strong>Oncology</strong>. 2000;14:1621-1633.<br />
19. Mount BM. Dealing with our losses. J Clin Oncol. 1986;4:1127-1134.<br />
20. Lyckholm L. Dealing with stress, burnout, and grief in the practice <strong>of</strong><br />
oncology. Lancet Oncol. 2001;2:750-755.<br />
21. Ramirez AJ, Graham J, Richards MA, et al. Mental health <strong>of</strong> hospital<br />
consultants: The effects <strong>of</strong> stress and satisfaction at work. Lancet. 1996;347:<br />
724-728.<br />
22. Block SD. Psychological considerations, growth, and transcendence at<br />
the end <strong>of</strong> life: the art <strong>of</strong> the possible. JAMA. 2001;285:2898-2905.<br />
23. Kearney MK, Weininger RB, Vachon MLS, et al. Self-care <strong>of</strong> physicians<br />
caring for patients at the end <strong>of</strong> life: “Being connected . . . a key to my<br />
survival”. JAMA. 2009;301:1155-1164.<br />
24. Brown SL, Nesse RM, Vinokur AD, et al. Providing social support may<br />
be more beneficial than receiving it: Results from a prospective study <strong>of</strong><br />
mortality. Psychol Sci. 2003;14:320-327.<br />
25. Prigerson HG, Jacobs SC. Caring for bereaved patients: “All the<br />
Doctors Just Suddenly Go”. JAMA. 2001;286:1369-1376.<br />
26. Smith L, Hough CL. Using death rounds to improve end-<strong>of</strong>-life education<br />
for internal medicine residents. J Palliative Medicine. 2011;14:55–58.<br />
27. Charon R. Narrative medicine: a model for empathy, reflection, pr<strong>of</strong>ession,<br />
and trust. JAMA. 2001;286:1897-1902.<br />
28. Siegel B, Schultz S. Crying in stairwells: How should we grieve for<br />
dying patients? JAMA. 1994;272:659.<br />
29. Gilewski T. The subtle power <strong>of</strong> compassion. JAMA. 2001;286:3052-<br />
3053.<br />
30. Penson RT, Green KM, Chabner BA, et al. When does the responsibility<br />
<strong>of</strong> our care end: Bereavement. Oncologist. 2002;7:251-258.<br />
31. Bedell SE, Cadenhead K, Graboys TB. The doctor’s letter <strong>of</strong> condolence.<br />
N Engl J Med. 2001;344:1162-1164.<br />
32. Chau NG, Zimmermann C, Ma C, et al. Bereavement practices <strong>of</strong><br />
physicians in oncology and palliative care. Arch Intern Med. 2009;169:963-<br />
971.
TALKING IN SYNC WITH PATIENTS ACROSS<br />
CULTURAL BARRIERS<br />
CHAIR<br />
Lidia Schapira, MD<br />
Massachusetts General Hospital<br />
Boston, MA<br />
SPEAKERS<br />
William J. Hicks, MD<br />
The Ohio State University<br />
Columbus, OH<br />
Louise Villejo, MPH<br />
University <strong>of</strong> Texas M. D. Anderson Cancer Center<br />
Houston, TX
New Insights in Cross-Cultural Communication<br />
Overview: Improving clinician-patient communication, improving<br />
clinical decision making, and eliminating mistrust<br />
have been identified as three key areas for reducing disparities<br />
in care. An important step is the training <strong>of</strong> cancer<br />
pr<strong>of</strong>essionals to deliver culturally competent care in clinical<br />
settings as well as increasing the proportion <strong>of</strong> underrepresented<br />
minorities in the health care workforce. Providing care<br />
that is attuned to the patient’s cultural preferences begins by<br />
talking to the patient about his or her cultural history and<br />
identifying the locus <strong>of</strong> decision making, preferences for<br />
disclosure <strong>of</strong> vital health information, and goals <strong>of</strong> care.<br />
Patients with low literacy and those with poor fluency <strong>of</strong> the<br />
THERE IS growing interest in providing culturally sensitive<br />
care, not only in the United States but also across<br />
the world. Awareness <strong>of</strong> the global repercussions <strong>of</strong> cancer<br />
and the enormous burden <strong>of</strong> suffering imposed mainly on<br />
those with little access to health care and few resources has<br />
kindled the imagination and determination <strong>of</strong> cancer physicians<br />
and led to new global initiatives to close the gap<br />
between low- and middle-income countries and those that<br />
are more affluent. 1 Although once considered a problem<br />
exclusive to industrialized and wealthy countries, cancer is a<br />
leading cause <strong>of</strong> disability and death in the developing<br />
world. 1 Multinational and interdisciplinary efforts with support<br />
from pr<strong>of</strong>essional organizations and industry are beginning<br />
to target global issues through innovative mechanisms<br />
that involve raising awareness and promoting universal<br />
access to cancer prevention, detection, and care. 1 And within<br />
countries, governments and nongovernmental organizations<br />
need to meet the challenge <strong>of</strong> expanding access for those<br />
who, for complex socioeconomic factors ranging from poverty<br />
to low health literacy, are deprived <strong>of</strong> access to optimal<br />
health services. Without such efforts, those with few resources<br />
will bear a disproportionate burden <strong>of</strong> misery from<br />
cancer.<br />
Many factors contribute to the unequal distribution and<br />
utilization <strong>of</strong> health services, and this, in turn, leads to<br />
differences in outcomes. A report commissioned by the U.S.<br />
Congress and published by the Institute <strong>of</strong> Medicine in 2002<br />
confirmed that members <strong>of</strong> groups identified as belonging to<br />
ethnic and racial minorities have worse health outcomes as<br />
a result <strong>of</strong> unequal treatment. 2 The report stressed the need<br />
for change and identified opportunities for interventions.<br />
Three key areas were identified as having clinical relevance<br />
and contributing to disparities in health care: providerpatient<br />
communication, clinical decision making, and mistrust.<br />
3 Recommendations for change addressed many<br />
aspects <strong>of</strong> both care and health services delivery, and three<br />
are worth highlighting: integrating cross-cultural education<br />
into the training <strong>of</strong> all health care pr<strong>of</strong>essionals, supporting<br />
the use <strong>of</strong> language interpretation services in clinical settings,<br />
and increasing the proportion <strong>of</strong> underrepresented<br />
minorities in the health care workforce. 2,3 Other factors that<br />
were identified as contributors to unequal outcomes are<br />
social determinants such as lower levels <strong>of</strong> education, overall<br />
lower socioeconomic status, unsafe housing, and exposure<br />
to environmental hazards. The report also stated that<br />
bias, prejudice, and stereotyping on the part <strong>of</strong> health care<br />
By Lidia Schapira, MD<br />
dominant language require additional services. Language interpretation<br />
by trained pr<strong>of</strong>essionals is fundamental to ensure<br />
that patients are able to provide informed consent for treatment.<br />
A working definition <strong>of</strong> culture involves multiple dimensions<br />
and levels and must be viewed as both dynamic and<br />
adaptive, rather than simply as a collection <strong>of</strong> beliefs and<br />
values. Effective cross-cultural education avoids stereotyping<br />
and promotes communication and negotiation to solve problems<br />
and minimize tension and conflict. Recent research has<br />
identified that unconscious biases held by clinicians affect<br />
their behavior and recommendations for treatment.<br />
pr<strong>of</strong>essionals exert important influences in their recommendations<br />
and treatment. 3 The inherent power asymmetry in<br />
the clinician-patient relationship coupled with unconscious<br />
biases can lead to differential treatment <strong>of</strong> those seeking<br />
care.<br />
Communication is a complex phenomenon that serves to<br />
transmit messages and construct and maintain relationships.<br />
It is one <strong>of</strong> the pillars <strong>of</strong> medicine and plays a<br />
fundamental role in the delivery <strong>of</strong> health care services. Now<br />
a focus <strong>of</strong> research in social sciences and medicine, it has<br />
gained the status <strong>of</strong> a necessary competency for medical<br />
trainees in the United States. Although there is consensus<br />
among experts that a basic skill set can be taught and<br />
evaluated, there is still no consistent way to address individual<br />
beliefs or deeply held biases that undoubtedly influence<br />
practice. Simply put, we can transmit and evaluate<br />
knowledge about communication, construct ethical frameworks,<br />
and even model ideal behaviors, but we cannot<br />
control or even modify deeply held beliefs that affect behavior.<br />
Evidence suggests that individual doctors’ responses tend<br />
to be similar with all patients and that moral judgments and<br />
decision making are driven by automatic emotional<br />
responses. 4-8 Rather than resorting to abstract reasoning,<br />
doctors and nurses respond to moral dilemmas with gut<br />
reactions. 5-8 Recognizing this fact, we need to address these<br />
instant and unconscious reactions, which most likely result<br />
from intergenerational transmission in families and communities,<br />
by designing courses that address cross-cultural communication.<br />
In a study <strong>of</strong> internal medicine and emergency<br />
medicine residents at four academic medical centers in<br />
Atlanta, GA, and Boston, MA, Green and colleagues found<br />
evidence <strong>of</strong> unconscious bias. 9 Most physicians in their<br />
survey did not admit to any racial biases explicitly. However,<br />
using validated tools that measure implicit measures,<br />
there were clear differences favoring one group over another,<br />
which influenced the physicians’ recommendations for<br />
treatment in a simulated scenario. 9 It seems plausible that<br />
From the Massachusetts General Hospital, Boston, MA.<br />
Author’s disclosures <strong>of</strong> potential conflicts <strong>of</strong> interest are found at the end <strong>of</strong> this article.<br />
Address reprint requests to Lidia Schapira, MD, Massachusetts General Hospital, 55<br />
Fruit Street, Boston, MA 02114; email: lschapira@partners.org.<br />
© <strong>2012</strong> by <strong>American</strong> <strong>Society</strong> <strong>of</strong> <strong>Clinical</strong> <strong>Oncology</strong>.<br />
1092-9118/10/1-10<br />
e85
unconscious likes and dislikes influence our daily practice.<br />
Feelings <strong>of</strong> disgust or kinship may affect our judgment and<br />
recommendations.<br />
Culture is a multilevel, multidimensional, dynamic, and<br />
adaptive system and not merely a collection <strong>of</strong> beliefs and<br />
values. 10 Kagawa-Singer reminds us that efforts to apply<br />
the concept <strong>of</strong> culture accurately and usefully in medical<br />
practice require an assessment at many different levels<br />
and across time in order to capture a composite <strong>of</strong> the<br />
lived world <strong>of</strong> the patient and family. Culture is a layered<br />
concept that is shaped by several components such as<br />
environment, economy, technology, religion and worldview,<br />
language, social structure, beliefs, and values. 10 From a<br />
purely practical perspective, we can assert that culture<br />
affects the entire way in which cancer is framed in meaning<br />
and response. 10 Culture affects the experience and expression<br />
<strong>of</strong> pain, the locus <strong>of</strong> decision making, the etiquette <strong>of</strong><br />
communication with clinicians, the preference for information<br />
disclosure and participation in decision making, the<br />
choices for end-<strong>of</strong>-life care, and expectations. 10-12 An appreciation<br />
<strong>of</strong> the role and influence <strong>of</strong> culture is fundamental in<br />
any model <strong>of</strong> care that is based on respect for individual<br />
patients.<br />
Societies differ in how they differentiate cultural groups<br />
and how they respond to those who are perceived to be<br />
different or not fully assimilated into the dominant culture.<br />
10 Some place great emphasis on religious differences or<br />
nationality. In the United States, <strong>of</strong>ten cultural differences<br />
are established by the proxies <strong>of</strong> racial and ethnic groups.<br />
The Office <strong>of</strong> Management and Budget Directive codified<br />
these categories but acknowledged that they are socialpolitical<br />
categories. 10 Just think about the broad category <strong>of</strong><br />
non-Hispanic whites, which lumps together immigrants<br />
from Iraq and Israel, from Italy and Russia. What sense can<br />
we possibly make <strong>of</strong> statistics that aggregate these groups<br />
that represent communities with vastly different historic<br />
traditions, cultures, and beliefs? And yet they have been<br />
used extensively in biomedical research and continue to be<br />
used in medicine.<br />
In a thoughtful essay that reviews the state <strong>of</strong> knowledge<br />
and thinking about race, ethnicity, and genetics, Francis<br />
Collins points out that “race” and “ethnicity” are terms<br />
without generally agreed-on definitions. 13 Both carry complex<br />
connotations that reflect culture, history, socioeconom-<br />
KEY POINTS<br />
● Integrating cross-cultural education into the training<br />
<strong>of</strong> cancer pr<strong>of</strong>essionals is crucial to reduce disparities.<br />
● Culture affects the way in which cancer is framed and<br />
affects the experience <strong>of</strong> patients and families.<br />
● Cross-cultural communication begins with selfawareness<br />
and respect for a patient’s worldview,<br />
beliefs, and values.<br />
● By talking with a patient about his or her cultural<br />
history we can establish that patient’s preference for<br />
decision making and desire for information.<br />
● Care that is culturally sensitive requires excellent<br />
communication and negotiation skills and expert consultation<br />
to mediate conflict.<br />
e86<br />
LIDIA SCHAPIRA<br />
ics, and political status, as well as a variably important<br />
connection to ancestral geographic origins. Those ancestral<br />
origins have at best a hazy connection to current issues <strong>of</strong><br />
health disparities, although they may well account for the<br />
unequal distribution <strong>of</strong> disease-associated alleles for certain<br />
recessive disorders or susceptibility to some cancers. 13 The<br />
debate intensifies because race and ethnicity in the view <strong>of</strong><br />
many are self-reported constructs and thus the relationship<br />
to disease or genetic risk is at best imperfect. Collins <strong>of</strong>fers<br />
sound and conciliatory advice: “Without discounting selfidentified<br />
race or ethnicity as a variable correlated with<br />
health, we must strive to move beyond these weak surrogate<br />
relationships and get to the root causes <strong>of</strong> health and<br />
disease, be they genetic, environmental, or both.” 13 We can<br />
appreciate and respect the interdisciplinary inquiry and<br />
research into race and ethnicity and move past this debate<br />
in order to focus on interventions that have the potential <strong>of</strong><br />
reducing barriers and enhancing the experience <strong>of</strong> patients<br />
considered “underserved.”<br />
Cross-cultural communication begins with awareness <strong>of</strong><br />
our own moral code and beliefs. As long as medical care is<br />
delivered by humans rather than robots or computers,<br />
communication models need to recognize that health care<br />
pr<strong>of</strong>essionals are not interchangeable figurines. A patient’s<br />
experience <strong>of</strong> care is a composite <strong>of</strong> many different meetings<br />
influenced by emotion and mediated by individual<br />
communication styles. From the patient’s perspective, the<br />
medical culture itself is challenging to decipher! We may<br />
take for granted a shared respect for the scientific<br />
method and for recommendations based on evidence and are<br />
accustomed to pr<strong>of</strong>essional hierarchies that vary between<br />
specialties and settings. Our patients, however, even those<br />
with superb education and intellectual abilities, are sometimes<br />
surprised by our protocols, rituals, guidelines, and<br />
practices.<br />
Cultural competence begins with respect, curiosity, humility,<br />
and empathy. The ideal mindset is that <strong>of</strong> a first grader,<br />
open to wonder. Specific knowledge about cultures is very<br />
valuable but <strong>of</strong>ten unavailable, especially at short notice.<br />
We must rely on learned communication skills to create a<br />
respectful environment where information is exchanged and<br />
dialogue is used to negotiate solutions that serve the patient<br />
well while minimizing conflict, frustration, and tension. In<br />
situations where cultures clash, the best strategy is to<br />
involve cultural brokers who can mediate and help clarify<br />
goals and expectations. Thus doctors need more than a skill<br />
set and ethical framework, they also need inspiration and<br />
imagination to respond to calamity and support patients in<br />
difficult situations so that the emphasis is on what really<br />
matters to the patient. 14<br />
Providing care that is attuned to the patient’s cultural<br />
preferences begins by talking to a patient about his or her<br />
cultural history. This entails eliciting the patient’s understanding<br />
<strong>of</strong> his or her illness and its causation and course,<br />
his or her preference for involvement in decision making,<br />
and his or her need for varying amounts <strong>of</strong> information.<br />
Through this process one can also identify the locus <strong>of</strong><br />
decision making, which <strong>of</strong>ten involves other family members.<br />
By showing respect and interest and by providing a<br />
comforting presence, clinicians can build trust in complicated<br />
relationships. Without trust, we remain limited in our<br />
ability to provide guidance in times <strong>of</strong> crisis. Consider for<br />
example that preferences for end-<strong>of</strong>-life care are <strong>of</strong>ten dic-
CROSS-CULTURAL COMMUNICATION<br />
Sidebar. Tips for Practice to Overcome Language<br />
and Literacy Barriers<br />
To overcome language barriers, obtain support from<br />
colleagues or institutional leadership to provide<br />
pr<strong>of</strong>essional interpreter services. There are national<br />
standards for interpreters, but the quality <strong>of</strong><br />
services and depth <strong>of</strong> training are quite variable.<br />
You may want to consider <strong>of</strong>fering training for<br />
interpreters to familiarize them with terms <strong>of</strong>ten<br />
used in complex oncology consultations and in clinical<br />
research. Remember that using minors for<br />
interpretation is illegal and that family members<br />
<strong>of</strong>ten distort the content <strong>of</strong> conversations. In the<br />
United States, health care organizations are obligated<br />
by law to provide language assistance services<br />
either in person or via remote interpretation<br />
services at no cost to patients with limited English<br />
pr<strong>of</strong>iciency. 22-24<br />
To overcome literacy barriers, check the patient’s<br />
reading level and explore any possible learning<br />
disabilities that could interfere with proper use <strong>of</strong><br />
medication or compromise the ability to discuss<br />
options for treatment. About one-half <strong>of</strong> <strong>American</strong>s<br />
are considered to have limited health literacy,<br />
which affects their understanding <strong>of</strong> basic medical<br />
terms, their ability to follow directions for diagnostic<br />
procedures and therapies, give their consent for<br />
research, and engage in a real dialogue about treatment<br />
options.<br />
tated by cultural norms and influenced by beliefs in an<br />
afterlife. Knowing and understanding these issues in advance<br />
allow physicians to provide direction as death<br />
approaches. Trust has been described as an iterative process,<br />
requiring steadiness and honesty. 15 In fact, distrust is<br />
<strong>of</strong>ten the cause <strong>of</strong> misunderstandings or miscommunication<br />
and may contribute to disparate outcomes. 15 Breaches <strong>of</strong><br />
trust committed decades ago by medical researchers in the<br />
United States still haunt relationships with African Ameri-<br />
Author’s Disclosures <strong>of</strong> Potential Conflicts <strong>of</strong> Interest<br />
Author<br />
Lidia Schapira*<br />
*No relevant relationships to disclose.<br />
Employment or<br />
Leadership<br />
Positions<br />
Consultant or<br />
Advisory Role<br />
1. Farmer P, Frenk J, Knaul FM, et al. Expansion <strong>of</strong> cancer care and<br />
control in low-income and middle-income countries: a call to action. Lancet.<br />
2010;376:9747.<br />
2. Smedley BD, Stith AY, Nelson AR. Unequal Treatment: Confronting<br />
Ethnic and Racial Disparities in Health Care. Washington, DC: Institute <strong>of</strong><br />
Medicine; 2002.<br />
3. Betancourt JR, Renfrew MR. Unequal treatment in the US: lessons and<br />
recommendations for cancer care internationally. J Pediatr Hematol Oncol.<br />
2011;33:S149-S153 (suppl 2).<br />
4. Fallowfield L. Truth sometimes hurts but deceit hurts more. AnnNY<br />
Acad Sci. 1997;809:525-536.<br />
5. Greene JD, Cushman FA, Stewart LE, et al. Pushing moral buttons: the<br />
cans patients, who may harbor suspicions about the integrity<br />
<strong>of</strong> research and refuse to participate in clinical trials. 16<br />
Similar doubts about the motives <strong>of</strong> researchers are also<br />
expressed by patients who have experienced discrimination<br />
or harbor concerns regarding the privacy <strong>of</strong> personal health<br />
information.<br />
There has been considerable interest among philosophers,<br />
psychologists, anthropologists, and ethicists in studying<br />
truth-telling and disclosure <strong>of</strong> both diagnosis and prognosis.<br />
In the United States and many Western countries, patient<br />
autonomy and involvement in medical decision making<br />
remains the key driver for full disclosure <strong>of</strong> health information.<br />
Autonomy trumps other ethical and social concerns. 17<br />
Research and practice have shown that most patients can<br />
cope with grim information and, with guidance and support,<br />
come to a resolution <strong>of</strong> their emotional pain. 18-20<br />
Physicians vary in their level <strong>of</strong> comfort with such disclosure,<br />
and despite the increasing availability <strong>of</strong> communication<br />
skills training, practices are <strong>of</strong>ten informed by<br />
instinct and shades <strong>of</strong> paternalism. 12 The same is true <strong>of</strong><br />
patients and family members, who <strong>of</strong>ten keep important<br />
information from each other. Negotiating these fundamentally<br />
private issues remains an important task for<br />
clinicians.<br />
At Harvard Medical School, instruction in cross-cultural<br />
communication begins in the first week <strong>of</strong> the first year. In<br />
small groups, students discuss their reactions to film clips<br />
and texts chosen to highlight the plight <strong>of</strong> new immigrants<br />
and the challenges <strong>of</strong> delivering care in a pluralistic society.<br />
Using case-based learning and with expert facilitation, they<br />
are guided to articulate and identify solutions for complex<br />
communication problems. Courses are designed and delivered<br />
by multidisciplinary faculty to drive home the important<br />
message that the lessons learned are fundamental for<br />
good practice and broadly applicable. Didactic sessions follow<br />
Kleinman’s explanatory model <strong>of</strong> illness, which stimulates<br />
inquiry and patient-centeredness and avoids<br />
stereotyping. 21 Lessons in cross-cultural communication are<br />
later inserted into other course materials to reinforce the<br />
message that good clinical care is based on understanding<br />
what is fundamentally important to each patient.<br />
Stock<br />
Ownership Honoraria<br />
REFERENCES<br />
Research<br />
Funding<br />
Expert<br />
Testimony<br />
Other<br />
Remuneration<br />
interaction between personal force and intention in moral judgment. Cognition.<br />
2009;111:364-371.<br />
6. Greene JD, Morelli SA, Lowenberg K, et al. Cognitive load selectively<br />
interferes with utilitarian moral judgment. Cognition. 2008;107:1144-1154.<br />
7. Greene J. From neural ‘is’ to moral ‘ought’: what are the moral<br />
implications <strong>of</strong> neuroscientific moral psychology? Nat Rev Neurosci. 2003;<br />
4:846-849.<br />
8. Greene J, Haidt J. How (and where) does moral judgment work?<br />
Trends Cogn Sci. 2002;6:517-523.<br />
9. Green AR, Carney DR, Pallin DJ, et al. Implicit bias among physicians<br />
and its prediction <strong>of</strong> thrombolysis decisions for black and white<br />
patients. J Gen Intern Med. 2007;22:1231-1238.<br />
e87
10. Kagawa-Singer M. Impact <strong>of</strong> culture on health outcomes. J Pediatr<br />
Hematol Oncol. 2011;33:S90-S95.<br />
11. Surbone A. Telling the truth to patients with cancer: what is the<br />
truth? Lancet Oncol. 2006;7:944-950.<br />
12. Cherny NI. Controversies in oncologist-patient communication: a<br />
nuanced approach to autonomy, culture, and paternalism. <strong>Oncology</strong>. <strong>2012</strong>;<br />
26:37-46.<br />
13. Collins F. What we do and don’t know about ‘race’, ‘ethnicity’, genetics<br />
and health at the dawn <strong>of</strong> the genome era. Nat Genet. 2004;36:S13-S15.<br />
14. Kleinman A. What Really Matters. New York: Oxford University Press,<br />
2006.<br />
15. Corbie-Smith G, Thomas SB, St George DM. Distrust, race, and<br />
research. Arch Intern Med. 2002;162:2458-2463.<br />
16. Corbie-Smith G, Thomas SB, Williams MV, et al. Attitudes and beliefs<br />
<strong>of</strong> African <strong>American</strong>s toward participation in medical research. J Gen Intern<br />
Med. 1999;14:537-546.<br />
17. Schneider C. The Practice <strong>of</strong> Autonomy: Patients, Doctors, and Medical<br />
Decisions. New York: Oxford University Press, 1998.<br />
e88<br />
LIDIA SCHAPIRA<br />
18. Parkes CM. Psychological aspects. In Saunders CM (ed). The Management<br />
<strong>of</strong> Terminal Disease. London: Edward Arnold, 1978;44-64.<br />
19. Smith TJ, Low LA, Virago E, et al. Giving honest information to<br />
patients with advanced cancer maintains hope. <strong>Oncology</strong> (Williston Park).<br />
2010;24:521-525.<br />
20. Mack JW, Wolfe J, Cok ER, et al. Hope and prognostic disclosure. J Clin<br />
Oncol. 2007;25:5636-5642.<br />
21. Kleinman A, Eisenberg L, Good B. Culture, illness, and care: clinical<br />
lessons from anthropologic and cross-cultural research. Ann Intern Med.<br />
1978;88:251-258.<br />
22. National Standards on Culturally and Linguistically Appropriate Standards.<br />
http://minorityhealth.hhs.gov/templates/browse.aspx?lvl�2&lvl1D�<br />
15. Accessed March 15, <strong>2012</strong>.<br />
23. Butow PN, Goldstein D, Bell ML, et al. Interpretation in consultations<br />
with immigrant patients with cancer: how accurate is it? J Clin Oncol.<br />
2011;29:2801-2807.<br />
24. Donelan K, Hobrecker K, Schapira L, et al. Medical interpreter knowledge<br />
<strong>of</strong> cancer and cancer clinical trials. Cancer. 2009;115:3283-3292.
THE ONCOLOGIST, THE PATIENT, AND THE MEDIA<br />
CHAIR<br />
Paul R. Helft, MD<br />
Indiana University Simon Cancer Center<br />
Indianapolis, IN<br />
SPEAKERS<br />
Diane Blum, MSW<br />
Lymphoma Research Foundation<br />
New York, NY<br />
Jeremy Manier<br />
The University <strong>of</strong> Chicago<br />
Chicago, IL
Patients with Cancer, Internet Information,<br />
and the <strong>Clinical</strong> Encounter: A Taxonomy <strong>of</strong><br />
Patient Users<br />
Overview: The Internet has changed all <strong>of</strong> our lives forever<br />
and has certainly changed the way in which patients acquire<br />
information, share their stories, find others in similar circumstances,<br />
and analyze their medical situations. It is very clear<br />
that patients have widely adopted the use <strong>of</strong> online resources<br />
in the face <strong>of</strong> illness. Access to unfiltered information online<br />
clearly has positive and negative potential effects, and the<br />
introduction <strong>of</strong> Internet information into the physician-patient<br />
IT IS difficult to imagine life without the Internet. Not<br />
only would the practice <strong>of</strong> medicine be drastically different<br />
from the way it is today, but the effect <strong>of</strong> the Internet on<br />
all aspects <strong>of</strong> our lives has been so protean as to represent a<br />
true revolution. From the early days <strong>of</strong> rapid uptake <strong>of</strong> the<br />
Internet among U.S. adults in the 1990s, use <strong>of</strong> the Internet<br />
to obtain health information is one <strong>of</strong> the most common<br />
tasks for Internet users.<br />
Early reports on the use <strong>of</strong> the Internet among both<br />
patients <strong>of</strong> multiple specialtists as well as people with<br />
cancer showed that patients from disadvantaged backgrounds<br />
used the Internet to obtain health information less<br />
<strong>of</strong>ten than patients with higher literacy and socioeconomic<br />
status. For example, in a study our group conducted in the<br />
early 2000s among disadvantaged patients at an urban<br />
public hospital oncology clinic, we found that only 10% <strong>of</strong><br />
patients with cancer reported looking up cancer information<br />
on the Internet. 1 More recent studies have suggested that<br />
the proportion <strong>of</strong> patients with cancer who use the Internet<br />
to obtain information about their diseases is closer to twothirds.<br />
2<br />
National data from representative samples <strong>of</strong> U.S. adults<br />
aimed at documenting trends in Internet use have been<br />
continuously collected by the Pew Internet and <strong>American</strong><br />
Life Project for more than a decade. In 2011, the Pew survey<br />
found that nearly 80% <strong>of</strong> all adults now use the Internet.<br />
Among them, 83% reported using the Internet to look up<br />
health information. Important trends documented by this<br />
survey have included a shrinking divide between black,<br />
white, and Hispanic patients in their rates <strong>of</strong> use; shrinking<br />
gaps between richer and poorer individuals as well as those<br />
who have more or less formal education; and smaller disparities<br />
in use between rural and urban/suburban users. 3<br />
Early conceptual papers about the effects that access to<br />
Internet information would have on patients hypothesized<br />
that the effects might be pr<strong>of</strong>ound. It was hypothesized that<br />
in many instances, patients might have as much access to<br />
peer reviewed and other information about their illness as<br />
their physicians would have. Many investigators evaluating<br />
the effects <strong>of</strong> Internet information have identified the inconsistent<br />
quality <strong>of</strong> Internet information as the primary risk<br />
posed by burgeoning access to information. Indeed, most<br />
studies <strong>of</strong> the quality and accuracy <strong>of</strong> information both in<br />
cancer and noncancer health conditions have found substantive<br />
issues with health information on the Internet. 4,5 Several<br />
authors have speculated or attempted to measure the<br />
positive and negative effects Internet information might<br />
By Paul R. Helft, MD<br />
encounter may be managed in more or less productive ways.<br />
The means <strong>of</strong> managing such introductions <strong>of</strong> information<br />
should vary based on physicians’ analyses <strong>of</strong> patients’ information<br />
preferences and styles and their apparent reactions to<br />
the information. Managed well, knowledgeable patients can<br />
<strong>of</strong>fer important opportunities <strong>of</strong> informed and shared decision<br />
making.<br />
have when introduced into the physician-patient relationship.<br />
In a telephone survey, Murray and colleagues found<br />
that patients who had looked up information on the Internet<br />
and discussed it with their physicians were primarily seeking<br />
their physicians’ opinion about the information rather<br />
than requesting a specific intervention. 6 Waid and colleagues<br />
reported in their literature review that Internet<br />
information was <strong>of</strong>ten “triangulated” in the physicianpatient<br />
relationship and that Internet information might<br />
help patients make more informed decisions and strengthen<br />
shared decision making. The authors pointed out that the<br />
dangers include the threat such access to information poses<br />
to the traditional, authoritarian model <strong>of</strong> the physicianpatient<br />
relationship, however. 7<br />
Data from the Health Information National Trends Survey<br />
(HINTS), sponsored by the U.S. Department <strong>of</strong> Health<br />
and Human Services, the National Cancer Institute, the<br />
Health Communication and Informatics Research Branch,<br />
and the Division <strong>of</strong> Cancer Control and Population Sciences,<br />
have suggested that patients initially seek information online<br />
and talk with their physicians about the information to<br />
seek approval or assessment <strong>of</strong> the information. 8 Respondents<br />
to the HINTS survey between 2002 and 2008 appeared<br />
to have gained greater trust in information from health care<br />
pr<strong>of</strong>essionals and to have lost trust in health information<br />
from the Internet. 9<br />
A less-studied, but inevitably important, area <strong>of</strong> Internet<br />
use among patients with cancer is the use <strong>of</strong> blogs and<br />
social networking sites (e.g., www.caringbridge.com; www.<br />
facebook.com). One study in which this issue was examined<br />
showed that, in a small sample <strong>of</strong> patients with cancer and<br />
their companions, four areas emerged as gratifications <strong>of</strong><br />
blog use: prevention and care, problem-solving, emotion<br />
management, and information-sharing. 10 This area is certainly<br />
important and growing and merits larger and more<br />
systematic study using methods developed to preserve confidentiality<br />
<strong>of</strong> information among subjects.<br />
When our group surveyed U.S. oncologists who were<br />
ASCO members in the early 2000s, respondents’ median<br />
From the Division <strong>of</strong> Hematology/<strong>Oncology</strong>, Indiana University School <strong>of</strong> Medicine,<br />
Indianapolis, IN.<br />
Author’s disclosures <strong>of</strong> potential conflicts <strong>of</strong> interest are found at the end <strong>of</strong> this article.<br />
Address reprint requests to Paul R. Helft, MD, 1800 N. Capitol Ave, Noyes E-130,<br />
Indianapolis, IN 46202; email: phelft@iupui.edu.<br />
© <strong>2012</strong> by <strong>American</strong> <strong>Society</strong> <strong>of</strong> <strong>Clinical</strong> <strong>Oncology</strong>.<br />
1092-9118/10/1-10<br />
e89
estimate <strong>of</strong> the proportion <strong>of</strong> their patients using the Internet<br />
to obtain cancer information was 30%. Oncologists<br />
reported that, on average, 10 minutes were added to each<br />
patient encounter in which Internet information was discussed,<br />
and that use <strong>of</strong> the Internet had the ability to<br />
simultaneously make patients more hopeful, confused, anxious,<br />
and knowledgeable. 11 One <strong>of</strong> the striking findings from<br />
this study and others, however, was the suggestion that the<br />
proportion <strong>of</strong> patients who use the Internet for cancer<br />
information is higher than the proportion who actually<br />
discuss the information they have read with their oncologists.<br />
However, there is little doubt that all practicing<br />
oncologists routinely see patients who use the Internet to<br />
obtain information and discuss that information in the<br />
context <strong>of</strong> clinical encounters.<br />
A Taxonomy <strong>of</strong> Patients Who Use the Internet<br />
Although the literature on Internet use among patients<br />
with cancer has begun to outline the effects that Internet<br />
information may have on patients’ decision making, the<br />
relationship with their oncologists, clinical trials enrollment,<br />
and on some emotional and psychological outcomes,<br />
the state <strong>of</strong> the science is not such that the existing literature<br />
might be synthesized into an uncontroversial “users<br />
guide” applicable to patient care. So instead <strong>of</strong> <strong>of</strong>fering a<br />
systematic analysis or synthesis <strong>of</strong> the existing literature, I<br />
would instead <strong>of</strong>fer suggestions regarding the management<br />
<strong>of</strong> various situations in which patients introduce Internet<br />
information into the clinical encounter. Clearly, patients’<br />
purposes for doing so, as well as their specific needs from a<br />
respondent, are likely to be as different as each individual<br />
patient. However, my own clinical experience suggests that<br />
Internet information tends to arise in patterned ways. In<br />
order to discuss each <strong>of</strong> these tendencies, I will <strong>of</strong>fer a<br />
stereotypical patient description, which is in no way meant<br />
to suggest that patients or their information needs or styles<br />
are reducible to simple formulas. Rather, my hope is that<br />
KEY POINTS<br />
● Internet information is pervasive and may have both<br />
positive and negative effects on the clinical encounter<br />
and the physician-patient relationship.<br />
● An individualized approach to information-sharing<br />
and analysis should be based on an explicit assessment<br />
<strong>of</strong> patients’ information preferences.<br />
● When patients raise questions driven by Internet<br />
information, recognition and validation that the issues<br />
raised are important can be an opportunity to<br />
subtly reinforce the physician’s competence.<br />
● Among patients with high health literacy who seek<br />
Internet information, the main challenge is that they<br />
may raise questions that even expert physicians are<br />
not fully capable <strong>of</strong> responding to without some preparation.<br />
● Appearing to demonstrate pervasive misunderstanding<br />
or misinterpretation <strong>of</strong> Internet information or<br />
clinging only to positive information may be signs <strong>of</strong><br />
substantial emotional distress that warrants attention.<br />
e90<br />
these categories will permit patterns to emerge for the sake<br />
<strong>of</strong> discussion and analysis. These four Internet informationusing<br />
patient archetypes are the Worrier, the Seeker, the<br />
Knower, and the Misunderstander.<br />
The Worrier<br />
PAUL R. HELFT<br />
Mr. A is a 58-year-old man with newly diagnosed pancreatic<br />
cancer. When he was first seen, he had pain in the<br />
abdomen and back and was jaundiced. A work-up revealed a<br />
pancreatic adenocarcinoma with involvement <strong>of</strong> the surrounding<br />
vasculature and no evidence <strong>of</strong> distant spread.<br />
Mr. A, as well as his wife and two children, actively sought<br />
information about pancreatic cancer on the Internet. He<br />
confesses that he stopped looking after the first 24 hours<br />
because the information was “too depressing.”<br />
Research on information acquisition and information preferences<br />
is far too vast to review here, but for the sake <strong>of</strong><br />
simplicity, there is a body <strong>of</strong> research indicating that,<br />
although some patients try to avoid or minimize obtaining<br />
stressful medical information, others seem to search for it<br />
and are highly sensitive to it. 12,13 Information avoidance<br />
and information-seeking might thereby be considered two<br />
different psychological strategies for managing anxiety: either<br />
by avoiding information or by seeking it out in order to<br />
gain knowledge and control over it. The first mode has been<br />
called monitoring (attention to, scanning for, and amplification<br />
<strong>of</strong> threatening cues), and denotes the extent to which<br />
individuals are alert for and sensitized to the negative,<br />
potentially painful, or dangerous aspects <strong>of</strong> information and<br />
experience. The second mode, blunting (avoidance <strong>of</strong> threatening<br />
cues), is characterized by individuals distracting<br />
themselves from such information. 12,13<br />
In this way, Mr. A might be understood simplistically to be<br />
a “blunter,” a person whose natural tendency in the face <strong>of</strong><br />
threatening or anxiety-provoking information is to avoid<br />
information. Instead <strong>of</strong> feeling the relief that many people<br />
gain when coming to understand information about their<br />
condition better, Mr. A’s attempt to seek information conflicted<br />
with his personal psychology, and he found it extremely<br />
anxiety-provoking. In my own experience, such<br />
patients frequently go through the course <strong>of</strong> their illness and<br />
ask few questions and engage in little discussion about<br />
potentially worrisome or threatening information. Instead<br />
they speak little or talk about other, less-threatening things.<br />
In dealing with a patient such as Mr. A, I would <strong>of</strong>fer three<br />
suggestions based on the important piece <strong>of</strong> information he<br />
provided to his oncologist about his experience with Internet<br />
information. First, evidence suggests that such information<br />
styles are deep-seated and pervasive. Being aware <strong>of</strong> a<br />
patient’s tendency to avoid threatening information as a<br />
coping mechanism should alert the physician that instances<br />
in which information—particularly negative information—is<br />
to be communicated and shared should be preceded<br />
by frequent assessments <strong>of</strong> the patient’s state <strong>of</strong> mind and<br />
preferences for such information at any given time; for<br />
example, “I have a few difficult things to talk with you about<br />
today, Mr. A. How do you feel about tackling some <strong>of</strong> them<br />
today?” Second, “blunters” are <strong>of</strong>ten surrounded by “monitors,”<br />
family members or loved ones who cope with the stress<br />
<strong>of</strong> the patient’s situation by seeking out as much information<br />
as possible. Although it is clearly not the physician’s role to<br />
manage all related individuals’ information-seeking behaviors,<br />
knowledge that their needs may be different from the
THE INTERNET AND THE CLINICAL ENCOUNTER<br />
patients allows the physician to seek clarification about who<br />
needs or wants to know what when. Finally, because the<br />
information-seeking preferences <strong>of</strong> “blunters” may cause<br />
them to avoid information, a strategy for managing critical<br />
information sharing needs to be negotiated early in the<br />
course <strong>of</strong> the patient’s illness trajectory. For example, asking<br />
a patient how he or she wants to discuss stressful information<br />
ahead <strong>of</strong> time may help the patient to feel more in<br />
control <strong>of</strong> information delivery events.<br />
The Seeker<br />
Mr. B is a 63-year-old man with newly diagnosed intrahepatic<br />
cholangiocarcinoma. When he was first seen, he had a<br />
palpable mass; he was subsequently found to have a large<br />
adenocarcinoma replacing much <strong>of</strong> his atrophied left lobe <strong>of</strong><br />
the liver, with extension into the right lobe. As time passes<br />
during his treatment, to which he has a major partial<br />
response, he raises issues he has encountered on the Internet,<br />
mostly after prompts from fellow patients he communicates<br />
with on the Cholangiocarcinoma.org message boards.<br />
One <strong>of</strong> the potential advantages <strong>of</strong> access to the Internet<br />
in the context <strong>of</strong> illness is that patients with uncommon<br />
cancers can form relationships online with other patients<br />
despite geographical and other barriers. For many patients,<br />
this ability to share experiences with other patients with the<br />
same disease is valuable and can provide an important<br />
source <strong>of</strong> emotional support. That said, I have found such<br />
Internet sources frequently lead to questions from patients<br />
that suggest they have difficulty sorting out what applies to<br />
them and what does not, having heard suggestions from<br />
other patients with the same diagnosis but whose circumstances<br />
may be radically different. Mr. B thus asks questions<br />
at each clinic encounter such as, “One <strong>of</strong> the patients<br />
from my message board had a liver transplant for his<br />
cholangiocarcinoma. Can I have a liver transplant?” At other<br />
times, he raises questions about clinical trials, hepatic<br />
intra-arterial infusion therapy, and surgical resection, all<br />
stimulated by comments from other patients. Each time he<br />
considers new information, such comments lead him to seek<br />
further information from other Internet sources and he asks<br />
clarification <strong>of</strong> his oncologist.<br />
Suggestions for management <strong>of</strong> Mr. B’s continued questions<br />
include focusing on the treatment options he raises and<br />
how they apply to patients in different circumstances, and<br />
explaining how these options have already been considered<br />
and why they do or do not specifically apply to him. The<br />
important message from both <strong>of</strong> these communicative suggestions<br />
is that the physician recognizes and validates the<br />
importance <strong>of</strong> the issues to him (and thus the discussion <strong>of</strong><br />
them validates the sense that the physician is willing to<br />
explore other options with him). This discussion can be an<br />
opportunity to subtly reinforce the physician’s competence,<br />
as such questions enable the physician to show that he or<br />
she has already carefully considered the options raised.<br />
Finally, recognizing the importance <strong>of</strong> such online experiences<br />
for the patient can represent a meaningful occasion for<br />
validating the patient’s efforts to continue to investigate and<br />
understand his disease and options fully.<br />
The Knower<br />
Ms. C is a 52-year-old former medical social worker who<br />
was seen for vague abdominal symptoms. A computerized<br />
tomography scan identifies three separate liver lesions in the<br />
right lobe <strong>of</strong> the liver. A biopsy is consistent with a low-grade<br />
neuroendocrine carcinoma. She undergoes a complete resection<br />
<strong>of</strong> disease, including an ileal primary tumor. During<br />
multiple clinic encounters, she asks many informed questions<br />
about tumor biology, clinical and serum biomarkers, genetic<br />
evaluation, surveillance measures, and potential future therapies.<br />
This patient represents the stereotypical highly-educated,<br />
informed, and savvy patient who has benefitted greatly from<br />
access to information. To pick up on an earlier theme, she<br />
would likely be characterized as a “monitor,” in that part <strong>of</strong><br />
her coping strategy for dealing with her illness is to learn as<br />
much as possible about her disease and its management.<br />
This patient’s information preferences represent both a<br />
challenge and an opportunity, however. The main challenge<br />
is that such patients, who have both access to information<br />
and the knowledge and health literacy to understand the<br />
information at a deep level, may at times raise questions<br />
that even expert physicians are not fully capable <strong>of</strong> responding<br />
to without some preparation. The important opportunity<br />
that such patients represent is that they can become sophisticated<br />
and rewarding true partners for an interested physician<br />
because they are capable <strong>of</strong> engaging in truly<br />
informed and shared decision making.<br />
An important communicative principle to keep in mind<br />
when interacting with such a patient is that the physician<br />
must make clear that he or she is open to having the patient<br />
engage in such information seeking and that her participation<br />
is welcome and encouraged. It is acceptable to ask the<br />
patient explicitly what she hopes to get out <strong>of</strong> any discussion<br />
that follows from her questions related to the information<br />
she finds (e.g., clarification? perspective? personal application?).<br />
Seeking explicit understanding about the patient’s<br />
goals and then trying to achieve them is a means <strong>of</strong> trust<br />
building, as there is evidence that trust “points” are acquired<br />
through behaviors that demonstrate true interest in<br />
and caring for the patient. It is also appropriate as a<br />
time-management strategy to ask the patient to identify the<br />
three or four most important questions for each visit, and to<br />
even to send them ahead <strong>of</strong> time if possible if the physician<br />
finds that answering a dozen such questions extends the<br />
clinical encounter beyond the feasible limits.<br />
The building <strong>of</strong> trust through openness to and respect for<br />
a patient’s dedicated effort to understand his or her disease<br />
as much as possible can be reinforced by comments that<br />
recognize or praise the patient’s efforts. This trust can<br />
become important for times when the patient may raise a<br />
question for which the physician does not have an immediate<br />
answer or informed perspective. This is especially true<br />
for the rare cases in which a patient introduces such questions<br />
as a kind <strong>of</strong> “test” <strong>of</strong> the physician’s competence or<br />
knowledge.<br />
The Misunderstander<br />
Mrs. D is a 72-year-old woman with newly diagnosed<br />
metastatic pancreatic adenocarcinoma. She has had a difficult<br />
time accepting her diagnosis and its associated prognosis,<br />
as has her husband and four grown sons. She encounters<br />
another patient on a message board who had “exactly the<br />
same thing as her” but who had intensive chemotherapy and<br />
underwent surgery to remove the primary tumor and liver<br />
metastases. She becomes convinced that she needs to undergo<br />
e91
exactly the same treatment, even though her oncologist has<br />
assessed her performance status and associated problems<br />
and found her not to be a candidate for intensive therapy. At<br />
each clinical encounter, she raises the same questions over<br />
and over, driven by continued Internet searches and interactions<br />
with other patients online.<br />
In many ways, the archetype that Mrs. D represents is the<br />
most challenging <strong>of</strong> the four oversimplified patient types<br />
presented. She is challenging because such patients appear<br />
to place more credence in information they find online than<br />
they do in their oncologist, even if they feel they have a good<br />
relationship with him or her. This is plausibly driven by<br />
psychological distress and the need to be reassured by<br />
seeking out positive information preferentially in the face <strong>of</strong><br />
such a devastating diagnosis.<br />
Several important issues should shape the approach to<br />
such a patient. The first is to recognize that the primary<br />
driver <strong>of</strong> this patient’s reaction to information (much <strong>of</strong><br />
which she may not fully understand or accept) is distress<br />
about her illness. In this respect, it would be valuable to<br />
pursue efforts at helping to alleviate her psychological<br />
distress (e.g., counseling, cognitive behavior therapy, medication,<br />
supportive interventions). The second is to take into<br />
account that such patients’ apparent lack <strong>of</strong> understanding<br />
or acceptance <strong>of</strong> information from the physician that contradicts<br />
online information is most likely due to distress and<br />
not to cognitive impairment, low health literacy, or low level<br />
<strong>of</strong> education (although all <strong>of</strong> these issues could also have a<br />
role). Thus, as a management strategy, continued explanations<br />
and attempts to correct misunderstandings are likely<br />
not to be productive in the long term. Stated differently,<br />
after an initial attempt to correct misinformation or misunderstandings,<br />
further long, cognitive explanations are not<br />
likely to change the patient’s mind and, indeed, might even<br />
appear adversarial. A different strategy I would recommend<br />
is to implement a supportive listening approach, in which<br />
Author’s Disclosures <strong>of</strong> Potential Conflicts <strong>of</strong> Interest<br />
Author<br />
Employment or<br />
Leadership<br />
Positions<br />
Consultant or<br />
Advisory Role<br />
the physician listens supportively and actively to the patient’s<br />
stories about information acquired on the Internet<br />
and acknowledges the patient’s efforts but does not attempt<br />
to correct the perceptions on a repeated basis, except when<br />
specifically asked to render an opinion. This approach recognizes<br />
the “cognitive block” that may be affecting the<br />
patient’s perceptions and the fact that further logic-driven<br />
discussion is not likely to lead to greater understanding. At<br />
the same time, this approach allows the physician to engage<br />
in relationship-building behavior (active listening). This<br />
issue is difficult, however, driven as we are to logic and<br />
explanation as a means <strong>of</strong> effective communication and so<br />
requires some “letting go” <strong>of</strong> those goals <strong>of</strong> communication.<br />
In my own experience with patients who do not appear<br />
emotionally prepared to have their misunderstandings and<br />
misperceptions corrected. I have found that shifting the<br />
focus <strong>of</strong> our interactions to supportive listening and abandoning<br />
attempts at contradictory argument is a reasonably<br />
productive approach.<br />
Conclusion<br />
The Internet has changed all <strong>of</strong> our lives forever and has<br />
certainly changed the way in which patients acquire information,<br />
share their stories, find others in similar circumstances,<br />
and analyze their medical situations. It is clear that<br />
patients have widely adopted the use <strong>of</strong> online resources in<br />
the face <strong>of</strong> illness. Access to unfiltered information online<br />
clearly has positive and negative potential effects, and the<br />
introduction <strong>of</strong> Internet information into the physicianpatient<br />
encounter may be managed in more or less productive<br />
ways. The means <strong>of</strong> managing such introductions <strong>of</strong><br />
information should vary based on the physician’s analysis <strong>of</strong><br />
a patient’s information preferences and styles and his or her<br />
apparent reactions to the information. Managed well,<br />
knowledegable patients can <strong>of</strong>fer important opportunities <strong>of</strong><br />
informed and shared decision making.<br />
Stock<br />
Ownership Honoraria<br />
Paul R. Helft Genentech; Lilly-<br />
Amylin (I);<br />
Merck (I)<br />
1. Helft PR, Eckles RE, Johnson-Calley CS, Daugherty CK. Use <strong>of</strong> the<br />
internet to obtain cancer information among cancer patients at an urban<br />
county hospital. J Clin Oncol. 2005;23:4954-4962.<br />
2. Castleton K, Fong T, Wang-Gillam A, et al. A survey <strong>of</strong> internet<br />
utilization among patients with cancer. Support Care Center. 2011;19:1183-<br />
1190.<br />
3. Pew Internet and <strong>American</strong> Life Project. Demographics <strong>of</strong> internet<br />
users. http://pewinternet.org/Static-Pages/Trend-Data/Whos-Online.aspx. Accessed<br />
Feb 13, <strong>2012</strong>.<br />
4. Eysenbach G, Powell J, Kuss O, Sa ER. Empirical studies assessing the<br />
quality <strong>of</strong> health information for consumers on the world wide web: a<br />
systematic review. JAMA. 2002;287:2691-2700.<br />
5. Lawrentschuk N, Sasges D, Tasevski R, et al. <strong>Oncology</strong> ealth information<br />
auality on the Internet: a multilingual evaluation. Ann Surg Oncol. Epub<br />
2011 Dec 7.<br />
6. Murray E, Lo B, Pollack L, et al. The impact <strong>of</strong> health information on the<br />
internet on the physician-patient relationship: patient perceptions. Arch<br />
Intern Med. 2003;163:1727-1734.<br />
7. Wald HS, Dube CE, Anthony DC. Untangling the Web—the impact <strong>of</strong><br />
e92<br />
REFERENCES<br />
Research<br />
Funding<br />
Expert<br />
Testimony<br />
PAUL R. HELFT<br />
Other<br />
Remuneration<br />
Internet use on health care and the physician-patient relationship. Patient<br />
Educ Couns. 2007;68:218-224.<br />
8. Hesse BW, Nelson DE, Kreps GL, et al. Trust and sources <strong>of</strong> health<br />
information: the impact <strong>of</strong> the Internet and its implications for health care<br />
providers: findings from the first Health Information National Trends Survey.<br />
Arch Intern Med. 2005;165:2618-2624.<br />
9. Hesse BW, Moser RP, Rutten LJ. Surveys <strong>of</strong> physicians and electronic<br />
health information. N Engl J Med. 2010;362:859-860.<br />
10. Chung DS, Kim S. Blogging activity among cancer patients and their<br />
companions: Uses, gratifications, and predictors <strong>of</strong> outcomes. J Am Soc<br />
Information Sci Technol. 2008;59:297-306.<br />
11. Helft PR, Hlubocky F, Daugherty CK. <strong>American</strong> oncologists’ views <strong>of</strong><br />
internet use by cancer patients: a mail survey <strong>of</strong> <strong>American</strong> <strong>Society</strong> <strong>of</strong> <strong>Clinical</strong><br />
<strong>Oncology</strong> members. J Clin Oncol. 2003;21:942-947.<br />
12. Miller SM, Combs C, Stoddard E. Information coping and control in<br />
patients undergoing surgery and stressful medical procedures. In: Steptoe A,<br />
Appels A (Eds.) Stress, Personal Control, and Health. Chichester, England;<br />
John Wiley & Sons, 1989;107-130.<br />
13. Miller SM. Cognitive informational styles in the process <strong>of</strong> coping with<br />
threat and frustration. Adv Behav Res Ther. 1989;11:223-234.
GLOBAL VARIATIONS IN ACCESS TO NEW<br />
THERAPIES FOR SARCOMAS AND<br />
GASTROINTESTINAL STROMAL TUMOR<br />
CHAIR<br />
George D. Demetri, MD<br />
Dana-Farber Cancer Institute/Harvard Medical School<br />
Boston, MA<br />
SPEAKERS<br />
Stefan Sleijfer, MD, PhD<br />
Erasmus University Medical Center, Daniel den Hoed Cancer Center<br />
Rotterdam, Netherlands<br />
Ian Judson, MD<br />
Royal Marsden Hospital and NHS Foundation Trust<br />
London, United Kingdom
How to Decide Whether to Offer and Use<br />
“Nonstandard” Therapies in Patients with<br />
Advanced Sarcomas and Gastrointestinal<br />
Stromal Tumors: Global Variations in <strong>Clinical</strong><br />
Practice, Assessment, and Access to<br />
Therapies in Diseases with Limited<br />
Incidence and Data<br />
By Stefan Sleijfer, MD, PhD, Ian Judson, MD, and George D. Demetri, MD<br />
Overview: As cancer is more generally recognized as a<br />
collection <strong>of</strong> various rare diseases rather than a homogeneous<br />
illness, sarcomas have become a model for the manner in<br />
which data can and cannot be used to drive clinical decision<br />
making. In this article, we explore the limitations <strong>of</strong> data<br />
generated in rare diseases such as sarcomas to provide an<br />
evidence base for clinical practice. How should patients be<br />
treated if there is no “standard” that <strong>of</strong>fers “pro<strong>of</strong> ” <strong>of</strong> clinical<br />
benefit? By asking this question, we also raise the issue <strong>of</strong><br />
CANCER IS increasingly recognized as a complex group<br />
<strong>of</strong> hundreds if not thousands <strong>of</strong> different and very<br />
diverse diseases. Even the most common forms <strong>of</strong> cancer are<br />
evolving into subsets <strong>of</strong> rare diseases. As a paradigmatic<br />
example <strong>of</strong> this fragmentation <strong>of</strong> neoplastic diseases based<br />
on improved clinicopathologic, molecular, and genomic subclassifications,<br />
the rare diagnostic group <strong>of</strong> s<strong>of</strong>t-tissue sarcomas<br />
(STS) comprises tumors in which more than 50<br />
different histologic subtypes can be distinguished. Despite<br />
substantial differences between the diverse subtypes in<br />
genetic aberrations, natural course, and sensitivity to systemic<br />
drugs, patients with advanced STS were until recently<br />
treated similarly, with doxorubicin and ifosfamide being the<br />
most commonly applied compounds.<br />
This changed after the importance <strong>of</strong> subtyping in<br />
STS was underlined by the finding that receptor tyrosine<br />
kinase inhibitors, such as imatinib or sunitinib, dramatically<br />
improved the prognosis <strong>of</strong> patients with advanced<br />
gastrointestinal stromal tumors (GISTs) and dermat<strong>of</strong>ibrosarcoma<br />
protuberans based on a rational understanding<br />
<strong>of</strong> the fundamental pathophysiologic abnormalities in<br />
these diseases. Importantly, the level <strong>of</strong> antitumor activity<br />
in other STS subtypes with imatinib was quite low, and<br />
anecdotal evidence <strong>of</strong> activity with the more broadly<br />
acting sunitinib has never been fully validated in large<br />
studies. Certain studies have been performed in specific<br />
STS subtypes (such as tenosynovial giant cell tumor or<br />
alveolar s<strong>of</strong>t part sarcoma), but the data have remained<br />
quite limited.<br />
Performing very selective, subtype-specific trials is the<br />
most rigorous and logical way to proceed in STS clinical<br />
research. However, an obvious consequence <strong>of</strong> doing studies<br />
in very rare subgroups <strong>of</strong> a rare disease like STS is the<br />
practical logistical and technical difficulty to perform clinical<br />
trials that are “adequately powered” by conventional statistical<br />
standards. As a result, many recent publications describe<br />
relatively small, single-arm phase II studies in<br />
specific tumor types from which it is difficult, if not impossible,<br />
to draw definitive conclusions to guide evidence-based<br />
decision making for daily clinical practice.<br />
what constitutes “clinical benefit”—and how to measure that—<br />
for patients with sarcomas and other rare diseases. As physicians<br />
become more accountable for decisions—and yet are<br />
always accountable to the patients and families who rely on<br />
them to provide the best and most appropriate care—oncologists<br />
must be cognizant <strong>of</strong> the limitations <strong>of</strong> data in rare<br />
diseases and be ready to justify actions that are in the best<br />
medical and social interests <strong>of</strong> patients.<br />
Although this will perhaps become a problem for all forms<br />
<strong>of</strong> cancer, as biologic understanding fragments even the<br />
most common cancers into groupings <strong>of</strong> increasingly rare<br />
subtypes (witness the 3% <strong>of</strong> non–small cell lung carcinoma<br />
with ALK genomic aberrancies, or the 10% to 20% incidence<br />
<strong>of</strong> HER2 overexpression in a subset <strong>of</strong> gastric carcinomas),<br />
the international sarcoma community may be in the unenviable<br />
position to manage this challenge now. There is a<br />
tension between wanting to provide each individual patient<br />
with the best possible and most appropriate care<br />
options and the lack <strong>of</strong> “pro<strong>of</strong> ” on which to base such<br />
judgments about what is truly “best possible” or even “appropriate.”<br />
One first important step is to agree internationally<br />
on parameters and thresholds for activity to define<br />
clinical benefit and to implement these in clinical trials,<br />
taking into account the rarity <strong>of</strong> STS subgroups. Another<br />
important issue is how to treat patients for whom no<br />
standard treatment is available. Are outcomes <strong>of</strong> singlearm<br />
studies with drugs enough to justify <strong>of</strong>f-label use <strong>of</strong><br />
these drugs with all the potential financial, regulatory,<br />
legal, and practical consequences? And what should physicians<br />
do with a drug that is considered to yield clinical<br />
benefit for a particular patient population but which is not<br />
considered cost-effective by administrative bodies (such as<br />
governmental agencies that decide public policy or private<br />
payers)?<br />
This article addresses how physicians in different parts <strong>of</strong><br />
the world involved in the treatment <strong>of</strong> STS patients struggle<br />
with, and manage, these issues.<br />
From the Erasmus MC Daniel Den Hoed Cancer Center, Rotterdam, Netherlands; CR UK<br />
Centre for Cancer Therapeutics, Sutton, United Kingdom; Ludwig Center and Sarcoma<br />
Center, Dana-Farber Cancer Institute and Harvard Medical School, Boston, MA.<br />
Authors’ disclosures <strong>of</strong> potential conflicts <strong>of</strong> interest are found at the end <strong>of</strong> this article.<br />
Address reprint requests to Stefan Sleijfer, MD, PhD, Erasmus MC Daniel Den Hoed<br />
Cancer Center, Medical <strong>Oncology</strong>, Groene Hilledijk 301, Rotterdam 3075 EA, Netherlands;<br />
email: s.sleijfer@erasmusmc.nl.<br />
© <strong>2012</strong> by <strong>American</strong> <strong>Society</strong> <strong>of</strong> <strong>Clinical</strong> <strong>Oncology</strong>.<br />
1092-9118/10/1-10<br />
645
What Metrics Should Be Used to Define <strong>Clinical</strong> Benefit<br />
in Sarcomas? Is There Consensus Internationally?<br />
The concept <strong>of</strong> “clinical benefit” appears—on the surface—as<br />
intuitively clear and simple: that which helps a<br />
patient feel, function, or survive better. The challenge is that<br />
measuring anything other than survival is impossibly complex<br />
and very insensitive to variations that may be highly<br />
relevant and desirable to patients. Even “survival” is a<br />
complex and composite measure dependent on many variables;<br />
as such, survival does not directly measure a single<br />
intervention such as an individual new drug or therapy. One<br />
could keep a patient alive longer than the patient might<br />
wish, for example, using heroic measures that do not provide<br />
a desirable or reasonable quality <strong>of</strong> life; alternatively, some<br />
patients who lose hope might withdraw and avoid potentially<br />
life-prolonging standard therapy options. The challenge<br />
for physicians caring for patients as well as for<br />
physicians serving as regulators (or advisors to regulatory<br />
bodies) is to focus on the needs <strong>of</strong> the patients in a given<br />
clinical context.<br />
This critically important element <strong>of</strong> clinical context is<br />
<strong>of</strong>ten glossed over or ignored entirely in discussions <strong>of</strong><br />
clinical benefit. The fact that people with different types <strong>of</strong><br />
illnesses and with different risks <strong>of</strong> morbidity or mortality<br />
may choose to make different decisions represents one clear<br />
and compelling piece <strong>of</strong> evidence that clinical context truly<br />
matters. The risk <strong>of</strong> life-altering symptoms, such as pain or<br />
severe dyspnea from massive lesions pressing on nerves or<br />
major bronchi, is a major concern to many patients with<br />
advanced sarcomas, given the fact that sarcomas can occur<br />
in young, otherwise healthy individuals and can grow to<br />
KEY POINTS<br />
● It is impossible to generate statistical pro<strong>of</strong> <strong>of</strong> beneficial<br />
activity in every rare disease for every possible<br />
therapy.<br />
● Physicians must learn to read between the lines and<br />
make clinical decisions on the basis <strong>of</strong> best available<br />
evidence and incomplete information.<br />
● All drugs have side effects and risks, but pr<strong>of</strong>essional<br />
judgment can and must be used in discussions with<br />
patients about management options once standard<br />
“proven” therapy has failed.<br />
● <strong>Clinical</strong> context is very important for weighing risks<br />
and benefits, and patient preferences must also be<br />
integrated into such clinical context for decision making.<br />
● Best supportive care may sometimes be the most<br />
appropriate option for patients with incurable diseases<br />
and with no available therapy <strong>of</strong> proven value,<br />
but such decisions also should take into account other<br />
lines <strong>of</strong> evidence as well as patient preferences.<br />
● Assessments <strong>of</strong> cost-effectiveness require clinical expert<br />
opinions, and clinical experts have an important<br />
role in advocacy for patient preferences and patient<br />
needs, as well as rigorously interpreting limited data<br />
by conventional methods <strong>of</strong> statistical pro<strong>of</strong>.<br />
646<br />
SLEIJFER, JUDSON, AND DEMETRI<br />
significant bulk before interfering with function in these<br />
patients. However, once that tipping point has been reached,<br />
the patients may have a limited ability to move from<br />
“adequate function” to “decreased function with symptoms”<br />
to “severe dysfunctional status” or even death. This contextual<br />
distance is highly relevant for patients with STS, who<br />
risk living with the burden <strong>of</strong> very bulky disease, <strong>of</strong>ten far<br />
more bulky than patients with other cancers.<br />
This is directly applicable to a controversy that is perhaps<br />
one <strong>of</strong> the most contentious in clinical oncology research<br />
today: what is the value to patients <strong>of</strong> maintaining patients<br />
in a progression-free state? This is particularly vexing in<br />
rare-disease research, where it may be difficult or impossible<br />
to prove any effect on overall survival (OS). This is<br />
exemplified by two large phase III clinical trials that enrolled<br />
patients with a variety <strong>of</strong> STS subtypes in different<br />
clinical settings. PALETTE was a phase III international<br />
trial in which a limited spectrum <strong>of</strong> patients with STS<br />
were randomly selected to receive either placebo or a multikinase<br />
inhibitor, pazopanib, in the setting <strong>of</strong> metastatic<br />
disease that was progressing despite prior standard chemotherapy.<br />
1 In contrast, SUCCEED was a phase III international<br />
trial in which patients with a broader array <strong>of</strong><br />
sarcomas were randomly selected to receive either placebo<br />
or a mammalian target <strong>of</strong> rapamycin inhibitor, ridaforolimus,<br />
in the setting <strong>of</strong> metastatic disease that was stable or<br />
in objective remission (partial or complete) following previous<br />
standard chemotherapy. 2 These trials both met statistically<br />
predefined endpoints that demonstrated the beneficial<br />
effects <strong>of</strong> the two drugs under study (pazopanib or ridaforolimus,<br />
respectively) compared with placebo. However,<br />
neither study was able to prove a statistically significant<br />
effect <strong>of</strong> either drug on OS. Nonetheless, these trials highlight<br />
the challenge: in patients with metastatic sarcomas,<br />
what is the metric <strong>of</strong> clinical benefit? Is this metric constant<br />
across clinical settings? In both studies, patients with metastatic,<br />
incurable sarcomas were enrolled and studied. However,<br />
in PALETTE, these patients had progressive disease<br />
that had already progressed despite prior standard systemic<br />
therapy; in SUCCEED, these patients entered the trial with<br />
evidence <strong>of</strong> disease control from prior standard therapy.<br />
These differences in clinical context may or may not be<br />
important in weighing the clinical benefit <strong>of</strong> a new treatment,<br />
especially in the palliative setting such as metastatic<br />
sarcomas.<br />
The relative risks and benefits <strong>of</strong> any new treatment are<br />
clearly important, but the relative risks <strong>of</strong> uncontrolled<br />
progressive sarcoma are also important to keep in mind.<br />
This is especially true in diseases such as sarcomas that<br />
tend to strike a somewhat younger and overall healthier<br />
population than many cancers. Finally, the lack <strong>of</strong> tools to<br />
dissect out subtle variations and differences in patient<br />
preferences as well as patient experiences with disease and<br />
with treatments also complicates the problems <strong>of</strong> assessing<br />
clinical benefit in this field.<br />
The rarity <strong>of</strong> sarcomas only adds to the complexities <strong>of</strong><br />
interpreting such data and making clinical extrapolations to<br />
interpret true clinical benefit for patients. Ultimately, clinical<br />
experience, expert judgment, and the totality <strong>of</strong> available<br />
evidence must be used to assess the clinical benefits to<br />
patients with advanced sarcomas.
NONSTANDARD THERAPIES FOR ADVANCED SARCOMAS AND GISTS<br />
Decision Making in Patients with Sarcomas Following<br />
Failure <strong>of</strong> Conventional Therapies—Off-Label Drug<br />
Use or Hospice?<br />
At a certain point in time, virtually all patients with<br />
advanced STS will eventually be confronted with the fact<br />
that no standard treatment options with solid evidence <strong>of</strong><br />
clinical benefit are available. This is always a difficult<br />
situation for both patients and physicians, with the most<br />
important question being how to proceed. (Although we<br />
focus on drug use here, we note that the same logic can also<br />
apply to decisions regarding application <strong>of</strong> surgical or radiotherapeutic<br />
interventions, perhaps with even greater risk <strong>of</strong><br />
morbidity and financial implications than that <strong>of</strong> drug use.)<br />
Many patients may be in a poor clinical condition after<br />
their sarcomas fail to react to standard treatments. Importantly,<br />
the strongest prognostic factor for early mortality <strong>of</strong><br />
patients with advanced STS is performance status. A recent<br />
analysis <strong>of</strong> the European Organisation for Research and<br />
Treatment <strong>of</strong> Cancer S<strong>of</strong>t Tissue and Bone Sarcoma Group<br />
(EORTC-STBSG) database <strong>of</strong> patients with advanced STS<br />
treated with systemic treatments in the context <strong>of</strong> clinical<br />
trials showed that the mortality rate within 90 days after<br />
treatment start was 3.3%, 9.4%, and 25.5% for patients with<br />
a World Health Organization (WHO) performance <strong>of</strong> 0, 1,<br />
and, 2 or more, respectively. 3 This study concerned only<br />
patients treated in the first-line setting, and it is likely that<br />
the outcomes would be worse for patients following failure <strong>of</strong><br />
prior systemic regimens. This observation, together with the<br />
risk <strong>of</strong> toxicities accompanying systemic treatments, justify<br />
best supportive care for patients with advanced STS following<br />
failure <strong>of</strong> prior therapies and who have a frail clinical<br />
status (certainly, those with a WHO performance status <strong>of</strong><br />
3or4).<br />
For pretreated patients who still have a good performance<br />
status, have good organ functions, and who strongly wish to<br />
be treated, the situation is more complex. Roughly, in the<br />
absence <strong>of</strong> available studies, there are two options: best<br />
supportive care or <strong>of</strong>f-label use <strong>of</strong> drugs. Off-label use <strong>of</strong><br />
drugs refers to those situations in which drugs are used<br />
outside their <strong>of</strong>ficial labeling. This applies to the use <strong>of</strong> drugs<br />
in diseases other than mentioned in the labeling <strong>of</strong> that<br />
particular drug, but also to those situations when a drug is<br />
given on a different schedule than mentioned in the <strong>of</strong>ficially<br />
approved language <strong>of</strong> the labeling text. Examples <strong>of</strong> the<br />
latter include other doses, administration as combination<br />
therapy if the drug was only approved as a single agent, or<br />
an alternative duration or route <strong>of</strong> administration.<br />
There are relatively few academic studies on the frequency<br />
<strong>of</strong> <strong>of</strong>f-label drug use in sarcoma care, or even in<br />
oncology overall. Some studies estimate the frequency <strong>of</strong><br />
<strong>of</strong>f-label drug use in oncology at approximately 30% to 50%<br />
<strong>of</strong> the prescribed medications. 4 Although there are no specific<br />
studies on the exact frequency <strong>of</strong> <strong>of</strong>f-label use in<br />
advanced STS, it is clear from our collective experience that<br />
<strong>of</strong>f-label use in this setting is exceedingly common. Only a<br />
very limited number <strong>of</strong> drugs have been approved for advanced<br />
STS by regulatory agencies. For example, the drugs<br />
that are approved by the U.S. Food and Drug Administration<br />
(FDA) comprise doxorubicin for adult and childhood<br />
sarcomas, imatinib and sunitinib for GISTs, and vincristine<br />
for rhabdomyosarcomas. 5 That means that a drug such as<br />
ifosfamide, which is commonly applied in advanced STS, is<br />
technically used <strong>of</strong>f-label when used in STS. It is important,<br />
however, to recognize the totality <strong>of</strong> evidence used in making<br />
therapeutic decisions, and that is why inclusion in compendia<br />
such as the National Comprehensive Cancer Network<br />
practice guidelines for sarcomas is based on criteria other<br />
than formal and full regulatory approval.<br />
These are several important issues that influence <strong>of</strong>f-label<br />
use. These include access to <strong>of</strong>f-label drugs, legal responsibilities<br />
<strong>of</strong> physicians and pharmacists, and costs. In some<br />
countries, <strong>of</strong>f-label use <strong>of</strong> drugs is even illegal. 4 All these<br />
aspects may create major difficulties for patients and physicians,<br />
which should be taken into account when considering<br />
prescribing a drug <strong>of</strong>f-label. Since legislation and<br />
regulations concerning <strong>of</strong>f-label use vary largely by country,<br />
these will not be discussed here in detail.<br />
Another important issue is the evidence underlying <strong>of</strong>flabel<br />
use, which greatly differs per drug. Whereas large<br />
randomized studies have established the clinical value <strong>of</strong><br />
drugs such as ifosfamide in adult STS 6 and etoposide in<br />
small blue round cell sarcomas, 7 and therefore have been<br />
incorporated in guidelines, for many other drugs used <strong>of</strong>flabel<br />
the evidence is far less robust. Given the lack <strong>of</strong><br />
publications on this topic specifically in advanced STS, it is<br />
difficult to discern exactly which drugs are frequently used<br />
<strong>of</strong>f-label in advanced STS and for which strong evidence is<br />
lacking. One example is sorafenib, a drug approved by<br />
regulatory agencies for hepatocellular carcinoma and renal<br />
cell carcinoma, but which has been used <strong>of</strong>f-label in GISTs.<br />
A recent study by Italiano and colleagues investigated the<br />
patterns <strong>of</strong> care in patients with metastatic GISTs after<br />
failure <strong>of</strong> the only two approved standard treatments imatinib<br />
and sunitinib. 8 Out <strong>of</strong> the 223 patients with GISTs,<br />
24.5% received sorafenib as third-line therapy. Other treatments<br />
administered to these 223 patients were (1) best<br />
supportive care, (2) rechallenge with imatinib, (3) imatinib<br />
combined with other drugs, (4) nilotinib, or (5) clinical trials<br />
with investigational products. In the group treated with<br />
sorafenib, 19% were reported to have exhibited a response;<br />
median progression-free survival (PFS) and OS were 4.9 and<br />
10.7 months, respectively. 8 Apart from this study, at the<br />
time <strong>of</strong> writing this article, there were no other full publications<br />
on the outcomes <strong>of</strong> patients with GISTs treated with<br />
sorafenib in this setting, except for two very small uncontrolled<br />
series <strong>of</strong> 38 and 32 patients, respectively, both<br />
roughly showing similar results as the study by Italiano and<br />
colleagues but as yet only presented in abstract form. 9,10<br />
Are these results sufficient to justify <strong>of</strong>f-label use <strong>of</strong><br />
sorafenib in GISTs? Single-arm studies or retrospective<br />
series are highly prone to all kinds <strong>of</strong> bias, as a consequence<br />
<strong>of</strong> which many treatments considered promising on the basis<br />
<strong>of</strong> such studies failed to demonstrate their clinical value in<br />
subsequent studies. Undoubtedly, sorafenib induces clinical<br />
benefit in a proportion <strong>of</strong> patients. But as long as we cannot<br />
identify the patients who are likely to benefit from sorafenib,<br />
some might question whether the benefits <strong>of</strong> sorafenib in a<br />
proportion <strong>of</strong> the patients outweigh the toxicities to which<br />
all patients are exposed. Sorafenib is a drug with wellknown<br />
toxicity necessitating a dose modification (e.g., interruption<br />
or reduction) in the majority <strong>of</strong> patients. 10 In patients<br />
with renal cell carcinoma, termination <strong>of</strong> sorafenib treatment<br />
because <strong>of</strong> toxicities or progressive disease substantially<br />
improved quality-<strong>of</strong>-life scores, 11 stressing the effect that<br />
these toxicities have on patient-reported quality <strong>of</strong> life.<br />
647
Altogether, there are large differences in the level <strong>of</strong><br />
evidence underlying the application <strong>of</strong> <strong>of</strong>f-label use in patients<br />
with advanced STS. It should be this evidence that<br />
determines whether a certain drug should be given <strong>of</strong>f-label<br />
to a particular patient category. Some might argue that<br />
physicians should be very cautious not to spoil the precious<br />
time <strong>of</strong> patients suffering from diseases with a short and<br />
dismal prognosis by exposing them to drugs with uncertain<br />
efficacy and certain toxicity. Others might note that for some<br />
patients, being denied a chance to benefit from uncertain<br />
drugs, even with the certainty <strong>of</strong> toxicity, could be unacceptably<br />
grim and could further affect negatively the patient’s<br />
own perceived quality <strong>of</strong> life. There are many individualspecific<br />
variables in these equations, and the lack <strong>of</strong> data<br />
makes “evidence-based” decision making challenging if not<br />
impossible.<br />
Apart from some rare exceptions such as imatinib as<br />
first-line treatment in advanced GISTs yielding unprecedented<br />
outcomes, whether a new regimen outperforms best<br />
supportive care in terms <strong>of</strong> OS and quality <strong>of</strong> life (the<br />
ultimate goals we should strive for in our patients) can only<br />
be determined in the context <strong>of</strong> adequately designed and<br />
rigorously conducted randomized studies. Additionally, <strong>of</strong>flabel<br />
use <strong>of</strong> drugs can be accompanied with issues <strong>of</strong> legal<br />
responsibilities and costs. If there is a reason to treat a<br />
patient with an <strong>of</strong>f-label drug and it is ensured that an<br />
individual patient can get access to the drug, a patient<br />
should be informed about these potential consequences. Also<br />
in health care systems where costs <strong>of</strong> drugs are covered by<br />
society and not by an individual patient, such as in the<br />
Netherlands, costs <strong>of</strong> <strong>of</strong>f-label drugs, in particular expensive<br />
ones, can be a major issue. In such situations, prescribing<br />
<strong>of</strong>f-label medications will eventually be at the expense <strong>of</strong> the<br />
whole system and subsequently will have an effect on the<br />
health care provided to other patients. But again, legislation<br />
on <strong>of</strong>f-label use greatly differs between countries.<br />
For patients for whom no standard therapies are available<br />
and who are candidates for further treatments, participation<br />
in a clinical trial rather than <strong>of</strong>f-label use should be the next<br />
step. Many physicians are convinced that sorafenib is potentially<br />
effective in patients with advanced GIST, yet no<br />
randomized trials have been developed to test this agent,<br />
which was already approved by the FDA in 2005. Clearly,<br />
there are currently too many barriers to set up such studies.<br />
At a global scale, patient advocacy groups, physicians, regulatory<br />
agencies, health insurance companies, and politicians<br />
should join forces to facilitate the efficient design,<br />
development, and deployment <strong>of</strong> such clinical trials. In this<br />
way, our field might determine more swiftly whether a<br />
promising drug should be implemented as an accepted<br />
standard in daily clinical practice. And if so, labeling <strong>of</strong> such<br />
a drug should be adjusted, thereby taking away all potential<br />
financial and legal burdens associated with the use <strong>of</strong> an<br />
<strong>of</strong>f-label drug.<br />
The Role <strong>of</strong> Physicians in Advocating for Patients<br />
When Cost-effectiveness Is Questioned by<br />
Administrative Bodies Such as NICE<br />
The Physician’s Duty as a <strong>Clinical</strong> Expert<br />
Physicians in certain countries may be called on to participate<br />
in the regulatory process by serving as clinical experts.<br />
The role <strong>of</strong> such clinical experts varies depending on the<br />
648<br />
SLEIJFER, JUDSON, AND DEMETRI<br />
regulatory setting: for example, the role <strong>of</strong> an expert physician<br />
on the <strong>Oncology</strong> Drugs Advisory Committee (ODAC) in<br />
the U.S. FDA process is significantly different from the role<br />
<strong>of</strong> an expert physician advising a governmental regulatory<br />
body such as the National Institute for Health and <strong>Clinical</strong><br />
Excellence (NICE) in the United Kingdom. With NICE, the<br />
focus is <strong>of</strong>ten on evaluating the cost-effectiveness <strong>of</strong> a new<br />
treatment, rather than the risk-benefit, which is <strong>of</strong>ten the<br />
subject <strong>of</strong> other regulatory bodies such as the FDA or the<br />
European Medicines Agency. The first thing to recognize<br />
when accepting the task <strong>of</strong> clinical expert in relation to<br />
determining the cost-effectiveness <strong>of</strong> a new treatment is that<br />
such physicians have a similar duty to that <strong>of</strong> an expert<br />
witness in court. In order to be credible, one needs to be<br />
(1) well-briefed, (2) prepared for robust cross-examination,<br />
(3) demonstrably independent <strong>of</strong> the pharmaceutical company<br />
whose product is the subject <strong>of</strong> discussion, and (4) able<br />
to argue effectively on behalf <strong>of</strong> the group <strong>of</strong> patients to<br />
whom the treatment relates.<br />
The United Kingdom’s NICE as an Example <strong>of</strong> a Societal Process<br />
in Drug<br />
NICE was set up to oversee the evaluation <strong>of</strong> new treatments.<br />
Its deliberations and standard operating procedures<br />
are being followed by other countries, which is why we chose<br />
to scrutinize the way that it operates. One <strong>of</strong> the stated aims<br />
<strong>of</strong> NICE was to reduce the inequalities <strong>of</strong> access to novel<br />
therapies according to the geographical location <strong>of</strong> the patient<br />
and the local financial arrangements. This was to be<br />
done by providing a robust method <strong>of</strong> determining costeffectiveness,<br />
and if a drug was deemed to be cost-effective,<br />
it would be made available everywhere in the country. In<br />
fact, within the United Kingdom this only applies to England<br />
and Wales; Scotland has a different system called the<br />
Scottish Medicines Consortium, which appears to have a<br />
more streamlined approach and usually produces a decision<br />
soon after licensing. Unfortunately, NICE has not succeeded<br />
in abolishing the “postcode lottery” for a variety <strong>of</strong> reasons,<br />
including not yet carrying out appraisals on certain drugs<br />
for rare cancers, and because it has always been possible to<br />
appeal to local funding bodies and more recently a regional<br />
Cancer Drugs Fund, in order to try and circumvent a<br />
negative NICE appraisal decision. Unfortunately, the application<br />
<strong>of</strong> these mechanisms is inconsistent across the<br />
country, hence the persistence <strong>of</strong> regional variations. Not<br />
surprisingly, this gives rise to considerable distress among<br />
patients. This pattern would most likely be applicable to<br />
other countries trying to introduce a similar mechanism <strong>of</strong><br />
health care rationing.<br />
The biggest source <strong>of</strong> controversy concerns the methods<br />
used to generate data on cost-effectiveness. The process <strong>of</strong><br />
model generation is open to tender and groups <strong>of</strong> health<br />
economists, generally based in academic institutions, bid for<br />
the job <strong>of</strong> evaluating the new agent. The models used are not<br />
placed in the public domain and hence cannot be directly<br />
challenged, and it is openly acknowledged in certain technology<br />
appraisal documents that many assumptions have<br />
been made owing to the lack <strong>of</strong> data on quality <strong>of</strong> life,<br />
survival <strong>of</strong> patients in disease subsets, etc. In some cases the<br />
results produced by different models for the same drug in<br />
different appraisals vary significantly.
NONSTANDARD THERAPIES FOR ADVANCED SARCOMAS AND GISTS<br />
Preparing a Pr<strong>of</strong>essional Organization Statement<br />
First, it is valuable to work closely with patient advocacy<br />
groups in order to understand the needs and concerns <strong>of</strong> the<br />
patients themselves. A carefully argued case in favor <strong>of</strong> a<br />
new treatment will be enhanced by the input <strong>of</strong> patients in<br />
the process <strong>of</strong> preparing the scientific and clinical case. It<br />
may even be appropriate on occasion to submit a joint report.<br />
You need to be aware <strong>of</strong> who you represent as a clinical<br />
expert and ensure that you have consulted widely before<br />
submission. This also lends the submission more weight.<br />
You are required to state that you are an expert in the<br />
treatment <strong>of</strong> patients who have the condition for which<br />
NICE is considering the new treatment, that you understand<br />
the scientific and clinical basis for the treatment, and<br />
that you are an employee <strong>of</strong> an organization representing<br />
clinicians who treat the condition in question. The framework<br />
for the submission takes into account a number <strong>of</strong><br />
specific issues that include the current therapy for the<br />
disease; any important variations in treatment, including<br />
the merits <strong>of</strong> alternative treatment, if any; whether there<br />
are any patient subgroups that might benefit particularly;<br />
how the treatment would be used; and the existence <strong>of</strong> any<br />
guidelines for the disease in question.<br />
You will need to explain the scientific basis for the treatment,<br />
especially the number and quality <strong>of</strong> clinical trials<br />
and how they might relate to everyday practice. Particular<br />
emphasis might be placed on side effects and toxicities, that<br />
is, the balance between benefit in terms <strong>of</strong> disease control<br />
and the potentially adverse effects <strong>of</strong> treatment on quality <strong>of</strong><br />
life.<br />
In practice this means having an in-depth knowledge <strong>of</strong><br />
the disease and its treatment, being familiar with the<br />
clinical trial evidence for the new treatment, and in particular<br />
being able to argue effectively that the clinical benefit<br />
outweighs the common adverse side effects.<br />
What Are the Key Issues?<br />
Cost-effectiveness in cancer treatment is determined by<br />
the cost per quality adjusted life year (QALY) gained. The<br />
calculations take into account the total number <strong>of</strong> patients<br />
treated relative to the number who might live longer. Hence<br />
if the percentage <strong>of</strong> patients who benefit is small and the cost<br />
is high, a drug could be deemed not to be cost-effective even<br />
if the benefit is prolonged for the few for whom it works. For<br />
example, in Technology Assessment T209 on high-dose imatinib,<br />
it is stated that the estimated median increase in life<br />
expectancy for imatinib 800 mg compared with best supportive<br />
care (which included imatinib 400 mg) was 4.2 months,<br />
but cost-effectiveness for high-dose imatinib was not demonstrated<br />
compared with sunitinib. 12 This was despite the<br />
fact that no comparison <strong>of</strong> sunitinib with imatinib 800 mg<br />
has been performed. In addition, in this particular assessment,<br />
no data on the benefit <strong>of</strong> imatinib 800 mg in patients<br />
with KIT exon 9 mutant GISTs were regarded as permissible<br />
for evaluation because the assessment focused exclusively<br />
on patients whose cancers had progressed on imatinib<br />
400 mg. The data on the PFS for patients with KIT exon 9<br />
came from randomized studies (62005 and S0033), in which<br />
patients received either 400 mg or 800 mg imatinib from day<br />
1. These data were <strong>of</strong> course evaluated in a meta-analysis<br />
demonstrating a significant benefit in terms <strong>of</strong> PFS for 800<br />
mg for the exon 9 mutant group (p � 0.012), 13 but because<br />
they did not address the limited scope <strong>of</strong> this appraisal they<br />
were not considered.<br />
Quality adjustment is <strong>of</strong>ten contentious, largely because<br />
there are few data available and the tools to assess it are<br />
somewhat crude. Essentially, the assumption is that patients<br />
with advanced cancer already have an impaired<br />
quality <strong>of</strong> life and hence any prolongation <strong>of</strong> life is weighted<br />
according to the disease itself and its likely effect on quality<br />
<strong>of</strong> life and further weighted by the side effects <strong>of</strong> treatment.<br />
In Technology Assessment 209, the “utility” <strong>of</strong> patients with<br />
GISTs on best supportive care is given as 0.577, based on a<br />
very small study. 12 In the technology assessment for trabectedin<br />
in the treatment <strong>of</strong> advanced STS, the “utility”<br />
values for progression-free and progressive disease health<br />
states are stated as 0.653 and 0.473, respectively. 14<br />
It must be said at this stage that utility in this context<br />
means the mean quality <strong>of</strong> life <strong>of</strong> patients relative to 1.00<br />
being perfect, at this stage <strong>of</strong> life, and hence a presumed<br />
reduction in the perceived value <strong>of</strong> any extension <strong>of</strong> life. This<br />
is an assumption that patients tend not to accept. However,<br />
it must be acknowledged that we do need better data on the<br />
effect <strong>of</strong> cancer on quality <strong>of</strong> life at different stages <strong>of</strong> disease<br />
in order to help make such assessments more realistic and<br />
meaningful.<br />
Incremental Improvements and the Comparator Approach<br />
In determining the value <strong>of</strong> a new therapy, the standard<br />
approach is to compare the new treatment against the<br />
current best available therapy and if there is no significant<br />
improvement in PFS or OS, then it is deemed ineffective, or<br />
if the improvement is small and it is expensive, then it is<br />
deemed cost-ineffective. However, no system has yet been<br />
devised to evaluate the cost-effectiveness <strong>of</strong> sequential effective<br />
therapies, each with an individual small incremental<br />
benefit but a significant combined effect. It is apparent that<br />
survival <strong>of</strong> patients with many cancer types is increasing<br />
because <strong>of</strong> earlier diagnosis, more effective primary treatment,<br />
and more effective therapy for advanced disease.<br />
What is less easy to quantify is the effects <strong>of</strong> new targeted<br />
therapies on survival and whether sequential therapy with<br />
effective agents, each with a limited duration <strong>of</strong> benefit, is<br />
truly beneficial or not. The current methods used by regulators<br />
and health care utility bodies such as NICE do not<br />
address this issue.<br />
The Question <strong>of</strong> Balance<br />
There is no publicly funded health care system in the<br />
world with unlimited resources, and methods need to be<br />
developed that are transparent and respected to restrict<br />
health care costs while ensuring value for money. No one<br />
questions the value <strong>of</strong> imatinib for the treatment <strong>of</strong> GISTs or<br />
chronic myeloid leukemia. However, a number <strong>of</strong> high-cost<br />
new drugs have been licensed for the treatment <strong>of</strong> common<br />
cancers, which have a statistically significant, but nevertheless<br />
limited effect on PFS or OS, or indeed have yet to<br />
demonstrate any effect on survival. Notwithstanding the<br />
issue <strong>of</strong> cumulative incremental benefit, this needs to be<br />
kept in mind when preparing evidence to funding bodies or<br />
health care evaluation organizations such as NICE. In<br />
addition, it is clearly easier to demonstrate statistically<br />
significant differences in PFS or OS in common cancers,<br />
where large trials can be performed, compared with rare<br />
cancers, even if the benefits are greater.<br />
649
Authors’ Disclosures <strong>of</strong> Potential Conflicts <strong>of</strong> Interest<br />
Author<br />
Employment or<br />
Leadership<br />
Positions<br />
Consultant or<br />
Advisory Role<br />
Stefan Sleijfer GlaxoSmithKline<br />
(U); Novartis<br />
Ian Judson GlaxoSmithKline;<br />
Morphotek;<br />
Novartis; Pfizer;<br />
PharmaMar;<br />
Threshold<br />
George D. Demetri Amgen; ARIAD;<br />
Daiichi Sankyo;<br />
Genentech;<br />
GlaxoSmithKline;<br />
Idera<br />
Pharmaceuticals;<br />
Infinity; Johnson<br />
& Johnson;<br />
Kolltan<br />
Pharmaceuticals;<br />
Merck Serono;<br />
Momenta<br />
Pharmaceuticals;<br />
Novartis; Pfizer;<br />
Plexxikon;<br />
Ziopharm<br />
<strong>Oncology</strong><br />
Stock<br />
Ownership Honoraria<br />
Champions<br />
Biotechnology;<br />
EmergingMed;<br />
Kolltan<br />
Pharmaceuticals;<br />
Plexxikon<br />
1. Van Der Graaf WT, Blay J, Chawla SP, et al. PALETTE: A randomized,<br />
double-blind, phase III trial <strong>of</strong> pazopanib versus placebo in patients (pts) with<br />
s<strong>of</strong>t-tissue sarcoma (STS) whose disease has progressed during or following<br />
prior chemotherapy—an EORTC STBSG Global Network Study (EORTC<br />
62072). J Clin Oncol. 2011;29:605s (suppl; abstr LBA10002).<br />
2. Chawla SP, Blay J, Ray-Coquard IL, et al. Results <strong>of</strong> the phase III,<br />
placebo-controlled trial (SUCCEED) evaluating the mTOR inhibitor ridaforolimus<br />
(R) as maintenance therapy in advanced sarcoma patients (pts)<br />
following clinical benefit from prior standard cytotoxic chemotherapy (CT).<br />
J Clin Oncol. 2011;29:606s (suppl; abstr 10005).<br />
3. Penel N, Glabbeke MV, Mathoulin-Pelissier S, et al. Performance status<br />
is the most powerful risk factor for early death among patients with advanced<br />
s<strong>of</strong>t tissue sarcoma: the European Organisation for Research and Treatment<br />
<strong>of</strong> Cancer-S<strong>of</strong>t Tissue and Bone Sarcoma Group (STBSG) and French Sarcoma<br />
Group (FSG) study. Br J Cancer. 2011;104:1544-1550.<br />
4. Leveque D. Off-label use <strong>of</strong> anticancer drugs. Lancet Oncol. 2008;9:1102-<br />
1107.<br />
5. National Cancer Institute. Drugs Approved for S<strong>of</strong>t Tissue Sarcoma.<br />
http://cancer.gov/cancertopics/druginfo/s<strong>of</strong>t-tissue-sarcoma. Accessed February<br />
6, <strong>2012</strong>.<br />
6. Lorigan P, Verweij J, Papai Z, et al. Phase III trial <strong>of</strong> two investigational<br />
schedules <strong>of</strong> ifosfamide compared with standard-dose doxorubicin in advanced<br />
or metastatic s<strong>of</strong>t tissue sarcoma: a European Organisation for<br />
Research and Treatment <strong>of</strong> Cancer S<strong>of</strong>t Tissue and Bone Sarcoma Group<br />
Study. J Clin Oncol. 2007;25:3144-3150.<br />
7. Grier HE, Krailo MD, Tarbell NJ, et al. Addition <strong>of</strong> ifosfamide and<br />
etoposide to standard chemotherapy for Ewing’s sarcoma and primitive<br />
neuroectodermal tumor <strong>of</strong> the bone. N Engl J Med. 2003;348:694-701.<br />
650<br />
REFERENCES<br />
Research<br />
Funding<br />
GlaxoSmithKline;<br />
Johnson &<br />
Johnson;<br />
Novartis; Pfizer<br />
Novartis; Pfizer GlaxoSmithKline;<br />
Novartis; Pfizer<br />
Novartis; Pfizer Amgen; ARIAD;<br />
Bristol-Myers<br />
Squibb; Daiichi<br />
Sankyo;<br />
Genentech;<br />
Infinity; Johnson<br />
& Johnson;<br />
Novartis; Pfizer;<br />
PharmaMar<br />
SLEIJFER, JUDSON, AND DEMETRI<br />
Expert<br />
Testimony<br />
ARIAD (U);<br />
Infinity (U);<br />
Johnson &<br />
Johnson (U);<br />
Novartis (U);<br />
Pfizer (U);<br />
PharmaMar (U)<br />
Other<br />
Remuneration<br />
8. Italiano A, Ci<strong>of</strong>fi A, Maki RG, et al. Patterns <strong>of</strong> care, prognosis, and<br />
survival in patients with metastatic gastrointestinal stromal tumors (GIST)<br />
refractory to first-line imatinib and second-line sunitinib. J Clin Oncol.<br />
2011;19:615s (suppl; abstr 10044).<br />
9. Reichardt P, Montemurro H, Gelderblom H, et al. Sorafenib fourth-line<br />
treatment in imatinib-, sunitinib-, and nilotinib-resistant metastatic GIST: a<br />
retrospective analysis. J Clin Oncol. 2009;27:15s (suppl; abstr 10564).<br />
10. Kindler HL, Campbell K, Wroblewski R, et al. Sorafenib (SOR) in<br />
patients (pts) with imatinib (IM) and sunitinib (SU)-resistant (RES) gastrointestinal<br />
stromal tumors (GIST): final results <strong>of</strong> a University <strong>of</strong> Chicago<br />
phase II consortium trial. J Clin Oncol. 2011;29:607s (suppl; abstr 10009).<br />
11. Cella D, Escudier B, Rini BI, et al. Patient-reported outcomes (PROs) in<br />
a phase III AXIS trial <strong>of</strong> axitinib versus sorafenib as second-line therapy for<br />
metastatic renal cell carcinoma (mRCC). J Clin Oncol. 2011;29:290s (suppl;<br />
abstr 4504).<br />
12. National Institute for Health and <strong>Clinical</strong> Excellence. Imatinib for the<br />
Treatment <strong>of</strong> Unresectable and/or Metastatic Gastrointestinal Stromal Tumours.<br />
NICE Technology Appraisal 209. http://nice.org.uk/guidance/TA209.<br />
Accessed February 5, <strong>2012</strong>.<br />
13. Gastrointestinal Stromal Tumor Meta-Analysis Group (MetaGIST).<br />
Comparison <strong>of</strong> two doses <strong>of</strong> imatinib for the treatment <strong>of</strong> unresectable or<br />
metastatic gastrointestinal stromal tumors: a meta-analysis <strong>of</strong> 1,640 patients.<br />
J Clin Oncol. 2010;28:1247-1253.<br />
14. National Institute for Health and <strong>Clinical</strong> Excellence. Trabectedin for<br />
the Treatment <strong>of</strong> Advanced S<strong>of</strong>t Tissue Sarcoma. NICE Technology Appraisal<br />
185. http://nice.org.uk/guidance/TA185. Accessed February 5, <strong>2012</strong>.
TARGETED THERAPIES IN TARGETED OR<br />
UNTARGETED SARCOMAS<br />
CHAIR<br />
Richard F. Riedel, MD<br />
Duke Cancer Institute, Duke University Medical Center<br />
Durham, NC<br />
SPEAKERS<br />
Andrew J. Wagner, MD, PhD<br />
Dana-Farber Cancer Institute<br />
Boston, MA<br />
Robert G. Maki, MD, PhD<br />
Mount Sinai School <strong>of</strong> Medicine<br />
New York, NY
Targeted Therapy in Sarcoma: Should We Be<br />
Lumpers or Splitters?<br />
By Richard F. Riedel, MD, Robert G. Maki, MD, PhD, and Andrew J. Wagner, MD, PhD<br />
Overview: The identification <strong>of</strong> KIT as a critical driver in the<br />
pathogenesis <strong>of</strong> GI stromal tumor (GIST), and its subsequent<br />
inhibition with imatinib, have resulted in tremendous efforts to<br />
identify other potential therapeutic targets for the heterogeneous<br />
group <strong>of</strong> diseases known as sarcomas. Because <strong>of</strong> the<br />
rarity <strong>of</strong> sarcoma and the <strong>of</strong>ten limited number <strong>of</strong> patients per<br />
individual sarcoma subtype, clinical trials to date have <strong>of</strong>ten<br />
utilized unselected patient populations including multiple subtypes.<br />
Although this strategy increases the ease with which a<br />
particular trial may accrue patients, statistically significant<br />
therapeutic responses across an unselected patient popula-<br />
SARCOMAS REPRESENT a diverse group <strong>of</strong> mesenchymal<br />
neoplasm affecting approximately 13,000 individuals<br />
each year. 1 The treatment <strong>of</strong> advanced disease is<br />
particularly challenging because <strong>of</strong> the limited number <strong>of</strong><br />
effective systemic therapies. As a result <strong>of</strong> the success <strong>of</strong><br />
tyrosine kinase inhibitors (TKIs), such as imatinib in the<br />
management <strong>of</strong> GIST, investigation <strong>of</strong> other targeted therapies<br />
for this rare group <strong>of</strong> diseases has become an active<br />
and ongoing area <strong>of</strong> research. Some <strong>of</strong> the most active<br />
investigations have centered on inhibiting angiogenesis as<br />
well as the phosphoinositide 3-kinase (PI3K), Akt/mammalian<br />
target <strong>of</strong> rapamycin (mTOR), and insulin-like growth<br />
factor 1 receptor (IGF1R) oncogenic pathways. Although a<br />
number <strong>of</strong> multitargeted TKIs (MTKIs) and pathwayspecific<br />
therapies have been explored, response rates have<br />
generally been disappointing and few have been investigated<br />
in the phase III setting.<br />
Whereas a mesenchymal origin is shared between histologic<br />
subtypes <strong>of</strong> sarcoma, the molecular drivers <strong>of</strong> tumor<br />
initiation, growth, and maintenance are more varied and<br />
complex. Critical to our success in further understanding<br />
the key signals and pathways that drive pathogenesis is<br />
active communication between basic scientists and clinical<br />
investigators. Although the traditional knowledge flow has<br />
been from the laboratory to the clinic, observations noted<br />
from clinical trials have, in fact, created fertile ground for<br />
scientific exploration and discovery in the laboratory setting.<br />
This review will highlight recent data from both the<br />
laboratory and clinical settings, including recently reported<br />
phase III studies, to support the ongoing investigation <strong>of</strong><br />
targeted therapies in the <strong>of</strong> patients with sarcoma. In<br />
addition, the rationale behind the use <strong>of</strong> specific therapies in<br />
more targeted populations <strong>of</strong> patients, with specific attention<br />
to unique sarcoma subtypes, will be provided.<br />
Targeted Therapy in Unselected Subtypes<br />
Because <strong>of</strong> the rarity <strong>of</strong> sarcoma and the desire to accrue<br />
patients in a timely fashion, prospective clinical trials have<br />
traditionally been more encompassing rather than more<br />
restrictive with respect to the inclusion <strong>of</strong> histologic subtypes.<br />
The result is <strong>of</strong>ten a mixture <strong>of</strong> histologies i n different<br />
proportions from one study to the next, including<br />
translocation-associated (e.g., synovial sarcoma) and aneuploid<br />
subtypes (e.g., leiomyosarcoma and undifferentiated<br />
pleomorphic sarcoma), with heterogeneous biologic mecha-<br />
652<br />
tion are <strong>of</strong>ten limited. Furthermore, in the absence <strong>of</strong> biologic<br />
correlatives, the identification <strong>of</strong> significant activity and subsequent<br />
interpretation <strong>of</strong> clinical trial results utilizing targeted<br />
therapies for this patient population have been challenging.<br />
However, hints have emerged, on the basis <strong>of</strong> preclinical and<br />
clinical observations, to suggest that certain targeted therapeutic<br />
approaches are appropriate in select histologic subtypes.<br />
This brief review will highlight data supporting the use<br />
<strong>of</strong> targeted therapy in both unselected and selected sarcoma<br />
patient populations.<br />
nisms driving their formation and growth. To date, the most<br />
robust areas <strong>of</strong> investigation have included therapies targeting<br />
mTOR, angiogenesis, and IGF1R.<br />
mTOR Inhibition and Sarcomas<br />
The mTOR pathway plays an important role in cell growth<br />
and proliferation and is an active target for developing<br />
cancer therapeutics. Activation <strong>of</strong> the mTOR pathway has<br />
been identified in a number <strong>of</strong> human cancers, including<br />
sarcoma. 2 A variety <strong>of</strong> mTOR inhibitors have been developed<br />
and are in active investigation.<br />
A phase II study <strong>of</strong> ridaforolimus, a nonprodrug rapamycin<br />
analog, in 212 heavily pretreated patients with advanced<br />
s<strong>of</strong>t tissue and bone sarcomas was reported. 3 Patients received<br />
ridaforolimus 12.5 mg intravenously once daily for 5<br />
days every 2 weeks. A “benefit rate” <strong>of</strong> 29% (complete<br />
response [CR], partial response [PR], or stable disease [SD]<br />
lasting 16 weeks or longer) with five PRs was observed. The<br />
median progression-free survival (PFS) and overall survival<br />
(OS) were 15.3 and 40 weeks, respectively.<br />
An oral formulation <strong>of</strong> ridaforolimus was also developed<br />
and subsequently moved into the clinical trial setting as part<br />
a maintenance strategy in sarcoma. Initial results from the<br />
randomized, multicenter, phase III SUCCEED (Sarcoma<br />
Multi-Center <strong>Clinical</strong> Evaluation <strong>of</strong> the Efficacy <strong>of</strong> Ridaforolimus)<br />
trial were reported at ASCO’s 47 th Annual Meeting<br />
(June 4–8, 2011, Chicago, IL). 4 Patients with advanced s<strong>of</strong>t<br />
tissue or bone sarcomas were randomly assigned 1:1 to<br />
receive oral ridaforolimus (40 mg daily for 5 days each week)<br />
or placebo as a maintenance approach in the setting <strong>of</strong> SD or<br />
responsive disease after at least four cycles but no more than<br />
12 months <strong>of</strong> chemotherapy. This was one <strong>of</strong> the largest<br />
randomized controlled trials performed in sarcoma to date<br />
and included over 700 patients. Median PFS, the primary<br />
outcome <strong>of</strong> interest, was 17.7 weeks with ridaforolimus<br />
compared with 14.6 weeks with placebo (hazard ratio [HR]<br />
� 0.72, p � 0.0001), as assessed by independent review. PFS<br />
From the Duke Cancer Institute, Duke University Medical Center, Durham, NC; Mount<br />
Sinai School <strong>of</strong> Medicine, New York, NY; Dana Farber Cancer Institute, Boston, MA.<br />
Authors’ disclosures <strong>of</strong> potential conflicts <strong>of</strong> interest are found at the end <strong>of</strong> this article.<br />
Address reprint requests to Richard F. Riedel MD, Box 3198 DUMC, Durham, NC 27710;<br />
e-mail: richard.riedel@duke.edu.<br />
© <strong>2012</strong> by <strong>American</strong> <strong>Society</strong> <strong>of</strong> <strong>Clinical</strong> <strong>Oncology</strong>.<br />
1092-9118/10/1-10
TARGETED THERAPY IN SARCOMA<br />
rates at 3- and 6-months were 70% and 34%, respectively,<br />
with ridaforolimus compared with 54% and 23%, respectively,<br />
with placebo. The “benefit rate” (CR, PR, and SD<br />
lasting 4 months or longer) was 41% with ridaforolimus<br />
versus 29% with placebo (p � 0.0009), and best target<br />
response was a mean decrease <strong>of</strong> 1.3% with ridaforolimus<br />
compared with a mean increase <strong>of</strong> 10.3% with placebo (p �<br />
0.0001). A nonsignificant trend in OS was seen favoring<br />
ridaforolimus (21.4 vs. 19.2 months, HR � 0.88, p � 0.23).<br />
Although the results from this study suggest that mTOR<br />
maintenance therapy may be considered an alternative to<br />
the more traditional “watch and wait” approach, it is important<br />
to consider associated adverse events (AEs) with this<br />
medication. In the data presented, nearly two-thirds <strong>of</strong><br />
patients experienced at least 1 grade 3 or worse AE, including<br />
stomatitis (9%), infection (6%), thrombocytopenia (10%),<br />
anemia (7%), and hyperglycemia (7%). When considering all<br />
grades <strong>of</strong> AEs in enrolled patients, 61% experienced stomatitis;<br />
52% infections; approximately one-third <strong>of</strong> patients<br />
experienced fatigue, thrombocytopenia, diarrhea, cough,<br />
and rash; and approximately one-quarter <strong>of</strong> patients experienced<br />
anemia, hypertriglyceridemia, and hypercholesterolemia.<br />
Even though these numbers represent events<br />
occurring in a minority <strong>of</strong> patients, the potential impact <strong>of</strong><br />
lesser grade 1 to 2 events occurring in the context <strong>of</strong> the<br />
maintenance setting should not be overlooked and results<br />
from quality-<strong>of</strong>-life questionnaires obtained as part <strong>of</strong> exploratory<br />
analyses are anticipated.<br />
In an effort to identify predictors <strong>of</strong> response to ridaforolimus,<br />
additional exploratory analyses were performed and<br />
recently reported. 5 Post hoc analyses revealed greater median<br />
PFS in those patients who experienced grade 2 or worse<br />
stomatitis compared with those who did not, or those who<br />
received placebo. In addition, an analysis <strong>of</strong> the cohort <strong>of</strong><br />
patients with SD at study entry suggested an increased<br />
benefit in those patients with truly stable or early growth<br />
compared with those with minor responses.<br />
The role <strong>of</strong> mTOR inhibition in the treatment <strong>of</strong> metastatic<br />
sarcoma continues to be developed. Combination therapy<br />
with conventional cytotoxics, as well as combination<br />
with other targeted therapies, is being considered. Assessing<br />
alternative pathway activation in mTOR inhibitor–resistant<br />
disease will be important, and will likely guide future trials.<br />
KEY POINTS<br />
● Sarcomas are a rare, heterogeneous cancer in need <strong>of</strong><br />
new therapies.<br />
● <strong>Clinical</strong> trials to date have focused largely on unselected<br />
patient populations including multiple subtypes.<br />
● A better understanding <strong>of</strong> the molecular mechanisms<br />
driving the pathogenesis <strong>of</strong> sarcoma is needed.<br />
● <strong>Clinical</strong> trials with integrated biologic correlatives<br />
are more likely to identify subtypes likely to benefit<br />
from targeted therapy.<br />
● Collaboration between basic scientists and clinical<br />
investigators is critical.<br />
Angiogenesis Inhibition and Sarcomas<br />
The use <strong>of</strong> antiangiogenic agents in sarcoma is not a new<br />
concept. The mixed clinical results in cancer studies suggest<br />
that the road to understanding the key elements <strong>of</strong> angiogenesis<br />
and where one may safely intervene in the patient<br />
with cancer is far from straightforward.<br />
Recent focus has been on therapies directed against vascular<br />
endothelial growth factor (VEGF) and its receptors<br />
(VEGFRs). The combination <strong>of</strong> anti-VEGF monoclonal antibody<br />
bevacizumab with doxorubicin was explored in a small<br />
phase II study. 6 The combination was not obviously more<br />
active than doxorubicin itself, and was associated with<br />
toxicity. A more recent phase I study <strong>of</strong> gemcitabine and<br />
docetaxel with bevacizumab showed a 31% RECIST (Response<br />
Evaluation Criteria in Solid Tumors) response rate. 7<br />
It is not clear whether this response rate reflects the activity<br />
<strong>of</strong> the gemcitabine-docetaxel combination alone, or synergy<br />
with bevacizumab. We hope this question will be answered<br />
for leiomyosarcoma in a randomized phase III clinical trial<br />
from the Gynecologic <strong>Oncology</strong> Group (clinicaltrials.gov<br />
identifier NCT01012297) comparing gemcitabine and docetaxel<br />
with or without bevacizumab. For angiosarcoma, the<br />
situation may be slightly different, with greater RECIST<br />
response rates (12%) observed for bevacizumab. 8<br />
MTKIs have shown modest benefit in specific non-GIST<br />
sarcomas as single agents, but their multitargeted nature<br />
makes ascribing the activity <strong>of</strong> one kinase or another to the<br />
growth <strong>of</strong> a specific sarcoma difficult. There are few data to<br />
suggest that any MTKI has substantial activity in sarcomas<br />
such as leiomyosarcoma or unclassified pleomorphic sarcoma/malignant<br />
fibrous histiocytoma (UPS/MFH). The phase<br />
II studies <strong>of</strong> sunitinib, sorafenib, and pazopanib yielded very<br />
low RECIST response rates across an unselected patient<br />
population. 9-12 Similar to bevacizumab, a higher response<br />
rate (14%) for sorafenib was observed in angiosarcomas.<br />
Nonetheless, MTKIs may have measurable activity, as<br />
demonstrated in a randomized clinical trial <strong>of</strong> pazopanib<br />
compared with placebo in advanced s<strong>of</strong>t tissue sarcoma<br />
subtypes. 13 The PALETTE (Pazopanib Explored in S<strong>of</strong>t-<br />
Tissue Sarcoma) study explored 369 patients randomly<br />
assigned to receive pazopanib 800 mg daily or placebo until<br />
disease progression. Median PFS was 20 weeks for pazopanib<br />
compared with 7 weeks for placebo (HR � 0.31, p � 0.0001). An<br />
interim analysis showed a nonsignificant improvement in<br />
median OS for pazopanib (11.9 vs. 10.4 months, HR � 0.83,<br />
p � 0.18), in this study with a cross-over design.<br />
Similarly, specific sarcoma subtypes may demonstrate<br />
sensitivity to antiangiogenic agents. In addition to angiosarcoma<br />
and VEGF inhibitors, alveolar s<strong>of</strong>t-part sarcoma can<br />
respond to VEGFR inhibitors cediranib 14 or sunitinib, 15 and<br />
solitary fibrous tumors demonstrate at least hints <strong>of</strong> activity<br />
to sunitinib 16 or the combination <strong>of</strong> bevacizumab and temozolomide.<br />
17 The precise mechanism <strong>of</strong> antiangiogenic agents<br />
in these sarcoma subtypes remains unknown.<br />
If mTOR can be construed to be a proangiogenic kinase, 18<br />
then the results <strong>of</strong> phase II and phase III studies are<br />
less-than-convincing evidence for the use <strong>of</strong> first-generation<br />
mTOR inhibitors as antiangiogenic agents. 19 These agents<br />
do not inhibit the reflex activation <strong>of</strong> Akt by mTOR complex<br />
TORC2 that is common to all presently available allosteric<br />
TORC1 complex inhibitors. As a result, newer direct TORC1<br />
653
and TORC2 inhibitors could potentially demonstrate greater<br />
activity as antiangiogenic compounds.<br />
Other antiangiogenic agents have shown hints <strong>of</strong> activity<br />
in sarcomas and other cancers, without convincing survival<br />
improvement. The thrombospondin mimetic ABT-510 demonstrated<br />
only one RECIST PR in 82 patients receiving<br />
treatment, although time to progression suggested some<br />
activity <strong>of</strong> the compound. 20 Vascular-disrupting agent combretastatin<br />
A4, a tubulin antagonist and antiangiogenic<br />
agent, showed only modest clinical activity as a single agent<br />
and in combination with cytotoxic chemotherapy agents in<br />
phase I studies, although it appeared to have biologic effects<br />
by dynamic contrast-enhanced magnetic resonance imaging<br />
and other correlative studies. 21,22 New derivatives <strong>of</strong> combretastatin<br />
are being examined preclinically as well as in<br />
phase I studies. A newer approach involves targeting <strong>of</strong><br />
other kinases involved in angiogenesis, including TIE1 and<br />
TIE2, with monoclonal antibodies, peptide-antibody conjugates,<br />
or small molecules, some <strong>of</strong> which have reached phase<br />
I clinical trials.<br />
Targeted Therapy in Selected Subtypes<br />
Although the clinical classifications <strong>of</strong> sarcomas can provide<br />
prognostic and predictive information, they rarely include<br />
the emerging molecular biologic data that may lead to<br />
advances in tumor-specific treatments. Several findings<br />
from laboratory-based studies have recently been brought to<br />
clinical investigation in specific sarcoma subtypes for which<br />
there are few known effective therapies. Although the true<br />
efficacy <strong>of</strong> these novel approaches is still unknown, the<br />
pairing <strong>of</strong> genetic and biochemical alterations with targeted<br />
therapies has generated pro<strong>of</strong>-<strong>of</strong>-concept evidence <strong>of</strong> activity<br />
in laboratory and early clinical settings.<br />
IGF1R Inhibition in Ewing Sarcoma<br />
IGF1R has been recognized as a viable target for treatment<br />
<strong>of</strong> pediatric sarcomas for 20 years. 23 For example, the<br />
role <strong>of</strong> IGF1R signaling to maintain activity <strong>of</strong> the Ewing<br />
sarcoma EWSR1-FLI1 translocation product indicated that<br />
blockade <strong>of</strong> this pathway would consistently block Ewing<br />
sarcoma growth. It was not until there was availability <strong>of</strong><br />
agents more potent than somatostatin analogs at blocking<br />
the IGF1R signaling axis that there has been an opportunity<br />
to examine this hypothesis. In what is a common story<br />
regarding the sophistication <strong>of</strong> cancer, IGF1R blockade,<br />
which appeared to be nearly a perfect tumor-specific target<br />
in Ewing sarcoma, has become a more complicated story.<br />
The avalanche <strong>of</strong> data regarding Ewing sarcoma and<br />
IGF1R inhibitors has come from the observation <strong>of</strong> responses<br />
from limited number <strong>of</strong> patients in the phase I<br />
development <strong>of</strong> the monoclonal antibodies targeting the<br />
receptor. Despite differences in specific antibody characteristics,<br />
a consistent RECIST response rate has been observed.<br />
For example, AMG479 IV every 2 weeks yielded a long-term<br />
RECIST CR and one PR in Ewing sarcoma and a PR in a<br />
patient with a neuroendocrine tumor, although 10 other<br />
Ewing sarcoma patients did not experience response. 24 Two<br />
<strong>of</strong> 13 assessable patients with Ewing sarcoma experienced<br />
RECIST PR while receiving intravenous figitumumab in a<br />
phase I expanded cohort study; others showed disease<br />
shrinking but did not experience PR. 25 A phase I/II study <strong>of</strong><br />
cixutumumab in a pediatric population showed two (10%) <strong>of</strong><br />
654<br />
RIEDEL, MAKI, AND WAGNER<br />
20 patients with a RECIST PR at the highest dose studied (9<br />
mg/kg/week), but the median PFS for this cohort was<br />
slightly longer than 6 weeks. 26<br />
Further study <strong>of</strong> figitumumab in a pediatric Ewing sarcoma<br />
population (10 to 18 years) in the phase II portion <strong>of</strong> a<br />
phase I/II study demonstrated a 14% RECIST response rate<br />
(15 PR in 106 patients). 27 In this study, a higher baseline<br />
total or free IGF1 level was associated with longer survival<br />
than was low baseline free or total IGF1. A population <strong>of</strong><br />
patients with very high baseline IGF1 fared as poorly as<br />
those with the lowest levels <strong>of</strong> IGF1, however. A similar<br />
phase II study performed with IGF1R monoclonal antibody<br />
yielded a RECIST response rate <strong>of</strong> 10% (11 <strong>of</strong> 115 patients,<br />
10 <strong>of</strong> whom had had a bony primary site). 28 In addition,<br />
among patients with IGF1 samples available at week 6 <strong>of</strong><br />
the study, total serum IGF1 greater than 110 ng/mL at 6<br />
weeks <strong>of</strong> therapy and a greater proportional increase in total<br />
IGF1 from baseline to week 6 both predicted for improved<br />
survival.<br />
The rapid development <strong>of</strong> progressive disease observed in<br />
some patients, after evidence <strong>of</strong> early treatment benefit, 29,30<br />
speaks against the development <strong>of</strong> secondary mutations in<br />
IGF1R. Other mechanisms <strong>of</strong> resistance include heterodimerization<br />
with other receptor tyrosine kinases 31 and<br />
nuclear translocation <strong>of</strong> adapter proteins such as IRS1, 32,33<br />
and IGF1R itself, which is mediated by SUMOylation, a<br />
post-translational modification. 34 The low but consistent<br />
response rate seen for IGF1R inhibitors increases our appreciation<br />
for the multiple levels <strong>of</strong> regulation that exist to<br />
ensure cellular homeostasis. At the same time, new mechanisms<br />
complementary to kinase blockade are needed to yield<br />
better combinations <strong>of</strong> anticancer therapeutics.<br />
ALK Inhibition in Inflammatory My<strong>of</strong>ibroblastic Tumors<br />
Inflammatory my<strong>of</strong>ibroblastic tumors (IMTs) are a rare<br />
subset <strong>of</strong> sarcomas <strong>of</strong> intermediate biologic potential with a<br />
predilection for local recurrence and, much less commonly,<br />
metastasis. 35 Recurrent cytogenetic alterations involving<br />
the short arm <strong>of</strong> chromosome 2 and immunohistochemical<br />
overexpression <strong>of</strong> the ALK gene that resides in this chromosome<br />
36 has led to the identification <strong>of</strong> translocations <strong>of</strong> ALK<br />
in approximately 50% <strong>of</strong> IMTs. 37 In IMTs, ALK encodes a<br />
receptor tyrosine kinase that can be fused to one <strong>of</strong> at least five<br />
other gene partners. ALK expression in IMT reliably predicts<br />
ALK gene rearrangement. Tumors without ALK rearrangement<br />
appear to have a higher propensity for distant<br />
metastases without an increased risk <strong>of</strong> local recurrence. 38<br />
Chromosomal translocations <strong>of</strong> ALK have been identified<br />
in a subset <strong>of</strong> non–small cell lung cancer (NSCLC), 39 and<br />
point mutations in ALK have been reported in neuroblastoma.<br />
40 The discovery <strong>of</strong> activating structural rearrangements<br />
and point mutations among different lineages<br />
strongly supported the oncogenic role <strong>of</strong> ALK, which was<br />
corroborated in a variety <strong>of</strong> in vitro models as well. Concurrent<br />
with many <strong>of</strong> these discoveries was the early-phase<br />
clinical development <strong>of</strong> crizotinib (formerly PF-02341066),<br />
an orally administered inhibitor <strong>of</strong> the MET and ALK<br />
tyrosine kinases. Dramatic clinical activity was observed in<br />
patients with ALK-rearranged NSCLC 41 ultimately leading<br />
to regulatory approval for this indication.<br />
Two patients with IMTs were enrolled in a phase I study<br />
<strong>of</strong> crizotinib. 42 A sustained PR was observed in a patient<br />
with an ALK-rearranged tumor, whereas no activity was
TARGETED THERAPY IN SARCOMA<br />
seen in the patient without the translocation. Of note, a<br />
F1174L point mutation was observed in the first patient,<br />
who developed secondary resistance to crizotinib. This point<br />
mutation was previously identified in neuroblastoma specimens<br />
and demonstrated in vitro resistance to crizotinib but<br />
sensitivity to TAE684, a chemically distinct ALK inhibitor<br />
that has not yet been clinically tested in patients with<br />
IMT. 43<br />
Together, these genetic, pathologic, and limited clinical<br />
findings demonstrate that a subset <strong>of</strong> IMT have activating<br />
rearrangements <strong>of</strong> ALK, and provide preliminary evidence<br />
that IMT patients with ALK rearrangements may derive<br />
clinical benefit from ALK inhibitors. Further clinical experience<br />
with ALK inhibitors in this disease will strengthen<br />
the translation <strong>of</strong> these observations.<br />
mTOR Activation in Perivascular Epithelioid<br />
Cell Tumor<br />
The perivascular epithelioid cell tumor (PEComa) family<br />
consists <strong>of</strong> mesenchymal neoplasms with epithelioid morphology<br />
and myomelanocytic differentiation. 44 These tumors<br />
present as small pulmonary nodules with cystic lung<br />
destruction in lymphangioleiomyomatosis (LAM), benign<br />
renal masses with vascular, myogenic, and adipocytic differentiation<br />
in angiomyolipoma (AML), or PEComa, an epithelioid<br />
tumor with nests and sheets <strong>of</strong> cells with clear-togranular<br />
eosinophilic cytoplasm arising most commonly in<br />
the uterus, retroperitoneum, and somatic s<strong>of</strong>t tissues. Although<br />
LAM, AML, and PEComa may arise sporadically,<br />
LAM and AML in particular are also found in patients with<br />
tuberous sclerosis, an autosomal dominant syndrome caused<br />
by inherited mutations in the TSC1 or TSC2 tumor suppressor<br />
genes. Inactivation <strong>of</strong> the TSC1/2 complex leads to hyperactivation<br />
<strong>of</strong> mTOR and dysregulated cellular proliferation.<br />
An evaluation <strong>of</strong> the mTOR inhibitor sirolimus in patients<br />
AML or LAM 45 showed modest reduction in size <strong>of</strong> AML and<br />
improvement in lung function, supporting the mechanistic<br />
role <strong>of</strong> mTOR activity in these diseases. On the basis <strong>of</strong><br />
the morphologic similarity <strong>of</strong> PEComa to AML and LAM,<br />
biochemical evidence <strong>of</strong> enhanced mTOR activity in<br />
PEComa, 46,47 and chromosomal abnormalities at the TSC1<br />
or TSC2 loci, the effects <strong>of</strong> sirolimus was explored in three<br />
consecutive patients with malignant PEComa. 48 All three<br />
patients experienced clinical benefit, but with duration varying<br />
from several months to several years. Retrospective<br />
analysis <strong>of</strong> tumor specimens demonstrated loss <strong>of</strong> expression<br />
<strong>of</strong> TSC2 in all tumors, but not in healthy tissues, and loss <strong>of</strong><br />
both copies <strong>of</strong> TSC1 in one tumor. These findings support<br />
the use <strong>of</strong> mTOR inhibitors in this disease. Similar clinical<br />
responses to mTOR inhibitors have been reported in other<br />
case reports, 49,50 but not all patients have derived such<br />
benefit. Further studies identifying predictive factors and<br />
mediators <strong>of</strong> resistance are needed.<br />
Hedgehog Signaling in Chondrosarcoma<br />
Chondrosarcomas are malignant neoplasms that most<br />
commonly arise in bone and have cartilaginous differentiation.<br />
No effective systemic therapies have been described for<br />
unresectable or metastatic disease. The hedgehog (HH)<br />
signaling pathway is implicated in normal chondrocyte development<br />
as well as in malignant transformation. Oncogenic<br />
activation <strong>of</strong> the HH pathway can result from multiple<br />
insults, including loss-<strong>of</strong>-function mutations in Patched (ligand)<br />
that lead to constitutive activation <strong>of</strong> the receptor,<br />
Smoothened (SMO) as well as dysregulated expression <strong>of</strong><br />
HH ligands leading to paracrine or autocrine tumor cell<br />
stimulation.<br />
Constitutively activated HH signaling has been identified<br />
in chondrosarcoma tumors. 51 IPI-926, an orally available<br />
SMO antagonist, has been evaluated on primary chondrosarcoma<br />
xenograft model. 52 Treatment <strong>of</strong> established xenografts<br />
with IPI-926 resulted in substantially smaller tumor<br />
sizes compared with chemotherapy- or control-treated animals.<br />
Consistent with HH pathway modulation, reduction in<br />
expression <strong>of</strong> target genes was also observed in both the<br />
tumor cells and the surrounding tumor-associated stroma.<br />
These laboratory findings formed the basis <strong>of</strong> an ongoing<br />
phase II study <strong>of</strong> IPI-926 compared with placebo in patients<br />
with unresectable or metastatic chondrosarcoma.<br />
Recently, mutations in the Krebs cycle enzymes isocitrate<br />
dehydrogenase 1 (IDH1) and isocitrate dehydrogenase 2<br />
(IDH2) have been identified in more than 50% <strong>of</strong> cartilaginous<br />
tumors. 53 Similar mutations have been suggested to<br />
confer an oncogenic neoactivity in gliomas and acute myelogenous<br />
leukemia. Whereas the precise role <strong>of</strong> IDH1 and IDH2<br />
mutations has yet to be described in chondrosarcoma behavior,<br />
these, too, may present an opportunity for targeted<br />
therapeutic approaches that can be explored when suitable<br />
drugs are available for clinical study.<br />
Conclusion<br />
The examples described in this article validate the concept<br />
that deeper understanding <strong>of</strong> the molecular mechanisms<br />
associated with the development <strong>of</strong> sarcoma positively affects<br />
patient outcomes. Until our understanding <strong>of</strong> individual<br />
sarcoma subtypes evolves further, it is essential to<br />
conduct clinical trials with biologic correlatives designed to<br />
gain a better understanding <strong>of</strong> potential subtypes likely to<br />
benefit from novel targeted therapies. The rarity and heterogeneity<br />
<strong>of</strong> s<strong>of</strong>t tissue and bone sarcoma necessitates that<br />
our basic scientists and clinical colleagues continue collaboration<br />
to bring observations from both preclinical and clinical<br />
realms to the forefront to more rapidly provide effective<br />
therapies to our patients.<br />
655
Authors’ Disclosures <strong>of</strong> Potential Conflicts <strong>of</strong> Interest<br />
Author<br />
Employment or<br />
Leadership<br />
Positions<br />
Consultant or<br />
Advisory Role<br />
Stock<br />
Ownership Honoraria<br />
Richard F. Riedel Merck ARIAD; Novartis;<br />
Pfizer<br />
Robert G. Maki Bayer; Eisai;<br />
Lilly; Novartis<br />
Andrew J. Wagner Astex<br />
Therapeutics (U);<br />
EMD Serono;<br />
Genentech;<br />
Infinity;<br />
Morphotek;<br />
Novartis; Pfizer;<br />
San<strong>of</strong>i<br />
1. Jemal A, Siegel R, Xu J, et al. Cancer statistics, 2010. CA Cancer J Clin.<br />
2010;60:277-300.<br />
2. Baird K, Davis S, Antonescu CR, et al. Gene expression pr<strong>of</strong>iling <strong>of</strong><br />
human sarcomas: insights into sarcoma biology. Cancer Res. 2005;65:9226-<br />
9235.<br />
3. Chawla SP, Staddon AP, Baker LH, et al. Phase II study <strong>of</strong> the<br />
mammalian target <strong>of</strong> rapamycin inhibitor ridaforolimus in patients with<br />
advanced bone and s<strong>of</strong>t tissue sarcomas. J Clin Oncol. <strong>2012</strong>;30:78-84.<br />
4. Chawla SP, Blay J, Ray-Coquard IL, et al. Results <strong>of</strong> the phase III,<br />
placebo-controlled trial (SUCCEED) evaluating the mTOR inhibitor ridaforolimus<br />
(R) as maintenance therapy in advanced sarcoma patients (pts)<br />
following clinical benefit from prior standard cytotoxic chemotherapy (CT).<br />
J Clin Oncol. 2011;29(suppl; abstr 10005).<br />
5. Demetri GD, Chawla SP, Ray-Coquard I, et al. Exploratory analysis <strong>of</strong><br />
potential predictive markers to identify sensitive/responder sarcoma patients<br />
with ridaforolimus in the phase 3 randomized placebo-controlled trial (SUC-<br />
CEED). Poster presented at: Combined Meeting <strong>of</strong> the Connective Tissue<br />
<strong>Oncology</strong> <strong>Society</strong>/Musculoskeletal Tumor <strong>Society</strong>; October 2011; Chicago, IL.<br />
6. D’Adamo DR, Anderson SE, Albritton K, et al. Phase II study <strong>of</strong><br />
doxorubicin and bevacizumab for patients with metastatic s<strong>of</strong>t-tissue sarcomas.<br />
J Clin Oncol. 2005;23:7135-7142.<br />
7. Verschraegen CF, Arias-Pulido H, Lee SJ, et al. Phase IB study <strong>of</strong> the<br />
combination <strong>of</strong> docetaxel, gemcitabine, and bevacizumab in patients with<br />
advanced or recurrent s<strong>of</strong>t tissue sarcoma: the Axtell regimen. Ann Oncol.<br />
Epub 2011 Jul 11.<br />
8. Agulnik M, Okuno SH, Von Mehren M, et al. An open-label multicenter<br />
phase II study <strong>of</strong> bevacizumab for the treatment <strong>of</strong> angiosarcoma. J Clin<br />
Oncol. 2009;27:XXX (suppl; abstr 10522).<br />
9. Maki RG, D’Adamo DR, Keohan ML, et al. Phase II study <strong>of</strong> sorafenib in<br />
patients with metastatic or recurrent sarcomas. J Clin Oncol. 2009;27:3133-<br />
3140.<br />
10. George S, Merriam P, Maki RG, et al. Multicenter phase II trial <strong>of</strong><br />
sunitinib in the treatment <strong>of</strong> nongastrointestinal stromal tumor sarcomas.<br />
J Clin Oncol. 2009;27:3154-3160.<br />
11. Sleijfer S, Ray-Coquard I, Papai Z, et al. Pazopanib, a multikinase<br />
angiogenesis inhibitor, in patients with relapsed or refractory advanced s<strong>of</strong>t<br />
tissue sarcoma: a phase II study from the European organisation for research<br />
and treatment <strong>of</strong> cancer-s<strong>of</strong>t tissue and bone sarcoma group (EORTC study<br />
62043). J Clin Oncol. 2009;27:3126-3132.<br />
12. Grignani G, Palmerini E, Dileo P, et al. A phase II trial <strong>of</strong> sorafenib in<br />
relapsed and unresectable high-grade osteosarcoma after failure <strong>of</strong> standard<br />
multimodal therapy: an Italian Sarcoma Group study. Ann Oncol. <strong>2012</strong>;23:<br />
508-516.<br />
13. Van Der Graaf WT, Blay J, Chawla SP, et al. PALETTE: a randomized,<br />
double-blind, phase III trial <strong>of</strong> pazopanib versus placebo in patients with<br />
s<strong>of</strong>t-tissue sarcoma whose disease has progressed during or following prior<br />
chemotherapy—an EORTC STBSG Global Network Study (EORTC 62072).<br />
J Clin Oncol. 2011;29 (suppl; abstr LBA002).<br />
14. Gardner K, Judson I, Leahy M, et al. Activity <strong>of</strong> cediranib, a highly<br />
potent and selective VEGF signaling inhibitor, in alveolar s<strong>of</strong>t part sarcoma.<br />
J Clin Oncol. 2009;27:15s (suppl; abstr 10523)<br />
15. Stacchiotti S, Negri T, Zaffaroni N, et al. Sunitinib in advanced alveolar<br />
s<strong>of</strong>t part sarcoma; evidence <strong>of</strong> a direct antitumor effect. Ann Oncol. 2011;22:<br />
1682-1690.<br />
16. Stacchiotti S, Negri T, Palassini E, et al. Sunitinib malate and figitu-<br />
656<br />
REFERENCES<br />
Novartis;<br />
Ziopharm<br />
<strong>Oncology</strong><br />
Research<br />
Funding<br />
ARIAD; Novartis;<br />
Ziopharm<br />
<strong>Oncology</strong><br />
Pfizer; Roche<br />
ArQule;<br />
Genentech;<br />
Prolexys<br />
RIEDEL, MAKI, AND WAGNER<br />
Expert<br />
Testimony<br />
Other<br />
Remuneration<br />
mumab in solitary fibrous tumor; patterns and molecular bases <strong>of</strong> tumor<br />
response. Mol Cancer Ther. 2010;9:1286-1297.<br />
17. Park MS, Patel SR, Ludwig JA, et al. Activity <strong>of</strong> temozolomide and<br />
bevacizumab in the treatment <strong>of</strong> locally advanced, recurrent, and metastatic<br />
hemangiopericytoma and malignant solitary fibrous tumor. Cancer. 2011;117:<br />
4939-4947.<br />
18. Miriuka SG, Rao V, Peterson M, et al. mTOR inhibition induces<br />
endothelial progenitor cell death. Am J Transplant. 2006;6:2069-2079.<br />
19. Dancey JE, Monzon J. Ridaforolimus: a promising drug in the treatment<br />
<strong>of</strong> s<strong>of</strong>t-tissue sarcoma and other malignancies. Future Oncol. 2011;7:<br />
827-839.<br />
20. Baker LH, Rowinsky EK, Mendelson D, et al. Randomized, phase II<br />
study <strong>of</strong> the thrombospondin-1-mimetic angiogenesis inhibitor ABT-510 in<br />
patients with advanced s<strong>of</strong>t tissue sarcoma. J Clin Oncol. 2008;26:5583-5588.<br />
21. Dowlati A, Robertson K, Cooney M, et al. A phase I pharmacokinetic<br />
and translational study <strong>of</strong> the novel vascular targeting agent combretastatin<br />
a-4 phosphate on a single-dose intravenous schedule in patients with advanced<br />
cancer. Cancer Res. 2002;62:3408-3416.<br />
22. Bilenker JH, Flaherty KT, Rosen M, et al. Phase I trial <strong>of</strong> combretastatin<br />
a-4 phosphate with carboplatin. Clin Cancer Res. 2005;11:1527-1533.<br />
23. van Valen F, Winkelmann W, Jurgens H. Type I and type II insulin-like<br />
growth factor receptors and their function in human Ewing’s sarcoma cells. J<br />
Cancer Res Clin Oncol. 1992;118:269-275.<br />
24. Tolcher AW, Sarantopoulos J, Patnaik A, et al. Phase I, pharmacokinetic,<br />
and pharmacodynamic study <strong>of</strong> AMG 479, a fully human monoclonal<br />
antibody to insulin-like growth factor receptor 1. J Clin Oncol. 2009;27:5800-<br />
5807.<br />
25. Olmos D, Postel-Vinay S, Molife LR, et al. Safety, pharmacokinetics,<br />
and preliminary activity <strong>of</strong> the anti-IGF-1R antibody figitumumab (CP-<br />
751,871) in patients with sarcoma and Ewing’s sarcoma: a phase 1 expansion<br />
cohort study. Lancet Oncol. 2010;11:129-135.<br />
26. Malempati S, Weigel B, Ingle AM, et al. Phase I/II trial and pharmacokinetic<br />
study <strong>of</strong> cixutumumab in pediatric patients with refractory solid<br />
tumors and Ewing sarcoma: a report from the Children’s <strong>Oncology</strong> Group.<br />
J Clin Oncol. <strong>2012</strong>;30:256-262.<br />
27. Juergens H, Daw NC, Geoerger B, et al. Preliminary efficacy <strong>of</strong><br />
the anti-insulin-like growth factor type 1 receptor antibody figitumumab<br />
in patients with refractory Ewing sarcoma. J Clin Oncol. 2011;29:4534-<br />
4540.<br />
28. Pappo AS, Patel SR, Crowley J, et al. R1507, a monoclonal antibody to<br />
the insulin-like growth factor 1 receptor, in patients with recurrent or refractory<br />
Ewing sarcoma family <strong>of</strong> tumors: Results <strong>of</strong> a phase II Sarcoma Alliance for<br />
Research through Collaboration study. J Clin Oncol. 2011;29:4541-4547.<br />
29. Huang F, Greer A, Hurlburt W, et al. The mechanisms <strong>of</strong> differential<br />
sensitivity to an insulin-like growth factor-1 receptor inhibitor (BMS-536924)<br />
and rationale for combining with EGFR/HER2 inhibitors. Cancer Res. 2009;<br />
69:161-170.<br />
30. Potratz JC, Saunders DN, Wai DH, et al. Synthetic lethality screens<br />
reveal RPS6 and MST1R as modifiers <strong>of</strong> insulin-like growth factor-1 receptor<br />
inhibitor activity in childhood sarcomas. Cancer Res. 2010;70:8770-8781.<br />
31. Nahta R, Yuan LX, Zhang B, et al. Insulin-like growth factor-I receptor/<br />
human epidermal growth factor receptor 2 heterodimerization contributes to<br />
trastuzumab resistance <strong>of</strong> breast cancer cells. Cancer Res. 2005;65:11118-<br />
11128.<br />
32. Del Valle L, Wang JY, Lassak A, et al. Insulin-like growth factor I
TARGETED THERAPY IN SARCOMA<br />
receptor signaling system in JC virus T antigen-induced primitive neuroectodermal<br />
tumors-medulloblastomas. J Neurovirol. 2002;2:138-147 (suppl).<br />
33. Lassak A, Del Valle L, Peruzzi F, et al. Insulin receptor substrate 1<br />
translocation to the nucleus by the human JC virus T-antigen. J Biol Chem.<br />
2002;277:17231-17238.<br />
34. Sehat B, T<strong>of</strong>igh A, Lin Y, et al: SUMOylation mediates the nuclear<br />
translocation and signaling <strong>of</strong> the IGF-1 receptor. Sci Signal. 2010;3:ra10<br />
35. Gleason BC, Hornick JL. Inflammatory my<strong>of</strong>ibroblastic tumours:<br />
where are we now? J Clin Pathol. 2008;61:428-437.<br />
36. Griffin CA, Hawkins AL, Dvorak C, et al. Recurrent involvement <strong>of</strong> 2p23<br />
in inflammatory my<strong>of</strong>ibroblastic tumors. Cancer Res. 1999;59:2776-2780.<br />
37. Lawrence B, Perez-Atayde A, Hibbard MK, et al. TPM3-ALK and<br />
TPM4-ALK oncogenes in inflammatory my<strong>of</strong>ibroblastic tumors. Am J Pathol.<br />
2000;157:377-384.<br />
38. C<strong>of</strong>fin CM, Hornick JL, Fletcher CD. Inflammatory my<strong>of</strong>ibroblastic<br />
tumor: Comparison <strong>of</strong> clinicopathologic, histologic, and immunohistochemical<br />
features including ALK expression in atypical and aggressive cases. Am J<br />
Surg Pathol. 2007;31:509-520.<br />
39. Soda M, Choi YL, Enomoto M, et al. Identification <strong>of</strong> the transforming<br />
EML4-ALK fusion gene in non-small-cell lung cancer. Nature. 2007;448:561-<br />
566.<br />
40. Janoueix-Lerosey I, Lequin D, Brugieres L, et al. Somatic and germline<br />
activating mutations <strong>of</strong> the ALK kinase receptor in neuroblastoma. Nature.<br />
2008;455:967-970.<br />
41. Shaw AT, Yeap BY, Mino-Kenudson M, et al. <strong>Clinical</strong> features and<br />
outcome <strong>of</strong> patients with non-small-cell lung cancer who harbor EML4-ALK.<br />
J Clin Oncol. 2009;27:4247-4253.<br />
42. Butrynski JE, D’Adamo DR, Hornick JL, et al. Crizotinib in ALKrearranged<br />
inflammatory my<strong>of</strong>ibroblastic tumor. N Engl J Med. 2010;363:<br />
1727-1733.<br />
43. Sasaki T, Okuda K, Zheng W, et al. The neuroblastoma-associated<br />
F1174L ALK mutation causes resistance to an ALK kinase inhibitor in<br />
ALK-translocated cancers. Cancer Res. 2010;70:10038-10043.<br />
44. Hornick JL, Fletcher CD. PEComa: What do we know so far? Histopathology.<br />
2006;48:75-82.<br />
45. Bissler JJ, McCormack FX, Young LR, et al. Sirolimus for angiomyolipoma<br />
in tuberous sclerosis complex or lymphangioleiomyomatosis. N Engl<br />
J Med. 2008;358:140-151.<br />
46 Kenerson H, Folpe AL, Takayama TK, et al. Activation <strong>of</strong> the mTOR<br />
pathway in sporadic angiomyolipomas and other perivascular epithelioid cell<br />
neoplasms. Hum Pathol. 2007;38:1361-1371.<br />
47. Pan CC, Chung MY, Ng KF, et al. Constant allelic alteration on<br />
chromosome 16p (TSC2 gene) in perivascular epithelioid cell tumour<br />
(PEComa): genetic evidence for the relationship <strong>of</strong> PEComa with angiomyolipoma.<br />
J Pathol. 2008;214:387-393.<br />
48. Wagner AJ, Malinowska-Kolodziej I, Morgan JA, et al. <strong>Clinical</strong> activity<br />
<strong>of</strong> mTOR inhibition with sirolimus in malignant perivascular epithelioid cell<br />
tumors: targeting the pathogenic activation <strong>of</strong> mTORC1 in tumors. J Clin<br />
Oncol. 2010;28:835-840.<br />
49. Italiano A, Delcambre C, Hostein I, et al. Treatment with the mTOR<br />
inhibitor temsirolimus in patients with malignant PEComa. Ann Oncol.<br />
2010;21:1135-1137.<br />
50. Wolff N, Kabbani W, Bradley T, et al. Sirolimus and temsirolimus for<br />
epithelioid angiomyolipoma. J Clin Oncol. 2010;28:e65-e68.<br />
51. Tiet TD, Hopyan S, Nadesan P, et al. Constitutive hedgehog signaling<br />
in chondrosarcoma up-regulates tumor cell proliferation. Am J Pathol. 2006;<br />
168:321-330.<br />
52. Campbell VT, Nadeson PP, Wang Y, et al. Direct targeting <strong>of</strong> the<br />
Hedgehog pathway in primary chondrosarcoma xenografts with the Smoothened<br />
inhibitor IPI-926. Presented at: Paper presented at: 102 nd Annual<br />
Meeting <strong>of</strong> the <strong>American</strong> Association for Cancer Research; April 2011;<br />
Orlando, FL.<br />
53. Amary MF, Bacsi K, Maggiani F, et al. IDH1 and IDH2 mutations<br />
are frequent events in central chondrosarcoma and central and periosteal<br />
chondromas but not in other mesenchymal tumours. J Pathol. 2011;224:334-343.<br />
657
WHAT THE BUSY ONCOLOGIST NEEDS TO KNOW<br />
ABOUT GASTROINTESTINAL STROMAL TUMOR<br />
CHAIR<br />
Jaap Verweij, MD, PhD<br />
Erasmus University Medical Center<br />
Rotterdam, Netherlands<br />
SPEAKERS<br />
William D. Tap, MD<br />
University <strong>of</strong> California, Los Angeles<br />
Santa Monica, CA<br />
Michael C. Heinrich, MD<br />
Oregon Health & Science University Knight Cancer Institute<br />
Portland, OR
Adjuvant Treatment <strong>of</strong> Gastrointestinal<br />
Stromal Tumor: The Pro<strong>of</strong>, The Pro, and<br />
the Practice<br />
Overview: Gastrointestinal stromal tumors (GIST) are rare<br />
tumors, but they are the most common mesenchymal tumor <strong>of</strong><br />
the gastrointestinal tract, driven by mutations in KIT and<br />
PDGF. The KIT and PDGF inhibiting agent imatinib has been<br />
tested as adjuvant postsurgery in GIST patients with an<br />
intermediate or high risk <strong>of</strong> relapse. Two <strong>of</strong> three prospective,<br />
randomized controlled studies have meanwhile been reported.<br />
The <strong>American</strong> College <strong>of</strong> Surgical <strong>Oncology</strong> Group (ACOSOG)<br />
in 713 patients reported a relapse-free survival benefit for<br />
adjuvant imatinib given for 1 year, but not an overall survival<br />
benefit. The Scandinavian Sarcoma Group (SSG) performed a<br />
study comparing 1 year <strong>of</strong> imatinib to 3 years <strong>of</strong> imatinib. At 3<br />
GASTROINTESTINAL STROMAL tumors are rare tumors,<br />
but they are the most common mesenchymal<br />
tumor <strong>of</strong> the gastrointestinal tract. 1 Activating mutations in<br />
the KIT gene and, to a lesser extent, the platelet-derived<br />
growth factor receptor (PDGFR)-alpha gene, are considered<br />
the key drivers <strong>of</strong> tumor growth in GIST, 2 and the immunohistochemical<br />
assessment <strong>of</strong> overexpression <strong>of</strong> KIT has become<br />
common practice for an appropriate diagnosis.<br />
The gold standard treatment for a localized primary GIST<br />
is radical surgical resection. However, with the possible<br />
exception <strong>of</strong> very small tumors (�1 cm) that are usually<br />
incidental findings, all GISTs have the potential to relapse<br />
even after radical surgery. 2,3 Most relapses occur within the<br />
first 5 years after surgery, but late relapses after 20 years<br />
and beyond have been reported. 4<br />
The estimation <strong>of</strong> the risk <strong>of</strong> recurrence in an individual<br />
patient has been difficult. Initial prognostic scores were<br />
based on tumor size, mitotic activity, and tumor site, 5-8 but<br />
none were validated or provided a more quantifiable risk.<br />
More recently two formally validated risk-assessment systems<br />
after radical surgery were published. 4,9 Gold and<br />
colleagues 9 used data on 127 patients who underwent complete<br />
resection <strong>of</strong> a localized primary GIST without adjuvant<br />
therapy to develop a nomogram to predict RFS, which was<br />
subsequently validated on two independent data sets <strong>of</strong> 212<br />
and 148 patients, respectively. Within each data set, an<br />
expert pathologist measured tumor size either before or<br />
after formalin fixation, confirmed the diagnosis <strong>of</strong> GIST, and<br />
calculated the mitotic index (number <strong>of</strong> mitoses per 50<br />
randomly selected high-power microscopic fields [HPFs]).<br />
The nomogram is based on tumor size (cm), primary tumor<br />
site (stomach, small intestine, colon/rectum, or other), and<br />
mitotic index (�5or�5 mitoses per 50 HPFs) (Table 1). 9 The<br />
relapse rate for high-risk GIST in older criteria roughly<br />
corresponds to a greater than 50% relapse chance at 5 years<br />
according to Gold’s nomogram that thus adds the tumor<br />
primary site to the equation. Joensuu and colleagues 4 pooled<br />
data on 2,560 patients from 10 multicenter and international<br />
series, adding presence <strong>of</strong> tumor rupture and male<br />
gender to the model. They subsequently validated their<br />
model in an independent set <strong>of</strong> 920 patients. The estimated<br />
15-year recurrence-free survival (RFS) after surgery was<br />
59.9%. This model seems to provide better differentiation for<br />
By Jaap Verweij, MD, PhD<br />
years the overall survival (OS) in patients with 3 years <strong>of</strong><br />
imatinib therapy was similar to the OS in those with 1 year <strong>of</strong><br />
imatinib 96% and 94% respectively, while at 5 years these<br />
numbers were 92% and 82% (HR: 0.45; 95% CI [0.22–0.89]; p �<br />
0.019). Data from the largest study, conducted by the European<br />
Organisation for Research and Treatment <strong>of</strong> Cancer<br />
(EORTC) in 908 patients randomly assigned to receive either 2<br />
years <strong>of</strong> adjuvant imatinib or no imatinib, have not yet been<br />
reported. Based on the current evidence, 3 years <strong>of</strong> imatinib at<br />
a daily dose <strong>of</strong> 400 mg should be considered in patients with a<br />
50% or higher risk <strong>of</strong> relapse within 5 years after surgery. The<br />
evidence and the remaining caveats are discussed.<br />
low to intermediate risks, but has the disadvantage <strong>of</strong> being<br />
more complex than the Gold nomogram.<br />
The relevance <strong>of</strong> these risk-stratification models is that<br />
they enable distinction between patients who are likely to be<br />
cured by surgery alone and those that in theory may benefit<br />
from adjuvant treatments. The randomized studies on adjuvant<br />
imatinib have all included patients with intermediate<br />
and high-risk tumors, although the criteria as used in these<br />
studies were based on nonvalidated models.<br />
Adjuvant Use <strong>of</strong> Imatinib: Aims, History, and<br />
Nonrandomized Studies<br />
To improve the outcome rates after surgery, adjuvant<br />
treatments are applied. It is important to try to define the<br />
main aim <strong>of</strong> such treatments. Since early application <strong>of</strong><br />
drugs in theory could also lead to early acquired tumor cell<br />
resistance, and thus have a detrimental effect, overall survival<br />
(and thus cure) seems to be the most favored endpoint.<br />
However, interestingly, a standard and commonly accepted<br />
endpoint seems to be lacking (Table 2).<br />
Imatinib, a tyrosine kinase inhibitor <strong>of</strong> the KIT and<br />
PDGFR receptors, which drive the GIST phenotype, has<br />
become the backbone <strong>of</strong> treatment for metastatic GIST. 10-12<br />
Given the magnitude <strong>of</strong> effect in that setting, there was an<br />
obvious rationale to explore imatinib in the adjuvant setting.<br />
Before proper studies were done, physicians in the United<br />
States started to use adjuvant imatinib. In a retrospective<br />
analysis <strong>of</strong> the National Cancer Data Base spanning 2001 to<br />
2007, Bilimoria and colleagues 13 reported that from 2001 to<br />
2003 and from 2006 to 2007, the adjuvant use <strong>of</strong> imatinib<br />
increased from 29% to 47%, without any formal evidence<br />
supporting this use and before U.S. Food and Drug Administration<br />
approval for this indication.<br />
Nonrandomized studies on adjuvant treatments cannot be<br />
From the Department <strong>of</strong> Medical <strong>Oncology</strong>, Erasmus University Medical Center-Daniel<br />
den Hoed Oncologic Center, Rotterdam, Netherlands.<br />
Author’s disclosures <strong>of</strong> potential conflicts <strong>of</strong> interest are found at the end <strong>of</strong> this article.<br />
Address reprint requests to Jaap Verweij, MD, PhD, Department <strong>of</strong> Medical <strong>Oncology</strong>,<br />
Erasmus University Medical Center-Daniel den Hoed Oncologic Center, ‘s-Gravendijkwal<br />
230, 3015 CE Rotterdam, Netherlands; email: j.verweij@erasmusmc.nl.<br />
© <strong>2012</strong> by <strong>American</strong> <strong>Society</strong> <strong>of</strong> <strong>Clinical</strong> <strong>Oncology</strong>.<br />
1092-9118/10/1-10<br />
659
Table 1. Risk Criteria for Relapse after Primary Surgery for GIST<br />
Author<br />
No. Patients<br />
Initial<br />
used for decision making since this design is unable to<br />
exclude major selection bias. This is best exemplified by the<br />
study <strong>of</strong> Li and colleagues 14 who compared a group <strong>of</strong> 56<br />
patients self-selected for 3 years <strong>of</strong> treatment with imatinib<br />
postsurgery to a group <strong>of</strong> self-selected patients that rejected<br />
the <strong>of</strong>fer <strong>of</strong> postoperative imatinib. The latter control<br />
group 14 performed much worse than the control group in<br />
either <strong>of</strong> the two prospectively randomized studies that will<br />
be discussed. Here, I will therefore not discuss the other<br />
performed nonrandomized studies.<br />
Adjuvant Imatinib: Randomized Studies<br />
Three randomized phase III studies on the adjuvant use <strong>of</strong><br />
imatinib have now been performed (Table 3), two <strong>of</strong> which<br />
have been published, either as a full paper 15 or abstract. 16<br />
DeMatteo and colleagues, through ACOSOG (trial Z9001) 15<br />
performed a randomized phase III, double-blind, placebocontrolled<br />
trial in patients after complete resection <strong>of</strong> a<br />
primary GIST at least 3 cm in diameter and positive for the<br />
KIT protein by immunohistochemistry. Seven hundred<br />
seventy-eight patients were registered, but because <strong>of</strong> randomization<br />
problems only 713 were formally randomly assigned<br />
to either 400 mg imatinib daily (359 patients) or<br />
placebo (354 patients) daily for 1 year. Patients and investigators<br />
were blinded to the treatment group. Cross-over to<br />
or re-treatment with imatinib was allowed in case <strong>of</strong> tumor<br />
recurrence. The primary endpoint was RFS. Accrual in this<br />
study was terminated early on recommendation <strong>of</strong> the Independent<br />
Data Monitoring Committee (IDMC) because the<br />
trial results crossed the interim analysis efficacy boundary<br />
for RFS. The intention to treat analysis is taken for the<br />
purpose <strong>of</strong> this review.<br />
At a median follow-up <strong>of</strong> 19.7 months, 70 patients (20%) in<br />
the placebo group as compared with 30 (8%) in the imatinib<br />
group had tumor recurrence or had died. Imatinib significantly<br />
improved 1-year RFS compared with placebo (98% vs.<br />
83%; HR: 0.35, p � 0.0001). Overall survival, however, was<br />
fully identical in both arms up to 48 months. 15<br />
A later analyses <strong>of</strong> the trial by mutation subtype held up<br />
KEY POINTS<br />
No. Patients<br />
Validation Criteria<br />
Gold 9 127 360 Size, site, mitotic rate<br />
Joensuu 4 2,560 920 Size, site, mitotic rate, rupture, gender<br />
● Validated models for postsurgery risk assessment are<br />
now available.<br />
● Adjuvant therapy aims to improve survival.<br />
● Adjuvant imatinib given for 3 years at a daily dose <strong>of</strong><br />
400 mg improves overall survival in patients at high<br />
risk <strong>of</strong> relapse.<br />
● Evidence is currently based on relatively limited<br />
numbers.<br />
● Data from future studies should be used when available,<br />
to re-evaluate the evidence for adjuvant imatinib<br />
use.<br />
660<br />
Table 2. Suggested Aims for Adjuvant Treatment<br />
Internet lay sites: Adjuvant treatment is aimed to improve overall treatment<br />
outcome, and together with the initial treatment forms the curative therapy<br />
Oncoline: Adjuvant systemic drug therapy is given after primary local treatment<br />
with the aim to eradicate possible occult metastases and increase cure rates<br />
ASCO guidelines: Adjuvant systemic therapy is aimed to improve clinically<br />
meaningful outcomes (i.e., disease-free survival, overall survival, quality <strong>of</strong> life,<br />
and toxicity) when compared with and/or added to other therapies<br />
NCCN guidelines: No definition <strong>of</strong> the aim <strong>of</strong> adjuvant therapies given<br />
ESMO guidelines: Aim differs per tumor type. Mostly disease-free survival<br />
Abbreviations: ASCO, <strong>American</strong> <strong>Society</strong> <strong>of</strong> <strong>Clinical</strong> <strong>Oncology</strong>; NCCN, National<br />
Comprehensive Cancer Network; ESMO, European <strong>Society</strong> <strong>of</strong> Medical <strong>Oncology</strong>.<br />
the statistically significant difference in RFS after 2 years<br />
for patients with KIT exon 11 and PDGFRA mutations<br />
(excluding PDGFRA D842V), but not for KIT wild-type and<br />
exon 9 mutations. 17 A further update is expected this summer.<br />
The SSG and the German Working Group on Medical<br />
<strong>Oncology</strong> (Arbeitsgemeinschaft Internistische Onkologie)<br />
have recently reported results <strong>of</strong> a smaller (400 patients)<br />
randomized controlled trial comparing adjuvant imatinib for<br />
3 years with adjuvant imatinib for 1 year, both at a daily<br />
dose <strong>of</strong> 400 mg, in high-risk GIST. 16 At a median follow-up<br />
<strong>of</strong> 54 months, 42% <strong>of</strong> patients in the 1-year treatment group<br />
had GIST recurrence, compared with 25% <strong>of</strong> patients in the<br />
3-year treatment group. 16 In the intention to treat analysis<br />
at 3 years, the RFS for the 3 years <strong>of</strong> imatinib therapy was<br />
87% compared with 60% for those given the drug for 1 year,<br />
which translated into 66% and 48%, respectively, at 5 years<br />
(HR: 0.46; 95% CI [0.32, 0.65]; p � 0.0001). 16 At 3 years the<br />
OS in patients with 3 years <strong>of</strong> imatinib therapy was similar<br />
to the OS in those with 1 year <strong>of</strong> imatinib (96% and 94%,<br />
respectively), while at 5 years these numbers were 92% and<br />
82% (HR: 0.45; 95% CI [0.22–0.89]; p � 0.019). 16 Although<br />
statistically the analysis was robust, we still should realize<br />
that because <strong>of</strong> the relatively small size <strong>of</strong> the study the<br />
difference in OS is based on seven more patients dying in the<br />
1-year treatment group. Death due to GIST occurred in 14<br />
patients (7%) in the 1-year group and in 7 (4%) in the 3-year<br />
group. The numbers at 5 years were even smaller and at<br />
that point the difference is based on three events.<br />
The third study, which has not yet been reported, is also<br />
the largest study, accruing 908 patients and coordinated by<br />
EORTC, in an intergroup setting with the Austral-Asian<br />
Gastro-Intestinal Tumor Group, the French Sarcoma Group<br />
(FSG), the Italian Sarcoma Group, and the Grupo Espanola<br />
Investigaciones Sarcomas. This study randomly selected<br />
patients at intermediate or high risk <strong>of</strong> relapse to either<br />
receive no further adjuvant therapy or 2 years <strong>of</strong> daily<br />
imatinib at 400 mg. The primary endpoint was overall<br />
survival. The study was closed to accrual in October 2008<br />
Table 3. Randomized Studies on Adjuvant Imatinib in GIST<br />
Group Randomization<br />
JAAP VERWEIJ<br />
No. <strong>of</strong><br />
Patients<br />
Ineligible<br />
(%)<br />
ASCOSOG 1 yr 400 mg dd versus control 713 9.1<br />
SSG 1 yr 400 mg dd versus 3 yr 400 mg dd 400 9.8<br />
EORTC 2 yr 400 mg dd versus control 908 NR<br />
Abbreviations: ACOSOG, <strong>American</strong> College <strong>of</strong> Surgical <strong>Oncology</strong> Group; dd,<br />
daily dose; SSG, Scandinavian Sarcoma Group; EORTC, European Organisation<br />
for Research and Treatment <strong>of</strong> Cancer; NR, not yet reported.
ADJUVANT THERAPY IN HIGH-RISK GASTROINTESTINAL STROMAL TUMORS<br />
and has thereafter been reviewed by the EORTC IDMC on a<br />
yearly basis. Based on IDMC recommendations and the<br />
recent awareness that late imatinib given at progression <strong>of</strong><br />
disease could in theory make up for early adjuvant imatinib<br />
without the theoretical disadvantage <strong>of</strong> inducing early resistance<br />
and that in the overall assessment, retreatment with<br />
imatinib also has to be taken into account, it was decided to<br />
change the primary endpoint <strong>of</strong> this study to time to imatinib<br />
failure: the start <strong>of</strong> any new systemic therapy other<br />
than imatinib or death due to any cause. The first analysis<br />
based on this new endpoint is currently expected in the first<br />
half <strong>of</strong> <strong>2012</strong>.<br />
What Do These Studies Tell Us?<br />
Importantly, the data details <strong>of</strong> the ACOSOG and SSG<br />
studies have not been publically presented in the same way.<br />
This renders it difficult to assess potential effects <strong>of</strong> differences<br />
in populations in prognostic parameters. Yet the<br />
majority <strong>of</strong> patients in the studies had tumors with high risk<br />
<strong>of</strong> relapse so the observations likely hold for this population.<br />
Both studies indicate that adjuvant imatinib given for either<br />
1 or 3 years is able to improve RFS. But as indicated this is<br />
not the primary aim <strong>of</strong> adjuvant therapy, and in an editorial<br />
Hohenberger has even described the use <strong>of</strong> RFS in this<br />
setting as a self-fulfilling prophecy. 18 Bearing in mind that<br />
there may be differences in the study populations, it is<br />
interesting to analyze the RFS curves. It is remarkable that<br />
the RFS in the patients given 1 year <strong>of</strong> imatinib in the SSG<br />
study is similar to the RFS <strong>of</strong> the control group in the<br />
ACOSOG study, while the curves for the respective 3-year<br />
and 1-year groups also seem to overlap. Of concern is the fact<br />
that at the time <strong>of</strong> reporting both studies showed bananashaped<br />
curves, suggesting that we can yet not exclude that<br />
the beneficial effects, if any, may be temporary.<br />
Importantly, if we also project in the curves (both for RFS<br />
and OS) <strong>of</strong> the nonrandomized comparative study 14 it shows<br />
that the control group in that study performed worst <strong>of</strong> all<br />
control groups, while the treated group outperformed all<br />
treatment groups. This in itself supports the notion that<br />
nonrandomized studies can be heavily biased by patient<br />
selection and should be ignored in deciding on practice<br />
recommendations.<br />
Plotting the data <strong>of</strong> the two published studies from the<br />
time that patients went <strong>of</strong>f drug, there again are remarkable<br />
differences but also some similarities. The SSG study arms<br />
seem to perform worse than those <strong>of</strong> the ACOSOG study, but<br />
all arms suggest a steady drop-<strong>of</strong>f rate, and the curves run in<br />
parallel. This seems to suggest that stopping imatinib in<br />
some patients leads to rapid relapse. This would be in line<br />
with the observations in the FSG study in metastatic disease<br />
19 that has shown that patients in whom imatinib was<br />
withheld while disease was not progressing rapidly experienced<br />
relapse. Yet, while in metastatic disease these data<br />
support the continuation <strong>of</strong> imatinib as long as patients<br />
seem to benefit, the absence <strong>of</strong> indisputable overall survival<br />
benefit in the adjuvant setting should lead to a conservative<br />
approach to the duration <strong>of</strong> adjuvant imatinib treatment.<br />
For OS a limitation <strong>of</strong> all studies is that they could not<br />
exclude cross-over to or retreatment with imatinib at recurrence,<br />
which may affect OS results. This is why the EORTC<br />
decided to change its primary endpoint after the study had<br />
fully accrued. The ACOSOG study 15 has only presented OS<br />
up to 4 years, while the SSG study 16 has presented longer<br />
follow-up. Again, remarkably, at 4 years the 1-year treated<br />
group in the SSG study seems to have worse outcome than<br />
any <strong>of</strong> the ACOSOG study arms. This may point to differences<br />
in the study populations. The ACOSOG study has not<br />
yet shown any benefit in overall survival. The SSG study, as<br />
indicated, shows a statistically significant difference in favor<br />
<strong>of</strong> the 3-year treatment. If the ACOSOG study does not show<br />
a difference between no treatment and 1-year treatment, in<br />
theory this 1-year treatment is an adequate control in the<br />
SSG study. However, if the above-mentioned differences<br />
between the 1-year groups in the two studies cannot be<br />
explained on population differences, the observed difference<br />
in the SSG study might solely be based on a worse performance<br />
<strong>of</strong> the 1-year group. Yet, since the statistics support<br />
this, for the time being we can consider the outcome <strong>of</strong> the<br />
SSG study as evidence supporting the use <strong>of</strong> 3 years <strong>of</strong><br />
adjuvant imatinib in the high-risk population. Still, caution<br />
needs to be taken in interpreting the outcome <strong>of</strong> the study<br />
since it is relatively small and differences are based on small<br />
numbers. For the same reason <strong>of</strong> small patient numbers in<br />
individual subgroups, further investigations are still warranted.<br />
One could wonder why the IDMC <strong>of</strong> the EORTC<br />
study has not decided to unblind the data <strong>of</strong> that study now<br />
that the SSG study shows a survival benefit. In theory, this<br />
could mean the study does not show a survival benefit and<br />
data need to further mature. It will be very interesting to see<br />
the report <strong>of</strong> that study later this year or early next year.<br />
Conclusion and Recommendation<br />
With all <strong>of</strong> the caveats mentioned, for the time being<br />
adjuvant imatinib can be considered the standard <strong>of</strong> care<br />
in patients with a high risk <strong>of</strong> GIST relapse. The most<br />
important support for this statement comes from the overall<br />
survival benefit suggested in the SSG trial. The current<br />
consensus seems to be to consider patients with a greater<br />
than 50% chance <strong>of</strong> relapse within 5 years according to the<br />
Gold nomogram as candidates for such treatment. Although<br />
not fully based on evidence, because <strong>of</strong> their rare occurrence,<br />
patients with neur<strong>of</strong>ibromatosis type 1–associated GIST,<br />
which strongly expresses wild-type KIT and has a very<br />
indolent behavior, and those with PDGFR-alpha D842V<br />
mutations, which are known to be insensitive to imatinib,<br />
17,20 may be excluded from this recommendation. Patients<br />
with tumor rupture may best be considered as having<br />
metastatic disease and be treated accordingly.<br />
Longer-term follow-up data as well as publication <strong>of</strong> the<br />
results <strong>of</strong> the EORTC study are eagerly awaited and should<br />
lead to a critical reconsideration <strong>of</strong> the above-mentioned<br />
recommendation.<br />
Based on the studies, adjuvant imatinib should be initiated<br />
within 3 months after definitive surgery and preferably<br />
be handled by an expert team after multidisciplinary discussion.<br />
Since all studies have applied a dose <strong>of</strong> 400 mg daily<br />
regardless <strong>of</strong> the mutation status <strong>of</strong> the primary tumor, this<br />
is also the dose recommended for standard application <strong>of</strong><br />
adjuvant imatinib. Based on the data <strong>of</strong> the SSG study, this<br />
dose should be given for 3 years. In GIST cases where<br />
preoperative adjuvant imatinib therapy is used for localized<br />
disease for the purpose <strong>of</strong> organ-sparing tumor shrinkage or<br />
patient refusal <strong>of</strong> surgery, the duration <strong>of</strong> neoadjuvant<br />
therapy should be considered as part <strong>of</strong> the whole adjuvant<br />
treatment duration <strong>of</strong> 3 years.<br />
Based on the performed studies, during adjuvant treat-<br />
661
ment, patients should be assessed for blood cell counts and<br />
blood chemistry analysis at 3 monthly intervals. Abdominal<br />
imaging is recommended every 3 to 6 months during adjuvant<br />
imatinib therapy and after cessation <strong>of</strong> imatinib ther-<br />
Author’s Disclosures <strong>of</strong> Potential Conflicts <strong>of</strong> Interest<br />
Author<br />
Employment or<br />
Leadership<br />
Positions<br />
Consultant or<br />
Advisory Role<br />
Jaap Verweij Abbott<br />
Laboratories;<br />
Boehringer<br />
Ingelheim;<br />
Enzon;<br />
GlaxoSmithKline;<br />
InteRNA;<br />
Johnson &<br />
Johnson;<br />
Novartis; Otsuka;<br />
Pfizer; San<strong>of</strong>i;<br />
Teva<br />
1. Miettinen M, Lasota J. Gastrointestinal stromal tumors: Review on<br />
morphology, molecular pathology, prognosis, and differential diagnosis. Arch<br />
Pathol Lab Med. 2006;130:1466-1478.<br />
2. Rubin BP, Heinrich MC, Corless CL. Gastrointestinal stromal tumour.<br />
Lancet. 2007;369:1731-1741.<br />
3. DeMatteo RP, Lewis JJ, Leung D, et al. Two hundred gastrointestinal<br />
stromal tumors: recurrence patterns and prognostic factors for survival. Ann<br />
Surg. 2000;231:51-58.<br />
4. Joensuu H, Vehtari A, Riihimäki J, et al. Risk <strong>of</strong> gastrointestinal<br />
stromal tumour recurrence after surgery: an analysis <strong>of</strong> pooled populationbased<br />
cohorts. Lancet <strong>Oncology</strong>. Epub 2011 December 7.<br />
5. Miettinen M, El-Rifai W, L HLS, et al. Evaluation <strong>of</strong> malignancy and<br />
prognosis <strong>of</strong> gastrointestinal stromal tumors: a review. Hum Pathol. 2002;33:<br />
478-483.<br />
6. Nilsson B, Bumming P, Meis-Kindblom JM, et al. Gastrointestinal<br />
stromal tumors: the incidence, prevalence, clinical course, and prognostication<br />
in the preimatinib mesylate era—a population-based study in western<br />
Sweden. Cancer. 2005;103:821-829.<br />
7. Huang HY, Li CF, Huang WW, et al. A modification <strong>of</strong> NIH consensus<br />
criteria to better distinguish the highly lethal subset <strong>of</strong> primary localized<br />
gastrointestinal stromal tumors: a subdivision <strong>of</strong> the original high-risk group<br />
on the basis <strong>of</strong> outcome. Surgery. 2007;141:748-756.<br />
8. Goh BK, Chow PK, Yap WM, et al. Which is the optimal risk stratification<br />
system for surgically treated localized primary GIST? Comparison <strong>of</strong><br />
three contemporary prognostic criteria in 171 tumors and a proposal for a<br />
modified Armed Forces Institute <strong>of</strong> Pathology risk criteria. Ann Surg Oncol.<br />
2008;15:2153-2163.<br />
9. Gold JS, Gonen M, Gutierrez A, et al. Development and validation <strong>of</strong> a<br />
prognostic nomogram for recurrence-free survival after complete surgical<br />
resection <strong>of</strong> localised primary gastrointestinal stromal tumour: a retrospective<br />
analysis. Lancet Oncol. 2009;10:1045-1052.<br />
10. Verweij J, Casali P, Zalcberg J, et al. Progression-free survival in<br />
gastrointestinal stromal tumours with high-dose imatinib: randomised trial.<br />
Lancet. 2004;364:1127-1134.<br />
11. Blanke CD, Rankin C, Demetri GD, et al. Phase III randomized,<br />
intergroup trial assessing imatinib mesylate at two dose levels in patients<br />
662<br />
apy every 3 months, for a 2-year period, every 6 months for<br />
the 3 years thereafter, and annually subsequently up to year<br />
10. Recurrence thereafter is rare 4 and thus the benefits <strong>of</strong><br />
imaging no longer outweigh the risks.<br />
Stock<br />
Ownership Honoraria<br />
Abbott<br />
Laboratories;<br />
Boehringer<br />
Ingelheim; Eisai;<br />
Enzon;<br />
GlaxoSmithKline;<br />
InteRNA;<br />
Johnson &<br />
Johnson; Merck<br />
Serono; MSD<br />
<strong>Oncology</strong>;<br />
Novartis; Otsuka;<br />
Pfizer; San<strong>of</strong>i;<br />
Sun Pharma;<br />
Synthon; Teva<br />
REFERENCES<br />
Research<br />
Funding<br />
Expert<br />
Testimony<br />
JAAP VERWEIJ<br />
Other<br />
Remuneration<br />
with unresectable or metastatic gastrointestinal stromal tumors expressing<br />
the kit receptor tyrosine kinase: S0033. J Clin Oncol. 2008;26:626-632.<br />
12. Blanke CD, Demetri GD, von Mehren M, et al. Long-term results from<br />
a randomized phase II trial <strong>of</strong> standard- versus higher-dose imatinib mesylate<br />
for patients with unresectable or metastatic gastrointestinal stromal tumors<br />
expressing KIT. J Clin Oncol. 2008;26:620-625.<br />
13. Bilimoria KY, Wayne JD, Merkow RP, et al. Incorporation <strong>of</strong> adjuvant<br />
therapy into the multimodality management <strong>of</strong> gastrointestinal stromal<br />
tumors <strong>of</strong> the stomach in the United States. Ann Surg Oncol. <strong>2012</strong>;19:184-<br />
191.<br />
14. Li J, Gong JF, Wu AW, Shen L. Post-operative imatinib in patients with<br />
intermediate or high risk gastrointestinal stromal tumor. Eur J Surg Oncol.<br />
2011;37:319-324.<br />
15. Dematteo RP, Ballman KV, Antonescu CR, et al. Adjuvant imatinib<br />
mesylate after resection <strong>of</strong> localised, primary gastrointestinal stromal tumour:<br />
a randomised, double-blind, placebo-controlled trial. Lancet. 2009;373:<br />
1097-1104.<br />
16. Joensuu H, Eriksson M, Hartmann J, et al. Twelve versus 36 months <strong>of</strong><br />
adjuvant imatinib (IM) as treatment <strong>of</strong> operable GIST with a high risk <strong>of</strong><br />
recurrence: final results <strong>of</strong> a randomized trial (SSGXVIII/AIO). J Clin Oncol.<br />
2011;29 (suppl, abstr LBA1).<br />
17. Corless CL, Ballman KV, Antonescu C, et al. Relation <strong>of</strong> tumor<br />
pathologic and molecular features to outcome after surgical resection <strong>of</strong><br />
localized primary gastrointestinal stromal tumor (GIST): Results <strong>of</strong> the<br />
intergroup phase III trial ACOSOG Z9001. J Clin Oncol. 2010;28 (suppl, abstr<br />
10006).<br />
18. Hohenberger P. Adjuvant imatinib in GIST: a self-fulfilling prophecy,<br />
or more? Lancet. 2009;373:1058-1060.<br />
19. Le Cesne A, Ray-Coquard I, Bui B, et al. Continuous versus interruption<br />
<strong>of</strong> imatinib (IM) in responding patients with advanced GIST after three<br />
years <strong>of</strong> treatment: a prospective randomized phase III trial <strong>of</strong> the French<br />
Sarcoma Group. J Clin Oncol. 2007;25 (suppl, abstr 10005).<br />
20. Biron P, Cassier Ph.A, Fumagalli E, et al. Outcome <strong>of</strong> patients with<br />
PDGFRA D842V mutant gastrointestinal stromal tumor treated with imatinib<br />
for advanced disease. J Clin Oncol. 2010;28:710s (suppl, abstr 10051).
Management <strong>of</strong> Tyrosine Kinase<br />
Inhibitor–Resistant Gastrointestinal<br />
Stromal Tumors<br />
By Christine M. Barnett, MD, and Michael C. Heinrich, MD<br />
Overview: The treatment <strong>of</strong> gastrointestinal stromal tumors<br />
(GISTs) has been a model for targeted cancer therapy. The<br />
discovery <strong>of</strong> driver somatic mutations in the KIT and PDGFRA<br />
receptor tyrosine kinases led to a shift <strong>of</strong> therapy from<br />
conventional cytotoxic chemotherapy to inhibitors <strong>of</strong> these<br />
receptors. Targeted molecular therapy <strong>of</strong> GIST has markedly<br />
increased the overall survival <strong>of</strong> patients with advanced dis-<br />
GISTS ARE the most common sarcoma arising in the<br />
abdominal cavity. The annual incidence in the United<br />
States is approximately 4,000 to 6,000 new cases per year.<br />
The stomach is the most common site <strong>of</strong> origin, but GISTs<br />
can arise anywhere in the gastrointestinal tract. 1 Localized<br />
GIST is primarily managed with surgical resection, but<br />
these tumors can recur locally or metastasize, typically to<br />
the liver or peritoneum. Increasingly, GIST is managed with<br />
systemic therapy utilizing tyrosine kinase inhibitor (TKI)<br />
therapy, even in the case <strong>of</strong> localized disease.<br />
GIST Biology<br />
Only in the last 15 years has GIST been recognized as a<br />
distinct pathologic entity. Histologically, there is an overlap<br />
<strong>of</strong> smooth muscle and neural elements, which historically<br />
made classification <strong>of</strong> GIST difficult. GISTs are now recognized<br />
as a malignancy <strong>of</strong> the interstitial cells <strong>of</strong> Cajal (ICC)<br />
which serve as “pacemaker” cells <strong>of</strong> the gastrointestinal<br />
tract. KIT activity is required for development <strong>of</strong> certain<br />
types <strong>of</strong> ICCs, 1 and in addition, 95% <strong>of</strong> GISTs are positive for<br />
KIT expression by immunohistochemistry. 2 Notably, the<br />
types <strong>of</strong> ICCs that are developmentally dependent on KIT<br />
are also the types that can give rise to GIST.<br />
KIT (CD117) is a type III receptor tyrosine kinase. This<br />
family also includes platelet-derived growth factor receptors<br />
A and B (PDGFRA/B), CSF-1R, and FLT3. Binding <strong>of</strong> stem<br />
cell factor results in homodimerization <strong>of</strong> the KIT receptor<br />
with subsequent activation <strong>of</strong> the intracellular tyrosine<br />
kinase domain. 3 Tyrosine phosphorylation <strong>of</strong> KITinteracting<br />
proteins leads to activation <strong>of</strong> multiple signal<br />
transduction pathways that stimulate cell proliferation and<br />
resistance to apoptosis. Under normal circumstances, the<br />
KIT receptor is autoinhibited by the juxtamembrane domain,<br />
which prevents the protein activation loop from assuming<br />
the necessary conformation for kinase activity.<br />
Oncogenic Mutations in GIST<br />
Mutations in KIT and PDGFRA result in ligandindependent<br />
kinase activation. 70% to 80% <strong>of</strong> GISTs have<br />
mutations in KIT, 1 making this the single most important<br />
therapeutic target for this tumor. The majority <strong>of</strong> KIT<br />
mutations affect the juxtamembrane domain encoded by<br />
exon 11, and up to two-thirds <strong>of</strong> GISTs harbor a mutation<br />
<strong>of</strong> this exon. Exon 11 mutations disrupt the autoinhibitory<br />
properties <strong>of</strong> the juxtamembrane domain, thus allowing<br />
kinase activation. 3<br />
ease. However, the ability <strong>of</strong> kinase therapy to control metastatic<br />
disease is ultimately limited by the ability <strong>of</strong> these<br />
agents to overcome intrinsic or acquired resistance mechanisms.<br />
Ongoing basic and clinical research is focusing on<br />
identifying new agents to inhibit KIT/PDGFRA kinase activity<br />
and/or other novel molecular targets in GIST.<br />
7% to 10% <strong>of</strong> GISTs have mutations <strong>of</strong> exon 9, which<br />
encodes the membrane-proximal most portion <strong>of</strong> the extracellular<br />
domain. These mutations lead to an extracellular<br />
conformational change that is similar to that created by<br />
ligand binding, resulting in kinase activation. 3 Exon 9<br />
mutations are more commonly seen in GISTs <strong>of</strong> the small<br />
and large bowel origin and are rarely seen in gastric GISTs.<br />
There are also less common primary mutations in the<br />
activation loop (exon 17), which stabilize the active conformation<br />
<strong>of</strong> the receptor, and the ATP binding region (exon<br />
13), which may affect the inhibitory function <strong>of</strong> the juxtamembrane<br />
domain. 3<br />
Approximately 10% <strong>of</strong> patients with GIST have tumors<br />
with activating mutations <strong>of</strong> PDGFRA, 1 which is closely<br />
related to KIT. KIT and PDGFRA mutations are mutually<br />
exclusive. The downstream pathways that are activated by<br />
PDGFRA are very similar to those activate by KIT. Most<br />
PDGFRA mutations occur in exon 18, which encodes the<br />
activation loop <strong>of</strong> the kinase. In addition, approximately 1%<br />
<strong>of</strong> GISTs have mutations <strong>of</strong> the PDGFRA juxtamembrane or<br />
ATP-binding domains.<br />
The remaining 10% to 15% <strong>of</strong> GISTs lack mutations in<br />
PDGFRA or KIT and have been traditionally designated<br />
“wild type” because <strong>of</strong> a lack <strong>of</strong> mutated kinase targets for<br />
therapy. Interestingly, KIT is still strongly phosphorylated<br />
in these tumors, indicating the presence <strong>of</strong> activated KIT. 4<br />
The mechanism responsible for this activation remains<br />
undefined. Increasingly, alternative oncogenic drivers have<br />
been identified in tumors lacking KIT or PDGFRA mutations.<br />
These include BRAF V600E mutations or, less commonly,<br />
mutations <strong>of</strong> RAS family members. 3 These make up<br />
an estimated 10% <strong>of</strong> wild-type GISTs, or 1% to 2% <strong>of</strong> GISTs<br />
overall. Rare loss-<strong>of</strong>-function mutations <strong>of</strong> the succinate<br />
dehydrogenase (SDH) complex have been identified in some<br />
familial syndromes associated with GIST (e.g., Carney-<br />
Stratakis). 3 SDH is important in the regulation <strong>of</strong> HIF1alpha,<br />
and loss <strong>of</strong> SDH complex function has similar biologic<br />
consequences as loss <strong>of</strong> VHL protein (the primary molecular<br />
From the Division <strong>of</strong> Hematology and Medical <strong>Oncology</strong>, Portland VA Medical Center<br />
and the Knight Cancer Institute, Oregon Health & Science University, Portland, OR.<br />
Authors’ disclosures <strong>of</strong> potential conflicts <strong>of</strong> interest are found at the end <strong>of</strong> this article.<br />
Address reprint requests to Michael C. Heinrich, MD, Division <strong>of</strong> Hematology/<strong>Oncology</strong>,<br />
Departments <strong>of</strong> Medicine and Cell and Developmental Biology, Portland VA Medical Center<br />
and OHSU Knight Cancer Institute, Oregon Health & Science University, R&D-19 3710<br />
U.S. Veterans Hospital Road, Portland, OR 97239; email: heinrich@ohsu.edu.<br />
© <strong>2012</strong> by <strong>American</strong> <strong>Society</strong> <strong>of</strong> <strong>Clinical</strong> <strong>Oncology</strong>.<br />
1092-9118/10/1-10<br />
663
defect in most cases <strong>of</strong> renal cell carcinoma). In addition, loss<br />
<strong>of</strong> SDHB protein expression is seen in pediatric GISTs and<br />
some GISTs with pediatric features that arise in adult<br />
patients. The underlying molecular defect in these wild-type<br />
GISTs likely has significant influence on the response (or<br />
lack there<strong>of</strong>) to TKI therapy.<br />
Initial Treatment with TKIs<br />
Before 2000, patients were treated with conventional<br />
chemotherapy, resulting in very poor response rates and no<br />
effect on overall survival. In 2000, this situation was dramatically<br />
altered by the first trials <strong>of</strong> TKI therapy for<br />
advanced or unresectable GIST.<br />
Imatinib was originally developed as a treatment for<br />
chronic myelogenous leukemia (CML). The fusion protein<br />
BCR-ABL is the target for imatinib in CML, 5 and the<br />
knowledge that ABL is structurally similar to KIT led to<br />
preclinical investigation <strong>of</strong> the ability <strong>of</strong> imatinib to inhibit<br />
KIT signaling. 6 In these studies, imatinib was shown to<br />
inhibit both wild-type and mutant KIT protein, including<br />
KIT exon 11 or 13 mutations that are found in GIST. 6,7 In<br />
addition, imatinib was also found to inhibit PDGFRA/B<br />
kinase activity.<br />
Based in part on these preclinical studies, imatinib was<br />
used to treat a single patient with widely metastatic disease.<br />
Based on the remarkable tumor response in this patient,<br />
imatinib was further tested in a phase I study <strong>of</strong> patients<br />
with GIST or other s<strong>of</strong>t tissue sarcoma. In the European<br />
Organisation for Research and Treatment <strong>of</strong> Cancer<br />
(EORTC) phase I study, 36 patients with GIST were treated<br />
with imatinib doses ranging from 400 to 800 mg. 8 Notably,<br />
only four patients demonstrated progressive disease; all<br />
others demonstrated response or stable disease (by RECIST<br />
criteria).<br />
This very successful phase I trial lead to two phase II<br />
trials, one by the EORTC and another by a U.S.-Finnish<br />
group. 9 Both trials compared doses <strong>of</strong> 400 mg compared with<br />
600 mg <strong>of</strong> imatinib. Overall, 89% <strong>of</strong> patients in the EORTC<br />
trial and 87% <strong>of</strong> patients in the U.S.-Finnish trial had stable<br />
KEY POINTS<br />
● Imatinib is the front-line treatment for patients with<br />
unresectable or advanced GIST.<br />
● Primary imatinib resistance is usually related to the<br />
pretreatment biology <strong>of</strong> the GIST and is more common<br />
in patients whose tumor has one <strong>of</strong> the following<br />
genotypes: wild-type, KIT exon 9-mutant, or PDG-<br />
FRA D842V.<br />
● Secondary imatinib resistance is typically due to the<br />
expansion <strong>of</strong> tumor clones with acquired kinase mutations<br />
that confer drug resistance.<br />
● Sunitinib treatment is indicated for patients with<br />
imatinib intolerance and/or imatinib-resistant tumors.<br />
● The optimal treatment for patients with sunitinibresistant<br />
tumors is unknown. Treatment options include<br />
participation in a clinical study, <strong>of</strong>f-label use<br />
<strong>of</strong> other KIT tyrosine kinase inhibitors, or palliative<br />
surgery.<br />
664<br />
disease or partial response as their best response to therapy.<br />
One notable difference between the trials was that patients<br />
in the U.S.-Finnish trial were allowed to cross over to the<br />
600 mg arm if they progressed on 400 mg. 30% <strong>of</strong> patients<br />
who progressed on the 400 mg dose who crossed over to the<br />
600 mg dose obtained a partial response or stable disease<br />
status.<br />
Following these pivotal trials, two phase III trials were<br />
performed, both similarly designed with the intent to combine<br />
them into a meta-analysis. 10 These trials compared 400<br />
mg daily with 800 mg daily <strong>of</strong> imatinib. Both trials allowed<br />
cross-over to 800 mg on progression during treatment with<br />
400 mg per day <strong>of</strong> imatinib. In EORTC 62005, only 18% <strong>of</strong><br />
patients had progressive disease, while in the S0033 study<br />
26% <strong>of</strong> patients had progressive disease as the best response<br />
to treatment. In the Meta-GIST analysis <strong>of</strong> these two studies,<br />
there was a small progression-free survival (PFS) seen<br />
with the 800 mg dose <strong>of</strong> imatinib, but this was almost<br />
exclusively due to the superior results seen with the higher<br />
dose among patients with KIT exon 9-mutant GIST. In<br />
contrast, there was no improvement in PFS for GISTs with<br />
non–exon-9 mutations treated with high-dose imatinib. Notably,<br />
there was no improvement in overall survival in<br />
patients treated with high-dose imatinib.<br />
Disease Persistence<br />
Despite high rates <strong>of</strong> response, it appears that continuous<br />
therapy is required for optimal disease control. A randomized<br />
study <strong>of</strong> patients who had disease control after 3 years<br />
<strong>of</strong> imatinib demonstrated a high rate <strong>of</strong> progression following<br />
interruption <strong>of</strong> treatment. 11 The 2-year PFS was only<br />
16% in those who stopped imatinib compared with 80% in<br />
patients who continued on imatinib.<br />
So why doesn’t TKI therapy completely eradicate (cure)<br />
GIST? There is evidence that unlike mature GIST cells,<br />
GIST stem and progenitor cells are immune to KIT inhibition.<br />
3 Therefore, KIT inhibitors such as imatinib can inhibit<br />
and potentially kill mature cells, but a pool <strong>of</strong> quiescent stem<br />
and progenitor cells will remain. With cessation <strong>of</strong> treatment,<br />
this pool can relatively quickly repopulate tumors<br />
with differentiated GIST cells. This phenomenon is referred<br />
to as disease persistence and is distinct from clinical resistance<br />
that is discussed below.<br />
Imatinib Resistance<br />
BARNETT AND HEINRICH<br />
Based on the phase II studies, imatinib was approved by<br />
the U.S. Food and Drug Administration (FDA) for first-line<br />
treatment <strong>of</strong> advanced GIST. Despite these promising results,<br />
treatment with imatinib is not curative and GIST<br />
lesions can develop drug resistance after varying intervals <strong>of</strong><br />
time. Resistance to imatinib can be divided into two types,<br />
primary resistance for the tumors that progress within the<br />
first 6 months <strong>of</strong> imatinib treatment, and secondary resistance<br />
in tumors that progress after this period. Notably,<br />
distinct molecular mechanisms are associated with primary<br />
versus secondary resistance.<br />
The median PFS on front-line imatinib is in the range <strong>of</strong><br />
20 to 24 months. In the U.S.-Finnish phase II study cited<br />
above, approximately 18% <strong>of</strong> patients remain on front-line<br />
imatinib a decade later. Thus, the majority <strong>of</strong> imatinibtreated<br />
patients will eventually develop resistant disease. 12
TKI-RESISTANT GIST<br />
Primary Imatinib Resistance<br />
Approximately 10% <strong>of</strong> treated GIST patients experience<br />
primary tumor resistance to imatinib therapy. The most<br />
common cause <strong>of</strong> early imatinib resistance is a primary<br />
resistance mutation in PDGFRA or KIT. 13 KIT exon 9<br />
mutations have decreased in vitro sensitivity to imatinib,<br />
and larger clinical doses <strong>of</strong> imatinib may be required to<br />
achieve a response. 10 Therefore, early progression on 400 mg<br />
<strong>of</strong> imatinib is common for patients with KIT exon 9 mutant<br />
GIST. In addition, the most common PDGFRA mutation,<br />
D842V, is strongly resistant to imatinib in vitro. This<br />
mutation changes the receptor to the active conformation to<br />
which imatinib cannot bind. Recently, mutations <strong>of</strong> downstream<br />
signaling pathways (KRAS, BRAF, or PIK3CA) have<br />
been reported to coexist with KIT mutations in a minority <strong>of</strong><br />
patients. These mutations could reduce or eliminate the<br />
efficacy <strong>of</strong> upstream KIT inhibition by maintaining signaling<br />
through KIT-dependent downstream signaling pathways. 3<br />
Wild-type GISTs (i.e., no KIT or PDGFRA mutation) also<br />
have a higher rate <strong>of</strong> primary resistance to imatinib than<br />
the typical KIT exon 11-mutant GIST. The rate <strong>of</strong> response<br />
likely correlates with the underlying molecular defect, but<br />
definitive genotype/outcome data are not available. For<br />
instance, BRAF exon 15 mutations have been found in this<br />
wild-type group, and these mutations are not sensitive to<br />
imatinib. Additionally, those GISTs with abnormalities in<br />
the SDH complex appear to be less responsive to imatinib. 3<br />
In addition to the effect <strong>of</strong> the underlying tumor biology,<br />
there is controversy about whether a threshold drug level<br />
<strong>of</strong> imatinib is needed to see a response in GIST. There is no<br />
prospective data looking at the proper imatinib levels<br />
needed to achieve a response in GIST. However, in a<br />
retrospective subgroup analysis <strong>of</strong> one <strong>of</strong> the phase II<br />
studies, a minimum trough level <strong>of</strong> 1,100 ng/mL <strong>of</strong> imatinib<br />
was required for optimal results. 14 However, the utility <strong>of</strong><br />
imatinib levels to adjust dosing remains unproven.<br />
Secondary Imatinib Resistance<br />
Eventual progression <strong>of</strong> disease is an issue for the vast<br />
majority <strong>of</strong> patients treated with imatinib. This is likely due<br />
to expansion <strong>of</strong> GIST clones with secondary kinase mutations.<br />
15 Secondary KIT or PDGFRA mutations have only<br />
been found in patients who had a primary mutation, that is,<br />
not in wild-type patients. In addition, the secondary mutations<br />
occur in the same gene and on the same allele (in cis)<br />
as the primary mutation. 15 Tumors with a primary exon 11<br />
mutation are the most likely to develop a secondary KIT<br />
mutation. In the above mentioned U.S.-Finnish phase II<br />
trial <strong>of</strong> imatinib, two-thirds <strong>of</strong> patients with progression on<br />
imatinib were found to have secondary mutations in one or<br />
more post-treatment samples. These secondary mutations<br />
occur primarily in the activation loop (exons 17 and 18) or<br />
ATP binding pocket (exons 13 and 14). The situation is<br />
further complicated by multiple studies demonstrating significant<br />
intra- and interlesional heterogeneity <strong>of</strong> secondary<br />
mutations. 16 In vitro pr<strong>of</strong>iling <strong>of</strong> alternative KIT/PDGFRA<br />
TKIs (e.g., sunitinib) has shown that most inhibitors lack<br />
activity against all relevant secondary mutations. Thus<br />
heterogeneity <strong>of</strong> resistance mutations in imatinib-resistant<br />
GIST can result in differential responses <strong>of</strong> lesions to<br />
second- or third-line TKIs. In addition to complicating rou-<br />
tine clinical care, this situation also can obscure the potential<br />
benefit <strong>of</strong> new agents tested in clinical studies.<br />
With respect to the model <strong>of</strong> GIST stem and progenitor<br />
cells discussed above, secondary resistance is due to the<br />
appearance <strong>of</strong> clones with secondary mutations. These<br />
clones could arise from either stem/progenitor cells or from<br />
more differentiated cell populations. It is yet unclear if these<br />
secondary, more resistant, kinase mutations are acquired<br />
during drug therapy, or pre-exist and simply expand under<br />
the selective pressure <strong>of</strong> TKI therapy.<br />
Second-Line TKI Therapy<br />
For patients with documented primary or secondary<br />
imatinib-resistant GIST, the suggested initial intervention<br />
is to increase the dose <strong>of</strong> imatinib to 800 mg if the patient is<br />
on a lesser dose. Most patients will not benefit from imatinib<br />
dose escalation, with a median PFS following dose increase<br />
<strong>of</strong> only 3 months. However, 15% to 25% <strong>of</strong> patients may have<br />
more prolonged responses that can last for many months or<br />
years. 9 This may be due to inadequate drug levels with the<br />
original dose or in some cases may be due to overcoming<br />
secondary mutations with the higher imatinib dose.<br />
If there is progression on the 800 mg dose <strong>of</strong> imatinib or if<br />
the patient cannot tolerate the 800 mg total daily dose, the<br />
next FDA-approved treatment is to switch to sunitinib. In a<br />
phase III study, sunitinib-treated patients had a median<br />
time to progression <strong>of</strong> disease <strong>of</strong> 27.3 weeks compared with<br />
6.4 weeks with placebo. 17 These results were obtained using<br />
the typical sunitinib dosing schedule <strong>of</strong> 50 mg daily for 4<br />
weeks with 2 weeks <strong>of</strong>f. Notably, a phase II trial <strong>of</strong> continuous<br />
daily dosing <strong>of</strong> sunitinib at 37.5 mg produced a similar<br />
time to progression at 34 weeks. 18 Therefore, given better<br />
tolerability, many GIST experts will recommend starting<br />
patients on continuous daily dosing <strong>of</strong> sunitinib.<br />
Because sunitinib was the first medication to be tested in<br />
the imatinib-resistant population, it was the first to get FDA<br />
approval. However, with our increased understanding about<br />
secondary KIT and PDGFRA mutations, it is clear that the<br />
range <strong>of</strong> activity <strong>of</strong> sunitinib against these secondary mutations<br />
is suboptimal. Sunitinib has substantial activity<br />
against secondary mutations in exons 13 and 14, but cannot<br />
inhibit kinase activity in the presence <strong>of</strong> mutations in exons<br />
17 or 18. 19 Given that there is approximately equal frequency<br />
<strong>of</strong> these secondary mutations in most series, and the<br />
aforementioned heterogeneity <strong>of</strong> secondary mutations in a<br />
given patient, it is common to see mixed responses during<br />
sunitinib therapy. Sunitinib, unlike imatinib, has activity<br />
against VEGFR1/2. 20 However, it appears that most <strong>of</strong> the<br />
activity addition <strong>of</strong> this agent against imatinib-resistant<br />
GIST is due to inhibition <strong>of</strong> KIT rather than inhibition <strong>of</strong><br />
angiogenesis.<br />
Third-Line Therapy (And Beyond)<br />
Following progression on sunitinib, the optimal next step<br />
in therapy is unclear. Multiple TKIs have been studied in<br />
this setting, but there has yet to be a clear best therapy for<br />
imatinib- and sunitinib-resistant GIST. Therefore, patient<br />
participation in a clinical study should be strongly considered<br />
in this setting. 20<br />
There are multiple drugs being tested in clinical trials<br />
that directly or indirectly target KIT, KIT-dependent signaling<br />
(e.g., MEK or PI3K), or more remote downstream events<br />
665
Table 1. Investigational Therapies in GIST<br />
Drug Type Drug Target(s) Trial Information<br />
Tyrosine kinase<br />
inhibitors<br />
Imatinib KIT, PDGFR FDA approved<br />
Sunitinib KIT, PDGFR, VEGFR FDA approved<br />
Nilotinib KIT, PDGFR Phase III<br />
NCT00785785<br />
Dasatanib KIT, PDGFR Phase II<br />
NCT00568750<br />
Sorafenib KIT, PDGFR, VEGFR Phase II<br />
NCT01091207<br />
Regorafenib KIT, PDGFR, VEGFR Phase III<br />
NCT01271712<br />
Vatalanib KIT, PDGFR, VEGFR Phase II<br />
NCT00117299<br />
Masitinib<br />
(AB1010)<br />
KIT, PDGFR Phase III<br />
NCT00812240<br />
Pazopanib KIT, PDGFR, VEGFR Phase II<br />
NCT01323400<br />
Crenolanib PDGFR Phase II<br />
NCT01243346<br />
Dovitinib VEGFR, PDGFR, FGFR Phase II<br />
NCT01478373<br />
HSP90 inhibitors STA-9090 HSP90 Phase II<br />
NCT01039519<br />
AT-13387 HSP90 Phase II<br />
NCT01294202<br />
AUY922 HSP90 Phase II<br />
NCT01404650<br />
Monoclonal<br />
antibody<br />
IMC-3G3<br />
(Olaratumab)<br />
PDGFR Phase II<br />
NCT01316263<br />
Bevacizumab VEGFR Phase III<br />
NCT00324987<br />
mTOR inhibitor Everolimus mTOR Phase II<br />
NCT00510354<br />
Miscellaneous Perifosine AKT (PI3K pathway) Phase II<br />
NCT00455559<br />
BKM120 PI3K pathway Phase I<br />
NCT1468688<br />
TH-302 Tumor hypoxia Phase I<br />
NCT01381822<br />
(histone deacetylase inhibitors). These medications are<br />
sometimes used in combination with KIT inhibitors. For<br />
example, multiple heat shock protein 90 (Hsp90) inhibitors<br />
are in phase II trials in GIST. Hsp90 is an important<br />
chaperone protein for oncoproteins such as KIT. Inhibition<br />
<strong>of</strong> this target leads to preferential degradation <strong>of</strong> activated<br />
forms <strong>of</strong> KIT (as well as other activated kinases). There is<br />
strong preclinical data showing that Hsp90 inhibitors have<br />
activity against imatinib-resistant GIST in vitro; 21 however,<br />
there is not yet strong clinical data for the use <strong>of</strong> these<br />
medications in GIST. In addition, there is evidence to<br />
suggest that the downstream PI3K-mTOR pathway is perhaps<br />
the most important signaling pathway in GIST 3 and<br />
inhibitors <strong>of</strong> this pathway including everolimus, BKM120,<br />
and perifosine are currently in phase II clinical trials (Table<br />
1). In addition to clinical strategies to inhibit downstream<br />
pathways, multiple TKIs are currently in trials for TKIresistant<br />
GIST. Of note, a novel agent, crenolanib, is being<br />
tested in a phase II study for treatment <strong>of</strong> PDGFRA D842Vmutant<br />
GIST. As noted above, this particular mutation is<br />
strongly resistant to imatinib (and sunitinib as well). 22<br />
In terms <strong>of</strong> FDA-approved KIT/PDGFRA kinase inhibitors,<br />
a number <strong>of</strong> these have been tested in phase II studies<br />
<strong>of</strong> drug-resistant GIST. For example, nilotinib has been<br />
666<br />
studied in a several clinical studies against drug-resistant<br />
GIST. 23 Nilotinib is structurally very similar to imatinib<br />
and, like sunitinib, lacks the ability to overcome any <strong>of</strong> the<br />
activation loop mutations in exons 17 or 18. In addition, like<br />
imatinib, it also has limited potency against ATP binding<br />
pocket mutations. Because <strong>of</strong> this, nilotinib has shown<br />
limited activity in imatinib- and sunitinib-resistant GIST.<br />
In a phase III trial, 248 patients were randomly assigned to<br />
nilotinib versus best supportive care for third-line therapy.<br />
24 There was no statistical difference in overall survival<br />
between nilotinib and best supportive care. Sorafenib has<br />
more promising data for treatment <strong>of</strong> drug-resistant GIST,<br />
with an approximate 20-week PFS in the third-line, and<br />
even fourth-line setting. 25 This may be due to its apparent<br />
ability to overcome some exon 17 mutations in vitro. However,<br />
it appears that this agent has been supplanted in<br />
clinical development by its closely related analog, regorafenib.<br />
In a recent phase II trial, regorafenib showed a<br />
promising 10-month PFS in the third- or fourth-line setting.<br />
26 Based on these results, a randomized, double-blind,<br />
placebo-controlled phase III study was initiated for patients<br />
who were intolerant <strong>of</strong> or who had progression during prior<br />
first- and second-line therapy. The study completed enrollment<br />
in mid-2011 and the final study analysis is planned for<br />
sometime in early <strong>2012</strong>.<br />
The National Comprehensive Cancer Network guidelines<br />
suggest consideration <strong>of</strong> a clinical study or <strong>of</strong>f-label treatment<br />
with sorafenib, nilotinib, or dasatinib for physicians<br />
with patients with imatinib- and sunitinib-resistant GIST. 20<br />
In addition, reintroduction <strong>of</strong> imatinib can be considered.<br />
Fumagalli and colleagues have published data supporting<br />
rechallenging patients with imatinib or sunitinib after failure<br />
<strong>of</strong> standard and investigational agents. 27 Most GIST<br />
experts believe that life-long continuation <strong>of</strong> TKI therapy<br />
should be considered an essential component <strong>of</strong> best supportive<br />
care. As noted above, GIST patients <strong>of</strong>ten harbor<br />
deposits <strong>of</strong> quiescent, drug-sensitive tumor cells that will<br />
become clinically active following discontinuation <strong>of</strong> TKI<br />
therapy. These lesions can add to overall disease burden and<br />
patient symptoms. In addition, there is evidence to suggest<br />
that some secondary drug resistance mutations result in<br />
relative rather than absolute resistance to TKI therapy;<br />
continued TKI treatment may slow (but not arrest) such<br />
clones, potentially palliating symptoms.<br />
Role <strong>of</strong> Surgery in Advanced GIST<br />
BARNETT AND HEINRICH<br />
Given that there can be significant heterogeneity in secondary<br />
mutations and therefore mixed responses with TKI<br />
treatment <strong>of</strong> advanced GIST, many patients inquire about<br />
surgery to control their disease. Raut and colleagues published<br />
a study about their series <strong>of</strong> 69 patients who underwent<br />
surgery while receiving TKI therapy, with the majority<br />
receiving front-line imatinib (but a minority on second-line<br />
sunitinib). Notably, only those patients with overall stable<br />
disease (i.e., disease controlled by TKIs) strongly benefited<br />
from surgery in terms <strong>of</strong> disease control (Table 2). 28 In<br />
another study, 50 patients receiving second-line sunitinib<br />
underwent surgery for their disease and median PFS was a<br />
mere 5.8 months, and was independent <strong>of</strong> the immediate<br />
presurgical sunitinib clinical response status. That is,<br />
sunitinib-responding patients did not have better PFS after<br />
surgery than patients with frank progression on sunitinib<br />
before surgery. 29 In addition, over half <strong>of</strong> the patients had
TKI-RESISTANT GIST<br />
Table 2. <strong>Clinical</strong> outcomes following surgery for TKI treated patients (pts) with advanced GIST. Median progression-free and<br />
overall-survival data are tabulated based on the immediate pre-surgery response status <strong>of</strong> patients to TKI therapy. a<br />
Progression-Free Survival Overall Survival<br />
Study Pt # Treatment Status<br />
Stableb Limitc Gend Stableb Limitc Gend Raut (28) 69 IM 45 pts IM 3 SU 21 pts �40 mo. 8 mo. 3 mo. �40 mo. 30 mo. 6 mo.<br />
Dematteo (30) 40 IM 37 pts IM 3 SU 3 pts �48 mo. 12 mo. 3 mo. �48 mo. 19 mo. 11 mo.<br />
Raut (29) 50 IM 3 SU 50 pts 11 mo. 4 mo. 4 mo. �36 mo. 19 mo. 9 mo.<br />
a IM (surgery during front-line imatinib therapy). IM 3 SU (surgery during second-line sunitinib).<br />
b Stable or responsive disease prior to surgery.<br />
c Limited or focal disease progression prior to surgery.<br />
d Generalized disease progression prior to surgery.<br />
complications from surgery, and it is important to keep in<br />
mind that surgical complications usually lead to a delay in<br />
restarting systemic therapy. Therefore, only for patients with<br />
primary disease, or for symptom control in advanced disease,<br />
is surgery generally recommended as a treatment strategy.<br />
Conclusion<br />
The knowledge <strong>of</strong> the biology <strong>of</strong> GIST has made this<br />
cancer a model <strong>of</strong> targeted therapy with tyrosine kinase<br />
Authors’ Disclosures <strong>of</strong> Potential Conflicts <strong>of</strong> Interest<br />
Employment or<br />
Leadership<br />
Positions<br />
Consultant or<br />
Advisory Role<br />
Author<br />
Christine M. Barnett*<br />
Michael C. Heinrich MolecularMD;<br />
Novartis; Pfizer<br />
*No relevant relationships to disclose.<br />
1. Corless CL, Fletcher JA, Heinrich MC. Biology <strong>of</strong> gastrointestinal<br />
stromal tumors. J Clin Oncol. 2004;22:3813-3825.<br />
2. Miettinen M, Sobin LH, Sarlomo-Rikala M. Immunohistochemical spectrum<br />
<strong>of</strong> GISTs at different sites and their differential diagnosis with a<br />
reference to CD117 (KIT). Mod Pathol. 2000;13:1134-1142.<br />
3. Corless CL, Barnett CM, Heinrich MC. Gastrointestinal stromal tumours:<br />
origin and molecular oncology. Nat Rev Cancer. 2011;11:865-878.<br />
4. Duensing A, Joseph NE, Medeiros F, et al. Protein Kinase C theta<br />
(PKCtheta) expression and constitutive activation in gastrointestinal stromal<br />
tumors (GISTs). Cancer Res. 2004;64:5127-5131.<br />
5. Druker BJ, Tamura S, Buchdunger E, et al. Effects <strong>of</strong> a selective<br />
inhibitor <strong>of</strong> the Abl tyrosine kinase on the growth <strong>of</strong> Bcr-Abl positive cells. Nat<br />
Med. 1996;2:561-566.<br />
6. Heinrich MC, Griffith DJ, Druker BJ, et al. Inhibition <strong>of</strong> c-kit receptor<br />
tyrosine kinase activity by STI 571, a selective tyrosine kinase inhibitor.<br />
Blood. 2000;96:925-932.<br />
7. Tuveson DA, Willis NA, Jacks T, et al. STI571 inactivation <strong>of</strong> the<br />
gastrointestinal stromal tumor c-KIT oncoprotein: biological and clinical<br />
implications. Oncogene. 2001;20:5054-5058.<br />
8. van Oosterom AT, Judson I, Verweij J, et al. Safety and efficacy <strong>of</strong><br />
imatinib (STI571) in metastatic gastrointestinal stromal tumours: a phase I<br />
study. Lancet. 2001;358:1421-1423.<br />
9. Blanke CD, Demetri GD, von Mehren M, et al. Long-term results from a<br />
randomized phase II trial <strong>of</strong> standard- versus higher-dose imatinib mesylate<br />
for patients with unresectable or metastatic gastrointestinal stromal tumors<br />
expressing KIT. J Clin Oncol. 2008;26:620-625.<br />
10. Comparison <strong>of</strong> two doses <strong>of</strong> imatinib for the treatment <strong>of</strong> unresectable<br />
or metastatic gastrointestinal stromal tumors: A meta-analysis <strong>of</strong> 1,640<br />
patients. J Clin Oncol. 2010;28:1247-1253.<br />
11. Le Cesne A, Ray-Coquard I, Bui BN, et al. Discontinuation <strong>of</strong> imatinib<br />
in patients with advanced gastrointestinal stromal tumours after 3 years <strong>of</strong><br />
treatment: an open-label multicentre randomised phase 3 trial. Lancet Oncol.<br />
2010;11:942-949.<br />
12. von Mehren M, Heinrich MC, Joensuu H, et al. Follow-up results after<br />
9 years (yrs) <strong>of</strong> the ongoing, phase II B2222 trial <strong>of</strong> imatinib mesylate (IM) in<br />
inhibitors. Currently, the mainstay <strong>of</strong> treatment for this<br />
disease is TKI-based therapy with enrollment in a clinical<br />
trial if multiple TKIs fail to control disease. Understanding<br />
the mechanisms <strong>of</strong> disease persistence and drug resistance<br />
has helped identify new targets for therapy. Hopefully, our<br />
increased understanding <strong>of</strong> GIST biology will not only help<br />
us improve current treatment, but ultimately help create<br />
new treatments that might cure patients with advanced<br />
disease.<br />
Stock<br />
Ownership Honoraria<br />
Research<br />
Funding<br />
MolecularMD Novartis ARIAD; Arog;<br />
ImClone<br />
Systems;<br />
Novartis; Pfizer<br />
REFERENCES<br />
Expert<br />
Testimony<br />
Other<br />
Remuneration<br />
patients (pts) with metastatic or unresectable KIT� gastrointestinal stromal<br />
tumors (GIST). J Clin Oncol. 2011;29 (suppl; abstr 10016).<br />
13. Heinrich MC, Corless CL, Demetri GD, et al. Kinase mutations and<br />
imatinib response in patients with metastatic gastrointestinal stromal tumor.<br />
J Clin Oncol. 2003;21:4342-4349.<br />
14. Demetri GD, Wang Y, Wehrle E, et al. Imatinib plasma levels are<br />
correlated with clinical benefit in patients with unresectable/metastatic<br />
gastrointestinal stromal tumors. J Clin Oncol. 2009;27:3141-3147.<br />
15. Antonescu CR, Besmer P, Guo T, et al. Acquired resistance to imatinib<br />
in gastrointestinal stromal tumor occurs through secondary gene mutation.<br />
Clin Cancer Res. 2005;11:4182-4190.<br />
16. Liegl B, Kepten I, Le C, et al. Heterogeneity <strong>of</strong> kinase inhibitor<br />
resistance mechanisms in GIST. J Pathol. 2008;216:64-74.<br />
17. Demetri GD, van Oosterom AT, Garrett CR, et al. Efficacy and safety <strong>of</strong><br />
sunitinib in patients with advanced gastrointestinal stromal tumour after<br />
failure <strong>of</strong> imatinib: a randomised controlled trial. Lancet. 2006;368:1329-<br />
1338.<br />
18. George S, Blay JY, Casali PG, et al. <strong>Clinical</strong> evaluation <strong>of</strong> continuous<br />
daily dosing <strong>of</strong> sunitinib malate in patients with advanced gastrointestinal<br />
stromal tumour after imatinib failure. Eur J Cancer. 2009;45:1959-1968.<br />
19. Gramza AW, Corless CL, Heinrich MC. Resistance to tyrosine kinase<br />
inhibitors in gastrointestinal stromal tumors. Clin Cancer Res. 2009;15:7510-<br />
7518.<br />
20. Demetri GD, von Mehren M, Antonescu CR, et al. NCCN Task Force<br />
report: update on the management <strong>of</strong> patients with gastrointestinal stromal<br />
tumors. J Natl Compr Canc Netw. 2010;8 Suppl 2:S1-41; quiz S2-4.<br />
21. Bauer S, Yu LK, Demetri GD, et al. Heat shock protein 90 inhibition in<br />
imatinib-resistant gastrointestinal stromal tumor. Cancer Res. 2006;66:9153-<br />
9161.<br />
22. Heinrich MC, Corless CL, Duensing A, et al. PDGFRA activating<br />
mutations in gastrointestinal stromal tumors. Science. 2003;299:708-710.<br />
23. Demetri GD, Casali PG, Blay JY, et al. A phase I study <strong>of</strong> single-agent<br />
nilotinib or in combination with imatinib in patients with imatinib-resistant<br />
gastrointestinal stromal tumors. Clin Cancer Res. 2009;15:5910-5916.<br />
24. Reichardt P, Blay J, Gelderblom H, et al. Phase III trial <strong>of</strong> nilotinib in<br />
667
patients with advanced gastrointestinal stromal tumor (GIST): First results<br />
from ENEST g3. J Clin Oncol. 2010; 28 (suppl; abstr 10017).<br />
25. Wiebe L, Kasza KE, Maki RG, et al. Activity <strong>of</strong> sorafenib (SOR) in<br />
patients (pts) with imatinib (IM) and sunitinib (SU)-resistant (RES) gastrointestinal<br />
stromal tumors (GIST): a phase II trial <strong>of</strong> the University <strong>of</strong> Chicago<br />
Phase II Consortium. J Clin Oncol. 2008;26 (suppl; abstr 10502).<br />
26. George S, von Mehren M, Heinrich MC, et al. A multicenter phase II<br />
study <strong>of</strong> regorafenib in patients (pts) with advanced gastrointestinal stromal<br />
tumor (GIST), after therapy with imatinib (IM) and sunitinib (SU). J Clin<br />
Oncol. 29 (suppl; abstr 10007).<br />
27. Fumagalli E, Coco P, Morosi C, et al. Sunitinib rechallenge in two<br />
668<br />
BARNETT AND HEINRICH<br />
advanced GIST patients after third-line anti-tyrosine kinase therapy. J Clin<br />
Oncol. 2010;28 (suppl; abstr e20519).<br />
28. Raut CP, Posner M, Desai J, et al. Surgical management <strong>of</strong> advanced<br />
gastrointestinal stromal tumors after treatment with targeted systemic<br />
therapy using kinase inhibitors. J Clin Oncol. 2006;24:2325-2331.<br />
29. Raut CP, Wang Q, Manola J, et al. Cytoreductive surgery in patients<br />
with metastatic gastrointestinal stromal tumor treated with sunitinib<br />
malate. Ann Surg Oncol. 2010;17:407-415.<br />
30. DeMatteo RP, Maki RG, Singer S, et al. Results <strong>of</strong> tyrosine kinase<br />
inhibitor therapy followed by surgical resection for metastatic gastrointestinal<br />
stromal tumor. Ann Surg. 2007;245:347-352.
BIOLOGIC PRINCIPLES OF TARGETED<br />
COMBINATION THERAPY<br />
CHAIR<br />
Johann S. de Bono, MB, ChB, MSc, PhD<br />
Royal Marsden Hospital<br />
Sutton, United Kingdom<br />
SPEAKERS<br />
Antoni Ribas, MD<br />
University <strong>of</strong> California, Los Angeles<br />
Los Angeles, CA<br />
Ge<strong>of</strong>frey Shapiro, MD, PhD<br />
Dana-Farber Cancer Institute<br />
Boston, MA
Opportunities and Pitfalls <strong>of</strong> Targeted<br />
Therapeutic Combinations in Solid Tumors<br />
By Joaquin Mateo, MD, MSc, Michael Ong, BSc, MD,<br />
Timothy A. Yap, BSc, MBBS, MRCP, and Johann S. de Bono, MB ChB, MSc, PhD<br />
Overview: Recent advances in cancer biology have led to the<br />
discovery and development <strong>of</strong> chemical compounds and drugs<br />
that target specific cellular receptors, mediators, or effectors<br />
that are central to oncogenic survival, growth, and invasion.<br />
However, the complexity <strong>of</strong> tumor biology makes it is unrealistic<br />
to expect an antitumor therapeutic to be successful<br />
based on the inhibition <strong>of</strong> a single target or even a lone<br />
signaling pathway. Therefore, the potential success <strong>of</strong> such<br />
“targeted therapies” is likely to require the development <strong>of</strong><br />
multiagent combinations. Combination strategies have a<br />
greater likelihood <strong>of</strong> addressing issues with genetically complex<br />
tumors, potentially avoiding drug resistance mechanisms<br />
NOVEL DRUGS aim to selectively inhibit mechanisms<br />
essential to cancer pathogenesis, while limiting “<strong>of</strong>ftarget<br />
effects” and undesired toxicities. Despite the hype,<br />
only a limited number <strong>of</strong> targeted agents have been successful;<br />
this includes the inhibition <strong>of</strong> ABL and c-KIT by imatinib<br />
for chronic myeloid leukemia and in gastrointestinal<br />
stromal tumors (GIST). 1 However, primary and secondary<br />
resistance to therapy is common because <strong>of</strong> mechanisms<br />
such as novel mutations that decrease imatinib binding, or<br />
mutations in downstream proteins involved in the phosphatidylinositol<br />
3-kinase (PI3K) pathway that are c-KIT independent.<br />
2,3 Ultimately, the application <strong>of</strong> a single targeted<br />
agent even in these successful models has not abrogated<br />
the development <strong>of</strong> resistance, nor has it led to increased<br />
rates <strong>of</strong> cure. 4,5<br />
Opportunities with Targeted Combinations<br />
Numerous well-established examples exist in oncology<br />
where combination regimens <strong>of</strong> anticancer drugs are more<br />
successful than single agents. An example <strong>of</strong> a successful<br />
antitumor combination is 5-fluorouracil, leucovorin, and<br />
irinotecan (FOLFIRI) with bevacizumab in colorectal cancer.<br />
6 In addition, it is well known that diseases such as<br />
Hodgkin lymphoma require multiagent chemotherapy to<br />
achieve cure. This and our increased understanding <strong>of</strong><br />
disease biology indicate that in order to increase the odds <strong>of</strong><br />
successful anticancer strategies, specific targeted agents<br />
should be combined with either other selective inhibitors or<br />
chemotherapy.<br />
The development <strong>of</strong> combination regimens involving targeted<br />
therapies versus chemotherapy combinations may<br />
differ in that chemotherapy was traditionally studied in<br />
combination only once a single agent was approved and<br />
demonstrated clinical activity benefit. Conversely, targeted<br />
drugs may and probably should be tested in hypothesistesting<br />
rational combinations earlier in the drug development<br />
journey if tolerability data are available, assuming<br />
there is a strong biologic rationale to combine two drugs.<br />
Even if these agents demonstrate modest anticancer activity<br />
as single agents, with substantial pharmacodynamic inhibition<br />
<strong>of</strong> target and pathway, satisfactory pharmacokinetics<br />
and tolerability, combination studies may be warranted. We<br />
envision that earlier testing <strong>of</strong> such combinations in early-<br />
670<br />
through the inhibition <strong>of</strong> escape signaling pathways and<br />
slowing the development <strong>of</strong> newly resistant tumor cells. Combination<br />
regimes also have the potential <strong>of</strong> enhancing target<br />
inhibition through synergistic antitumor effects and minimizing<br />
drug-related toxicities to patients. However, numerous<br />
challenges to developing these combinations exist. This review<br />
will focus on the opportunities and pitfalls <strong>of</strong> developing<br />
novel targeted drug combinations, with a particular focus on<br />
early-phase drug development, where the greatest challenges<br />
exist, analyzing key points for the design and development <strong>of</strong><br />
clinical trials for combinations <strong>of</strong> targeted agents.<br />
phase trials to be increasingly desirable and an important<br />
opportunity to accelerate drug development.<br />
Before the clinical testing <strong>of</strong> a combination regimen, it is<br />
critical to understand the underlying biologic rationale and<br />
to have some preclinical evidence demonstrating synergistic<br />
or additive antitumor effects for the anticancer drugs being<br />
evaluated. Because targeted agents have been developed<br />
as potent inhibitors <strong>of</strong> select biologic targets, combination<br />
targeted therapy strategies can also seize the opportunity to<br />
selectively titrate effects on multiple targets using potent<br />
inhibitors. In this manner, combination dosages and schedules<br />
can be altered to optimize antitumor efficacy and<br />
minimize the toxicity pr<strong>of</strong>ile (Table 1).<br />
Examples <strong>of</strong> Strategies to Combine Targeted Agents<br />
Superinhibition <strong>of</strong> a single target. This strategy maximizes<br />
the inhibition <strong>of</strong> a single target, enhancing the “on<br />
target” effect. Examples include strategies <strong>of</strong> dual HER2<br />
blockade (e.g., trastuzumab with lapatinib or pertuzumab in<br />
combination). Several studies have demonstrated these<br />
strategies to be tolerable, with evidence <strong>of</strong> clinical benefit,<br />
in various stages <strong>of</strong> breast cancer. 7,8 This strategy may be<br />
best utilized in tumors that are highly dependent on a single<br />
gene or target, without significant resistance mechanisms<br />
that bypass the target. This strategy also has the disadvantage<br />
<strong>of</strong> possibly enhancing overlapping drug-related<br />
toxicities.<br />
Inhibition <strong>of</strong> several targets within a single pathway. This<br />
strategy takes an approach to inhibiting multiple substrates<br />
to maximize pathway inhibition while minimizing resistance<br />
mechanisms that may occur because <strong>of</strong> regulatory<br />
feedback loops. For example, there is evidence <strong>of</strong> upregulation<br />
<strong>of</strong> AKT phosphorylation after mTOR inhibition in several<br />
cell lines and human tumor types, 9 suggesting a role for<br />
From the Drug Development Unit, Royal Marsden NHS Foundation Trust, The Institute<br />
<strong>of</strong> Cancer Research, Downs Road, Sutton, Surrey, United Kingdom.<br />
Authors’ disclosures <strong>of</strong> potential conflicts <strong>of</strong> interest are found at the end <strong>of</strong> this article.<br />
Address reprint requests to Pr<strong>of</strong>essor Johann S. de Bono, MB ChB, MSc, PhD, Division<br />
<strong>of</strong> Cancer Therapeutics and Division <strong>of</strong> <strong>Clinical</strong> Sciences, The Institute <strong>of</strong> Cancer Research/<br />
Royal Marsden NHS Foundation Trust, Downs Road, Sutton, Surrey SM2 5PT, United<br />
Kingdom; email: johann.de-bono@icr.ac.uk<br />
© <strong>2012</strong> by <strong>American</strong> <strong>Society</strong> <strong>of</strong> <strong>Clinical</strong> <strong>Oncology</strong>.<br />
1092-9118/10/1-10
COMBINING TARGETED AGENTS IN CLINICAL TRIALS<br />
the rational combination <strong>of</strong> mTOR inhibitors with either AKT<br />
or PIK3CA inhibitors or the development <strong>of</strong> TOR kinase<br />
inhibitors. This strategy may be best utilized in tumors driven<br />
mainly by one pathway but which may fail if alternative<br />
pathways are prominent as a mechanism <strong>of</strong> resistance. This<br />
strategy also has the disadvantage <strong>of</strong> possible overlapping<br />
drug-related toxicities via inhibiting the same pathway.<br />
Targeting parallel or unrelated pathways. This strategy<br />
maximizes the likelihood <strong>of</strong> avoiding resistance mechanisms,<br />
especially those caused by the upregulation <strong>of</strong> parallel<br />
pathways, and also minimizes overlapping toxicities. The<br />
concomitant inhibition <strong>of</strong> two pathways may also result in<br />
selective tumor cell cytotoxicity, whereas each agent alone<br />
may not be active as a single agent.<br />
An example <strong>of</strong> this approach is the dual targeting <strong>of</strong> the<br />
PI3K and Ras/Raf/Mek/ERK pathways. Deregulation <strong>of</strong> the<br />
Ras/Raf/Mek/ERK pathway has been identified as a determinant<br />
<strong>of</strong> resistance to PI3K-Akt inhibitors. 10 There is also<br />
strong preclinical evidence that ERK-independent mechanisms<br />
<strong>of</strong> resistance to B-RAF and MEK inhibitors are<br />
mediated by the upregulation <strong>of</strong> the PIK3CA/Akt/mTOR<br />
pathway. 11 In murine models, double inhibition <strong>of</strong> PIK3CA<br />
and MEK shows synergistic antitumor activity for K-RAS<br />
mutant tumors, 12 and PI3K-Akt activation leads to MEKinhibitor<br />
resistance in K-RAS mutants, which is reversible<br />
when blocking both pathways. 13 Moreover, mutations in<br />
both pathways tend to coexist. 14<br />
<strong>Clinical</strong> trials are currently being conducted to evaluate<br />
this strategy <strong>of</strong> targeting parallel pathways. Recently, a<br />
first-in-human trial <strong>of</strong> the pan-AKT inhibitor MK-2206 has<br />
been reported; although strong pharmacodynamics evidence<br />
<strong>of</strong> AKT signaling blockade was demonstrated, minimal evidence<br />
<strong>of</strong> single-agent antitumor activity was observed. 15 A<br />
combination trial <strong>of</strong> MK-2206 with the MEK inhibitor<br />
AZD6244 is currently being evaluated (NCT01021748) and<br />
is demonstrating antitumor activity in KRAS-mutant cancers.<br />
Other similar early-phase trials <strong>of</strong> combinations based<br />
on this approach include drugs targeting MEK and mTOR<br />
(NCT01378377), or MEK and PIK3CA/mTOR inhibition<br />
KEY POINTS<br />
● The biology <strong>of</strong> cancer cells involves complex signaling<br />
networks, so anticancer strategies may not be successful<br />
if blocking single targets.<br />
● A combination <strong>of</strong> targeted agents will be necessary to<br />
maximize antitumor activity, counteracting primary<br />
and secondary resistance.<br />
● The rationale for evaluating combinations <strong>of</strong> drugs<br />
must rely on biology-driven hypothesis and smart<br />
screening strategies.<br />
● Different strategies include using two or more drugs<br />
against the same target, targeting different points<br />
upstream and downstream <strong>of</strong> one pathway, or targeting<br />
several pathways simultaneously on which the<br />
cancer cell is dependent.<br />
● Design <strong>of</strong> clinical trials with combinations <strong>of</strong> targeted<br />
agents should consider how to deal with the challenges<br />
pertaining to such studies.<br />
Table 1. Opportunities and Pitfalls <strong>of</strong> Combining<br />
Targeted Agents<br />
Opportunities Pitfalls<br />
1. Validate novel biologic hypotheses<br />
2. Synergize antitumor effect without<br />
synergizing toxicity: increasing the<br />
therapeutic window<br />
3. Further develop single agents that do<br />
not have activity as monotherapy:<br />
synthetic lethality<br />
4. Counteract primary and secondary<br />
resistance<br />
5. Develop novel indications for existing<br />
and/or approved drugs<br />
(NCT01390818). These data allow us to envision the testing<br />
<strong>of</strong> biology-driven hypotheses with multiple combinations;<br />
these may eventually require more drugs for optimal blockade<br />
<strong>of</strong> cancer growth. Combinations for the treatment <strong>of</strong><br />
advanced prostate cancer that we propose include abiraterone<br />
acetate and MDV3100; either <strong>of</strong> these agents with<br />
PI3K/AKT/TOR inhibitors; or possibly triple therapy with<br />
abiraterone, MDV3100, and a PI3K/TOR inhibitor.<br />
Choosing the Right Combination Therapy<br />
1. Unreliable preclinical models<br />
2. Optimal selection <strong>of</strong> drugs and<br />
targets to combine<br />
3. Optimal sequence and dosage <strong>of</strong><br />
the combination<br />
4. Risk <strong>of</strong> overlapping toxicities<br />
5. Lack <strong>of</strong> standardized design for<br />
phase I/II trials for combination<br />
targeted therapies<br />
6. Competing interests <strong>of</strong> different<br />
researchers, corporations, and/or<br />
institutions to combine therapies<br />
When combining two (or more) targeted therapies for the<br />
first time, each must be tested at various doses and dosing<br />
schedules in a phase I study. However, the process <strong>of</strong> dose<br />
finding, or defining the recommended phase II trial dose,<br />
can take 1 year or longer to complete depending on the<br />
complexity <strong>of</strong> the trial and compatibility <strong>of</strong> the agents, all at<br />
significant cost. Therefore, it is critical to utilize strategies<br />
that identify the best therapeutic combinations, and furthermore<br />
identify targets that cancer cells are reliant on so that<br />
when blocked, there is a lethal or at least cytostatic effect.<br />
This type <strong>of</strong> approach is based on concepts <strong>of</strong> oncogene<br />
addition, nononcogene addiction, and synthetic lethality<br />
that attack the “hallmarks” <strong>of</strong> cancer. 16<br />
Adopting systems-based approaches may help to address<br />
this complexity, such as utilizing information from deep<br />
transcriptomic, genomic, proteomic, and/or metabolic analyses.<br />
17 Although this “personalized” approach is attractive,<br />
it may <strong>of</strong>ten yield information on targets for which there is<br />
no active drug, for which there are only “passenger” mutations<br />
rather than drivers <strong>of</strong> oncogenesis, or for genes for<br />
which we have incomplete functional information. It has<br />
been suggested that systems-based approaches can be used<br />
to select targeted combinations 17 ; however, further development<br />
and application <strong>of</strong> bioinformatics is clearly necessary<br />
to process the high amounts <strong>of</strong> data derived from massive<br />
screening strategies.<br />
Another strategy is to identify potential drug combinations<br />
using RNA interference (RNAi) to simulate models <strong>of</strong><br />
pharmacologic inhibition. In this method, wild-type cell lines<br />
or cell lines with induced loss <strong>of</strong> function for a particular<br />
target are exposed to RNAi libraries with or without exposure<br />
to a particular drug to look for antitumoral synergism<br />
or drug sensitization. These RNAi screens can therefore<br />
select potential combination <strong>of</strong> targets to be tested further<br />
18,19 and ideally identify potential synthetic lethality<br />
effects when inhibiting two a priori nonlethal targets. How-<br />
671
Fig. 1. Schematic demonstrating the increased complexity <strong>of</strong> dose and schedule selection when combining targeted agents. Key factors that<br />
must be considered include the pharmacokinetic (PK) and pharmacodynamic (PD) properties <strong>of</strong> each single agent, potential PK and PD<br />
interactions between the agents, how exposure to the combination may result in antitumor synergism, and what toxicities may overlap.<br />
ever, current limitations <strong>of</strong> this approach include incomplete<br />
specificity <strong>of</strong> the RNAis used, the appearance <strong>of</strong> several<br />
phenotypes after the inhibition by RNAi, 20 and the issue<br />
that subsequent preclinical models are not reliable in predicting<br />
the ultimate sensitivity <strong>of</strong> tumor tissue to combination<br />
therapy. 21<br />
Ultimately, however, combinations <strong>of</strong> targeted agents<br />
must be selected for evaluation based on a strong biologic<br />
rationale. Preclinical evidence for targeted combinations is<br />
<strong>of</strong>ten lacking, but can be critical to take a proper combination<br />
forward. For example, Wam and colleagues observed an<br />
upregulation <strong>of</strong> AKT phosphorylation in rhabdomyosarcoma<br />
cell lines and xenografts that was dependent on insulin<br />
growth factor-1 receptor (IGF-1R) activation. 22 This codependence<br />
suggested a potential for combination targeted<br />
therapy using mTOR and IGF-1R blockade. Results reported<br />
from an early-phase trial combining ridaforolimus and an<br />
anti-IGF-1R antibody showed signs <strong>of</strong> promising clinical<br />
benefit. 23 Nonetheless, initiatives <strong>of</strong> sequential biopsies<br />
should be integral to clinical research for a better understanding<br />
<strong>of</strong> molecular mechanism <strong>of</strong> resistance to targeted<br />
agents 17 to better define whether combination treatments<br />
can address these issues.<br />
672<br />
Design <strong>of</strong> Early-Phase <strong>Clinical</strong> Trials for Combinations<br />
The design <strong>of</strong> trials evaluating combinations <strong>of</strong> targeted<br />
agents can be significantly more complex than trials <strong>of</strong><br />
single targeted agents. Successful combination strategies<br />
require not only validation <strong>of</strong> the biologic rationale but also<br />
determination <strong>of</strong> the optimal dosing and scheduling. Combinations<br />
<strong>of</strong> drugs may also be significantly affected by<br />
pharmacokinetic and pharmacodynamic interactions, and<br />
phase I or II trials may be designed to evaluate multiple<br />
schedules <strong>of</strong> interest to determine not only tolerable regimens<br />
but feasible dosing regimens (Fig. 1).<br />
Early valid pharmacodynamic biomarkers should be required<br />
for pro<strong>of</strong> <strong>of</strong> mechanism for the combination, ideally<br />
from both tumor and normal tissue biopsies when feasible.<br />
Design <strong>of</strong> early-phase clinical trials should be flexible<br />
enough to facilitate “real time” reverse translation between<br />
clinical and laboratory research.<br />
Dose Escalation in Combination Trials<br />
MATEO ET AL<br />
In single-agent phase I trials, dose escalation usually<br />
occurs by standard rules such as the 3�3 design. However,<br />
in combination trials, the maximally tolerated dose (MTD) <strong>of</strong><br />
each drug is likely to be well defined; therefore, evaluation <strong>of</strong>
COMBINING TARGETED AGENTS IN CLINICAL TRIALS<br />
the doses for combination can occur in many permutations.<br />
Dose-escalation rules for combination therapy must take into<br />
consideration preclinical data for the single agent and the<br />
combination. Depending on the stage <strong>of</strong> development <strong>of</strong> each<br />
agent, their mechanism <strong>of</strong> action and the potential overlapping<br />
toxicities, different scenarios are envisioned: dose escalation<br />
is probably best pursued sequentially, increasing the<br />
dose <strong>of</strong> one compound in every step; dose escalation <strong>of</strong> both<br />
drugs at the same time is not recommended; and to optimize<br />
antitumor cell kill there is also merit in fixing the drug<br />
targeting the driver pathway (e.g., BRAF/MEK inhibitor for<br />
BRAF mutation in melanoma) at the highest tolerable dose<br />
while dose escalating the combination drug (e.g., PI3K/AKT/<br />
TOR inhibitor). When combining two drugs with no pharmacokinetic<br />
interactions and nonoverlapping mechanisms<br />
<strong>of</strong> action and toxicity, and when we already have extensive<br />
experience regarding the tolerability pr<strong>of</strong>ile <strong>of</strong> each drug,<br />
fewer dose-escalation steps may be necessary to accelerate<br />
combination development, as recently reported in a trial<br />
exploring the combination <strong>of</strong> sunitinib and everolimus in<br />
metastatic renal cell carcinoma (mRCC). 24<br />
Novel combinatory designs have also been suggested as<br />
potential improvements over algorithmic approaches. These<br />
strategies try to pursue fewer levels <strong>of</strong> dose escalation,<br />
thereby reducing the rate <strong>of</strong> patients exposed to biologically<br />
inactive doses without exposing them to more toxicities, in<br />
contrast to more classical drug development designs. 25<br />
Novel models <strong>of</strong> multiple parallel cohorts permit escalating<br />
two or more drugs in a sequential/parallel manner based on<br />
mechanism-related toxicities. 26 Bayesian model-based designs<br />
are based on continual reassessment <strong>of</strong> risk <strong>of</strong> toxicities<br />
and have been applied to dose escalation <strong>of</strong> combination<br />
targeted agents. 27<br />
Flexible dose escalation and de-escalation rules and intense<br />
knowledge about the mechanisms <strong>of</strong> toxicity <strong>of</strong> each<br />
drug in the combination are required to maximize outcomes<br />
without compromising a potentially or already known active<br />
dose <strong>of</strong> a single agent. Despite the availability <strong>of</strong> these novel<br />
trial designs, they are not commonly used.<br />
Furthermore, although the rationale for the combination<br />
should consider which pathway is the main driver <strong>of</strong> the<br />
disease and therefore which drug is administered at a dose<br />
level closer to its single-agent MTD, investigators may have<br />
to be prepared to decrease the dose <strong>of</strong> the main drug if the<br />
biologic rationale <strong>of</strong> a trial suggests a likelihood that synergism<br />
and full-dose combination is not tolerable.<br />
Biomarker-Guided Selection <strong>of</strong> Patients<br />
Although the selection <strong>of</strong> patients in early-phase trials<br />
based on biomarkers may make trial recruitment more difficult,<br />
combinations <strong>of</strong> targeted agents based on hypotheses <strong>of</strong><br />
synergistic biologic effect will ultimately require patient selection<br />
using biomarkers that identify populations that may<br />
benefit. Unfortunately, <strong>of</strong>ten the development <strong>of</strong> valid biomarkers<br />
that predict for benefit are <strong>of</strong>ten slow to develop<br />
in both clinical and preclinical models, making patient selection<br />
in early-phase trials difficult. The development <strong>of</strong> predictive<br />
biomarkers for patient selection is crucial to the success <strong>of</strong><br />
targeted agents as well as combinations <strong>of</strong> targeted agents.<br />
Assessment <strong>of</strong> Tolerability in Combination Studies<br />
Since chronic dosing is usually indicated for these drug<br />
combinations, delayed and cumulative toxicities should be<br />
carefully considered when selecting the recommended dose<br />
for further development. As an example, a phase I trial <strong>of</strong><br />
the combination <strong>of</strong> bevacizumab and everolimus 28 escalated<br />
doses up to 10 mg/kg fortnightly and 10 mg/d po respectively,<br />
without observing dose-limiting toxicities during the<br />
first cycles <strong>of</strong> treatment. In a phase II trial in mRCC, 29 up to<br />
a quarter <strong>of</strong> the patients had grade 3 or 4 proteinuria with<br />
the combination, while grade 3 or 4 proteinuria with bevacizumab<br />
alone at the same dose in mRCC was below 8%. 30<br />
Trial designs that integrate information from late onset<br />
toxicities for deciding regarding dose selection may be warranted.<br />
31 In addition, clinical trials should aim to assess<br />
mechanistic data to define causal relationships with drug<br />
and observed toxicities to make better informed decisions<br />
regarding choices <strong>of</strong> combination targeted agents. Uncertainty<br />
regarding the mechanisms behind targeted agent<br />
toxicity and <strong>of</strong>f-target drug effects may negatively affect<br />
promising combinations.<br />
Regulatory Concerns and Legal Barriers<br />
The evaluation <strong>of</strong> different combinations <strong>of</strong> several targeted<br />
agents from multiple companies may sometimes be<br />
challenging because <strong>of</strong> competing interests, and may be<br />
further complicated if multiple sponsors are not fully committed<br />
to the trial. A clear example may be combining an<br />
established drug or chemotherapy regimen with a novel<br />
agent, as the outcome <strong>of</strong> the trial may not be equally<br />
pr<strong>of</strong>itable for both sponsors. Apart from intellectual property<br />
issues and the requirement <strong>of</strong> financial support, the<br />
risk <strong>of</strong> novel toxicities and adverse outcomes with concomitant<br />
administration might be considered a risky commercial<br />
decision for sponsors whose antitumor compound is already<br />
at a more advanced stage <strong>of</strong> development. Academic institutions,<br />
patient lobbies, and regulatory authorities may<br />
need to assume a championing role for some combinations<br />
with a strong biologic rationale that are perhaps less commercially<br />
attractive.<br />
Conclusion<br />
The advent <strong>of</strong> targeted therapies has shifted the emphasis<br />
in drug development from cytotoxic chemotherapy combinations<br />
to selective and potent molecular therapeutics. Such<br />
specific inhibitors may not result in significant or durable<br />
antitumor activity as single agents, and should therefore be<br />
given in combination with one or more drugs to molecularly<br />
defined populations <strong>of</strong> patients. We should now be focusing<br />
more efforts on combinatorial drug development strategies,<br />
perhaps even from the outset <strong>of</strong> first-in-human trials to<br />
maximize the likelihood <strong>of</strong> benefit to patients dying <strong>of</strong><br />
advanced cancer. Such an approach may be able to further<br />
accelerate the delivery <strong>of</strong> anticancer treatment to benefit<br />
and improve the lives <strong>of</strong> our patients with cancer.<br />
ACKNOWLEDGMENTS<br />
The Drug Development Unit <strong>of</strong> the Royal Marsden NHS<br />
Foundation Trust and The Institute <strong>of</strong> Cancer Research is<br />
supported in part by a program grant from Cancer Research<br />
United Kingdom. Support was also provided by the Experimental<br />
Cancer Medicine Centre (to The Institute <strong>of</strong> Cancer<br />
Research) and the National Institute for Health Research Biomedical<br />
Research Centre (jointly to the Royal Marsden NHS<br />
Foundation Trust and The Institute <strong>of</strong> Cancer Research).<br />
673
Authors’ Disclosures <strong>of</strong> Potential Conflicts <strong>of</strong> Interest<br />
Employment or<br />
Leadership<br />
Author<br />
Positions<br />
Joaquin Mateo*<br />
Michael Ong*<br />
Timothy A. Yap*<br />
Johann S. de Bono The Institute <strong>of</strong><br />
Cancer Research<br />
(L)<br />
*No relevant relationships to disclose.<br />
Consultant or<br />
Advisory Role<br />
AstraZeneca;<br />
Boehringer<br />
Ingelheim;<br />
Genentech;<br />
Johnson &<br />
Johnson<br />
1. Demetri GD, von Mehren M, Blanke CD, et al. Efficacy and safety <strong>of</strong><br />
imatinib mesylate in advanced gastrointestinal stromal tumors. N Engl<br />
J Med. 2002;347:472-480.<br />
2. Antonescu CR, Besmer P, Guo T, et al. Acquired resistance to imatinib<br />
in gastrointestinal stromal tumor occurs through secondary gene mutation.<br />
Clin Cancer Res. 2005;11:4182-4190.<br />
3. Yang J, Ikezoe T, Nishioka C, et al. Long-term exposure <strong>of</strong> gastrointestinal<br />
stromal tumor cells to sunitinib induces epigenetic silencing <strong>of</strong> the<br />
PTEN gene. Int J Cancer. <strong>2012</strong>;130:959-966.<br />
4. Nilsson B, Sjölund K, Kindblom L-G, et al. Adjuvant imatinib treatment<br />
improves recurrence-free survival in patients with high-risk gastrointestinal<br />
stromal tumors (GIST). Br J Cancer. 2007;96:1656-1658.<br />
5. Dematteo RP, Ballman KV, Antonescu CR, et al. Adjuvant imatinib<br />
mesylate after resection <strong>of</strong> localized, primary gastrointestinal stromal tumor:<br />
a randomised, double-blind, placebo-controlled trial. Lancet. 2009;373:1097-<br />
2104.<br />
6. Hurwitz H, Fehrenbacher L, Novotny W, et al. Bevacizumab plus<br />
irinotecan, fluorouracil, and leucovorin for metastatic colorectal cancer.<br />
N Engl J Med. 2004;350:2335-2342.<br />
7. Nahta R, Hung M-C, Esteva FJ. The HER-2-targeting antibodies trastuzumab<br />
and pertuzumab synergistically inhibit the survival <strong>of</strong> breast cancer<br />
cells. Cancer Res. 2004;64:2343-2346.<br />
8. Baselga J, Bradbury I, Eidtmann H, et al. Lapatinib with trastuzumab<br />
for HER2-positive early breast cancer (NeoALTTO): a randomized, openlabel,<br />
multicenter, phase 3 trial. Lancet. <strong>2012</strong>;379:633-640. Epub <strong>2012</strong> Jan 17.<br />
Erratum in: Lancet. <strong>2012</strong>;379:616.<br />
9. Tabernero J, Rojo F, Calvo E, et al. Dose- and schedule-dependent<br />
inhibition <strong>of</strong> the mammalian target <strong>of</strong> rapamycin pathway with everolimus: a<br />
phase I tumor pharmacodynamic study in patients with advanced solid<br />
tumors. J Clin Oncol. 2008;26:1603-1610.<br />
10. Yu K, Toral-Barza L, Shi C, et al. Response and determinants <strong>of</strong> cancer<br />
cell susceptibility to PI3K inhibitors: combined targeting <strong>of</strong> PI3K and Mek1 as<br />
an effective anticancer strategy. Cancer Biol Ther. 2008;7:307-315.<br />
11. Villanueva J, Vultur A, Lee JT, et al. Acquired resistance to BRAF<br />
inhibitors mediated by a RAF kinase switch in melanoma can be overcome by<br />
cotargeting MEK and IGF-1R/PI3K. Cancer Cell. 2010;18:683-695.<br />
12. Engelman JA, Chen L, Tan X, et al. Effective use <strong>of</strong> PI3K and MEK<br />
inhibitors to treat mutant Kras G12D and PIK3CA H1047R murine lung<br />
cancers. Nature Medicine. 2008;14:1351-1356.<br />
13. Wee S, Jagani Z, Xiang KX, et al. PI3K pathway activation mediates<br />
resistance to MEK inhibitors in KRAS mutant cancers. Cancer Res. 2009;69:<br />
4286-4293.<br />
14. Janku F, Lee JJ, Tsimberidou AM, et al. PIK3CA mutations frequently<br />
coexist with RAS and BRAF mutations in patients with advanced cancers.<br />
PLoS One. 2011;6e:22769.<br />
15. Yap TA, Yan L, Patnaik A, et al. First-in-man clinical trial <strong>of</strong> the oral<br />
674<br />
Stock<br />
Ownership Honoraria<br />
REFERENCES<br />
AstraZeneca;<br />
Boehringer<br />
Ingelheim;<br />
Genentech;<br />
GlaxoSmithKline;<br />
Johnson &<br />
Johnson; Pfizer<br />
Research<br />
Funding<br />
AstraZeneca;<br />
Immunicon<br />
Expert<br />
Testimony<br />
MATEO ET AL<br />
Other<br />
Remuneration<br />
pan-AKT inhibitor MK-2206 in patients with advanced solid tumors. J Clin<br />
Oncol. 2011;29:4688-4695.<br />
16. Hanahan D, Weinberg RA. The hallmarks <strong>of</strong> cancer. Cell. 2000;100:57-<br />
70.<br />
17. Sequist LV, Waltman BA, Dias-Santagata D, et al. Genotypic and<br />
histological evolution <strong>of</strong> lung cancers acquiring resistance to EGFR inhibitors.<br />
Sci Transl Med. 2011;3:75ra26.<br />
18. Whitehurst AW, Bodemann BO, Cardenas J, et al. Synthetic lethal<br />
screen identification <strong>of</strong> chemosensitizer loci in cancer cells. Nature. 2007;<br />
446:815-819.<br />
19. O’Grady M, Raha D, Hanson BJ, et al. Combining RNA interference<br />
and kinase inhibitors against cell signaling components involved in cancer.<br />
BMC Cancer. 2005;5:125.<br />
20. Iorns E, Lord CJ, Turner N, et al. Utilizing RNA interference to<br />
enhance cancer drug discovery. Nature Rev Drug Discov. 2007;6:556-568.<br />
21. Becher OJ, Holland EC. Genetically engineered models have advantages<br />
over xenografts for preclinical studies. Cancer Res. 2006;66:3355-3358;<br />
discussion 3358-3359.<br />
22. Wan X, Harkavy B, Shen N, et al. Rapamycin induces feedback<br />
activation <strong>of</strong> Akt signaling through an IGF-1R-dependent mechanism. Oncogene.<br />
2007;26:1932-1940.<br />
23. Di Cosimo S, Bendell J, Cervantes-Ruiperez A, et al. A phase I study <strong>of</strong><br />
the oral mTOR inhibitor ridaforolimus (RIDA) in combination with the<br />
IGF-1R antibody dalotozumab (DALO) in patients (pts) with advanced solid<br />
tumors. J Clin Oncol. 2010;28:15s (suppl; abstr 3008).<br />
24. Molina AM, Feldman DR, Voss RJ, et al. Phase 1 trial <strong>of</strong> everolimus<br />
plus sunitinib in patients with metastatic renal cell carcinoma. Cancer. Epub<br />
2011 Sept 6.<br />
25. Braun TM, Alonzo TA. Beyond the 3�3 method: Expanded algorithms<br />
for dose-escalation in Phase I oncology trials <strong>of</strong> two agents. Clin Trials.<br />
2011;8:247-259.<br />
26. Rodon J, Perez J, Kurzrock R. Combining targeted therapies: practical<br />
issues to consider at the bench and bedside. Oncologist. 2010;15:37-50.<br />
27. Yin G, Li Y, Ji Y. Bayesian dose-finding in phase I/II clinical trials<br />
using toxicity and efficacy odds ratios. Biometrics. 2006;62:777-784.<br />
28. Zafar Y, Bendell J, Lager J, et al. Preliminary results <strong>of</strong> a phase I study<br />
<strong>of</strong> bevacizumab (BV) in combination with everolimus (E) in patients with<br />
advanced solid tumors. J Clin Oncol. 2006;24:145s (suppl; abstr 3097).<br />
29. Hainsworth JD, Spigel DR, Burris HA, et al. Phase II trial <strong>of</strong> bevacizumab<br />
and everolimus in patients with advanced renal cell carcinoma. J Clin<br />
Oncol. 2010;28:2131-2136.<br />
30. Yang JC, Haworth L, Sherry RM, et al. A randomized trial <strong>of</strong> bevacizumab,<br />
an anti-vascular endothelial growth factor antibody, for metastatic<br />
renal cancer. N Engl J Med. 2003;349:427-434.<br />
31. Polley, MY. Practical modifications to the time-to-event continual<br />
reassessment method for phase I cancer trials with fast patient accrual and<br />
late-onset toxicities. Stat Med. 2011;30:2130-2143.
Combination Therapies Building on the<br />
Efficacy <strong>of</strong> CTLA4 and BRAF Inhibitors for<br />
Metastatic Melanoma<br />
Overview: The demonstration <strong>of</strong> improved survival with the<br />
anti-CTLA4 antibody ipilimumab and the BRAF inhibitor vemurafenib<br />
in patients with metastatic melanoma is arguably the<br />
most significant advance in the treatment for these patients in<br />
the last 30 years. However, the majority <strong>of</strong> patients will either<br />
not experience response, or will experience response and<br />
then progression, when receiving these therapies, so additional<br />
treatment options are required. Since these agents<br />
DURING 2011, two new single-agent therapies were<br />
approved by the U.S. Food and Drug Administration<br />
(FDA) for the treatment <strong>of</strong> patients with metastatic melanoma.<br />
The antitumor activity <strong>of</strong> these newly approved<br />
agents, ipilimumab (formerly MDX010) and vemurafenib<br />
(formerly PLX4032/RG7204), is based on improved means to<br />
stimulate immune responses to melanoma and to block<br />
oncogenic signaling in cancer cells, respectively. Ipilimumab<br />
was approved on the basis <strong>of</strong> two randomized clinical trials,<br />
one demonstrating improvement in survival over a peptide<br />
vaccine in second-line or later therapy, 1 and another in<br />
combination with the chemotherapy agent dacarbazine compared<br />
with dacarbazine and placebo. 2 Vemurafenib was<br />
approved based on the results <strong>of</strong> a randomized clinical trial<br />
demonstrating an early benefit in overall survival compared<br />
with dacarbazine in patients with BRAF V600 mutant metastatic<br />
melanoma. 3 Despite the unquestionable benefits <strong>of</strong><br />
these two therapies, the great majority <strong>of</strong> patients with<br />
metastatic melanoma continue to require improved treatments<br />
because the majority will not experience response to<br />
ipilimumab, and the majority with BRAF V600 mutations will<br />
experience response to vemurafenib but will have progression<br />
within a matter <strong>of</strong> months.<br />
These recent advances are based on targeted interventions<br />
that build on a detailed knowledge <strong>of</strong> what is being<br />
modulated, either the CTLA4 coinhibitory molecule for ipilimumab<br />
or the BRAF driver oncogene for vemurafenib.<br />
Their antitumor activity is based on a refined understanding<br />
<strong>of</strong> their mechanisms <strong>of</strong> action, allowing the rational advancement<br />
<strong>of</strong> therapies building on these two agents. Other<br />
single-agent therapies that have mechanisms <strong>of</strong> action <strong>of</strong><br />
immune modulatory effects, either inhibiting other negative<br />
immune regulators such as PD1, or activating immune<br />
coactivators such as CD40, CD137 or OX40, are being<br />
developed in addition to the already approved immune activating<br />
cytokines interferon-alpha (IFN-a) and interleukin-2<br />
(IL-2). Furthermore, other targeted driver oncogene inhibitors<br />
are in clinical testing blocking BRAF (such as dabrafenib,<br />
formerly GSK2118436), blocking ckit in a subset <strong>of</strong><br />
ckit mutant melanomas (primarily acral and mucosal melanomas)<br />
or the use <strong>of</strong> downstream MEK inhibitors for tumors<br />
with oncogenic mitogen-activated protein kinase (MAPK)<br />
signaling (Fig. 1).<br />
All <strong>of</strong> these agents with proven antitumor activity in<br />
melanoma are pointing the way <strong>of</strong> future research that<br />
addresses current limitations while building on their benefits.<br />
Being able to combine several <strong>of</strong> these active agents in<br />
By Antoni Ribas, MD<br />
have been developed with a refined understanding <strong>of</strong> their<br />
mechanism <strong>of</strong> action and mechanisms leading to resistance<br />
are being elucidated, then combination therapies building<br />
on these single-agent therapies can be designed rationally.<br />
Such combinations are being tested both preclinically and in<br />
the clinic, and provide a strong promise to improve on the<br />
current treatment approaches for patients with metastatic<br />
melanoma.<br />
a rational manner using in-depth understanding <strong>of</strong> the<br />
underlying biology is likely to continue to provide advances<br />
to the treatment options for advanced melanoma.<br />
Combination <strong>of</strong> Immunotherapies<br />
The immune system has multiple control mechanisms<br />
aimed at preventing an overt immune activation to selfantigens,<br />
which would lead to high frequency <strong>of</strong> autoimmune<br />
diseases. Therefore, an activated immune response<br />
to cancer requires fine tuning to balance between killing<br />
cellular targets but not inducing autoimmunity. With the<br />
improved understanding <strong>of</strong> the regulation <strong>of</strong> immune activation<br />
and the existence <strong>of</strong> multiple inhibitory immune<br />
pathways, combinations <strong>of</strong> immune activators, and blockers<br />
<strong>of</strong> immune inhibitors, are being tested in preclinical models<br />
and in the clinic. Among them are combinations being tested<br />
in the clinic building on anti-CTLA4 antibodies with the<br />
addition <strong>of</strong> immune activating cytokines such as IFN-a 4 or<br />
IL-2, 5 with dendritic cell vaccines, 6 and also earlier studies<br />
combining with other immune modulating antibodies such<br />
as anti-PD1.<br />
Combination <strong>of</strong> three or four immune modulating antibodies<br />
has been demonstrated to provide antitumor activity<br />
over single-agent or double combinations in mouse models. 7<br />
For example, high antitumor activity has been demonstrated<br />
combining anti-CD40, anti-CD137, and anti-TRAIL,<br />
with further enhancement with anti-CTLA4 antibodies. 7,8<br />
The testing <strong>of</strong> triplet immune-modulating monoclonal antibody<br />
therapy in humans is limited by the availability <strong>of</strong><br />
these antibodies and the unclear path for their combined<br />
clinical development. Restrictions are imposed by the regulatory<br />
process, which requires demonstrating the individual<br />
contribution to the antitumoral benefit while avoiding increased<br />
toxicities, as well as the anticipated high expense <strong>of</strong><br />
each one <strong>of</strong> these monoclonal antibodies. However, if the<br />
benefit is as predicted in the mouse models, then a higher<br />
rate <strong>of</strong> durable tumor responses than can be achieved<br />
From the Department <strong>of</strong> Medicine, Division <strong>of</strong> Hematology-<strong>Oncology</strong>, Jonsson Comprehensive<br />
Cancer Center at the University <strong>of</strong> California, Los Angeles, Los Angeles, CA.<br />
Author’s disclosures <strong>of</strong> potential conflicts <strong>of</strong> interest are found at the end <strong>of</strong> this article.<br />
Address reprint requests to Antoni Ribas, MD, Department <strong>of</strong> Medicine, Division <strong>of</strong><br />
Hematology-<strong>Oncology</strong>, Jonsson Comprehensive Cancer Center at the University <strong>of</strong> California,<br />
Los Angeles, 11-934 Factor Building, 10833 Le Conte Avenue, Los Angeles, CA<br />
90095-1782; email: aribas@mednet.ucla.edu.<br />
© <strong>2012</strong> by <strong>American</strong> <strong>Society</strong> <strong>of</strong> <strong>Clinical</strong> <strong>Oncology</strong>.<br />
1092-9118/10/1-10<br />
675
otherwise would be compelling to develop clinically and<br />
cost-effective by avoiding the need for further therapies.<br />
Combination <strong>of</strong> Targeted Therapies<br />
The description <strong>of</strong> molecular mechanisms <strong>of</strong> the development<br />
<strong>of</strong> acquired resistance to BRAF inhibitors, and <strong>of</strong><br />
toxicities with these agents, allow the rational development<br />
<strong>of</strong> combination targeted therapies for the treatment <strong>of</strong> advanced<br />
melanoma. Mechanisms <strong>of</strong> resistance are diverse<br />
and can be categorized between the ones that reactivate the<br />
MAPK pathway and the mechanisms that lead to a MAPK<br />
pathway-independent signaling that substitutes for the<br />
blocked driver oncogenic signal. MAPK reactivating mechanisms<br />
reported to date in patient-derived samples include<br />
truncations in the BRAF protein resulting in increased<br />
KEY POINTS<br />
● Single-agent ipilimumab and vemurafenib have been<br />
approved for clinical use after demonstration <strong>of</strong> improvement<br />
in the survival <strong>of</strong> patients with metastatic<br />
melanoma.<br />
● Combinations <strong>of</strong> immunotherapy approaches may<br />
increase the frequency <strong>of</strong> tumor responses while<br />
maintaining their long durability.<br />
● Combinations based on BRAF inhibitors can prevent<br />
or treat acquired resistance.<br />
● The combination <strong>of</strong> a BRAF and a MEK inhibitor can<br />
increase the response rate, extend the response duration<br />
and prevent the major skin toxicity <strong>of</strong> these<br />
agents.<br />
● Combinations <strong>of</strong> immunotherapy and BRAF-targeted<br />
therapy may allow maintaining the high response<br />
rate <strong>of</strong> the targeted therapy and increase its durability<br />
with immunotherapy.<br />
676<br />
ANTONI RIBAS<br />
Fig. 1. Schematic <strong>of</strong> the mechanism <strong>of</strong><br />
action <strong>of</strong> new treatments demonstrating patient<br />
benefit in advanced melanoma. The<br />
majority <strong>of</strong> melanomas demonstrate a constitutively<br />
active mitogen-activated protein<br />
kinase (MAPK) pathway leading to uncontrolled<br />
cell proliferation and avoidance <strong>of</strong><br />
apoptosis, which is to the result <strong>of</strong> the presence<br />
<strong>of</strong> mutually exclusive activating mutations<br />
in the receptor tyrosine kinase cKit, in<br />
NRAS or BRAF (red stars). Inhibitors <strong>of</strong> oncogenic<br />
driver mutations blocking cKit or<br />
BRAF result in high response rates, and this<br />
oncogenic pathway can also be blocked<br />
with MEK inhibitors. Melanoma can also be<br />
treated by activating an antitumor immune<br />
response, either by turning it on against<br />
cancer cells by the immune activating cytokines<br />
interleukin-2 (IL2) or interferon, or the<br />
immune activating antibodies to CD40,<br />
CD137 or OX40, or by administering antibodies<br />
blocking negative costimulatory signaling<br />
through CTLA4 or PD1.<br />
kinase activity, 9 amplifications <strong>of</strong> the mutant BRAF gene, 10<br />
secondary mutations in NRAS or MEK, 11,12 or overexpression<br />
<strong>of</strong> COT. 13 The mechanisms leading to MAPK-redundant<br />
pathway activation are induced by overexpression or overactivation<br />
<strong>of</strong> receptor tyrosine kinases such as the plateletderived<br />
growth factor receptor beta (PDGFRb) 11,14 or the<br />
insulin-like growth factor receptor 1 (IGF-1R), 15 leading<br />
oncogenic signaling through the phosphoinositide 3-kinase<br />
(PI3K)/AKT pathway that can be pharmacologically blocked<br />
in preclinical models with specific inhibitors <strong>of</strong> this<br />
pathway. 14,16-18<br />
Adding an MEK inhibitor to a BRAF inhibitor would be<br />
able to treat or prevent most <strong>of</strong> the MAPK-reactivation<br />
resistance mechanisms, and such combinations are already<br />
in advanced clinical testing. Early evidence suggests that<br />
some patients experiencing progression with single-agent<br />
BRAF inhibitors can have secondary responses when adding<br />
an MEK inhibitor to continued therapy with a BRAF inhibitor.<br />
19 On the contrary, and as suggested by preclinical<br />
modeling, 17 sequential therapy stopping a BRAF inhibitor<br />
and changing the therapy to a MEK inhibitor provides no<br />
benefit. 20 An even greater benefit may be derived from the<br />
initiation <strong>of</strong> therapy combining a BRAF and an MEK inhibitor<br />
as the first oncogene-targeted therapy for BRAF mutant<br />
melanoma, as has been done with the combination <strong>of</strong> dabrafenib<br />
and the MEK inhibitor GSK1120212. 19,21 This combination,<br />
compared with single-agent BRAF inhibitors, may<br />
result in higher antitumor activity by more pr<strong>of</strong>ound oncogenic<br />
MAPK inhibition, potentially leading to more durable<br />
responses by preventing the MAPK-dependent acquired<br />
resistance mechanisms. An additional benefit <strong>of</strong> the combination<br />
<strong>of</strong> BRAF and MEK inhibitors is the ability <strong>of</strong> the<br />
MEK inhibitors to block paradoxical MAPK activation leading<br />
to the development <strong>of</strong> the main grade 3 toxicity with the<br />
use <strong>of</strong> single-agent BRAF inhibitors, which is the development<br />
<strong>of</strong> secondary cutaneous squamous cell carcinomas. 22<br />
Therefore, this combination would be a rare example <strong>of</strong> two
COMBINATIONS FOR MELANOMA<br />
agents that, when combined, have increased antitumor<br />
activity with decreased toxicities.<br />
Despite the wealth <strong>of</strong> supportive preliminary evidence,<br />
14,16-18 studies combining a BRAF inhibitor with a<br />
PI3K or AKT inhibitor to treat or prevent MAPKindependent<br />
resistance are not in the clinic. However, there<br />
are several experiences combining MEK inhibitors with<br />
PI3K inhibitors that may provide the benefit <strong>of</strong> blocking<br />
both pathways. But this potential benefit has to be weighed<br />
against the expected increase in toxicities because neither<br />
agent would be directly blocking a mutated oncogene in this<br />
setting.<br />
Combination <strong>of</strong> Immunotherapies and<br />
Targeted Therapies<br />
Combinations <strong>of</strong> immunotherapy agents and oncogene<br />
driver inhibitors have the potential to merge the benefits <strong>of</strong><br />
both types <strong>of</strong> agents for patients with advanced melanoma.<br />
Targeted oncogene pathway inhibitors induce high rates <strong>of</strong><br />
tumor responses that tend to last months, whereas immunemodulating<br />
agents tend to have infrequent but highly durable<br />
tumor responses. The promise <strong>of</strong> combining them is to<br />
maintain the high rate <strong>of</strong> targeted therapy responses and<br />
make them more frequently durable with immunotherapies.<br />
23,24<br />
Key to the development <strong>of</strong> this type <strong>of</strong> combination is<br />
demonstrating that both approaches are not negatively<br />
interacting, although they can have synergistic effects by<br />
potentiating each other. In vitro studies have shown that<br />
Author’s Disclosures <strong>of</strong> Potential Conflicts <strong>of</strong> Interest<br />
Author<br />
Employment or<br />
Leadership<br />
Positions<br />
Consultant or<br />
Advisory Role<br />
Antoni Ribas Amgen; Bristol-<br />
Myers Squibb;<br />
Celgene;<br />
Genentech;<br />
Merck;<br />
Millennium;<br />
Roche<br />
1. Hodi FS, O’Day SJ, McDermott DF, et al. Improved survival with<br />
ipilimumab in patients with metastatic melanoma. New Engl J Med. 2010;<br />
363:711-723.<br />
2. Robert C, Thomas L, Bondarenko I, et al. Ipilimumab plus dacarbazine<br />
for previously untreated metastatic melanoma. New Engl J Med. 2011;364:<br />
2517-2526.<br />
3. Chapman PB, Hauschild A, Robert C, et al. Improved survival with<br />
vemurafenib in melanoma with BRAF V600E mutation. N Engl J Med.<br />
2011;364:2507-2716.<br />
4. Tarhini AA, Kirkwood JM. Tremelimumab and interferon combination<br />
in melanoma. Semin Oncol. In press.<br />
5. Maker AV, Phan GQ, Attia P, et al. Tumor regression and autoimmunity<br />
in patients treated with cytotoxic T lymphocyte-associated antigen<br />
4 blockade and interleukin 2: a phase I/II study. Ann Surg Oncol. 2005;12:<br />
1005-1016.<br />
6. Ribas A, Comin-Anduix B, Chmielowski B, et al. Dendritic cell vaccination<br />
combined with CTLA4 blockade in patients with metastatic melanoma.<br />
Clin Cancer Res. 2009;15:6267-6276.<br />
7. Uno T, Takeda K, Kojima Y, et al. Eradication <strong>of</strong> established tumors in<br />
mice by a combination antibody-based therapy. Nat Med. 2006;12:693-698.<br />
8. Takeda K, Kojima Y, Uno T, et al. Combination therapy <strong>of</strong> established<br />
tumors by antibodies targeting immune activating and suppressing molecules.<br />
J Immunol. 2010;184:5493-5501.<br />
BRAF inhibitors are not detrimental to human or murine<br />
lymphocyte function. 25,26 Furthermore, given the evidence<br />
that BRAF inhibitors can induce a paradoxical activation <strong>of</strong><br />
the MAPK pathway in cells that have strong RAS/GTP<br />
activity, 27-29 which would be expected to also occur in<br />
activated lymphocytes, it is certainly possible that at certain<br />
doses and concentrations BRAF inhibitors may actually<br />
potentiate the function <strong>of</strong> lymphocytes. Besides this potential<br />
beneficial effect <strong>of</strong> BRAF inhibitors directly in immune<br />
cells, these agents may benefit in combination by increasing<br />
expression or cross-presentation <strong>of</strong> tumor antigens to the<br />
immune system, or by releasing signals that lead to an<br />
adverse tumor environment limiting immune cell function.<br />
Conclusion<br />
Advances in the treatment <strong>of</strong> metastatic melanoma are<br />
based on an improved understanding <strong>of</strong> the immunobiology<br />
and oncogenic signaling in this disease. Single-agent therapies<br />
with ipilimumab and vemurafenib have demonstrated<br />
improvement in patient survival in randomized clinical<br />
trials. 1-3 However, the majority <strong>of</strong> patients still require<br />
additional treatment options. Understanding the mechanisms<br />
<strong>of</strong> action and resistance to the new single-agent<br />
therapies allows an opportunity to rationally design combination<br />
therapies that may provide additional benefits to<br />
patients. Such combinations are being developed preclinically<br />
and in the clinic, and are poised to change the current<br />
treatment approaches for patients with metastatic melanoma.<br />
Stock<br />
Ownership Honoraria<br />
REFERENCES<br />
Research<br />
Funding<br />
Expert<br />
Testimony<br />
Other<br />
Remuneration<br />
9. Poulikakos PI, Persaud Y, Janakiraman M, et al. RAF inhibitor resistance<br />
is mediated by dimerization <strong>of</strong> aberrantly spliced BRAF(V600E).<br />
Nature. 2011;480:387-390.<br />
10. Shi H, Moriceau G, Kong X, et al. Melanoma V600E BRAF amplification<br />
drives acquired resistance to vemurafenib. Nat Commun. In press.<br />
11. Nazarian R, Shi H, Wang Q, et al. Melanomas acquire resistance to<br />
B-RAF(V600E) inhibition by RTK or N-RAS upregulation. Nature. 2010;468:<br />
973-977.<br />
12. Wagle N, Emery C, Berger MF, et al. Dissecting therapeutic resistance<br />
to RAF inhibition in melanoma by tumor genomic pr<strong>of</strong>iling. J Clin Oncol.<br />
2011;29:3085-3096.<br />
13. Johannessen CM, Boehm JS, Kim SY, et al. COT drives resistance to<br />
RAF inhibition through MAP kinase pathway reactivation. Nature. 2010;468:<br />
968-672.<br />
14. Shi H, Kong X, Ribas A, et al. Combinatorial treatments that overcome<br />
PDGFR{beta}-driven resistance <strong>of</strong> melanoma cells to V600EB-RAF inhibition.<br />
Cancer Res. 2011;71:5067-5074.<br />
15. Villanueva J, Vultur A, Lee JT, et al. Acquired resistance to BRAF<br />
inhibitors mediated by a RAF kinase switch in melanoma can be overcome by<br />
cotargeting MEK and IGF-1R/PI3K. Cancer Cell. 2010;18:683-695.<br />
16. Jiang CC, Lai F, Thorne RF, et al. MEK-independent survival <strong>of</strong><br />
B-RAFV600E melanoma cells selected for resistance to apoptosis induced by<br />
the RAF inhibitor PLX4720. Clin Cancer Res. 2011;17:721-730.<br />
677
17. Atefi M, von Euw E, Attar N, et al. Reversing melanoma crossresistance<br />
to BRAF and MEK inhibitors by co-targeting the AKT/mTOR<br />
pathway. PloS One. 2011;6:e28973.<br />
18. Paraiso KH, Xiang Y, Rebecca VW, et al. PTEN loss confers BRAF<br />
inhibitor resistance to melanoma cells through the suppression <strong>of</strong> BIM<br />
expression. Cancer Res. 2011;71:2750-2760.<br />
19. Flaherty K, Infante JR, Falchook GS, et al. Phase I/II study <strong>of</strong> BRAFi<br />
GSK2118436 � MEKi GSK1120212 in patients with BRAF mutant metastatic<br />
melanoma who progressed on a prior BRAFi. Pigment Cell Melanoma<br />
Res. 2011;24:LBA1.<br />
20. Kim KB, Lewis KD, Pavlick AC, et al. A phase II study <strong>of</strong> the<br />
MEK1/MEK2 inhibitor GSK1120212 in metastatic BRAF-V600E or K mutant<br />
cutaneous melanoma patients previously treated with or without a BRAF<br />
inhibitor. Pigment Cell Melanoma Res. 2011;24:1021.<br />
21. Infante JR, Falchook GS, Lawrence DP, et al. Phase I/II study to assess<br />
safety, pharmacokinetics, and efficacy <strong>of</strong> the oral MEK 1/2 inhibitor<br />
GSK1120212 (GSK212) dosed in combination with the oral BRAF inhibitor<br />
GSK2118436 (GSK436). J Clin Oncol 2011;29 (suppl; abstr CRA8503).<br />
22. Su F, Viros A, Milagre C, et al. RAS mutations in cutaneous squamouscell<br />
carcinomas in patients treated with BRAF inhibitors. New Engl J Med.<br />
<strong>2012</strong>;366:207-215.<br />
678<br />
ANTONI RIBAS<br />
23. Begley J, Ribas A. Targeted therapies to improve tumor immunotherapy.<br />
Clin Cancer Res. 2008;14:4385-4391.<br />
24. Ribas A, Flaherty KT. BRAF targeted therapy changes the treatment<br />
paradigm in melanoma. Nat Rev Clin Oncol. 2011;8:426-433.<br />
25. Boni A, Cogdill AP, Dang P, et al. Selective BRAFV600E inhibition<br />
enhances T-cell recognition <strong>of</strong> melanoma without affecting lymphocyte function.<br />
Cancer Res. 2010;70:5213-5219.<br />
26. Comin-Anduix B, Chodon T, Sazegar H, et al. The oncogenic BRAF<br />
kinase inhibitor PLX4032/RG7204 does not affect the viability or function <strong>of</strong><br />
human lymphocytes across a wide range <strong>of</strong> concentrations. Clin Cancer Res.<br />
2010;16:6040-6048.<br />
27. Heidorn SJ, Milagre C, Whittaker S, et al. Kinase-dead BRAF and<br />
oncogenic RAS cooperate to drive tumor progression through CRAF. Cell.<br />
2010;140:209-221.<br />
28. Halaban R, Zhang W, Bacchiocchi A, et al. PLX4032, a selective<br />
BRAF(V600E) kinase inhibitor, activates the ERK pathway and enhances cell<br />
migration and proliferation <strong>of</strong> BRAF melanoma cells. Pigment Cell Melanoma<br />
Res. 2010;23:190-200.<br />
29. Poulikakos PI, Zhang C, Bollag G, et al. RAF inhibitors transactivate<br />
RAF dimers and ERK signalling in cells with wild-type BRAF. Nature.<br />
2010;464:427-430.
MECHANISMS OF RESISTANCE TO TARGETED<br />
ANTICANCER AGENTS<br />
CHAIR<br />
Levi A. Garraway, MD, PhD<br />
Dana-Farber Cancer Institute<br />
Boston, MA<br />
SPEAKERS<br />
Alberto Bardelli, PhD<br />
Istituto Ricerca Cura Cancro and FIRC<br />
Candiolo, Italy<br />
Neil P. Shah, MD, PhD<br />
University <strong>of</strong> California, San Francisco<br />
San Francisco, CA<br />
Alice Tsang Shaw, MD, PhD<br />
Massachusetts General Hospital<br />
Boston, MA
Mechanisms <strong>of</strong> Resistance to Mitogen-<br />
Activated Protein Kinase Pathway Inhibition<br />
in BRAF-Mutant Melanoma<br />
By Eva M. Goetz, PhD, and Levi A. Garraway, MD, PhD<br />
Overview: Anticancer drug resistance remains a crucial impediment<br />
to the care <strong>of</strong> many patients with cancer. Although<br />
the exact mechanisms <strong>of</strong> resistance may differ for each<br />
therapy, common mechanisms <strong>of</strong> resistance predominate,<br />
including drug inactivation or modification, mutation <strong>of</strong> the<br />
target protein, reduced drug accumulation, or bypass <strong>of</strong> target<br />
inhibition. With the discovery and use <strong>of</strong> targeted therapies<br />
(such as small-molecule kinase inhibitors), resistance has<br />
received renewed attention—especially in light <strong>of</strong> the dramatic<br />
responses that may emerge from such therapeutics in<br />
particular genetic or molecular contexts. Recently, the<br />
mitogen-activated protein kinase (MAPK) pathway has become<br />
exemplary in this regard, since it is activated in many<br />
different cancers. Drugs targeting RAF and MAPK kinase<br />
(MEK) are currently in clinical trials for the treatment <strong>of</strong><br />
several types <strong>of</strong> cancer. Vemurafenib, a selective RAF kinase<br />
MELANOMA IS among the most common cancers, and<br />
with early detection and surgical resection, the 5-year<br />
survival rate for melanoma is 98%. 1 Unfortunately, the<br />
5-year survival rate falls to 15% for individuals with metastatic<br />
disease. Even considering the recent spectacular therapeutic<br />
successes, relatively few treatment options exist for<br />
metastatic melanoma. Those in current use include chemotherapy<br />
(dacarbazine), immunotherapy (high-dose interferon<br />
or ipilimumab), and the newly US Food and Drug<br />
Administration (FDA)–approved targeted therapy vemurafenib.<br />
Therefore, new, targeted therapies are needed to<br />
improve these dire statistics.<br />
Approximately 40% to 60% <strong>of</strong> cutaneous melanomas contain<br />
a V600E mutation in BRAF that results in constitutive<br />
MAPK signaling, independent <strong>of</strong> RAS (Fig. 1). 2-4 These<br />
melanomas are dependent on MAPK, therefore, a series <strong>of</strong><br />
small molecule inhibitors targeting mutant BRAF have been<br />
developed and remain a topic <strong>of</strong> intense clinical investigation.<br />
Vemurafenib (PLX4032) is an FDA-approved inhibitor<br />
<strong>of</strong> BRAF(V600E), and although patients with BRAF mutation<br />
have enjoyed a clinical benefit with vemurafenib treatment,<br />
the fraction <strong>of</strong> patients who achieve a Response<br />
Evaluation Criteria in Solid Tumors (RECIST) partial response<br />
is 48%. 5 Furthermore, the average duration <strong>of</strong> response<br />
is only 7 months. 6 Thus, the magnitude and duration<br />
<strong>of</strong> clinical response is limited, and the development <strong>of</strong><br />
progressive disease after several months to approximately 1<br />
year <strong>of</strong> treatment is nearly universal. Since mutant BRAF<br />
melanoma is dependent on MAPK kinase/extracellular<br />
signal-regulated kinase (MEK/ERK) signaling, MEK inhibitors<br />
were predicted to be useful for treatment <strong>of</strong> metastatic<br />
melanoma. Early data from phase II clinical trials <strong>of</strong> MEK<br />
inhibitors were disappointing, however, the addition <strong>of</strong> MEK<br />
inhibitors to RAF inhibitors provides a promising investigational<br />
avenue, as described below. 7 Clearly, overcoming<br />
resistance to vemurafenib may enable additional clinical<br />
benefits to patients with BRAF(V600E) melanoma.<br />
The development <strong>of</strong> resistance to targeted anticancer<br />
therapy is typical in solid tumors. Several overarching<br />
680<br />
inhibitor recently approved for the treatment <strong>of</strong> BRAF(V600E)<br />
melanoma, shows strong efficacy initially; however, the<br />
development <strong>of</strong> resistance is nearly ubiquitous. In vitro testing<br />
and analysis <strong>of</strong> patient samples have uncovered several<br />
mechanisms <strong>of</strong> resistance to RAF inhibition. Surprisingly,<br />
mutations in the drug-binding pocket have not thus far been<br />
observed; however, other alterations at the level <strong>of</strong> RAF, as<br />
well as downstream activation <strong>of</strong> MEK and bypass <strong>of</strong> MEK/<br />
extracellular signal-regulated kinase (ERK) signaling altogether,<br />
confer resistance to vemurafenib. Looking forward,<br />
combined RAF and MEK inhibitor treatments may improve<br />
efficacy—yet we must anticipate mechanisms <strong>of</strong> resistance<br />
to this combination as well. Therefore, understanding and/or<br />
determining the mechanism <strong>of</strong> resistance are paramount to<br />
effective cancer treatment.<br />
mechanisms have been described for many targeted therapies,<br />
including erlotinib and gefitinib—epidermal growth<br />
factor receptor (EGFR) inhibitors in non–small cell lung<br />
cancer—and imatinib, which is used for BCR-ABL inhibition<br />
in leukemia and KIT blockade in gastrointestinal stromal<br />
tumors. 8-10 In studying resistance mechanisms, the overarching<br />
goal is the development and implementation <strong>of</strong><br />
therapeutic combinations that can anticipate and thwart<br />
this process at the onset <strong>of</strong> treatment, thereby increasing the<br />
magnitude and duration <strong>of</strong> clinical benefit. Taken to its<br />
logical conclusion, such understanding could pave the way to<br />
durable therapeutic control <strong>of</strong> cancers driven by “druggable”<br />
oncoprotein dysregulation.<br />
Mechanistic Categories <strong>of</strong> Resistance to RAF Inhibition<br />
Since resistance is so common with targeted therapies,<br />
studies <strong>of</strong> resistance to RAF inhibitors have been at the<br />
forefront <strong>of</strong> translational research in this area, even during<br />
preclinical development (Table 1). Given the importance <strong>of</strong><br />
MAPK signaling in BRAF-mutant melanoma, it stands to<br />
reason that any resistance mechanism leading to sustained<br />
or reactivated MEK/ERK signaling could in principle contribute<br />
to therapeutic resistance.<br />
BRAF Bypass<br />
In other genetically defined tumor subtypes, resistance<br />
has been achieved by the acquisition <strong>of</strong> point mutations<br />
within the oncoprotein target itself, such as the T315I<br />
mutation in BCR-ABL resulting in resistance to imatinib,<br />
nilotinib, and dasatinib and the T790M mutation in EGFR,<br />
From the Department <strong>of</strong> Medical <strong>Oncology</strong>, Dana-Farber Cancer Institute, Boston, MA;<br />
The Broad Institute <strong>of</strong> Harvard and MIT, Cambridge, MA.<br />
Authors’ disclosures <strong>of</strong> potential conflicts <strong>of</strong> interest are found at the end <strong>of</strong> this article.<br />
Address reprint requests to Levi A. Garraway, MD, PhD, Broad Institute <strong>of</strong> MIT<br />
and Harvard, Dana Building, Room 1542, 44 Binney St., Boston, MA 02115; email:<br />
levi_garraway@dfci.harvard.edu.<br />
© <strong>2012</strong> by <strong>American</strong> <strong>Society</strong> <strong>of</strong> <strong>Clinical</strong> <strong>Oncology</strong>.<br />
1092-9118/10/1-10
RESISTANCE MECHANISMS TO MAPK INHIBITION<br />
Fig 1. Mechanisms <strong>of</strong> resistance to vemurafenib.<br />
BRAF(V600E) mutant tumors have<br />
constitutive activation <strong>of</strong> MEK and ERK (MAPK)<br />
signaling, which is inhibited by vemurafenib.<br />
Activating mutations (*) in NRAS or KRAS (1) or<br />
CRAF amplification (2) lead to activation <strong>of</strong><br />
CRAF/MEK/ERK and resistance to RAF inhibition.<br />
p61BRAF(V600E) (3) forms constitutive<br />
dimers resulting in activation <strong>of</strong> MAPK signaling.<br />
COT amplification (4) leads to activation<br />
<strong>of</strong> MEK/ERK, independent <strong>of</strong> BRAF or CRAF.<br />
Mutations (*) in MEK (5) that activate ERK can<br />
lead to resistance. Bypass <strong>of</strong> MAPK signaling<br />
through RTKs, such as PDGFRß and IGF1R (6)<br />
have been shown to mediate resistance to<br />
vemurafenib.<br />
Abbreviations: ERK, extracellular signalregulated<br />
kinase; IGF1R, insulin-like growth<br />
factor-1 receptor; MAPK, mitogen-activated<br />
protein kinase; MEK, mitogen-activated protein<br />
kinase kinase; PDGFRß. platelet-derived<br />
growth factor-beta; RTK, receptor tyrosine<br />
kinase.<br />
which confer resistance to gefitinib and erlotinib. 8,10,11 Such<br />
point mutations lead to either diminished drug binding or<br />
stabilization <strong>of</strong> ATP binding. In striking contrast, secondary<br />
mutations in BRAF(V600E) have not been reported thus far<br />
in patients with melanoma treated with vemurafenib. However,<br />
bypass <strong>of</strong> BRAF(V600E) was observed through other<br />
RAF family members, initiating sustained MAPK signaling.<br />
Preclinical studies have demonstrated that either intrinsic<br />
CRAF overexpression or upstream alterations leading to<br />
CRAF activation (e.g., mutant NRAS or KRAS, or RAP-1<br />
expression) confer robust resistance to RAF inhibition. 12-14<br />
Mutations in RAS at hotspot codons 12, 13, and 61, found<br />
in vemurafenib-resistant cell lines, lead to constitutive activation<br />
<strong>of</strong> RAS. In turn, these mutant oncoproteins engage<br />
MAPK signaling through CRAF, thus obviating BRAF(V600E)<br />
inhibition (Table 1). Also, an activating mutation in KRAS<br />
at K117N is associated with resistance to vemurafenib<br />
through CRAF-mediated MAPK signaling. 15 Additionally,<br />
KEY POINTS<br />
● Resistance to inhibitors <strong>of</strong> mutant BRAF inhibitors<br />
in metastatic melanoma is nearly universal, even<br />
though initial response is observed for many patients.<br />
● Secondary mutations in BRAF have not been found in<br />
resistant patient samples.<br />
● Reactivation <strong>of</strong> the MAPK pathway by activating<br />
mutations in NRAS or KRAS, CRAF amplification, or<br />
COT/MAP3K8 amplification confer resistance to<br />
BRAF inhibitors.<br />
● Bypass <strong>of</strong> MAPK through alternative receptor tyrosine<br />
kinases, such as platelet-derived growth factor<br />
or insulin-like growth factor-1 receptor may also<br />
confer resistance.<br />
● <strong>Clinical</strong> mutations in MEK are cross-resistant to both<br />
BRAF and MEK inhibitors, suggesting a mechanism<br />
<strong>of</strong> resistance to any RAF/MEK inhibitor combination<br />
therapy.<br />
CRAF overexpression itself can lead to activation <strong>of</strong> MAPK<br />
signaling. These studies clearly show that CRAF dysregulation<br />
can lead to resistance to mutant BRAF inhibitors (Fig.<br />
1).<br />
In addition to upstream activation <strong>of</strong> CRAF, recent studies<br />
have uncovered an alternatively spliced BRAF(V600E)<br />
is<strong>of</strong>orm that mediates resistance. This alternative splicing<br />
yields a truncated BRAF variant, termed p61BRAF(V600E).<br />
The p61 protein lacks the RAS-binding domain and thus<br />
confers constitutive dimerization and sustained MEK/ERK<br />
signaling, even in the presence <strong>of</strong> vemurafenib. 16 This interesting<br />
mechanism generally affirms the importance <strong>of</strong><br />
RAF dimerization in production <strong>of</strong> paradoxical MEK/ERK<br />
activation in the RAS-activated setting and has been observed<br />
in several clinical specimens analyzed thus far. 17-19<br />
Thus, sustained “RAF-dependent” signaling yields a prominent<br />
resistance mechanism even though secondary<br />
BRAF(V600E) point mutations have not yet been detected<br />
clinically.<br />
In addition to RAF-dependent mechanisms described<br />
above, resistance to mutant BRAF inhibition may also arise<br />
by dysregulation <strong>of</strong> signaling effectors capable <strong>of</strong> circumventing<br />
RAF altogether. One example <strong>of</strong> this is COT, which<br />
is a MAP3 kinase known to direct MEK/ERK signaling in<br />
certain hematologic and inflammatory contexts. 12 COT was<br />
identified in a systematic functional screen for kinases and<br />
related proteins that promote resistance to PLX4720, a<br />
preclinical compound related to vemurafenib. Exogenous<br />
expression <strong>of</strong> COT conferred 100-fold resistance to<br />
BRAF(V600E)-expressing cell lines, and COT overexpression<br />
was confirmed in de novo resistant cell lines and tumors<br />
that were resistant to vemurafenib. Although MEK/ERK<br />
signaling was restored with COT overexpression in the<br />
presence <strong>of</strong> PLX4720, BRAF or CRAF was not required for<br />
COT-mediated MAPK activity. This result accorded well<br />
with the notion that COT acts downstream <strong>of</strong> RAF and<br />
directly activates MEK/ERK signaling. More generally,<br />
these studies suggest that reactivation <strong>of</strong> the MAPK pathway<br />
(or sustained MAPK signaling) in the setting <strong>of</strong> RAF<br />
inhibition may compose a broadly important category <strong>of</strong><br />
resistance to vemurafenib.<br />
681
Mechanism <strong>of</strong> Resistance Drug a<br />
MAPK Bypass<br />
In addition to persistent MAPK pathway activation, alternative<br />
signaling pathways may confer resistance, as evidenced<br />
by the discovery <strong>of</strong> several receptor tyrosine kinases<br />
(RTKs) associated with this phenomenon. In particular,<br />
the RTKs HER2, AXL, platelet-derived growth factor-beta<br />
(PDGFRß), and insulin-like growth factor-1 receptor<br />
(IGF1R) have been linked to vemurafenib resistance in<br />
preclinical studies. 12,14,20 In addition, PDGFRß and IGF1R<br />
were activated and phosphorylated in some tumors that<br />
developed resistance to vemurafenib. 14,20 These RTKs seem<br />
to operate independently <strong>of</strong> MAPK signaling (Fig. 1), since<br />
ectopic expression <strong>of</strong> these RTKs tended not to show MAPK<br />
pathway reactivation. The extent to which these RTK alterations<br />
are sufficient to confer clinical resistance remains<br />
unclear, because overexpression <strong>of</strong> PDGFRß and IGF1R did<br />
not confer resistance in some preclinical studies, nor did the<br />
administration <strong>of</strong> imatinib resensitize melanoma cells that<br />
exhibited PDGFRß upregulation. 12,21 Thus, their resistance<br />
effects may manifest in certain molecular contexts or in<br />
cooperation with as-yet unmeasured microenvironmental<br />
effects. The RTK activation phenomena may also relate to<br />
the relief <strong>of</strong> feedback inhibition that may result from therapeutic<br />
inhibition <strong>of</strong> activated oncoproteins. 22<br />
Although the majority <strong>of</strong> robust resistance mechanisms<br />
characterized thus far tend to occur through downstream<br />
pathway reactivation, there is some evidence that ERKindependent<br />
mechanisms may also play a role. The ERKindependent<br />
effects <strong>of</strong> RTKs such as PDGFR and IGF1R<br />
constitute a line <strong>of</strong> evidence in this regard. In addition,<br />
reports <strong>of</strong> disease progression with persistent suppression<br />
<strong>of</strong> MEK/ERK activity have begun to emerge. 23 The mechanisms<br />
<strong>of</strong> ERK bypass remain poorly characterized and may<br />
gain increased importance if MEK inhibitors gain a role in<br />
the treatment <strong>of</strong> BRAF-mutant melanoma.<br />
Anticipating Resistance to Combined<br />
RAF/MEK Inhibition<br />
The preponderance <strong>of</strong> MEK/ERK-dependent resistance<br />
mechanisms provides a strong rationale for the use <strong>of</strong> RAF<br />
and MEK inhibitors in combination. Indeed, clinical trials <strong>of</strong><br />
combined RAF/MEK inhibition are ongoing. Several resistance<br />
mechanisms described above should in principle exhibit<br />
sensitivity to MEK inhibition (e.g., NRAS mutation,<br />
Table 1. Summary <strong>of</strong> Studies Probing Resistance to BRAF Inhibitors<br />
Type <strong>of</strong> Study Confirmed In Vivo? Reference<br />
In Vitro Studies<br />
NRAS mutation (Q61K, G12D) PLX4032 Drug-resistant clones Yes 14<br />
KRAS mutation (K117N) PLX4032 Stepwise selection No 15<br />
p61BRAF(V600E) PLX4032 Drug-resistant clones Yes 16<br />
CRAF overexpression PLX4720 Systematic ORF screen No 12<br />
CRAF overexpression AZ628 Drug-resistant clones No 13<br />
COT/MAP3K8 amplification PLX4720 Systematic ORF screen Yes 12<br />
IGF1R activation SB-590885 Drug-resistant clones Yes 20<br />
PDGFR activation PLX4032 Drug-resistant clones Yes 14<br />
MEK mutation (many) AZD6244, PLX4032 Random mutagenesis No 25<br />
In Vivo Studies<br />
MEK mutation (C121S) PLX4032 Targeted massively parallel sequencing N/A 24<br />
Abbreviations: IGF1R, insulin-like growth factor-1 receptor; MEK, mitogen-activated protein kinase kinase; ORF, open reading frame; PDGFR, platelet-derived growth<br />
factor.<br />
a PLX4032 is vemurafenib. PLX4720 is the preclinical equivalent <strong>of</strong> PLX4032. AZ628 inhibits BRAF(V600E) and wild-type CRAF. SB-590885 is a BRAF(V600E) kinase<br />
inhibitor. AZD6244 is a MEK inhibitor. The generation <strong>of</strong> drug-resistant clones (continuous exposure at one dose) or stepwise selection (steadily increasing drug<br />
concentration) is described in the text.<br />
682<br />
GOETZ AND GARRAWAY<br />
COT/MAP3K8 overexpression/copy gain, BRAF alternative<br />
splicing). On the other hand, activating mutations in MEKs<br />
might be predicted to confer resistance to combined inhibitor<br />
exposure (Fig. 1). Indeed, clinical evidence has already<br />
begun to emerge in support <strong>of</strong> this hypothesis. A C121S<br />
mutation in MEK1 was discovered in a clinical specimen<br />
resistant to vemurafenib, and expression <strong>of</strong> C121S MEK1<br />
in a vemurafenib-sensitive cell line conferred resistance to<br />
both RAF and MEK inhibition. 24 Thus, mutations in MEK<br />
that occur clinically can confer resistance to both RAF and<br />
MEK inhibition—and additional preclinical studies have<br />
indentified several other mutations in MEK that were crossresistant<br />
to RAF and MEK inhibitors. 24,25 In principle,<br />
mutations in MEK would still be vulnerable to ERK inhibition.<br />
Toward this end, selective ERK inhibitors have<br />
recently entered clinical trials. The role (and safe administration)<br />
<strong>of</strong> such agents in either up-front or salvage therapy<br />
for BRAF-mutant melanoma will undoubtedly emerge as an<br />
area <strong>of</strong> active investigation.<br />
Methods <strong>of</strong> Therapeutic Resistance Discovery<br />
The most common method for examining resistance involves<br />
generation <strong>of</strong> drug-resistant subclones using cancer<br />
cell lines in vitro. This is accomplished by either slowly<br />
increasing drug concentrations over time (stepwise selection)<br />
or through continuous incubation in high doses <strong>of</strong> a<br />
drug (Table 1). After several months, the clones are analyzed<br />
for mechanisms <strong>of</strong> resistance. Toward this end, changes in<br />
gene expression, protein modification, or point mutations<br />
can all lead to development <strong>of</strong> resistance. These changes<br />
may sometimes be difficult to disambiguate; alternatively,<br />
multiple mechanisms may contribute. Moreover, stepwise<br />
selection may not accurately model the acquisition <strong>of</strong> resistance,<br />
since tumors are not exposed to an increasing gradient<br />
<strong>of</strong> a drug over a period <strong>of</strong> weeks or months. Therefore, an<br />
alternative method for in vitro selection <strong>of</strong> drug-resistant<br />
subclones involves continuous exposure at high drug doses.<br />
This approach has proved useful in studies <strong>of</strong> resistance to<br />
vemurafenib in vitro.<br />
Several other techniques, such as random mutagenesis<br />
and systematic screens, are useful to probe resistance in<br />
vitro. These methods can speed up development <strong>of</strong> resistance<br />
(weeks instead <strong>of</strong> months) and may inform distinct<br />
spectra <strong>of</strong> resistance mechanisms. Random mutagenesis is
RESISTANCE MECHANISMS TO MAPK INHIBITION<br />
beneficial for defining mutations within the drug target that<br />
may cause resistance. Here, a library <strong>of</strong> complementary<br />
DNA (cDNA) is generated using error-prone polymerase<br />
chain reaction (PCR) or by growth in mutagenic bacteria.<br />
The mutated library is introduced into cells, followed by<br />
incubation <strong>of</strong> these cells in the presence <strong>of</strong> high concentrations<br />
<strong>of</strong> the targeted agent. After resistance cells emerge,<br />
large-scale DNA sequencing is used to identify candidate<br />
resistance mutations, which are then validated experimentally.<br />
This technique has been used to successfully determine<br />
mutations in MEK1 that confer resistance to the MEK<br />
inhibitor AZD6244 (Table 1). 25 Chemical-based mutagenesis<br />
using alkylating agents has also been used to examine<br />
resistance to kinase inhibitors in Ba/F3 cells, a murine pro-B<br />
cell line. The growth <strong>of</strong> Ba/F3 cells is dependent on<br />
interleukin-3 (IL3), however, these cells can be engineered<br />
to be independent <strong>of</strong> IL3 by constitutive expression <strong>of</strong> a<br />
dominant oncogene such as BCR-ABL. 26 Chemical-based<br />
mutagenesis <strong>of</strong> Ba/F3 cells expressing BCR-ABL and grown<br />
in the presence <strong>of</strong> imatinib recapitulated mutations found in<br />
patients whose cancers confer resistance to imatinib. 27 Additionally,<br />
genome-wide short hairpin RNA or open reading<br />
frame (ORF) overexpression screens may be used to find<br />
synthetic lethal and resistance alleles in a systematic way.<br />
These methods may help define the spectrum <strong>of</strong> mechanisms<br />
through which a cancer cell may manifest resistance. A<br />
caveat to these methods is that they may not exactly<br />
recapitulate in vivo resistance, given the intrinsic artificial<br />
nature <strong>of</strong> cell culture experiments. Nonetheless, many <strong>of</strong> the<br />
mechanisms determined in vitro have been confirmed in vivo<br />
(Table 1).<br />
The use <strong>of</strong> animal models, in addition to cell culture<br />
models, to study drug resistance may also identify clinically<br />
relevant resistance mechanisms. In addition to confirmation<br />
<strong>of</strong> in vitro resistance mechanisms in animal models, the use<br />
Authors’ Disclosures <strong>of</strong> Potential Conflicts <strong>of</strong> Interest<br />
Author<br />
Employment or<br />
Leadership<br />
Positions<br />
Consultant or<br />
Advisory Role<br />
Eva M. Goetz*<br />
Levi A. Garraway Daiichi Sankyo;<br />
Foundation<br />
Medicine;<br />
Novartis<br />
*No relevant relationships to disclose.<br />
<strong>of</strong> animals can also inform novel mechanisms <strong>of</strong> resistance.<br />
Animals bearing tumor xenografts can be treated with<br />
targeted therapies, and, over time, drug-resistant tumors<br />
analyzed for DNA or protein changes. Additionally, genetically<br />
engineered mouse models have been used to probe<br />
resistance; a study examining the inhibition <strong>of</strong> vascular<br />
endothelial growth factor receptor in a model <strong>of</strong> pancreatic<br />
islet carcinogenesis led to the discovery that fibroblast<br />
growth factor induction conferred resistance. 28 Since these<br />
experiments are performed in vivo, they may be more likely<br />
to represent actual alterations found in patients, but they<br />
can also be costly and difficult to perform.<br />
After a drug has been used in clinical trials, analysis <strong>of</strong><br />
pretreatment and postrelapse patient biopsies will be effective<br />
means <strong>of</strong> determining resistance. Deep sequencing <strong>of</strong><br />
exomes or whole genomes will provide information on the<br />
alterations in DNA sequence. In addition, analysis <strong>of</strong> protein<br />
expression patterns or protein phosphorylation may also<br />
help elucidate mechanisms <strong>of</strong> resistance. Examining preand<br />
post-treatment patient samples will be the most accurate<br />
way to study mechanisms <strong>of</strong> resistance; however, determining<br />
the mechanisms <strong>of</strong> resistance during preclinical<br />
testing may be just as important as examining postrelapse<br />
tumors.<br />
Conclusion<br />
With every medical therapy used for cancer treatment, de<br />
novo or acquired resistance causes treatment failure in a<br />
subset <strong>of</strong> patients. Therefore, studying resistance during<br />
preclinical drug investigation in addition to after clinical<br />
testing has failed is paramount to developing effective therapies.<br />
In the future, rationally designed drug combinations<br />
will hopefully minimize the appearance <strong>of</strong> resistance tumors<br />
and enhance clinical response.<br />
Stock<br />
Ownership Honoraria<br />
Foundation<br />
Medicine<br />
1. <strong>American</strong> Cancer <strong>Society</strong>. Cancer Facts & Figures. Atlanta, GA: <strong>American</strong><br />
Cancer <strong>Society</strong>; <strong>2012</strong>.<br />
2. Wellbrock C, Ogilvie L, Hedley D, et al. V599EB-RAF is an oncogene in<br />
melanocytes. Cancer Res. 2004;64:2338-2342.<br />
3. Wan PT, Garnett MJ, Roe SM, et al. Mechanism <strong>of</strong> activation <strong>of</strong> the<br />
RAF-ERK signaling pathway by oncogenic mutations <strong>of</strong> B-RAF. Cell. 2004;<br />
116:855-867.<br />
4. Davies H, Bignell GR, Cox C, et al. Mutations <strong>of</strong> the BRAF gene in<br />
human cancer. Nature. 2002;417:949-954.<br />
5. Chapman PB, Hauschild A, Robert C, et al. Improved survival with<br />
vemurafenib in melanoma with BRAF V600E mutation. N Engl J Med.<br />
2011;364:2507-2516.<br />
6. Flaherty KT, Puzanov I, Kim KB, et al. Inhibition <strong>of</strong> mutated, activated<br />
BRAF in metastatic melanoma. N Engl J Med. 2010;363:809-819.<br />
7. Dummer R, Robert C, Chapman PB, et al. AZD6244 (ARRY-142886) vs<br />
temozolomide (TMZ) in patients (pts) with advanced melanoma: an openlabel,<br />
randomized, multicenter, phase II study. J Clin Oncol. 2008;26:491s<br />
(suppl; abstr 9033).<br />
REFERENCES<br />
Boehringer<br />
Ingelheim;<br />
Constellation<br />
Pharmaceuticals;<br />
Novartis<br />
Research<br />
Funding<br />
Novartis<br />
Expert<br />
Testimony<br />
Other<br />
Remuneration<br />
8. Kobayashi S, Boggon TJ, Dayaram T, et al. EGFR mutation and<br />
resistance <strong>of</strong> non-small-cell lung cancer to gefitinib. N Engl J Med. 2005;352:<br />
786-792.<br />
9. Antonescu CR, Besmer P, Guo T, et al. Acquired resistance to imatinib<br />
in gastrointestinal stromal tumor occurs through secondary gene mutation.<br />
Clin Cancer Res. 2005;11:4182-4190.<br />
10. Gorre ME, Mohammed M, Ellwood K, et al. <strong>Clinical</strong> resistance to<br />
STI-571 cancer therapy caused by BCR-ABL gene mutation or amplification.<br />
Science. 2001;293:876-880.<br />
11. O’Hare T, Walters DK, St<strong>of</strong>fregen EP, et al. In vitro activity <strong>of</strong> Bcr-Abl<br />
inhibitors AMN107 and BMS-354825 against clinically relevant imatinibresistant<br />
Abl kinase domain mutants. Cancer Res. 2005;65:4500-4505.<br />
12. Johannessen CM, Boehm JS, Kim SY, et al. COT drives resistance to<br />
RAF inhibition through MAP kinase pathway reactivation. Nature. 2010;468:<br />
968-972.<br />
13. Montagut C, Sharma SV, Shioda T, et al. Elevated CRAF as a potential<br />
mechanism <strong>of</strong> acquired resistance to BRAF inhibition in melanoma. Cancer<br />
Res. 2008;68:4853-4861.<br />
683
14. Nazarian R, Shi H, Wang Q, et al. Melanomas acquire resistance to<br />
B-RAF(V600E) inhibition by RTK or N-RAS upregulation. Nature. 2010;468:<br />
973-977.<br />
15. Su F, Bradley WD, Wang Q, et al. Resistance to selective BRAF<br />
inhibition can be mediated by modest upstream pathway activation. Cancer<br />
Res. <strong>2012</strong>;72:969-978. Epub 2011 Dec 28.<br />
16. Poulikakos PI, Persaud Y, Janakiraman M, et al. RAF inhibitor<br />
resistance is mediated by dimerization <strong>of</strong> aberrantly spliced BRAF(V600E).<br />
Nature. 2011;480:387-390.<br />
17. Poulikakos PI, Zhang C, Bollag G, et al. RAF inhibitors transactivate<br />
RAF dimers and ERK signalling in cells with wild-type BRAF. Nature.<br />
2010;464:427-430.<br />
18. Su F, Viros A, Milagre C, et al. RAS mutations in cutaneous squamouscell<br />
carcinomas in patients treated with BRAF inhibitors. N Engl J Med.<br />
<strong>2012</strong>;366:207-215.<br />
19. Oberholzer PA, Kee D, Dziunycz P, et al. RAS mutations are associated<br />
with the development <strong>of</strong> cutaneous squamous cell tumors in patients treated<br />
with RAF inhibitors. J Clin Oncol. <strong>2012</strong>;30:316-321.<br />
20. Villanueva J, Vultur A, Lee JT, et al. Acquired resistance to BRAF<br />
inhibitors mediated by a RAF kinase switch in melanoma can be overcome by<br />
cotargeting MEK and IGF-1R/PI3K. Cancer Cell. 2010;18:683-695.<br />
21. Shi H, Kong X, Ribas A, et al. Combinatorial treatments that overcome<br />
PDGFRbeta-driven resistance <strong>of</strong> melanoma cells to V600EB-RAF inhibition.<br />
Cancer Res. 2011;71:5067-5074.<br />
684<br />
GOETZ AND GARRAWAY<br />
22. Pratilas CA, Taylor BS, Ye Q, et al. (V600E)BRAF is associated with<br />
disabled feedback inhibition <strong>of</strong> RAF-MEK signaling and elevated transcriptional<br />
output <strong>of</strong> the pathway. Proc Natl Acad Sci USA. 2009;106:4519-<br />
4524.<br />
23. McArthur GA, Ribas A, Chapman PB, et al. Molecular analyses from a<br />
phase I trial <strong>of</strong> vemurafenib to study mechanism <strong>of</strong> action (MOA) and<br />
resistance in repeated biopsies from BRAF mutation-positive metastatic<br />
melanoma patients (pts). J Clin Oncol. 2011;29:526s (suppl; abstr 8502).<br />
24. Wagle N, Emery C, Berger MF, et al. Dissecting therapeutic resistance<br />
to RAF inhibition in melanoma by tumor genomic pr<strong>of</strong>iling. J Clin Oncol.<br />
2011;29:3085-3096.<br />
25. Emery CM, Vijayendran KG, Zipser MC, et al. MEK1 mutations confer<br />
resistance to MEK and B-RAF inhibition. Proc Natl Acad Sci U S A.<br />
2009;106:20411-20416.<br />
26. Warmuth M, Kim S, Gu XJ, et al. Ba/F3 cells and their use in kinase<br />
drug discovery. Curr Opin Oncol. 2007;19:55-60.<br />
27. Bradeen HA, Eide CA, O’Hare T, et al. Comparison <strong>of</strong> imatinib<br />
mesylate, dasatinib (BMS-354825), and nilotinib (AMN107) in an N-ethyl-Nnitrosourea<br />
(ENU)-based mutagenesis screen: high efficacy <strong>of</strong> drug combinations.<br />
Blood. 2006;108:2332-2338.<br />
28. Casanovas O, Hicklin DJ, Bergers G, et al. Drug resistance by evasion<br />
<strong>of</strong> antiangiogenic targeting <strong>of</strong> VEGF signaling in late-stage pancreatic islet<br />
tumors. Cancer Cell. 2005;8:299-309.
Mechanisms <strong>of</strong> Resistance to Targeted<br />
Therapies in Acute Myeloid Leukemia and<br />
Chronic Myeloid Leukemia<br />
By Catherine C. Smith, MD, and Neil P. Shah, MD, PhD<br />
Overview: Small molecule kinase inhibitors <strong>of</strong> BCR-ABL in<br />
chronic myeloid leukemia (CML) and <strong>of</strong> FMS-like tyrosine<br />
kinase 3 internal tandem duplication (FLT3-ITD) in acute<br />
myeloid leukemia (AML) have been successful at achieving<br />
remissions in these diseases as monotherapy, but these<br />
leukemias do not initially respond in a subset <strong>of</strong> patients<br />
(primary resistance) and they progress in an additional group<br />
CLINICALLY EFFECTIVE small molecule inhibitors<br />
that target pathologically activated tyrosine kinases<br />
have become increasingly prevalent as cancer therapeutics.<br />
Identification <strong>of</strong> resistance mechanisms in patients treated<br />
with clinically active targeted therapeutics can yield critical<br />
insights into the importance <strong>of</strong> specific pathways in human<br />
cancers. Though the first highly successful molecularly “targeted”<br />
therapeutic in leukemia was all-trans retinoic acid,<br />
which promotes the differentiation <strong>of</strong> acute promyelocytic<br />
leukemic blasts driven by the fusion protein PML-<br />
RARalpha, mechanisms <strong>of</strong> resistance responsible for clinical<br />
resistance to this agent remain poorly defined. Therefore,<br />
this review will focus on emerging data surrounding clinical<br />
resistance mechanisms to tyrosine kinase inhibitors (TKIs)<br />
in CML and AML, which target oncogenic tyrosine kinases.<br />
CML: A Model <strong>of</strong> Resistance to Targeted Therapy<br />
The remarkable success <strong>of</strong> the small molecule inhibitor<br />
imatinib, which achieves remissions in CML in all phases <strong>of</strong><br />
disease, ushered in a new era <strong>of</strong> molecularly targeted therapeutics<br />
and created a new paradigm for the treatment <strong>of</strong><br />
cancer. Though CML in a small percentage <strong>of</strong> patients with<br />
chronic phase CML (CP-CML) does not respond to imatinib<br />
initially (primary resistance), the overall hematologic and<br />
major cytogenetic response (MCyR) rates (95.3% and 85.2%,<br />
respectively) are exceedingly high in newly diagnosed patients.<br />
1 Response rates in the advanced phases <strong>of</strong> CML<br />
(accelerated and blast phases; AP-CML) are substantially<br />
lower, and remissions in AP-CML are typically shortlived.<br />
2,3 Additionally, a proportion <strong>of</strong> patients with CP-CML<br />
also experiences relapse despite continuation <strong>of</strong> treatment<br />
(secondary resistance). Elucidation <strong>of</strong> the molecular mechanisms<br />
responsible for clinical resistance to effective targeted<br />
therapeutics can lead to strategies to improve clinical outcomes<br />
and also provide important insights into the role <strong>of</strong><br />
specific molecular targets in disease pathogenesis. To this<br />
end, studies that interrogate primary patient samples are<br />
most likely to both affect clinical management and be most<br />
relevant to disease. Illustrative <strong>of</strong> this point, paradigmatic<br />
studies <strong>of</strong> mechanisms <strong>of</strong> relapse on imatinib in patients<br />
with CML confirmed the importance <strong>of</strong> BCR-ABL as critical<br />
for CML pathogenesis; facilitated the development <strong>of</strong> the<br />
effective second-generation inhibitors dasatinib and nilotinib,<br />
which are effective against most imatinib-resistant<br />
mutations 4,5 ; and informed studies <strong>of</strong> secondary resistance<br />
mechanisms to active targeted therapies in a variety <strong>of</strong><br />
malignancies. Resistance to targeted therapies can be conceptualized<br />
in two major categories, which can inform efforts<br />
<strong>of</strong> patients after an initial response (secondary resistance).<br />
Resistance to these agents can be divided into mechanisms<br />
that allow reactivation kinase activity and those that bypass<br />
reliance on oncogenic signaling mediated by the target kinase.<br />
Elucidation <strong>of</strong> clinical resistance mechanisms to targeted<br />
therapies for patients can provide important insights into<br />
disease pathogenesis and signaling.<br />
to override resistance: (1) “on-target” resistance, whereby<br />
the disease remains dependent on aberrant signaling mediated<br />
by the target oncogene, and (2) “<strong>of</strong>f-target” resistance,<br />
whereby activation <strong>of</strong> downstream or parallel signaling<br />
pathways bypasses tumor dependence on the target oncogene<br />
(Fig. 1).<br />
On-target Resistance: BCR-ABL–Dependent<br />
Mechanisms <strong>of</strong> Resistance<br />
In a seminal report, Gorre and colleagues evaluated 11<br />
patients with advanced phase Philadelphia (Ph) chromosome–positive<br />
CML or acute lymphoblastic leukemia (ALL)<br />
whose leukemias relapsed after initially responding to imatinib<br />
and found reactivation <strong>of</strong> BCR-ABL kinase activity at<br />
the time <strong>of</strong> relapse in all cases, most commonly by gene<br />
amplification (three patients) or point mutation in the BCR-<br />
ABL kinase domain (six patients), specifically a C3T substitution<br />
coding for an isoleucine substitution at Thr 315 (the<br />
“gatekeeper” residue) predicted to impair imatinib binding 6<br />
by blocking access to a deeper hydrophobic pocket at the<br />
ATP binding site in the kinase domain <strong>of</strong> BCR-ABL. This<br />
mutation has been subsequently shown to mediate resistance<br />
to dasatinib and nilotinib as well. 5,7 The findings <strong>of</strong><br />
Gorre and colleagues confirmed the continued central dependence<br />
<strong>of</strong> CML on BCR-ABL kinase activity. A second mechanism<br />
<strong>of</strong> clinical resistance provided by Gorre and<br />
colleagues was genomic amplification <strong>of</strong> the BCR-ABL<br />
genomic locus, with presumed overexpression <strong>of</strong> BCR-ABL<br />
mRNA and protein. Though others have reported clinical<br />
resistance to be associated with BCR-ABL overexpression or<br />
the acquisition <strong>of</strong> additional Ph chromosomes, 8 numerous<br />
studies have confirmed that BCR-ABL kinase domain mutations<br />
represent the most common cause <strong>of</strong> relapse in all<br />
phases <strong>of</strong> disease, 9-11 both for imatinib and the secondgeneration<br />
BCR-ABL inhibitors dasatinib and nilotinib.<br />
Although clinical resistance to imatinib has been associated<br />
with more than 90 different resistance-causing kinase<br />
domain mutations to date, 12 the second-generation inhibitors<br />
dasatinib and nilotinib are vulnerable to far fewer<br />
resistance-conferring mutations. Dasatinib has activity<br />
From the Division <strong>of</strong> Hematology/<strong>Oncology</strong>, University <strong>of</strong> California, San Francisco, CA.<br />
Authors’ disclosures <strong>of</strong> potential conflicts <strong>of</strong> interest are found at the end <strong>of</strong> this article.<br />
Address reprint requests to Neil P. Shah, MD, PhD, Division <strong>of</strong> Hematology/<strong>Oncology</strong>,<br />
UCSF, 505 Parnassus Avenue, Suite M1286, Box 1270, San Francisco, CA 94143; email:<br />
nshah@medicine.ucsf.edu.<br />
© <strong>2012</strong> by <strong>American</strong> <strong>Society</strong> <strong>of</strong> <strong>Clinical</strong> <strong>Oncology</strong>.<br />
1092-9118/10/1-10<br />
685
against most imatinib-resistant BCR-ABL mutants in<br />
vitro 4,13 and clinically. Notable exceptions are the highly<br />
resistant T315I mutation and the moderately resistant<br />
F317L mutation, 4 which is frequently associated with hematologic<br />
response, rarely with cytogenetic response, and occasionally<br />
with secondary resistance to dasatinib. 10 Other<br />
mutations associated with clinical resistance to dasatinib<br />
include V299L, T315A, and F317I, 4,10 which are largely<br />
imatinib-sensitive. Nilotinib is a structural derivative <strong>of</strong><br />
imatinib that is similarly clinically vulnerable to a small<br />
number <strong>of</strong> BCR-ABL kinase domain mutations, including<br />
T315I, 14 Y253H, E255V/K, and F359C/V. 15,16 Collectively,<br />
the three approved BCR-ABL kinase inhibitors appear to<br />
retain activity against all BCR-ABL kinase domain mutants<br />
associated with single nucleotide substitution, with the<br />
exception <strong>of</strong> BCR-ABL/T315I mutants. Both dasatinib and<br />
nilotinib have been approved for the treatment <strong>of</strong> CML in<br />
the second line, and more recently in the first line, as a<br />
result <strong>of</strong> improved cytogenetic and molecular response rates<br />
associated with these agents relative to imatinib. 17,18<br />
Commonly, the choice <strong>of</strong> a second-line BCR-ABL inhibitor<br />
in patients is guided by the specific mutation responsible for<br />
imatinib resistance, as well as comorbid conditions that may<br />
influence the choice <strong>of</strong> agent. Complex substitutions, including<br />
polyclonal (more than one mutation in a single patient)<br />
and compound mutations (more than one mutation on a<br />
single BCR-ABL allele), have also been reported to develop<br />
in patients whose leukemias relapse on sequential treatment<br />
with BCR-ABL inhibitors, 10,19 which makes the choice<br />
<strong>of</strong> subsequent inhibitors less clear. Until recently, the T315I<br />
mutation has remained a vexing problem. Encouragingly,<br />
the investigational third-generation BCR-ABL inhibitor<br />
KEY POINTS<br />
● <strong>Clinical</strong>ly relevant mechanisms <strong>of</strong> resistance to targeted<br />
therapies can provide insight into oncogenic<br />
signaling and disease pathogenesis.<br />
● Resistance can be divided into mechanisms that reactivate<br />
kinase activity (“on-target”), indicating<br />
maintained critical reliance on the target kinase, and<br />
those that bypass kinase activity (“<strong>of</strong>f-target”).<br />
● Secondary resistance to BCR-ABL and FMS-like tyrosine<br />
kinase 3 (FLT3) inhibitors is largely mediated<br />
by on-target mechanisms, specifically kinase domain<br />
mutations.<br />
● Increased levels <strong>of</strong> primary resistance in advanced<br />
phases <strong>of</strong> chronic myeloid leukemia and in FLT3–<br />
internal tandem duplication mutant acute myeloid<br />
leukemia may be indicative <strong>of</strong> cooperating lesions in<br />
these diseases that allow bypass <strong>of</strong> oncogenic signaling<br />
mediated by the target kinase.<br />
● As more potent second- and third-generation BCR-<br />
ABL and FLT3 inhibitors with broader activity<br />
against drug resistance–causing kinase domain mutations<br />
become available, it is anticipated that resistance<br />
will be increasingly associated with <strong>of</strong>f-target<br />
mechanisms <strong>of</strong> resistance.<br />
686<br />
ponatinib has demonstrated substantial clinical activity in<br />
patients with the T315I mutation in an interim analysis <strong>of</strong> a<br />
pivotal phase II study, 20 with 57% <strong>of</strong> patients with CP-CML<br />
with the T315I mutation achieving a MCyR, defined as 35%<br />
or less Philadelphia chromosome–positive bone marrow<br />
metaphases. Ponatinib also demonstrated clinical activity<br />
against other drug-resistant BCR-ABL mutant is<strong>of</strong>orms in<br />
this study. 20 Encouragingly, this agent appears to be highly<br />
invulnerable to kinase domain mutation as a mechanism <strong>of</strong><br />
resistance in vitro, 21 and translational studies that define<br />
mechanisms <strong>of</strong> resistance to this agent will be <strong>of</strong> particular<br />
interest. It is anticipated that the ability to clinically inhibit<br />
all BCR-ABL kinase domain mutant is<strong>of</strong>orms will likely<br />
render <strong>of</strong>f-target mechanisms <strong>of</strong> resistance increasingly relevant.<br />
Decreased expression <strong>of</strong> the drug transporter human<br />
organic cation transporter 1 (hOCT1), responsible for imatinib<br />
influx, has also been correlated with suboptimal response<br />
to imatinib, though this could be overcome with<br />
higher doses <strong>of</strong> imatinib. 22 Pretreatment expression <strong>of</strong><br />
hOCT1 has also been correlated with clinical outcomes to<br />
imatinib including response rate, progression-free survival<br />
(PFS), and overall survival (OS). 23 It is likely, however, with<br />
evidence <strong>of</strong> improved molecular response rates with more<br />
potent second-generation BCR-ABL inhibitors used in the<br />
first line 17,18 that this mechanism <strong>of</strong> clinical resistance will<br />
become less relevant.<br />
Off-target Resistance: BCR-ABL–Independent<br />
Mechanisms <strong>of</strong> Resistance<br />
SMITH AND SHAH<br />
To date, the molecular mechanisms <strong>of</strong> clinically relevant<br />
BCR-ABL–independent resistance remain largely uncharacterized.<br />
It is clear that certain additional clonal cytogenetic<br />
abnormalities are associated with progression from CP-CML<br />
to AP-CML, decreased TKI responsiveness, and poorer prognosis.<br />
In a large study <strong>of</strong> 1,151 patients treated with imatinib,<br />
“major-route” cytogenetic abnormalities defined as a<br />
second Ph chromosome, trisomy 8, isochromosome 17q, or<br />
trisomy 19 were clearly associated with longer times to<br />
achievement <strong>of</strong> complete cytogenetic response and major<br />
molecular response and overall decreased PFS and OS. 24 A<br />
recent report suggested an association between translocations<br />
involving ecotropic viral integration site 1 (EVI1) and<br />
TKI resistance in myeloid blast crisis disease. 25 Additionally,<br />
the substantially lower initial hematologic response<br />
rates to imatinib in Ph-positive ALL and AP-CML suggest<br />
that advanced phases <strong>of</strong> disease may have decreased reliance<br />
on BCR-ABL signaling. 2,3 The association between<br />
acquired clonal cytogenetic abnormalities and disease progression<br />
despite TKI therapy and the lower overall TKI<br />
response rates observed in advanced phases <strong>of</strong> disease<br />
suggests that select cooperating genetic alterations may<br />
functionally bypass dependence on BCR-ABL tyrosine kinase<br />
activity. Multiple in vitro studies have suggested that<br />
CML cells may bypass dependence on BCR-ABL signaling<br />
through aberrant activation <strong>of</strong> parallel or downstream signaling<br />
pathways by other means, including overexpression<br />
<strong>of</strong> SRC-family kinases, 26 overexpression <strong>of</strong> STAT5, 27 or<br />
activation <strong>of</strong> the PI3K/AKT pathway. 28 However, the specific<br />
molecular mechanisms mediating this process in patients<br />
remain poorly defined.
RESISTANCE TO TARGETED THERAPIES IN AML AND CML<br />
Fig. 1. Schematic depicting primary and secondary<br />
tyrosine kinase inhibitor resistance, and on-target/<strong>of</strong>ftarget<br />
resistance mechanisms.<br />
AML: Emerging Targets and Emerging Resistance<br />
The success <strong>of</strong> imatinib in CML has driven the desire to<br />
translate the paradigm <strong>of</strong> targeted therapy to other cancers,<br />
including other hematologic malignancies such as AML.<br />
Unlike CML however, the genetic drivers underlying AML<br />
pathogenesis are heterogeneous, and though multiple recurrent<br />
molecular abnormalities have emerged in recent years,<br />
to date, no therapies targeting these lesions have been<br />
approved. It is likely that, in contrast to the experience with<br />
BCR-ABL, many <strong>of</strong> these molecular lesions represent cooperating<br />
rather than initiating genetic events in disease<br />
pathogenesis, which common wisdom would deem to be<br />
unlikely therapeutic targets.<br />
Among the most common genetic lesions in AML are<br />
constitutively activating mutations <strong>of</strong> FLT3. ITD mutations<br />
in FLT3 are found in approximately 20% <strong>of</strong> patients with<br />
AML and are associated with poor prognosis. 29,30 Another<br />
5% to 10% <strong>of</strong> patients with AML express constitutively<br />
activating point mutations in the tyrosine kinase domain. 31<br />
Despite early enthusiasm, initial attempts to target FLT3<br />
with small molecule inhibitors were largely unsuccessful,<br />
resulting in only transient reductions in peripheral blasts<br />
with few bone marrow responses. 32,33 Emerging data have<br />
further suggested that FLT3 mutations are likely secondary<br />
rather than initiating genetic lesions in AML, hinting<br />
that FLT3 mutations may not be critical for disease<br />
pathogenesis. 34-36<br />
The recent success <strong>of</strong> the investigational FLT3 inhibitor<br />
AC220 in achieving a composite complete remission rate <strong>of</strong><br />
45% in an interim analysis <strong>of</strong> a multinational phase II study<br />
in patients with relapsed/refractory FLT3-ITD–positive<br />
AML has rekindled interest in FLT3 as a therapeutic target.<br />
The evolution <strong>of</strong> tyrosine kinase domain mutations in eight<br />
<strong>of</strong> eight patients whose leukemias relapsed after a bone<br />
marrow response on AC220 confirmed that clinical response<br />
in these patients was achieved by inhibition <strong>of</strong> FLT3, and<br />
that relapses were mediated by the reactivation <strong>of</strong> FLT3<br />
kinase activity. 37 Similar to BCR-ABL/T315I, F691, the<br />
“gatekeeper” residue in FLT3, was mutated in a subset <strong>of</strong><br />
patients in whom AC220-resistant disease developed (to<br />
leucine; 3 patients), but mutations at the activation loop<br />
residue D835 developed in the remaining six patients (including<br />
one patient in whom both a F691L and D835V<br />
mutation developed), suggesting that substitutions at the<br />
D835 residue may be a common cause <strong>of</strong> AC220 resistance.<br />
These findings validate FLT3-ITD as a therapeutic target<br />
and further suggest that the activity <strong>of</strong> “cooperating” genetic<br />
events that are believed to occur relatively late during<br />
evolution <strong>of</strong> the malignant clone, such as FLT3-ITD, can<br />
nonetheless be critically important to the survival <strong>of</strong> cells in<br />
which they arise. Though this limited study suggests that<br />
on-target resistance mechanisms are largely responsible for<br />
secondary resistance to AC220, the lower overall response<br />
rates to AC220 imply that <strong>of</strong>f-target mechanisms are likely<br />
to be important in both primary and secondary resistant<br />
cases, as has been similarly postulated in cases <strong>of</strong> AP-CML.<br />
It is hypothesized that cooperative genetic events in AP-<br />
CML and FLT3-ITD–positive AML can be placed into two<br />
general categories, those that preserve reliance on the activated<br />
oncogene and those that bypass it. The latter category<br />
may explain the lower initial response rates in blast phase<br />
CML and FLT3-ITD–positive AML when compared with<br />
chronic phase CML.<br />
It has been suggested that another potential clinically<br />
relevant FLT3-dependent mechanism <strong>of</strong> primary resistance<br />
may involve increases in FLT3 ligand levels following chemotherapy,<br />
which is predicted to increase the concentration<br />
<strong>of</strong> inhibitor required to effect cytotoxicity. 38 This mechanism<br />
may be more relevant in patients treated concurrently with<br />
FLT3 inhibitors and chemotherapy, and further studies are<br />
necessary to more accurately define its clinical importance.<br />
Other targeted therapies currently undergoing clinical<br />
development in AML include inhibitors <strong>of</strong> mammalian target<br />
<strong>of</strong> rapamycin, 39 PI3K, and MEK, 40 but the lack <strong>of</strong> clear<br />
clinical efficacy to date precludes the establishment <strong>of</strong> clinically<br />
relevant resistance mechanisms. As was the case with<br />
inhibitors <strong>of</strong> BCR-ABL and FLT3, elucidation <strong>of</strong> clinically<br />
relevant resistance mechanisms necessitates inhibitors that<br />
effect deep clinical responses when administered as monotherapy<br />
and detailed analysis <strong>of</strong> samples isolated from<br />
patients at the time <strong>of</strong> secondary evolution.<br />
687
Conclusion<br />
The overall success <strong>of</strong> small molecule inhibitors targeted<br />
against BCR-ABL in CML has spurred the development <strong>of</strong><br />
inhibitors targeting other pathologically activated kinases.<br />
The on-target BCR-ABL TKI resistance mechanism paradigm<br />
has been implicated as operative in a number <strong>of</strong> other<br />
malignancies associated with clinically active kinase<br />
inhibitors, 41-43 now including FLT3-ITD–positive AML. In<br />
particular, the concept that many relapses occur because <strong>of</strong><br />
Authors’ Disclosures <strong>of</strong> Potential Conflicts <strong>of</strong> Interest<br />
Author<br />
Employment or<br />
Leadership<br />
Positions<br />
Consultant or<br />
Advisory Role<br />
Catherine C. Smith Astellas<br />
Neil P. Shah ARIAD; Bristol-<br />
Myers Squibb;<br />
Novartis<br />
1. O’Brien SG, Guilhot F, Larson RA, et al. Imatinib compared with<br />
interferon and low-dose cytarabine for newly diagnosed chronic-phase chronic<br />
myeloid leukemia. N Engl J Med. 2003;348:994-1004.<br />
2. Sawyers CL, Hochhaus A, Feldman E, et al. Imatinib induces hematologic<br />
and cytogenetic responses in patients with chronic myelogenous leukemia<br />
in myeloid blast crisis: results <strong>of</strong> a phase II study. Blood. 2002;99:3530-<br />
3539.<br />
3. Ottmann OG, Druker BJ, Sawyers CL, et al. A phase 2 study <strong>of</strong> imatinib<br />
in patients with relapsed or refractory Philadelphia chromosome-positive<br />
acute lymphoid leukemias. Blood. 2002;100:1965-1971.<br />
4. Shah NP, Tran C, Lee FY, et al. Overriding imatinib resistance with a<br />
novel ABL kinase inhibitor. Science. 2004;305:399-401.<br />
5. Bradeen HA, Eide CA, O’Hare T, et al. Comparison <strong>of</strong> imatinib mesylate,<br />
dasatinib (BMS-354825), and nilotinib (AMN107) in an N-ethyl-Nnitrosourea<br />
(ENU)-based mutagenesis screen: high efficacy <strong>of</strong> drug combinations.<br />
Blood. 2006;108:2332-2338.<br />
6. Gorre ME, Mohammed M, Ellwood K, et al. <strong>Clinical</strong> resistance to<br />
STI-571 cancer therapy caused by BCR-ABL gene mutation or amplification.<br />
Science. 2001;293:876-880.<br />
7. Burgess MR, Skaggs BJ, Shah NP, et al. Comparative analysis <strong>of</strong> two<br />
clinically active BCR-ABL kinase inhibitors reveals the role <strong>of</strong> conformationspecific<br />
binding in resistance. Proc Natl Acad Sci USA. 2005;102:3395-3400.<br />
8. Hochhaus A, Kreil S, Corbin AS, et al. Molecular and chromosomal<br />
mechanisms <strong>of</strong> resistance to imatinib (STI571) therapy. Leukemia. 2002;16:<br />
2190-2196.<br />
9. Shah NP, Nicoll JM, Nagar B, et al. Multiple BCR-ABL kinase domain<br />
mutations confer polyclonal resistance to the tyrosine kinase inhibitor imatinib<br />
(STI571) in chronic phase and blast crisis chronic myeloid leukemia.<br />
Cancer Cell. 2002;2:117-125.<br />
10. Shah NP, Skaggs BJ, Branford S, et al. Sequential ABL kinase<br />
inhibitor therapy selects for compound drug-resistant BCR-ABL mutations<br />
with altered oncogenic potency. J Clin Invest. 2007;117:2562-2569.<br />
11. Soverini S, Colarossi S, Gnani A, et al. Contribution <strong>of</strong> ABL kinase<br />
domain mutations to imatinib resistance in different subsets <strong>of</strong> Philadelphiapositive<br />
patients: by the GIMEMA Working Party on Chronic Myeloid<br />
Leukemia. Clin Cancer Res. 2006;12:7374-7379.<br />
12. Soverini S, Hochhaus A, Nicolini FE, et al. BCR-ABL kinase domain<br />
mutation analysis in chronic myeloid leukemia patients treated with tyrosine<br />
kinase inhibitors: recommendations from an expert panel on behalf <strong>of</strong><br />
European LeukemiaNet. Blood. 2011;118:1208-1215.<br />
13. Talpaz M, Shah NP, Kantarjian H, et al. Dasatinib in imatinibresistant<br />
Philadelphia chromosome-positive leukemias. N Engl J Med. 2006;<br />
354:2531-2541.<br />
14. O’Hare T, Walters DK, St<strong>of</strong>fregen EP, et al. In vitro activity <strong>of</strong> Bcr-Abl<br />
inhibitors AMN107 and BMS-354825 against clinically relevant imatinibresistant<br />
Abl kinase domain mutants. Cancer Res. 2005;65:4500-4505.<br />
15. Kantarjian H, Giles F, Wunderle L, et al. Nilotinib in imatinibresistant<br />
CML and Philadelphia chromosome-positive ALL. N Engl J Med.<br />
2006;354:2542-2551.<br />
688<br />
the reactivation <strong>of</strong> oncogenic kinase activity mediated by<br />
on-target mutation, amplification, decreased drug influx, or<br />
increased ligand activation seems to be particularly clinically<br />
relevant and has led to rekindled interest in developing<br />
potent selective FLT3 inhibitors. As inhibitors with increased<br />
potency/selectivity and decreased vulnerability to<br />
resistance-conferring mutations become available, it is anticipated<br />
that <strong>of</strong>f-target mechanisms <strong>of</strong> resistance, such as<br />
activation <strong>of</strong> parallel or downstream signaling pathways,<br />
will be more prominent in both CML and AML.<br />
Stock<br />
Ownership Honoraria<br />
REFERENCES<br />
Research<br />
Funding<br />
Ambit; ARIAD;<br />
Bristol-Myers<br />
Squibb;<br />
Plexxikon<br />
Expert<br />
Testimony<br />
SMITH AND SHAH<br />
Other<br />
Remuneration<br />
16. Kantarjian HM, Giles FJ, Bhalla KN, et al. Update on imatinibresistant<br />
chronic myeloid leukemia patients in chronic phase (CML-CP) on<br />
nilotinib therapy at 24 months: clinical response, safety, and long-term<br />
outcomes. ASH Annual Meeting Abstracts. 2009;114 (abstr 1129).<br />
17. Kantarjian H, Shah NP, Hochhaus A, et al. Dasatinib versus imatinib<br />
in newly diagnosed chronic-phase chronic myeloid leukemia. N Engl J Med.<br />
2010;362:2260-2270.<br />
18. Saglio G, Kim DW, Issaragrisil S, et al. Nilotinib versus imatinib for<br />
newly diagnosed chronic myeloid leukemia. N Engl J Med. 2010;362:2251-<br />
2259.<br />
19. Smith CC, Brown M, Chin J, et al. Single Molecule Real Time<br />
(SMRT) sequencing sensitively detects polyclonal and compound BCR-ABL<br />
in patients who relapse on kinase inhibitor therapy. ASH Annual Meeting<br />
Abstracts. 2011;118 (abstr 3752).<br />
20. Cortes JE, Kim D-W, Pinilla-Ibarz J, et al. Initial findings from the<br />
PACE trial: a pivotal phase 2 study <strong>of</strong> ponatinib in patients with CML and<br />
Ph� ALL resistant or intolerant to dasatinib or nilotinib, or with the T315I<br />
mutation. ASH Annual Meeting Abstracts. 2011;118 (abstr 109).<br />
21. O’Hare T, Shakespeare WC, Zhu X, et al. AP24534, a pan-BCR-ABL<br />
inhibitor for chronic myeloid leukemia, potently inhibits the T315I mutant<br />
and overcomes mutation-based resistance. Cancer Cell. 2009;16:401-412.<br />
22. White DL, Saunders VA, Dang P, et al. Most CML patients who have a<br />
suboptimal response to imatinib have low OCT-1 activity: higher doses <strong>of</strong><br />
imatinib may overcome the negative impact <strong>of</strong> low OCT-1 activity. Blood.<br />
2007;110:4064-4072.<br />
23. Wang L, Giannoudis A, Lane S, et al. Expression <strong>of</strong> the uptake drug<br />
transporter hOCT1 is an important clinical determinant <strong>of</strong> the response to<br />
imatinib in chronic myeloid leukemia. Clin Pharmacol Ther. 2008;83:258-264.<br />
24. Fabarius A, Leitner A, Hochhaus A, et al. Impact <strong>of</strong> additional cytogenetic<br />
aberrations at diagnosis on prognosis <strong>of</strong> CML: long-term observation <strong>of</strong><br />
1151 patients from the randomized CML Study IV. Blood. 2011;118:6760-<br />
6768.<br />
25. Paquette RL, Nicoll J, Chalukya M, et al. Frequent EVI1 translocations<br />
in myeloid blast crisis CML that evolves through tyrosine kinase inhibitors.<br />
Cancer Genet. 2011;204:392-397.<br />
26. Donato NJ, Wu JY, Stapley J, et al. BCR-ABL independence and LYN<br />
kinase overexpression in chronic myelogenous leukemia cells selected for<br />
resistance to STI571. Blood. 2003;101:690-698.<br />
27. Warsch W, Kollmann K, Eckelhart E, et al. High STAT5 levels mediate<br />
imatinib resistance and indicate disease progression in chronic myeloid<br />
leukemia. Blood. 2011;117:3409-3420.<br />
28. Quentmeier H, Eberth S, Romani J, et al. BCR-ABL1-independent<br />
PI3Kinase activation causing imatinib-resistance. J Hematol Oncol. 2011;4:6.<br />
29. Abu-Duhier FM, Goodeve AC, Wilson GA, et al. FLT3 internal tandem<br />
duplication mutations in adult acute myeloid leukaemia define a high-risk<br />
group. Br J Haematol. 2000;111:190-195.<br />
30. Kottaridis PD, Gale RE, Frew ME, et al. The presence <strong>of</strong> a FLT3<br />
internal tandem duplication in patients with acute myeloid leukemia (AML)<br />
adds important prognostic information to cytogenetic risk group and response
RESISTANCE TO TARGETED THERAPIES IN AML AND CML<br />
to the first cycle <strong>of</strong> chemotherapy: analysis <strong>of</strong> 854 patients from the United<br />
Kingdom Medical Research Council AML 10 and 12 trials. Blood. 2001;98:<br />
1752-1759.<br />
31. Abu-Duhier FM, Goodeve AC, Wilson GA, et al. Identification <strong>of</strong> novel<br />
FLT-3 Asp835 mutations in adult acute myeloid leukaemia. Br J Haematol.<br />
2001;113:983-988.<br />
32. Stone RM, DeAngelo DJ, Klimek V, et al. Patients with acute myeloid<br />
leukemia and an activating mutation in FLT3 respond to a small-molecule<br />
FLT3 tyrosine kinase inhibitor, PKC412. Blood. 2005;105:54-60.<br />
33. Knapper S, Burnett AK, Littlewood T, et al. A phase 2 trial <strong>of</strong> the FLT3<br />
inhibitor lestaurtinib (CEP701) as first-line treatment for older patients with<br />
acute myeloid leukemia not considered fit for intensive chemotherapy. Blood.<br />
2006;108:3262-3270.<br />
34. Kelly LM, Liu Q, Kutok JL, et al. FLT3 internal tandem duplication<br />
mutations associated with human acute myeloid leukemias induce myeloproliferative<br />
disease in a murine bone marrow transplant model. Blood. 2002;<br />
99:310-318.<br />
35. Shih LY, Huang CF, Wu JH, et al. Internal tandem duplication <strong>of</strong> FLT3<br />
in relapsed acute myeloid leukemia: a comparative analysis <strong>of</strong> bone marrow<br />
samples from 108 adult patients at diagnosis and relapse. Blood. 2002;100:<br />
2387-2392.<br />
36. McCormick SR, McCormick MJ, Grutkoski PS, et al. FLT3 mutations<br />
at diagnosis and relapse in acute myeloid leukemia: cytogenetic and patho-<br />
logic correlations, including cuplike blast morphology. Arch Pathol Lab Med.<br />
2010;134:1143-1151.<br />
37. Smith CC, Chin J, Wang Q, et al. Validation <strong>of</strong> FLT3-ITD as a<br />
therapeutic target in human acute myeloid leukemia. ASH Annual Meeting<br />
Abstracts. 2011;118 (abstr 937).<br />
38. Sato T, Yang X, Knapper S, et al. FLT3 ligand impedes the efficacy <strong>of</strong><br />
FLT3 inhibitors in vitro and in vivo. Blood. 2011;117:3286-3293.<br />
39. Rizzieri DA, Feldman E, Moore JO, et al. A phase 2 clinical trial <strong>of</strong><br />
AP23573, an mTOR inhibitor, in patients with relapsed or refractory hematologic<br />
malignancies. ASH Annual Meeting Abstracts. 2005;106 (abstr 2980).<br />
40. Odenike O, Curran E, Iyengar N, et al. Phase II study <strong>of</strong> the oral MEK<br />
inhibitor AZD6244 in advanced acute myeloid leukemia (AML). ASH Annual<br />
Meeting Abstracts. 2009;114 (abstr 2081).<br />
41. Pao W, Miller VA, Politi KA, et al. Acquired resistance <strong>of</strong> lung<br />
adenocarcinomas to gefitinib or erlotinib is associated with a second mutation<br />
in the EGFR kinase domain. PLoS Med. 2005;2:e73.<br />
42. Cools J, DeAngelo DJ, Gotlib J, et al. A tyrosine kinase created by<br />
fusion <strong>of</strong> the PDGFRA and FIP1L1 genes as a therapeutic target <strong>of</strong> imatinib<br />
in idiopathic hypereosinophilic syndrome. N Engl J Med. 2003;348:1201-<br />
1214.<br />
43. Choi YL, Soda M, Yamashita Y, et al. EML4-ALK mutations in lung<br />
cancer that confer resistance to ALK inhibitors. N Engl J Med. 2010;363:<br />
1734-1739.<br />
689
TOWARD SUCCESSFUL TARGETING OF THE<br />
PI3 KINASE PATHWAY IN CANCER<br />
CHAIR<br />
Paul Workman, PhD<br />
The Institute <strong>of</strong> Cancer Research<br />
Sutton, United Kingdom<br />
SPEAKERS<br />
John C. Byrd, MD<br />
The Ohio State University Comprehensive Cancer Center<br />
Columbus, OH<br />
Carlos L. Arteaga, MD<br />
Vanderbilt University<br />
Nashville, TN
Translating PI3K-Delta Inhibitors to the<br />
Clinic in Chronic Lymphocytic Leukemia:<br />
The Story <strong>of</strong> CAL-101 (GS1101)<br />
By John C. Byrd, MD, Jennifer A. Woyach, MD, and Amy J. Johnson, PhD<br />
Overview: Targeted therapy with imatinib has transformed<br />
the treatment <strong>of</strong> chronic myeloid leukemia (CML). Unlike CML,<br />
chronic lymphocytic leukemia (CLL) lacks a common genetic<br />
aberration but does demonstrate constitutive activation <strong>of</strong><br />
PI3-kinase (PI3K) as compared to normal B cells. This constitutively<br />
active PI3K in CLL likely relates to tonic B-cell<br />
receptor signaling that is present across a wide variety <strong>of</strong><br />
B-cell malignancies. Although PI3K is quite proximal and<br />
represents an ideal target to pharmacologically modulate, the<br />
complexity <strong>of</strong> this pathway on which many normal functions<br />
are dependent had for many years been problematic. The p110<br />
THE SUCCESS <strong>of</strong> imatinib in CML has prompted great<br />
interest in developing targeted therapies for other<br />
types <strong>of</strong> blood cancers. Unlike CML, most other blood cancers<br />
lack a genetically altered common target that is amendable<br />
for therapeutic targeting. Instead, pathways that are<br />
active in normal cells at select times are <strong>of</strong>ten aberrantly<br />
continuously activated. An example <strong>of</strong> this is B-cell receptor<br />
(BCR) signaling in normal B cells, which facilitates proliferation,<br />
migration, and survival in the setting <strong>of</strong> activation by<br />
an antigen but which is promptly downregulated as exposure<br />
to this diminishes. In contrast, BCR signaling is <strong>of</strong>ten<br />
constitutively active in select B-cell malignancies, including<br />
CLL and NHL. The etiology <strong>of</strong> this pathway activation is<br />
multifactorial, as a result <strong>of</strong> enhanced microenvironment<br />
activation, mutation <strong>of</strong> activating genes, or silencing <strong>of</strong><br />
specific phosphatases involved in proximal B-cell receptor<br />
signaling. Although this pathway is quite complex and <strong>of</strong>fers<br />
several therapeutic targets for pharmacologic intervention,<br />
one <strong>of</strong> the more proximal kinases activated is PI3K.<br />
The PI3K Pathway<br />
The PI3K pathway is a key component <strong>of</strong> cell survival in<br />
many cancers, including CLL and low-grade NHL. PI3K is<br />
activated by receptors or the small GTPase Ras and includes<br />
multiple is<strong>of</strong>orms. 1 There are three classes <strong>of</strong> PI3K is<strong>of</strong>orms;<br />
however, only the class I is<strong>of</strong>orms phosphorylate inositol<br />
lipids to form second messenger phosphoinositides. Specifically,<br />
class I PI3K enzymes convert PtdIns(3,4)P 2 into<br />
PtdIns(3,4,5)P 3 that recruit downstream signaling proteins<br />
such as BTK, PDK, AKT, ILK, and Rac GEF 1,2 that can<br />
generate survival signals. Class I is<strong>of</strong>orms are made up <strong>of</strong><br />
two subsets (IA and IB). Class IA encompasses p110 alpha,<br />
p110 beta, and p110 delta (catalytic domains) that are bound<br />
by p85, p50, or p55 (regulatory domains). Class IB is made<br />
up solely <strong>of</strong> the p110 gamma (catalytic domain) bound by<br />
the regulatory domain p101. The p110 alpha and p110 beta<br />
is<strong>of</strong>orms are ubiquitously expressed, and knock-out mice for<br />
both are embryonic lethal. 3 It is thought that this widespread<br />
functionality <strong>of</strong> PI3K signaling is at least partially<br />
responsible for the significant cellular toxicity associated<br />
with pan-PI3K inhibitors such as LY294002. 4 However, in<br />
recent years it has been shown that these catalytic is<strong>of</strong>orms<br />
have different expression pr<strong>of</strong>iles in different cell types. 5-8<br />
delta is<strong>of</strong>orm <strong>of</strong> PI3K is relatively specific to hematopoietic<br />
cells, and elegant mouse studies where p110 delta was genetically<br />
inactivated demonstrated only a selective B-cell defect.<br />
Subsequent development <strong>of</strong> a potent, selective p110 delta<br />
inhibitor prompted translation into the clinic for the treatment<br />
<strong>of</strong> CLL and low-grade non-Hodgkin lymphoma (NHL). From the<br />
first patient treated where a dramatic early nodal response<br />
was noted, considerable excitement has developed for this<br />
class <strong>of</strong> drugs in CLL and NHL. We will summarize the<br />
development process <strong>of</strong> CAL-101 (now GS1101) in the treatment<br />
<strong>of</strong> chronic lymphoid malignancies such as CLL.<br />
The expression <strong>of</strong> PI3K-delta is generally restricted to hematopoietic<br />
cells. 9 Mice with deleted or kinase dead PI3Kdelta<br />
exhibit a B-cell defect, with a lack <strong>of</strong> B1 lymphocytes,<br />
decreased mature B-cell numbers, and impaired antibody<br />
production. 3,5,10 Biochemically, B cells derived from PI3Kdelta<br />
knock-out mice also demonstrate less AKT phosphorylation<br />
when activated, decreased PIP 3 levels, and<br />
decreased ability to phosphorylate a YXXM phospho-specific<br />
peptide. 3 These mouse studies suggest that specific targeting<br />
<strong>of</strong> the PI3K-delta is<strong>of</strong>orm may be cytotoxic to B cells with<br />
minimal toxicity to other hematopoietic cell types. Forced<br />
expression <strong>of</strong> PI3K-delta was shown to be transforming in<br />
cell lines. 11 These collective findings provided a rationale<br />
to pursue is<strong>of</strong>orm-specific PI3K inhibitors preclinically and<br />
clinically in hematologic malignancies, including CLL. The<br />
acquisition <strong>of</strong> a platform <strong>of</strong> novel PI3K-delta inhibitors by<br />
Calistoga, a small biotechnology company with a very collaborative<br />
research team, prompted rapid exploitation <strong>of</strong><br />
this potential target in a variety <strong>of</strong> hematologic malignancies<br />
including CLL, which we will describe in this review.<br />
Preclinical Targeting <strong>of</strong> the PI3K-delta<br />
Pathway in CLL<br />
With the goal <strong>of</strong> determining the feasibility <strong>of</strong> targeting<br />
the is<strong>of</strong>orm PI3K-delta in CLL, our group collaboratively<br />
worked with Calistoga to initially examine 20 patients with<br />
CLL for expression <strong>of</strong> the different PI3K is<strong>of</strong>orms. In these<br />
patient samples, abundant PI3K-delta and PI3K-gamma<br />
protein expression was found, but very modest PI3K-alpha<br />
and essentially no PI3K-beta. 12 Beyond this work, we confirmed<br />
a previous observation that CLL cells have increased<br />
resting activity <strong>of</strong> PI3K activity as compared to normal B<br />
cells. This finding was concurrently corroborated by Lan-<br />
From the Department <strong>of</strong> Internal Medicine, Division <strong>of</strong> Hematology, the Comprehensive<br />
Cancer Center at The Ohio State University, Columbus, OH; Division <strong>of</strong> Medicinal<br />
Chemistry, College <strong>of</strong> Pharmacy, and the Comprehensive Cancer Center at The Ohio State<br />
University, Columbus, OH.<br />
Authors’ disclosures <strong>of</strong> potential conflicts <strong>of</strong> interest are found at the end <strong>of</strong> this article.<br />
Address reprint requests to John C. Byrd, MD, OSUCCC Building, Room 445B, 410 West<br />
12th Ave., Columbus, OH 43210; email: john.byrd@osumc.edu.<br />
© <strong>2012</strong> by <strong>American</strong> <strong>Society</strong> <strong>of</strong> <strong>Clinical</strong> <strong>Oncology</strong>.<br />
1092-9118/10/1-10<br />
691
nutti and colleagues, who demonstrated downstream constitutive<br />
phosphorylation <strong>of</strong> Threonine-308 on AKT, suggesting<br />
activated PDK1 and PI3K in CLL cells as compared to<br />
normal cells. 13 Collectively, these findings demonstrated<br />
that CLL cells may depend on PI3K-delta for constitutive<br />
AKT activation and survival. To further confirm this finding,<br />
small interfering RNA targeting p110-delta was performed<br />
in CLL cells demonstrating knock-down <strong>of</strong> target<br />
protein, diminished AKT Thr-308 phosphorylation, and apoptosis.<br />
These findings provided rationale for further targeting<br />
<strong>of</strong> CLL with a potential is<strong>of</strong>orm-specific inhibitor <strong>of</strong><br />
PI3K.<br />
CAL-101 was an example <strong>of</strong> a PI3K-delta-specific inhibitor<br />
that is 100-fold more selective for PI3K-delta compared<br />
to other PI3K is<strong>of</strong>orms. 13 Whole blood studies in basophils<br />
demonstrated appropriate inhibition <strong>of</strong> PI3K-delta at 50 �M<br />
concentrations, whereas PI3K-gamma-related targets were<br />
not inhibited until �M concentrations were attained. Furthermore,<br />
a kinase screen demonstrated significant selectivity<br />
for PI3K. The selectivity for PI3K-delta and other<br />
favorable drug properties prompted the choice <strong>of</strong> CAL-101 as<br />
a lead candidate for clinical development in hematologic<br />
malignancies and initiation <strong>of</strong> preclinical studies with this<br />
agent. Here we found that CAL-101 exhibits dose-dependent<br />
induction <strong>of</strong> apoptosis in CLL tumor cells that is independent<br />
<strong>of</strong> patient genomic features, including immunoglobulin<br />
gene variable heavy gene (IVGH) mutational status and<br />
del(17p13.1). 12 In contrast, CAL-101 did not promote cytotoxicity<br />
toward T cells or natural killer cells. We next<br />
confirmed the effect <strong>of</strong> microenvironment stimuli on CLL<br />
cells, including that soluble factors such as CD40L, BAFF,<br />
and TNF-� and contact factors (fibronectin and stromal cell<br />
coculture) all can activate PI3K and downstream AKT,<br />
thereby promoting CLL survival. 12 However, treatment<br />
with CAL-101 in the context <strong>of</strong> any <strong>of</strong> these microenvironment<br />
protective features was shown to abrogate the AKT<br />
phosphorylation and protection from cell death. CAL-101<br />
modulation <strong>of</strong> microenvironment protection was confirmed<br />
later by the Burger laboratory that also demonstrated disrupted<br />
CLL cell migration, adhesion, and diminished CLL<br />
production <strong>of</strong> chemokines that recruit immune cells to tumor<br />
sites. 14 Similar findings in mantle cell lymphoma and other<br />
types <strong>of</strong> B-cell lymphoma, 13 as well as multiple myeloma, 15<br />
were also shown. Collectively, these in vitro studies provided<br />
a strong rationale for moving forward to the clinic with<br />
CAL-101 with particular focus on targeting CLL and other<br />
types <strong>of</strong> B-cell lymphomas as the lead study diseases.<br />
KEY POINTS<br />
● PI3-kinase-delta in genetic knock-out models has<br />
predominately a B-cell phenotype.<br />
● GS1101 is a PI3-kinase-delta-specific orally administered<br />
agent in clinical trials for chronic lymphocytic<br />
leukemia and nonHodgkin lymphoma.<br />
● GS1101 demonstrates a unique pattern <strong>of</strong> response<br />
with early lymphocytosis concomitant with reduction<br />
in lymph-node and spleen disease.<br />
● GS1101 has a favorable side-effect pr<strong>of</strong>ile with a<br />
dose-limiting side effect <strong>of</strong> transient transaminitis.<br />
692<br />
Additionally, these preclinical studies provided for biomarkers<br />
<strong>of</strong> target inhibition such as AKT Thr308 phosphorylation<br />
loss by flow cytometry, cytokine production by tumor and<br />
normal immune cells, and also chemokine production by<br />
tumor cells.<br />
CAL-101: Early <strong>Clinical</strong> Development<br />
The selectivity <strong>of</strong> CAL-101 and also a very favorable<br />
toxicity pr<strong>of</strong>ile observed in preclinical toxicology at doses<br />
where PI3K-delta was inhibited allowed an initial study in<br />
healthy volunteers. This phase I study demonstrated favorable<br />
pharmacokinetics during a week <strong>of</strong> CAL-101 oral administration<br />
with absent lymphocyte count changes or<br />
notable toxicity. A phase I study <strong>of</strong> CAL-101 in select<br />
lymphoid malignancies was initiated in July 2008 beginning<br />
at a 50-mg once daily dosage. Unlike many phase I studies<br />
that struggle through enrollment with only modest efficacy,<br />
dramatic reduction in tumor size was noted in the first<br />
patient at the 1-month response evaluation. Accrual to this<br />
phase I trial was quite rapid, with efficient data communication<br />
and extensive investigator involvement typical <strong>of</strong><br />
early studies performed with small biopharmaceutical companies.<br />
Dose escalation proceeded through several dose<br />
levels with a loss <strong>of</strong> linear pharmacokinetics at higher doses<br />
examined. The most predominate toxicity observed with<br />
CAL-101 was reversible transaminitis early during therapy,<br />
which usually resolved with temporary discontinuation <strong>of</strong><br />
the drug. Reinitiation <strong>of</strong> CAL-101 at lower doses generally<br />
was possible without recurrence <strong>of</strong> transaminitis. Perplexing,<br />
however, was the finding that this toxicity occurred<br />
much more frequently in patients with lymphoma rather<br />
than with CLL. This liver toxicity and the nonlinear pharmacokinetics<br />
observed with CAL-101 required exploration <strong>of</strong><br />
multiple dose levels that ultimately delayed development <strong>of</strong><br />
this compound proceeding to phase II formal testing. Results<br />
for the CLL and low-grade lymphoma cohort were<br />
most promising, prompting focused expansion in each <strong>of</strong><br />
these groups. In contrast, single-agent activity in aggressive<br />
large cell lymphoma, refractory acute myeloid leukemia, and<br />
multiple myeloma was not observed.<br />
Disease-Specific Activity and Toxicity<br />
<strong>of</strong> CAL-101 in CLL<br />
BYRD, WOYACH, AND JOHNSON<br />
With respect to CLL and its exploration in the initial<br />
phase I study, a dose <strong>of</strong> 150 mg every 12 hours was identified<br />
to be ideal for the phase II dose in CLL based on tolerability<br />
and minimal increase in plasma C max at increasing doses. Of<br />
54 patients with CLL enrolled in this trial, 84% achieved a<br />
decrease in lymph node and spleen size <strong>of</strong> 50% or greater. An<br />
increase in peripheral lymphocyte count <strong>of</strong> more than 50%<br />
was seen in 58% <strong>of</strong> patients, consistent with other BCR<br />
antagonists. Lymphocytosis peaked at 2 months and resolved<br />
over time in a subset <strong>of</strong> patients. As a consequence <strong>of</strong><br />
the lymphocyte count not falling to 50% <strong>of</strong> baseline, response<br />
across all patients enrolled was 24%. This response was<br />
independent <strong>of</strong> high-risk genetic features, bulky adenopathy,<br />
prior therapy, or presence <strong>of</strong> cytopenias. Median<br />
progression-free survival was 15 months, with 46% <strong>of</strong> patients<br />
continuing on therapy at the time <strong>of</strong> presentation. 16<br />
Side effects with this agent have been mild and have<br />
included rare cytopenias and pneumonia. Approximately
THE STORY OF CAL-101<br />
6% <strong>of</strong> patients developed grade 3 or 4 transient liver function<br />
abnormalities during the early phase, which was reversible<br />
with holding therapy and generally did not recur<br />
with resumption at a lower dose level. Several <strong>of</strong> the pharmacodynamic<br />
studies optimized with the preclinical CAL-<br />
101 work confirmed target inhibition <strong>of</strong> PI3K-delta in vivo in<br />
patients receiving this agent, including serial loss <strong>of</strong> AKT-<br />
Thr308 phosphorylation, diminished chemokine, and also<br />
cytokine levels in plasma. These data collectively provide<br />
support that CAL-101 has significant clinical activity in CLL<br />
and can be administered as continuous therapy for an<br />
extended period <strong>of</strong> time with very modest toxicity.<br />
Based on the favorable toxicity observed with CAL-101<br />
monotherapy, target inhibition <strong>of</strong> PI3K-delta, and early<br />
lymphocytosis observed that was believed to be representative<br />
<strong>of</strong> the ability <strong>of</strong> CAL-101 to mobilize CLL cells from<br />
protected stromal cell niches, combination studies with<br />
other therapies were clearly justified. CAL-101 studies with<br />
either monoclonal antibodies (rituximab or <strong>of</strong>atumumab) or<br />
chemotherapy agents (bendamustine, bendamustine and<br />
rituximab, or fludarabine) in the phase I setting have been<br />
pursued. In these studies, CAL-101 was dosed at 100 mg<br />
twice daily or 150 mg twice daily with other drugs administered<br />
as standard. Therapy was well-tolerated, with no<br />
toxicities in addition to those expected with the single<br />
agents. Response data have been presented for the CAL-101<br />
studies combined with rituximab, bendamustine, and bendamustine<br />
and rituximab. For patients receiving CAL-101<br />
with rituximab or bendamustine, 90% or more <strong>of</strong> patients<br />
achieved a reduction in lymph node size <strong>of</strong> 50% or more. For<br />
three patients treated with bendamustine, rituximab, and<br />
CAL-101, all patients achieved a lymph node response. 17<br />
Using traditional IWCLL 2008 CLL response criteria, more<br />
than 80% <strong>of</strong> patients receiving each <strong>of</strong> these three regimens<br />
met criteria for response. Collectively, these data provide<br />
evidence that CAL-101 can be safely combined with other<br />
therapies used in CLL and also can contribute to durable<br />
remissions with these agents.<br />
Authors’ Disclosures <strong>of</strong> Potential Conflicts <strong>of</strong> Interest<br />
Author<br />
John C. Byrd*<br />
Jennifer A. Woyach*<br />
Amy J. Johnson*<br />
*No relevant relationships to disclose.<br />
Employment or<br />
Leadership<br />
Positions<br />
Consultant or<br />
Advisory Role<br />
1. Cantrell DA. Phosphoinositide 3-kinase signalling pathways. J Cell Sci.<br />
2001;114:1439-1445.<br />
2. Wymann MP, Zvelebil M, Laffargue M. Phosphoinositide 3-kinase signalling—which<br />
way to target? Trends Pharmacol Sci. 2003;24:366-376.<br />
3. Vanhaesebroeck B, Ali K, Bilancio A, et al. Signalling by PI3K is<strong>of</strong>orms:<br />
insights from gene-targeted mice. Trends Biochem Sci. 2005;30:194-204.<br />
4. Xu GB, Lu XG, Luo LH, et al. Detection <strong>of</strong> soluble Apo-1/Fas in plasma,<br />
pleural and ascites fluid <strong>of</strong> malignant tumor patients and its clinical significance<br />
(in Chinese). Zhejiang Da Xue Xue Bao Yi Xue Ban. 2003;32:335-338.<br />
5. Jou ST, Carpino N, Takahashi Y, et al. Essential, nonredundant role for<br />
the phosphoinositide 3-kinase p110delta in signaling by the B-cell receptor<br />
complex. Mol Cell Biol. 2002;22:8580-8591.<br />
6. Fruman DA. Towards an understanding <strong>of</strong> is<strong>of</strong>orm specificity in phosphoinositide<br />
3-kinase signalling in lymphocytes. Biochem Soc Trans. 2004;<br />
32:315-319.<br />
Future Directions for CAL-101 and Other PI3K<br />
Inhibitors in CLL<br />
The documented success <strong>of</strong> CAL-101 has prompted transition<br />
<strong>of</strong> the compound from Calistoga, a very small biotechnology<br />
company, to Gilead, a much larger pharmaceutical<br />
company. This transition resulted in both a renaming <strong>of</strong><br />
CAL-101 to GS1101 and also the initiation <strong>of</strong> more definitive<br />
registration studies for eventual marketing approval in CLL<br />
and low-grade NHL. As an expected consequence <strong>of</strong> the<br />
success <strong>of</strong> CAL-101, other PI3K-delta specific inhibitors are<br />
now being explored in CLL and related B-cell malignancies.<br />
Additionally, trials with more broad PI3K inhibitors have<br />
also been initiated with documented clinical activity. In<br />
particular, PI3K inhibitors targeting the PI3K-alpha is<strong>of</strong>orm<br />
may prove worthwhile in CLL. Burger and colleagues<br />
have previously demonstrated that PI3K-alpha antagonists<br />
abrogate stromal microenvironment protection against<br />
chemotherapy-mediated death. 18 Additionally, the success<br />
<strong>of</strong> CAL-101 has brought forth inhibitors <strong>of</strong> other kinases<br />
downstream in the BCR signaling pathway including BTK,<br />
which have also demonstrated dramatic clinical responses.<br />
At this point, the field <strong>of</strong> CLL has been mesmerized by the<br />
durable activity <strong>of</strong> these BCR antagonizing agents and also<br />
the true potential <strong>of</strong> these therapeutics to favorably affect<br />
the natural history <strong>of</strong> CLL.<br />
Acknowledgements<br />
The authors wish to acknowledge the employees <strong>of</strong> Calistoga<br />
Pharmaceuticals intricately involved in CAL-101 development<br />
(Neill Giess, PhD, Brian Lannutti, PhD, Carol Gallagher, Albert<br />
Yu, MD, David Johnson, and Langdon Miller, MD); the early<br />
phase I investigators on this trial (Ian Flinn, MD, Brad Kahl,<br />
MD, Richard Furman, MD, Jennifer Brown, MD, and Nancy<br />
Bartlett, MD); and most importantly, the patients who enrolled<br />
in these trials. We are grateful for research support that has<br />
made possible our BCR kinase work from The Leukemia and<br />
Lymphoma <strong>Society</strong>, the National Cancer Institute (P50<br />
CA140158, PO1 CA95426, PO1 CA81534, 1K12 CA133250), Mr.<br />
and Mrs. Michael Thomas, The Harry Mangurian Foundation,<br />
and The D. Warren Brown Foundation.<br />
Stock<br />
Ownership Honoraria<br />
REFERENCES<br />
Research<br />
Funding<br />
Expert<br />
Testimony<br />
Other<br />
Remuneration<br />
7. Papakonstanti EA, et al. Distinct roles <strong>of</strong> class IA PI3K is<strong>of</strong>orms in<br />
primary and immortalised macrophages. J Cell Sci. 2008;121:4124-4133.<br />
8. Min SH, Abrams CS. Why do phosphatidylinositol kinases have so many<br />
is<strong>of</strong>orms? Biochem J. 2009;423:e5-e8.<br />
9. Vanhaesebroeck B, Welham MJ, Kotani K, et al. P110delta, a novel<br />
phosphoinositide 3-kinase in leukocytes. Proc Natl Acad Sci USA.1997;94:<br />
4330-4335.<br />
10. Okkenhaug K, Bilancio A, Farjot G, et al. Impaired B and T cell antigen<br />
receptor signaling in p110delta PI 3-kinase mutant mice. Science. 2002;297:<br />
1031-1034.<br />
11. Kang S, Denley A, Vanhaesebroeck B, et al. Oncogenic transformation<br />
induced by the p110beta, -gamma, and -delta is<strong>of</strong>orms <strong>of</strong> class I phosphoinositide<br />
3-kinase. Proc Natl Acad Sci USA.2006;103:1289-1294.<br />
12. Herman SE, Gordon AL, Wagner AJ, et al. Phosphatidylinositol<br />
3-kinase-delta inhibitor CAL-101 shows promising preclinical activity in<br />
693
chronic lymphocytic leukemia by antagonizing intrinsic and extrinsic cellular<br />
survival signals. Blood. 2010;116:2078-2088.<br />
13. Lannutti BJ, Meadows SA, Herman SE, et al. CAL-101, a p110delta<br />
selective phosphatidylinositol-3-kinase inhibitor for the treatment <strong>of</strong> B-cell<br />
malignancies, inhibits PI3K signaling and cellular viability. Blood. 2011;117:<br />
591-594.<br />
14. Hoellenriegel J, Meadows SA, Sivina M, et al. The phosphoinositide<br />
3�-kinase delta inhibitor, CAL-101, inhibits B-cell receptor signaling and chemokine<br />
networks in chronic lymphocytic leukemia. Blood. 2011;118:3603-3612.<br />
15. Ikeda H, Hideshima T, Fulciniti M, et al. PI3K/p110{delta} is a novel<br />
therapeutic target in multiple myeloma. Blood. 2010;116:1460-1468.<br />
16. Coutre SE, et al. Phase I study <strong>of</strong> CAL-101, an is<strong>of</strong>orm-selective<br />
694<br />
BYRD, WOYACH, AND JOHNSON<br />
inhibitor <strong>of</strong> phosphatidylinositol 3-kinase P110d, in patients with previously<br />
treated chronic lymphocytic leukemia. J Clin Oncol. 2011;29,451s (suppl;<br />
abstr 6631).<br />
17. Sharman J, de Vos S, Leonard JP, et al. A phase 1 study <strong>of</strong> the selective<br />
phosphatidylinositol 3-kinase-delta (PI3Kdelta) inhibitor, CAL-101 (GS-<br />
1101), in combination with rituximab and/or bendamustine in patients with<br />
relapsed or refractory chronic lymphocytic leukemia (CLL). Blood (ASH<br />
Annual Meeting Abstracts) 2011;118,?? (abstr 1787).<br />
18. Niedermeier M, Hennessy BT, Knight ZA, et al. Is<strong>of</strong>orm-selective<br />
phosphoinositide 3�-kinase inhibitors inhibit CXCR4 signaling and overcome<br />
stromal cell-mediated drug resistance in chronic lymphocytic leukemia: a<br />
novel therapeutic approach. Blood. 2009;113:5549-5557.
PI3 Kinase in Cancer: From Biology to Clinic<br />
Overview: The discovery and clinical development <strong>of</strong> smallmolecule<br />
inhibitors <strong>of</strong> the phosphatidylinositide 3-kinase (PI3<br />
kinase) family <strong>of</strong> lipid kinases have marked a remarkable<br />
20-year journey that follows the progressive developments in<br />
cancer biology over the last few decades: from hypothesisdriven,<br />
basic cancer research that began with viral oncogenesis<br />
and developed in the 1960s and 70s, through the<br />
discovery <strong>of</strong> individual mutated oncogenes and tumor suppressor<br />
genes in 1970 and 80s and the linkage <strong>of</strong> these cancer<br />
genes to signal transduction pathways in the 1990s, to all<br />
large-scale genome-wide sequencing, functional screening,<br />
and network biology efforts today. Thus, PI3 kinase research<br />
began with the discovery in 1985 <strong>of</strong> a new type <strong>of</strong> enzyme<br />
activity associated with viral oncogenesis. It benefited greatly<br />
from the discovery <strong>of</strong> wortmannin and LY294002 as PI3 kinase<br />
THE PI3 kinase family <strong>of</strong> lipid and protein kinases<br />
regulates an intracellular signaling network that controls<br />
many features <strong>of</strong> cell behavior, including growth,<br />
survival, motility, metabolism, and additional specialized<br />
functions (Fig. 1). There are four distinct PI3 kinase subfamilies<br />
that are categorized by their substrate specificities,<br />
primary structures, modes <strong>of</strong> regulation, and domain content.<br />
Of these, the class I is<strong>of</strong>orms (p110�, �, �, �) and class<br />
IV PI3 kinase-related protein kinase mTOR have been the<br />
most intensively explored as targets for small-molecule<br />
therapeutics. 1-3<br />
The rationale for the development <strong>of</strong> the class I PI3 kinase<br />
group <strong>of</strong> lipid kinases is the observation <strong>of</strong> frequent genetic<br />
and epigenetic alterations that result in activation <strong>of</strong> the PI3<br />
kinase pathway. 2,3 The class I PI3 kinases catalyze the<br />
addition <strong>of</strong> a phosphate group to the 3�-hydroxyl position <strong>of</strong><br />
the inositide ring present in membrane phosphatidylinositides.<br />
This produces products, the most notable <strong>of</strong> which is<br />
phosphatidylinositol-3,4,5-trisphosphate (PIP3), which acts<br />
as a second messenger that recruits PKB/AKT to the cell<br />
membrane (Fig. 1). The phosphatase PTEN, a negative<br />
regulator <strong>of</strong> PI3 kinase signaling that dephosphorylates<br />
PIP3, is one <strong>of</strong> the most commonly mutated tumorsuppressor<br />
proteins in human malignancy. 4 In contrast, the<br />
gene encoding PIK3CA, the p110� catalytic subunit, is<br />
amplified, overexpressed, and frequently mutated in many<br />
cancers. 5<br />
A greater understanding <strong>of</strong> the specific and geneticsdependent<br />
roles <strong>of</strong> the class I is<strong>of</strong>orms in tumorigenesis<br />
has been established. It has been shown that p110� is<br />
critical for the growth <strong>of</strong> tumors driven by PIK3CA mutations<br />
and is activated by KRAS and also by receptor tyrosine<br />
kinases that are in turn activated by ligand or by oncogenic<br />
mutation, amplification, or translocation. 2,3 In contrast,<br />
p110� has been identified as the principal is<strong>of</strong>orm mediating<br />
tumorigenesis in PTEN-deficient backgrounds. 6-8 The deltais<strong>of</strong>orm,<br />
p110�, has also emerged as a potential therapeutic<br />
target for hematological malignancies, notably acute myeloid<br />
leukemia, and perhaps in neuroblastoma, melanoma,<br />
and breast cancers also. 1,2,9,10 Finally, there is potential for<br />
all the class I PI3 kinases to be activated in cancer cells<br />
through mutation <strong>of</strong> the p85 regulatory subunits that recruit<br />
the catalytic subunit to growth factor receptors.<br />
By Paul Workman, PhD, and Paul Clarke, PhD<br />
inhibitors and chemical tools in late 1980s to mid-90s. Alongside<br />
these tools, genetic validation <strong>of</strong> PI3 kinase as a target<br />
initially involved activation by upstream oncogenic receptor<br />
tyrosine kinases and RAS mutation, together with overexpression<br />
and amplification <strong>of</strong> the p110� catalytic is<strong>of</strong>orm <strong>of</strong> PI3<br />
kinase and frequent loss <strong>of</strong> the tumor suppressor and negative<br />
regulator <strong>of</strong> PI3 kinase activity, PTEN. As PI3 kinase drug<br />
development began, further stimulus came from the discovery<br />
through genome sequencing <strong>of</strong> mutations in PIK3CA, which<br />
encodes p110� and is the most frequently mutated kinase in<br />
the human genome. From these beginnings, there are now<br />
many PI3 kinase inhibitors in clinical trials and more in<br />
preclinical development. We review progress, current challenges,<br />
and future opportunities in this article.<br />
Based on this broad platform <strong>of</strong> research validation,<br />
inhibition <strong>of</strong> the class I—and especially the class IA subclass<br />
(p110�, �, �)—has emerged as an important strategy for the<br />
development <strong>of</strong> novel molecular cancer therapeutics. Looking<br />
further ahead, PI3 kinase inhibitors are anticipated to<br />
have a significant impact on the discovery and development<br />
<strong>of</strong> new personalized medicines in the oncology setting as well<br />
as finding utility in other disease areas. 10<br />
Discovery <strong>of</strong> PI3 Kinase Inhibitors<br />
Since the elucidation <strong>of</strong> the mechanism <strong>of</strong> action <strong>of</strong> the<br />
natural product wortmannin and the discovery <strong>of</strong> the synthetic<br />
flavone LY294002, both <strong>of</strong> which inhibit the class I<br />
PI3 kinase is<strong>of</strong>orms and have served as valuable chemical<br />
tools, considerable progress has been made in the design <strong>of</strong> a<br />
plethora <strong>of</strong> small-molecule inhibitors. 10-12 High-throughput<br />
screening <strong>of</strong> compound collections supported by rational<br />
structure-based design combined with medicinal chemistry<br />
optimization has led to the discovery and development <strong>of</strong><br />
many chemically diverse inhibitors that possess a range <strong>of</strong><br />
PI3 kinase subtype–selectivity pr<strong>of</strong>iles.<br />
A number <strong>of</strong> these inhibitors, such as GDC-0941, GDC-<br />
0980, NVP-BEZ235, NVP-BMK120, GSK-2126458, GSK-<br />
1059615, PF-04691502, XL147, XL765, CAL-101, and<br />
derivatives <strong>of</strong> wortmannin and LY294002 (as reviewed by<br />
Shuttleworth and colleagues 10 ), have entered early-phase<br />
human trials over recent years and are in some cases<br />
progressing to phase Ib expansion cohort and phase II<br />
single-agent efficacy studies and early combination trials.<br />
A paradigm for the rational discovery and development <strong>of</strong><br />
a potent and selective PI3 kinase inhibitor is GDC-0941 that<br />
has a well-defined and informative case-history. 13 In 2007<br />
Hayakawa and colleagues reported the identification <strong>of</strong> a<br />
pyrid<strong>of</strong>uropyrimidine lead and improved chemical tool com-<br />
From the Cancer Research UK Cancer Therapeutics Unit, Division <strong>of</strong> Cancer Therapeutics,<br />
The Institute <strong>of</strong> Cancer Research, Sutton, Surrey UK.<br />
Authors’ disclosures <strong>of</strong> potential conflicts <strong>of</strong> interest are found at the end <strong>of</strong> this article.<br />
Address reprint requests to Paul Workman, PhD, Cancer Research UK Cancer Therapeutics<br />
Unit, The Institute <strong>of</strong> Cancer Research, Haddow Laboratories, 15 Cotswold Road,<br />
Sutton, Surrey, SM2 5NG UK; email: paul.workman@icr.ac.uk.<br />
© <strong>2012</strong> by <strong>American</strong> <strong>Society</strong> <strong>of</strong> <strong>Clinical</strong> <strong>Oncology</strong>.<br />
1092-9118/10/1–10<br />
e93
pound. 14 This chemical probe, PI-103, exhibited advantages<br />
over wortmannin and LY294002, with excellent potency and<br />
selectivity for class I PI3 kinases compared with a panel <strong>of</strong><br />
other protein and lipid kinases. PI-103 exhibited therapeutic<br />
activity against human tumor xenograft models representing<br />
various cancer types, including several with PIK3CA<br />
mutations or loss <strong>of</strong> PTEN, which are predicted to be<br />
“addicted” to the PI3 kinase pathway. 15 The antitumor<br />
activity <strong>of</strong> PI-103 was also associated with effects on pharmacodynamic<br />
pathway biomarkers, including reduced phosphorylation<br />
<strong>of</strong> downstream substrates such as AKT,<br />
providing evidence <strong>of</strong> target engagement. The biomarker<br />
results were fully consistent with the therapeutic mechanism<br />
being PI3 kinase inhibition, and, importantly, clearly<br />
demonstrating pro<strong>of</strong>-<strong>of</strong>-concept for targeting this pathway.<br />
At the same time, it was also recognized that PI-103 had<br />
some liabilities, specifically limited aqueous solubility and<br />
extensive metabolism, which were considered serious drawbacks<br />
in terms <strong>of</strong> its drug-like properties and thus its<br />
suitability for clinical development<br />
Subsequently, a large number <strong>of</strong> chemical analogs from<br />
two chemical starting points—the pyrid<strong>of</strong>uropyrimidines<br />
exemplified by PI-103 and the thienopyrimidine ‘compound<br />
12’—were designed, synthesized, and tested in an iterative<br />
fashion. 14,16 These integrated medicinal chemistry, and<br />
biogical evaluation efforts identified the more advanced<br />
thienopyrimidine lead compounds, PI-540 and PI-620, which<br />
retained the potency and selectivity <strong>of</strong> PI-103 but with much<br />
improved aqueous solubility and reduced in vitro metabolism.<br />
17 Although the improved absorbtion, distribution, metabolism,<br />
and excretion properties <strong>of</strong> these compounds<br />
resulted in correspondingly greater antitumor activity as<br />
compared with PI-103, in vivo metabolism caused by conjugation<br />
<strong>of</strong> a phenolic hydroxyl group remained an obstacle<br />
to the clinical development <strong>of</strong> this series. 17 Indazoles were<br />
investigated as replacements for the problematic phenolic<br />
hydroxyl, with the intention that they might markedly<br />
reduce metabolism while maintaining the necessary target<br />
interaction with PI3 kinase, as revealed by protein-ligand<br />
X-ray cocrystal structures (Fig. 2). This strategy paid <strong>of</strong>f and<br />
GDC-0941 (PI-728) was identified as a potent and selective<br />
KEY POINTS<br />
● The PI3K pathway is frequently activated by oncogenic<br />
mutations in many different cancers.<br />
● Inhibiting the PI3K pathway is an attractive approach<br />
to exploit the resulting oncogenic addictions<br />
and vulnerabilities.<br />
● A number <strong>of</strong> potent and selective PI3K inhibitors<br />
have been discovered and are now in clinical studies.<br />
● Although fit-for-purpose target engagement biomarkers<br />
are now available, identifying predictive biomarkers<br />
for patient stratification remains a major<br />
challenge; current “predictive” biomarkers are useful<br />
for enrichment <strong>of</strong> early clinical trials with patients<br />
more likely to respond.<br />
● Other key challenges being addressed in the clinic are<br />
to define optimal PI3 kinase selectivity pr<strong>of</strong>iles and to<br />
explore rational drug combinations.<br />
e94<br />
WORKMAN AND CLARKE<br />
inhibitor pan-class I PI3 kinase inhibitor with excellent<br />
drug-like properties, including good oral bioavailability, a<br />
major advantage over the previous analogs. 18 Interestingly,<br />
GDC-0941 retained full therapeutic activity, at least in the<br />
cancer models tested, despite lacking mTOR inhibition,<br />
which was in contrast to previous analogs such as PI-103. 17<br />
A subsequent structurally related clinical candidate GDC-<br />
0980 does inhibit mTOR, providing an interesting comparator.<br />
19<br />
The improved pharmacokinetic pr<strong>of</strong>ile <strong>of</strong> GDC-0941 resulted<br />
in pr<strong>of</strong>ound and prolonged target engagement, demonstrated<br />
by inhibition <strong>of</strong> PI3 kinase pathway biomarkers in<br />
human cancer xenografts, consistent with the sustained<br />
tumor drug exposure (Fig. 3). In these experiments, tumor<br />
drug levels were well above antiproliferative GI 50 concentrations<br />
for at least 6 hours, and decreased phosphorylation<br />
<strong>of</strong> AKT together with other pathway markers was maintained<br />
for at least 8 hours (Fig. 3). 17<br />
As a result <strong>of</strong> these enhanced pharmacokinetic and pharmacodynamic<br />
properties, GDC-0941 exhibited excellent<br />
dose-dependent therapeutic activity in the PTEN null, PI3<br />
kinase pathway-addicted U87MG human glioblastoma<br />
xenograft model (Fig. 3). Activity was also demonstrated in<br />
another “PI3 kinase-addicted” tumor, the IGROV-1 ovarian<br />
cancer carrying both an activating PIK3CA mutation and<br />
lacking PTEN expression and also in additional human<br />
tumor xenografts. 17,20<br />
The linkage <strong>of</strong> pharmacokinetic exposure, biomarker<br />
changes in the PI3 kinase pathway, and therapeutic tumor<br />
response provided a convincing pharmacologic audit trail<br />
(PhaT 21 ) for GDC-0941, which formed a solid platform for<br />
subsequent clinical studies. On the basis <strong>of</strong> its very attractive<br />
molecular, pharmacologic, and therapeutic pr<strong>of</strong>ile, together<br />
with its excellent selectivity pr<strong>of</strong>ile for class I PI3<br />
kinase compared with a wide range <strong>of</strong> other kinases, GDC-<br />
0941 was selected for clinical development. As mentioned<br />
earlier, there are now a considerable number <strong>of</strong> PI3 kinase<br />
inhibitors undergoing clinical trials and still more in preclinical<br />
discovery and development. 10<br />
Current Challenges and Future Opportunities<br />
It is important to assess the critical challenges that we<br />
face in the development <strong>of</strong> PI3 kinase inhibitors for cancer<br />
treatment. 22 A major question is certainly the desired specificity<br />
toward the various individual PI3 kinase family<br />
members. There has been some real progress, particularly in<br />
designing and evaluating inhibitors with different is<strong>of</strong>orm<br />
selectivities, <strong>of</strong>ten facilitated by structure-based design (Fig.<br />
2). However, the definition <strong>of</strong> the optimal PI3 kinase is<strong>of</strong>orm<br />
selectivity pr<strong>of</strong>ile or pr<strong>of</strong>iles—which may well depend on<br />
tumor genetics—is still a major ongoing research activity.<br />
For example, we remain unsure about whether the addition<br />
<strong>of</strong> mTOR inhibitory activity to a class I PI3 kinase inhibitor<br />
activity is advantageous, e.g., for overcoming feedback loops.<br />
In addition, it is also not yet clearly established whether<br />
individual compared with combinatorial inhibition <strong>of</strong> class I<br />
is<strong>of</strong>orms is advantageous—especially given that most clinical<br />
candidates inhibit all 4 class I is<strong>of</strong>orms with low nanomolar<br />
potency and appear to be generally well tolerated,<br />
at least in the short to medium term. The preferred pr<strong>of</strong>iles<br />
will emerge with time from clinical experience coupled with<br />
mechanistic preclinical studies.<br />
The issue <strong>of</strong> biomarkers is a critically important one. The
PI3 KINASE IN CANCER<br />
Fig. 1. The PI3 kinase network. Class I PI3 kinases are activated to varying extents by receptor tyrosine kinases (RTKs) and G-protein coupled<br />
receptors (GPCRs). PI3 kinases phosphorylate the 3�-hydroxy position <strong>of</strong> inositol ring <strong>of</strong> phosphatidylinositides, yielding lipid products that<br />
include phosphatidylinositol-3,4,5-trisphosphate (PIP3), generated from phosphatidylinositol-4,5-bisphosphate (PIP2). PIP3 is a second messenger<br />
molecule that recruits certain protein kinases with PH domains, such as PDK1 and AKT, to the cell membrane (dashed line), resulting in<br />
their activation and stimulation <strong>of</strong> the PI3 kinase signaling network. PI3 kinase may promote cancer through both AKT-dependent and<br />
AKT-independent mechanisms, the latter via PDK1 and the serine/threonine-protein kinase 3 (SGK3). 31 Activation <strong>of</strong> the PI3 kinase pathway is<br />
very common in a wide range <strong>of</strong> cancers. PIK3CA, which encodes the p110� catalytic subunit <strong>of</strong> PI3 kinase, appears to be the most commonly<br />
mutated kinase in the entire human cancer genome (12% <strong>of</strong> all cancers) and is also amplified in some tumors. PTEN, which encodes the opposing<br />
phosphatase to PI3 kinase, is the second most commonly affected tumor-suppressor gene after p53. 32 Activation <strong>of</strong> PI3 kinase-signaling in<br />
cancer also occurs as a result <strong>of</strong> mutation or increased expression <strong>of</strong> RTKs, AKT, and RAS. Activation <strong>of</strong> PI3 kinase signaling has been shown to<br />
reduce cellular dependence on growth factors, attenuate apoptosis, and facilitate tumor growth and invasiveness. Class I PI3 kinases are shown<br />
here as molecular targets <strong>of</strong> PI3 kinase inhibitors. Figure adapted with permission from Clarke and Workman. © <strong>2012</strong> <strong>American</strong> <strong>Society</strong> <strong>of</strong><br />
<strong>Clinical</strong> <strong>Oncology</strong>. All rights reserved. 22<br />
use <strong>of</strong> pro<strong>of</strong>-<strong>of</strong>-mechanism pharmacodynamic biomarkers to<br />
demonstrate target and pathway modulation in both the<br />
preclinical discovery phase and the early clinical development<br />
<strong>of</strong> PI3 kinase inhibitors is crucial for the implementation<br />
<strong>of</strong> the PhaT. 21 This enables rational optimization <strong>of</strong><br />
dose and schedule <strong>of</strong> administration as well as go/no-go<br />
decision making. 23 Available pro<strong>of</strong>-<strong>of</strong>-mechanism biomarkers<br />
<strong>of</strong> signaling output, such as phosphorylation <strong>of</strong> ribosomal<br />
protein S6, PRAS40, 4EBP1, and AKT have been used as<br />
indicators <strong>of</strong> PI3 kinase-pathway inhibition in both tumor<br />
and surrogate tissues and certainly appear to be fit for<br />
purpose. 10<br />
In contrast, we do not yet have biomarkers that are<br />
robustly predictive for sensitivity and resistance in the<br />
clinic. Several nonclinical studies have defined intrinsic<br />
factors such as HER2 amplification, PIK3CA mutations, or<br />
combined defects in these genes as factors that may influence<br />
the sensitivity <strong>of</strong> breast tumors to PI3 kinase inhibitors.<br />
10 On the other hand, the presence <strong>of</strong> a KRAS mutation<br />
may be a dominant determinant <strong>of</strong> resistance in other types<br />
<strong>of</strong> cancer. 10,13 The currently available therapeutic response<br />
biomarkers are probably best viewed as enrichment biomarkers,<br />
which can be used in early clinical trials for<br />
selecting patients with tumors having molecular characteristics<br />
that make them somewhat more likely to respond.<br />
It is clear that much more work needs to be done to<br />
identify, validate, and clinically qualify biomarkers that<br />
may be truly predictive. Additional biomarkers may well be<br />
needed for drugs with different PI3 kinase-selectivity pr<strong>of</strong>iles.<br />
For example, tumor cells harboring activating mutations<br />
in the PIK3CA gene are potentially dependent on or<br />
“addicted to” this is<strong>of</strong>orm, whereas cancer cells deficient in<br />
PTEN expression can exhibit a dependency on p110�. 1<br />
As a further complication, it should also be considered<br />
that in addition to their direct therapeutic action on cancer<br />
cells, PI3 kinase inhibitors can exert effects on tumor angiogenesis,<br />
immune cells, and other tumor microenvironmental<br />
interactions. Hence, it is quite conceivable that there may<br />
e95
Fig. 2. Structural biology <strong>of</strong> PI3 kinase and inhibitors. The solution <strong>of</strong> the X-ray crystal structures <strong>of</strong> PI3 kinases in complex with ATP or<br />
inhibitors acting at the ATP site has facilitated the understanding <strong>of</strong> their potency and selectivity and educated the prospective design <strong>of</strong> new<br />
ones. a) Ribbon diagram <strong>of</strong> human p110� (protein data bank [PDB] code 3dbs) illustrating the five-domain structure <strong>of</strong> the enzyme. The<br />
Ras-binding domain (RBD) domain is shown in red, the C2 domain is depicted in yellow, the helical domain is illustrated in green, and the bilobal<br />
catalytic domain is displayed in purple. The N-terminal adaptor binding domain (ABD) preceding the RBD-domain is shown in orange. The<br />
ATP-binding site is located in between the N- and C-terminal lobes <strong>of</strong> the catalytic domain, as indicated by the arrow. b) Binding <strong>of</strong> ATP to porcine<br />
p110� (PDB code 1e8�), showing the hydrogen bond interactions <strong>of</strong> ATP with the hinge region. The protein is shown in pink and ATP is displayed<br />
in light blue. The hydrogen bonding interactions are shown as dotted lines. Shown in gray are the two metal ions coordinating the phosphate<br />
groups <strong>of</strong> ATP. c) Binding <strong>of</strong> LY294002 to porcine p110� (PDB code 1e7v) illustrates the crucial hydrogen bonding interaction (dotted line)<br />
between the oxygen <strong>of</strong> morpholino group <strong>of</strong> this weakly potent and relatively unselective inhibitor and the PI3 kinase hinge region. The protein<br />
is shown in blue with key PI3 kinase residues displayed and labeled. LY294002 is shown in orange. Despite the fact that LY294002 extends into<br />
the affinity pocket (right-hand side), it does not fill the pocket so efficiently as does the indazole portion <strong>of</strong> the GDC-0941 molecule (see d).<br />
d) Binding <strong>of</strong> GDC-0941 to human p110� showing the hydrogen bond interactions (dotted lines) that anchor into the inhibitor deeply in the<br />
ATP-site. p110� is displayed in purple, and GDC-0941 is depicted in light green. e) Ribbon diagram <strong>of</strong> the p110�/p85� structure, showing p110�<br />
in yellow and the p85 niSH2 domain in blue (PDB code 2rd0). The ATP from the ATP-bound p110� structure superimposed on the p110�/p85�<br />
complex is shown. ATP is depicted in cylinder representation with its semi-transparent molecular surface superimposed in blue. The oncogenic<br />
mutation hotspots Glu 545 and His1047 are highlighted in red. f) Superposition <strong>of</strong> the GDC-0941-bound p110� structure and the human<br />
p110�/p85� structure. p110� is displayed with its solvent-accessible surface colored yellow and GDC-0941 molecule is shown with its<br />
semi-transparent surface in light green. It can be seen that based on this superposition, the GDC-0941 fits extremely well into the p110� ATP site,<br />
showing only a minor clash <strong>of</strong> the indazole ring with the wall <strong>of</strong> the ATP site (right-hand side). Figure reprinted with permission from Workman<br />
and colleagues 2010. 13<br />
not be a single biomarker <strong>of</strong> sensitivity but instead a<br />
predictive molecular signature, perhaps involving a geneexpression<br />
pr<strong>of</strong>ile. Studies in preclinical systems, including<br />
large molecularly characterized cancer-cell panels 24-26 and<br />
human tumor xenografts, together with genetically engineered<br />
mouse models, may contribute to the definition <strong>of</strong><br />
candidate biomarkers. However, it is likely that clinically<br />
effective predictive biomarkers will emerge from molecularpathology<br />
pr<strong>of</strong>iling <strong>of</strong> clinical tumor material—including<br />
global cancer genome sequencing, gene expression analysis,<br />
e96<br />
WORKMAN AND CLARKE<br />
and proteomics—and the correlation <strong>of</strong> these with therapeutic<br />
response and outcome. In addition, continued longitudinal<br />
pr<strong>of</strong>iling will also be essential to understand the basis for<br />
mechanisms <strong>of</strong> drug resistance that, based on experience<br />
with protein kinase inhibitors, is likely to develop with<br />
prolonged exposure to inhibitors.<br />
Another major challenge that we now face is the identification<br />
and development <strong>of</strong> rational mechanism-based combinatorial<br />
treatments for optimal therapeutic efficacy with<br />
PI3 kinase inhibitors. In general, PI3 kinase inhibitors are
PI3 KINASE IN CANCER<br />
Fig. 3. In vivo oral antitumor activity<br />
and associated pharmacokinetic and pharmacodynamic<br />
properties <strong>of</strong> GDC-0941 in<br />
the PTEN null U87MG human glioblastoma<br />
xenograft model. a) Chemical structure <strong>of</strong><br />
GDC-0941. b&c)Relative mean tumor volume<br />
(percentage <strong>of</strong> initial volume before<br />
therapy) and mean final tumor weights following<br />
doses <strong>of</strong> 25, 50, 100, and 150<br />
mg/kg daily oral doses <strong>of</strong> GDC-0941 administered<br />
for 19 days. Tumors were<br />
grown subcutaneously bilaterally in female<br />
NCr athymic mice, and controls received<br />
equivalent volumes <strong>of</strong> drug vehicle.<br />
d) Tumors from the same efficacy study<br />
were taken 4 and 8 hours following the<br />
final dose. AKT SER473 phosphorylation and<br />
total AKT were measured by a quantitative<br />
electrochemiluminescent immunoassay.<br />
For therapy data, results are mean<br />
standard error (SE) <strong>of</strong> 16 mice. For pharmacodynamic<br />
target engagement biomarker<br />
data, results are mean SE <strong>of</strong> 6 mice.<br />
Adapted with permission from Raynaud<br />
and colleagues 2009. 17<br />
not strongly proapoptotic but rather induce a strong and<br />
durable effect on cell-cycle progression. 15,24 More pr<strong>of</strong>ound<br />
therapeutic activity is likely to require combinatorial treatments<br />
to facilitate the death <strong>of</strong> cancer cells. Many combination<br />
studies are underway, but because <strong>of</strong> the sheer number<br />
<strong>of</strong> potential combinations, it will certainly take some considerable<br />
time to identify optimal combinations; both preclinical<br />
and clinical studies will contribute to this. Certain<br />
rationally based combinations, for example, using PI3 kinase<br />
together with MEK inhibitors that block the important<br />
RAS-RAF-MEK MAP kinase pathway, have obvious mechanistic<br />
rationale and these are being given a high priority in<br />
clinical studies. 25 Both “horizontal” pathway targeting and<br />
also “vertical” targeting within the PI3 kinase pathway are<br />
being investigated. Another interesting opportunity is the<br />
combination <strong>of</strong> PI3 kinase inhibitors with agents targeted to<br />
the androgen receptor in prostate cancer. 29 In addition to<br />
combinations with other molecularly targeted agents, use <strong>of</strong><br />
PI3 kinase inhibitors together with cytotoxic agents is also<br />
being pursued.<br />
The picture that we currently see emerging from the clinic<br />
is that PI3 kinase inhibitors are generally well-tolerated<br />
agents and show promising signs <strong>of</strong> biologic and clinical<br />
activity. Concerns about potential effects on glucose metab-<br />
Authors’ Disclosures <strong>of</strong> Potential Conflicts <strong>of</strong> Interest<br />
Author<br />
Employment or<br />
Leadership<br />
Positions<br />
Consultant or<br />
Advisory Role<br />
Paul Workman Chroma<br />
Therapeutics;<br />
Nextech Invest;<br />
Piramed (Roche);<br />
Wilex<br />
olism appear to have been alleviated, with only mild and<br />
manageable changes being seen, at least with the doses and<br />
schedules used to date. 27 The main side effects commonly<br />
include skin rash and urticaria. Encouragingly, there is<br />
early evidence <strong>of</strong> single agent therapeutic activity in the<br />
current clinical trials in patients with cancer, 30 although<br />
stable disease is <strong>of</strong>ten seen and it may require combination<br />
studies to reveal the full therapeutic potential.<br />
Concluding Remarks<br />
The genetic validation provided by the frequency <strong>of</strong> PI3<br />
kinase pathway deregulation in human cancers and evidence<br />
from mouse models encourages continued efforts to<br />
pursue inhibitors <strong>of</strong> PI3 kinases. The dramatic increase in<br />
the number <strong>of</strong> PI3 kinase inhibitors that display fascinatingly<br />
distinct selectivity pr<strong>of</strong>iles now provides us with many<br />
powerful tools with which to further investigate disease<br />
mechanisms and assess therapeutic potential in different<br />
pathogenetic contexts. Continued efforts to define clinically<br />
useful predictive biomarkers for patient stratification, together<br />
with rational drug combination studies, will allow us<br />
to uncover the full clinical potential <strong>of</strong> PI3 kinase inhibitors<br />
in cancer treatment.<br />
Stock<br />
Ownership Honoraria<br />
Research<br />
Funding<br />
Piramed (Roche) Piramed (Roche);<br />
Yamanouchi<br />
Expert<br />
Testimony<br />
Other<br />
Remuneration<br />
Paul Clarke Piramed Pharma Institute <strong>of</strong><br />
Cancer<br />
Research<br />
e97
1. Vanhaesebroeck B, Guillermet-Guibert J, Graupera M, et al. The emerging<br />
mechanisms <strong>of</strong> is<strong>of</strong>orm-specific PI3K signalling. Nat Rev Mol Cell Biol.<br />
2010;11:329-341.<br />
2. Yuan TL, Cantley LC. PI3K pathway alterations in cancer: variations on<br />
a theme. Oncogene. 2008;27:5497-5510.<br />
3. Liu P, Cheng H, Roberts TM, et al. Targeting the phosphoinositide<br />
3-kinase pathway in cancer. Nat Rev Drug Discov. 2009;8:627-644.<br />
4. Chalhoub N, Baker SJ. PTEN and the PI3-Kinase Pathway in Cancer.<br />
Annu Rev Pathol Mech Disease. 2009;4:127-150.<br />
5. Samuels Y, Diaz LA Jr., Schmidt-Kittler O, et al. Mutant PIK3CA<br />
promotes cell growth and invasion <strong>of</strong> human cancer cells. Cancer Cell.<br />
2005;7:561-573.<br />
6. Vogt P, Gymnopolous M, Hart JR. PI 3-kinase and cancer: changing<br />
accents. Curr Opin Genetics and Dev. 2009;19:1-6.<br />
7. Wee S, Wiederschain D, Maira SM, et al. PTEN-deficient cancers depend<br />
on PIK3CB. Proc Natl Acad Sci USA.2008;105:13057-13062.<br />
8. Jia S, Roberts TM, Zhao JJ. Should individual PI3 kinase is<strong>of</strong>orms be<br />
targeted in cancer? Curr Opin Cell Biol. 2009;21:199-208.<br />
9. Billottet C, Grandage VL, Gale RE, et al. A selective inhibitor <strong>of</strong> the<br />
p110� is<strong>of</strong>orm <strong>of</strong> PI3-kinase inhibits AML cell proliferation and survival and<br />
increases the cytotoxic effects <strong>of</strong> VP16. Oncogene. 2006;25:6648-6659.<br />
10. Shuttleworth SJ, Silva FA, Cecil AR, et al. Progress in the preclinical<br />
discovery and clinical development <strong>of</strong> class I and dual class I/IV phosphoinositide<br />
3-kinase (PI3K) inhibitors. Curr Med Chem. 2011;18:2686-2714.<br />
11. Arcaro A, Wymann MP: Wortmannin is a potent phosphatidylinositol<br />
3-kinase inhibitor. The role <strong>of</strong> phosphatidylinositol 3,4,5-trisphosphate in<br />
neutrophil response. Biochem J. 1993;296:297-301.<br />
12. Vlahos CJ, Matter WF, Hui KY, et al. A specific inhibitor <strong>of</strong> phosphatidylinositol<br />
3-kinase, 2-(4-morpholinyl)-8-phenyl-4H-1-benzopyran-4-one<br />
(LY294002). J Biol Chem. 1994;269:5241-5248.<br />
13. Workman P, Clarke PA, Raynaud FI, et al. Drugging the PI3 kinome:<br />
from chemical tools to drugs in the clinic. Cancer Res. 2010;70:2146-2157.<br />
14. Hayakawa M, Kaizawa H, Moritomo H, et al. Synthesis and biological<br />
evaluation <strong>of</strong> pyrido[3�,2�:4,5]furo[3,2-d]pyrimidine derivatives as novel PI3<br />
kinase p110� inhibitors. Bioorg Med Chem Lett. 2007;17:2438-2442.<br />
15. Raynaud FI, Eccles S, Clarke PA, et al. Pharmacologic characterization<br />
<strong>of</strong> a potent inhibitor <strong>of</strong> class I phosphatidylinositide 3-kinases. Cancer Res.<br />
2007;67:5840-5850.<br />
16. Hayakawa M, Kaizawa H, Moritomo H, et al. Synthesis and biological<br />
evaluation <strong>of</strong> 4-morpholino-2-phenylquinazolines and related derivatives as<br />
novel PI3 kinase p110� inhibitors. Bioorg Med Chem. 2006;14:6847-6858.<br />
17. Raynaud FI, Eccles SA, Patel S, et al. Biological properties <strong>of</strong> potent<br />
inhibitors <strong>of</strong> class I phosphatidylinositide 3-kinases: From PI-103 through<br />
PI-540, PI-620 to the oral agent GDC-0941. Mol Cancer Ther. 2009;8:1725-<br />
1738.<br />
e98<br />
REFERENCES<br />
WORKMAN AND CLARKE<br />
18. Folkes AJ, Ahmadi K, Alderton WK, et al. The identification <strong>of</strong> 2-(1Hindazol-4-yl)-6-(4-methanesulfonyl-piperazin-1-ylmethyl)-4-morpholin-4-ylthieno[3,2d]pyrimidine<br />
(GDC-0941) as a potent, selective, orally bioavailable<br />
inhibitor <strong>of</strong> class I P13 kinase for the treatment <strong>of</strong> cancer. J Med Chem.<br />
2008;51:5522-5532.<br />
19. Sutherlin DP, Bao L, Berry M, et al. Discovery <strong>of</strong> a potent, selective,<br />
and orally available class I phosphatidylinositol 3-kinase (PI3K)/mammalian<br />
target <strong>of</strong> rapamycin (mTOR) kinase inhibitor (GDC-0980) for the treatment <strong>of</strong><br />
cancer. J Med Chem. 2011;54:7579-7587.<br />
20. Friedman L. GDC-0941, a potent selective orally bioavailable inhibitor<br />
<strong>of</strong> class I PI3K. Proc <strong>American</strong> Assoc Cancer Res 2008;LB-110.<br />
21. Yap TA, Workman P. Exploiting the cancer genome: Strategies for the<br />
discovery and clinical development <strong>of</strong> targeted molecular therapeutics. Annu<br />
Rev Pharmacol Toxicol. <strong>2012</strong>;52:549-573.<br />
22. Clarke PA, Workman P. Phosphatidylinositide-3-kinase inhibitors:<br />
Addressing questions <strong>of</strong> is<strong>of</strong>orm selectivity and pharmacodynamic/predictive<br />
biomarkers in early clinical trials. J Clin Oncol. <strong>2012</strong>;30:331-333.<br />
23. Yap TA, Sandhu SK, Workman P, et al. Envisioning the future <strong>of</strong> early<br />
anticancer drug development. Nat Rev Cancer. 2010;10:514-523.<br />
24. Barretina J, Caponigro G, Stransky N, et al. The Cancer Cell Line<br />
Encyclopedia enables predictive modelling <strong>of</strong> anticancer drug sensitivity.<br />
Nature. <strong>2012</strong>;483:603-607.<br />
25. Garnett MJ, Edelman EJ, Heidorn SJ, et al. Systematic identification<br />
<strong>of</strong> genomic markers <strong>of</strong> drug sensitivity in cancer cells. Nature. <strong>2012</strong>;483:570-<br />
575.<br />
26. Workman P, Clarke PA, Al-Lazikani B. Personalized medicine: Patientpredictive<br />
panel power. Cancer Cell. <strong>2012</strong>; in press.<br />
27. Guillard S, Clarke PA, Te Poele R, et al. Molecular pharmacology <strong>of</strong><br />
phosphatidylinositol 3-kinase inhibition in human glioma. Cell Cycle. 2009;<br />
8:443-453.<br />
28. Hoeflich KP, Merchant M, Orr C, et al. Intermittent administration <strong>of</strong><br />
MEK inhibitor GDC-0973 plus PI3K inhibitor GDC-0941 triggers robust<br />
apoptosis and tumor growth inhibition. Cancer Res. <strong>2012</strong>;72:210-219.<br />
29. Carver BS, Chapinski C, Wongvipat J, et al. Reciprocal feedback<br />
regulation <strong>of</strong> PI3K and androgen receptor signaling in PTEN-deficient prostate<br />
cancer. Cancer Cell. 2011;19:575-586.<br />
30. Bendell JC, Rodon J, Burris HA, et al. Phase I, dose-escalation study <strong>of</strong><br />
BKM120, an oral pan-class I PI3K inhibitor, in patients with advanced solid<br />
tumors. J Clin Oncol. <strong>2012</strong>;30:282-290.<br />
31. Vasudevan KM, Arbie DA, Davies MA, et al. AKT-independent signaling<br />
downstream <strong>of</strong> oncogenic PIK3CA mutations in human cancer. Cancer<br />
Cell. 2009;16:21-32.<br />
32. Forbes SA, Tang G, Bindal N, et al. COSMIC (the Catalogue <strong>of</strong> Somatic<br />
Mutations in Cancer): A resource to investigate acquired mutations in human<br />
cancer. Nucleic Acids Res. 2010;38:D652-D657.
This publication is supported by an educational<br />
donation provided by:<br />
Support for this program is funded through
Change language