Annual Meeting Proceedings Part 1 - American Society of Clinical ...
Annual Meeting Proceedings Part 1 - American Society of Clinical ...
Annual Meeting Proceedings Part 1 - American Society of Clinical ...
Create successful ePaper yourself
Turn your PDF publications into a flip-book with our unique Google optimized e-Paper software.
656s Tumor Biology<br />
10500 Oral Abstract Session, Mon, 8:00 AM-11:00 AM<br />
Gene expression-based predictors <strong>of</strong> chemotherapy response in basal-like<br />
breast cancer. Presenting Author: Aleix Prat, Lineberger Comprehensive<br />
Cancer Center, University <strong>of</strong> North Carolina at Chapel Hill, Chapel Hill, NC<br />
Background: The Basal-like subtype is generally associated with high<br />
chemo-sensitivity, but not all tumors respond and/or benefit to the same<br />
extend. In this study, we sought to identify gene expression predictors <strong>of</strong><br />
neoadjuvant chemotherapy sensitivity in Basal-like breast cancer. Methods:<br />
Expression <strong>of</strong> 542 genes was measured using the Nanostring nCounter<br />
platform from 69 FFPE pre-treated samples <strong>of</strong> the GEICAM/2006-03<br />
phase II trial, which were treated with epirrubicin/cyclophosphamide<br />
followed by docetaxel�/-carboplatin. Research-based PAM50 and Claudinlow<br />
predictors were also evaluated. The association between response<br />
(Miller-Payne criteria) and gene/signature expression was assessed by<br />
multivariable ordinal logistic regression. Significant findings were evaluated<br />
in 109 independent triple-negative and Basal-like tumors treated with<br />
anthracycline/taxane-based chemotherapy (Hatzis et al.). Finally, interaction<br />
tests were performed to identify genes/signatures associated with<br />
carboplatin response. Results: In GEICAM/2006-03, 61/69 (88%) tumors<br />
were identified as Basal-like by PAM50. High correlation to the Basal-like<br />
centroid, or high expression <strong>of</strong> proliferation-related genes (i.e. FANCA),<br />
were found to be significantly associated with high chemo-sensitivity,<br />
whereas high expression <strong>of</strong> genes associated with mesenchymal/stem cell<br />
biological processes (i.e. SNAI1 and IL6) and/or luminal differentiation<br />
(i.e. MUC1 and FOXA1) were significantly associated with chemoresistance;<br />
similar findings were observed in Hatzis et al. Finally, high<br />
expression <strong>of</strong> genes associated with proliferation/DNA-repair (i.e. ATR) and<br />
tight junctions (i.e. CLDN3/4/7) were found associated with carboplatin<br />
response, whereas expression <strong>of</strong> the Claudin-low signature was found<br />
associated with carboplatin resistance. Conclusions: High expression <strong>of</strong><br />
Basal-like and/or proliferation-related genes and low expression <strong>of</strong> luminal/<br />
mesenchymal/stem cell-like biological processes were consistently identified<br />
as predictive <strong>of</strong> chemotherapy response. Our data suggests that gene<br />
expression pr<strong>of</strong>iling might help shed light into the biological and clinical<br />
heterogeneity <strong>of</strong> Basal-like breast cancer.<br />
10502 Oral Abstract Session, Mon, 8:00 AM-11:00 AM<br />
A heuristic platform for clinical interpretation <strong>of</strong> cancer genome sequencing<br />
data. Presenting Author: Eliezer Mendel Van Allen, Dana-Farber Cancer<br />
Institute, Boston, MA<br />
Background: The ability to identify and effectively sort the full spectrum <strong>of</strong><br />
biologically and therapeutically relevant genetic alterations identified by<br />
massively parallel sequencing may improve cancer care. A major challenge<br />
involves rapid and rational categorization <strong>of</strong> data-intensive output, including<br />
somatic mutations, insertions/deletions, copy number alterations, and<br />
rearrangements into ranked categories for clinician review. Methods: A<br />
database <strong>of</strong> clinically actionable alterations was created, consisting <strong>of</strong> over<br />
100 annotated genes known to undergo somatic genomic alterations in<br />
cancer that may impact clinical decision-making. A heuristic algorithm was<br />
developed, which selectively identifies somatic alterations based on the<br />
clinically actionable alterations database. Remaining variants are sorted<br />
based on additional heuristics, including high priority alterations based on<br />
presence in the Cancer Gene Census, biologically significant cancer genes<br />
based on presence in COSMIC or MSigDB, and low priority alterations in the<br />
same gene family as biologically significant cancer genes. The heuristic<br />
algorithm was applied to whole exome sequencing data <strong>of</strong> clinical samples<br />
and whole genome sequencing data from a cohort <strong>of</strong> prostate cancer<br />
samples processed using established Broad Institute pipelines. Results:<br />
Application <strong>of</strong> the heuristic algorithm to the prostate cancer whole genome<br />
rearrangement data identified 172 (out <strong>of</strong> 5978) rearrangements involving<br />
actionable genes (averaging 2-3 events per tumor). Furthermore, two<br />
clinical samples processed prospectively were analyzed, yielding three<br />
potentially actionable alterations for clinical review. Conclusions: The<br />
heuristic model for clinical interpretation <strong>of</strong> next generation sequencing<br />
data may facilitate rapid analysis <strong>of</strong> tumor genomic information for<br />
clinician review by identifying and prioritizing alterations that can directly<br />
impact care. Our platform can also be applied to research data to<br />
prospectively explore clinically relevant findings from existing cohorts.<br />
Future analytical approaches using heuristic or probabilistic algorithms<br />
should underpin a robust prospective assessment <strong>of</strong> clinical cancer<br />
genome data.<br />
10501 Oral Abstract Session, Mon, 8:00 AM-11:00 AM<br />
Estrogen receptor alpha (ESR1) gene amplification status and clinical<br />
outcome in tamoxifen-treated postmenopausal patients with endocrineresponsive<br />
early breast cancer: An analysis <strong>of</strong> the prospective ABCSG-6<br />
trial. Presenting Author: Christian F. Singer, Medical University <strong>of</strong> Vienna,<br />
Department <strong>of</strong> OB/GYN, Vienna, Austria<br />
Background: Estrogen receptor alpha (ER�) expression is a prognostic<br />
parameter in breast cancer and predicts response to endocrine therapy.<br />
One <strong>of</strong> the factors important for protein expression is amplification <strong>of</strong> its<br />
encoding gene ESR1. We have investigated the value <strong>of</strong> ESR1 amplification<br />
in predicting the long-term clinical outcome in tamoxifen-treated<br />
postmenopausal women with endocrine-responsive breast cancer. Methods:<br />
394 patients who had been randomized into the tamoxifen-only arm <strong>of</strong> the<br />
prospectively designed endocrine ABCSG-06 trial and in whom FFPE tumor<br />
tissue was available were included in this analysis. Immunohistochemical<br />
ER� expression was evaluated both locally and centrally using the Allred<br />
score, while ESR1 gene amplification status was evaluated by FISH<br />
analysis using the ESR1/CEN6 ratio. Results: ESR1 copy number gains<br />
were detected in 187 <strong>of</strong> 394 (47%) tumor specimen and was associated<br />
with favorable clinical outcome. At a median follow-up <strong>of</strong> 10 years, women<br />
with intratumoral ESR1 copy number gains had a significantly longer<br />
distant recurrence-free survival (adjusted HR for relapse 0.48; 95% CI<br />
0.28-0.83; p�0.009) and breast cancer-specific survival (adjusted HR for<br />
death 0.46; CI 0.46-0.71; p�0.006) when compared to women with<br />
normal ESR1 copy numbers. Immunohistochemical ER� protein expression,<br />
evaluated by Allred score, was significantly correlated with ESR1 copy<br />
number alterations (p�0.0001; Chi-Square test), but did itself not allow to<br />
discriminate between patients with poor and good prognosis. Conclusions:<br />
ESR1 amplification status is an independent and powerful predictor for<br />
long-term distant recurrence-free and breast cancer-specific survival in<br />
postmenopausal women with endocrine-responsive early-stage breast cancer<br />
who received 5 years <strong>of</strong> tamoxifen.<br />
10503 Oral Abstract Session, Mon, 8:00 AM-11:00 AM<br />
Circulating tumor cell counts (CTC) as prognostic <strong>of</strong> overall survival (OS) in<br />
SWOG S0421-docetaxel with or without atrasentan for metastatic castration<br />
resistant prostate cancer (mCRPC). Presenting Author: Nicholas J.<br />
Vogelzang, Comprehensive Cancer Centers <strong>of</strong> Nevada and US Oncology<br />
Research, Las Vegas, NV<br />
Background: CTC are promising biomarkers in mCRPC but have not been<br />
prospectively validated for docetaxel treatment (Rx). Using CellSearch<br />
technology (J&J), we enumerated CTC in this Phase 3 trial & assessed<br />
prognostic value for OS. The aim <strong>of</strong> this correlative study nested within<br />
0421 was to compare CTC via CellSearch technology to newer micr<strong>of</strong>ilter<br />
technologies (Cote & Goldkorn PIs RO1 CA141077). Comparative data<br />
analysis is ongoing. Methods: CTC were drawn at baseline (d1) & pre-cycle 2<br />
(d21) <strong>of</strong> Rx & shipped overnight to a central site for enumeration (CTC/7.5<br />
ml). Cox regression evaluated the association between OS and (i) baseline<br />
CTC counts & (ii) CTC dynamics (d1 to d21) in pts with good (�5) vs. poor<br />
(��5) baseline CTC counts. Receiver operator characteristic (ROC) analysis<br />
and Characteristics & Regression Trees (CART) were used to explore<br />
further prognostic CTC cutpoints for 2-yr survival. Results: Of 263 patients<br />
(pts) consented, 238 were evaluable at d1 & 232 at d21. At d1 median<br />
CTC was 5 (range 0-5916) & d1 CTC � vs. �� 5 was associated with<br />
baseline PSA (mean 99 vs. 320 ng/ml, p�0.004) and worse bone pain<br />
(36% vs. 51%, p�0.03). There was a significant difference in OS for d1<br />
CTC � vs. ��5, with a hazard ratio (HR) <strong>of</strong> 2.92 (95% CI 1.92-4.43,<br />
p�0.001) after adjustment for PSA & other factors. In pts with low d1 CTC<br />
(� 5), an increase in CTC was associated with shorter OS, HR 4.04 (95%CI<br />
1.56-10.44, p�0.004); in pts with high d1 CTC (��5), a ��2-fold<br />
decrease in CTC was associated with longer OS, HR 0.45 (95%CI<br />
0.24-0.84, p�0.012); adjusting for risk factors. D1 CTC and 2-year<br />
survival had ROC AUC <strong>of</strong> 0.781. CART analysis identified prognostic<br />
subgroups based on CTC <strong>of</strong> 0, 1-5, 6-53, and �53: (HR 0.36, 0.77,1.3<br />
and 2.8). Conclusions: In this phase 3 trial, d1 CTC was prognostic <strong>of</strong> OS<br />
after risk factor adjustment. CTC dynamics from d1 to d21 were also<br />
prognostic <strong>of</strong> OS. These data are an exploratory subset analysis <strong>of</strong> the<br />
overall study. Yet, they comprise the largest docetaxel-based prospective<br />
cohort to date, which validates a 5 CTC prognostic threshold for OS &<br />
identifies new potential prognostic subgroups that may extend the clinical<br />
utility <strong>of</strong> CTC enumeration in mCRPC.<br />
Visit abstract.asco.org and search by abstract for the full list <strong>of</strong> abstract authors and their disclosure information.