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.
658s Tumor Biology<br />
10508 Poster Discussion Session (Board #1), Tue, 8:00 AM-12:00 PM and<br />
11:30 AM-12:30 PM<br />
Next-generation RNA sequencing reveals transcriptomic changes after<br />
brief exposure to preoperative nab-paclitaxel, bevacizumab, and trastuzumab.<br />
Presenting Author: Vinay Varadan, Philips Research North America,<br />
Briarcliff Manor, NY<br />
Background: Identification <strong>of</strong> differentially expressed transcripts after brief<br />
exposure to preoperative therapy can help determine likely response<br />
markers. We quantify and compare differential gene and is<strong>of</strong>orm expression<br />
using RNA-seq on patient samples with 10 day exposure to one dose <strong>of</strong><br />
trastuzumab, bevacizumab or nab-paclitaxel. Methods: We sequenced<br />
transcriptomes <strong>of</strong> 23 pairs <strong>of</strong> core biopsy RNA from breast cancers pre/post<br />
10 day exposure to therapy. Paired-end sequencing was done on the<br />
Illumina GAII platform using amplified total RNA with 74bp read length,<br />
yielding data on transcript abundance for a total <strong>of</strong> 22,160 genes and<br />
34,449 transcripts. Differential expression <strong>of</strong> transcripts between pre/post<br />
samples was estimated assuming Poisson-distributed read-counts, followed<br />
by multiple testing correction and enrichment analysis <strong>of</strong> 185 KEGG<br />
pathways. Results: PAM50-based clustering showed individual samples<br />
cluster together, demonstrating that tumor subtypes do not change over the<br />
10-day treatment (SABCS 2011). We identified genes that were significantly<br />
differentially expressed (p�0.05; FDR�0.1) in at least 60% <strong>of</strong><br />
samples within each therapy arm: 780 genes in trastuzumab, 302 in<br />
bevacizumab, and 176 in nab-paclitaxel. Surprisingly, only THAP11 and<br />
TINF2 were common amongst them. THAP11 is involved in stem cell<br />
maintenance and TINF2 is important for regulation <strong>of</strong> telomere length.<br />
Immune system and metabolism-related pathways were commonly affected<br />
(p�0.05) across all arms. The bevacizumab arm showed significant<br />
down-regulation <strong>of</strong> angiogenesis-associated genes: ESM1 and VEGFR2 in<br />
� 80% <strong>of</strong> samples. The nab-paclitaxel arm exhibited changes in TGF-beta<br />
signaling, Nod-like receptor and Wnt signaling. The trastuzumab arm<br />
exhibited consistent alteration <strong>of</strong> ErbB2 and mTOR pathways, with SOX11<br />
and TOP2B downregulated in every sample. Conclusions: This is the first<br />
study to compare gene expression with brief exposure across therapies<br />
using RNA-seq technology. The unique aspects <strong>of</strong> transcriptional response<br />
to each treatment underscore the need for specific markers <strong>of</strong> therapeutic<br />
response to nab-paclitaxel, bevacizumab and trastuzumab.<br />
10510 Poster Discussion Session (Board #3), Tue, 8:00 AM-12:00 PM and<br />
11:30 AM-12:30 PM<br />
PTEN assessment and PI3K/mTOR inhibitors: Importance <strong>of</strong> simultaneous<br />
assessment <strong>of</strong> MAPK pathway aberrations. Presenting Author: Filip Janku,<br />
Department <strong>of</strong> Investigational Cancer Therapeutics (Phase I Program),<br />
University <strong>of</strong> Texas M. D. Anderson Cancer Center, Houston, TX<br />
Background: Loss <strong>of</strong> PTEN function increases PI3K/mTOR signaling.<br />
Preclinical data suggest that PTEN aberrations can predict sensitivity to<br />
drugs targeting the PI3K/mTOR pathway. Methods: Tumors from 461<br />
patients with diverse cancers referred to the <strong>Clinical</strong> Center for Targeted<br />
Therapy were tested for cytoplasmic immunohistochemical expression <strong>of</strong><br />
PTEN using the following scoring system: negative (no staining); positive<br />
(intact staining); reduced staining, with stroma serving as an internal<br />
positive control. Whenever possible, tumors were also tested for mutations<br />
in the MAPK pathway (KRAS, NRAS, and BRAF). We analyzed outcomes <strong>of</strong><br />
patients treated with PI3K /mTOR with respect to their PTEN status.<br />
Results: Of 461 patients, 47 (10%) had negative PTEN expression, 168<br />
(36.5%) had positive PTEN expression, and 246 (53.5%) had reduced<br />
PTEN. In tumor types with �10 patients tested, negative PTEN was most<br />
frequent in uterine (10/33, 33%), renal (3/11, 27%), salivary gland (2/10,<br />
20%), colorectal (15/82, 18%), breast (2/15, 13%), pancreatobiliary<br />
(3/24, 13%), and prostate cancers (2/19, 11%). Of 206 patients tested for<br />
MAPK (KRAS, NRAS, and BRAF) pathway mutations, patients with either<br />
negative or reduced PTEN had more MAPK mutations than patients with<br />
positive PTEN expression (72/144, 50% vs. 20/62, 32%, p�0.02). A total<br />
<strong>of</strong> 153 (33%) patients received therapies with PI3K /mTOR inhibitors and<br />
17 (11%, 95% CI 0.07-0.17) achieved a partial response (PR). Proportions<br />
<strong>of</strong> PRs were 2/13 (15%) for PTEN negative, 3/50 (6%) for PTEN<br />
positive, and 12/90 (13%) for reduced PTEN. There was no significant<br />
difference in PR rate amongst patients with positive PTEN expression<br />
compared to patients with negative or reduced PTEN (3/50, 6% vs.<br />
14/103, 14%; p�0.27). Conclusions: Complete loss <strong>of</strong> PTEN expression<br />
was found in 10% <strong>of</strong> patients. Positive PTEN expression was found in<br />
36.5% <strong>of</strong> patients. PTEN status did not predict response to PI3K/mTOR<br />
inhibitors, perhaps because patients with negative and reduced PTEN<br />
expression had a higher incidence <strong>of</strong> simultaneous MAPK (KRAS, NRAS,<br />
BRAF) mutations.<br />
10509 Poster Discussion Session (Board #2), Tue, 8:00 AM-12:00 PM and<br />
11:30 AM-12:30 PM<br />
Early change in 18-fluorodeoxyglucose (FDG) uptake on positron emission<br />
tomography (PET) to predict response to preoperative systemic therapy<br />
(PST) in HER2-negative primary operable breast cancer: Translational<br />
breast cancer research consortium (TBCRC008). Presenting Author: Roisin<br />
M. Connolly, Sidney Kimmel Comprehensive Cancer Center, Johns Hopkins<br />
School <strong>of</strong> Medicine, Baltimore, MD<br />
Background: PST allows for improved surgical outcomes and response<br />
assessment without compromising long term outcomes. PST is an attractive<br />
model for assessing surrogate markers <strong>of</strong> response to therapy. We<br />
hypothesized that changes in tumor standardized uptake values corrected<br />
for lean body mass (SUL) max on FDG- PET by cycle 1 day 15 (C1D15) <strong>of</strong><br />
therapy would predict pathological complete response (pCR) to PST in<br />
women with stage 2-3, grade 2-3, HER2-negative breast cancer. Methods:<br />
TBCRC008 is a multicenter placebo-controlled trial that investigates pCR<br />
following 12 weeks <strong>of</strong> preoperative carboplatin and albumin-bound paclitaxel<br />
with or without vorinostat. FDG-PET followed by tumor biopsies were<br />
performed at baseline and C1D15. We correlated % reduction in SULmax<br />
on FDG-PET (PERCIST 1.0; Wahl RL, J Nuc Med 2009) with pCR (no<br />
invasive cancer in breast/axilla). We compared % reduction in SULmax<br />
between responders (pCR) and non responders (no pCR) using nonparametric<br />
Wilcoxon rank sum test. We explored association <strong>of</strong> % reduction in<br />
SULmax at pre-specified cut<strong>of</strong>f with response using Fisher’s exact test and<br />
logistic regression. We correlated baseline, C1D15, and % change in Ki67<br />
at C1D15 with pCR. Results: Accrual is complete. Of 62 women enrolled<br />
(10/2009-11/2011), 40 have completed study PST and surgery (median<br />
age 47.5 [range 30-68], ER-positive 67%). Overall pCR was 26%. In an<br />
intent to treat analysis (n�39), we observed a significant difference in<br />
median % reduction in SULmax between responders vs not (66.6% vs<br />
32.4%, p �0.001). We observed a higher proportion <strong>of</strong> reduction in<br />
SULmax � 60% in responders vs not (80% vs 3.5%, p �0.001). The<br />
differences in baseline, C1D15 and % change in Ki67 were not significant<br />
between responders and non-responders. Conclusions: Change in SULmax<br />
on FDG-PET 15 days after initiating PST was significantly greater in<br />
patients with pCR versus no pCR. Future studies will determine whether<br />
altering therapy based on early changes in SULmax will improve pCR.<br />
Unblinded data from all participants will be presented at the meeting.<br />
10511 Poster Discussion Session (Board #4), Tue, 8:00 AM-12:00 PM and<br />
11:30 AM-12:30 PM<br />
Analysis <strong>of</strong> the intratumoral heterogeneity <strong>of</strong> PIK3CA mutant alleles in<br />
breast cancer (BC): Implications for the luminal (LUM) phenotype. Presenting<br />
Author: Leticia De Mattos-Arruda, Medical Oncology Department, Vall<br />
d’Hebron University Hospital, Barcelona, Spain<br />
Background: The hyperactivation <strong>of</strong> the phosphatidylinositol 3-kinase<br />
(PI3K) pathway may confer endocrine therapy resistance and is an<br />
attractive target for LUM patients (pts). However, PI3KCA mutations could<br />
be heterogeneously distributed within the tumor as different genetic clones<br />
and this could have therapeutic implications. Our aim was to assess the<br />
frequency <strong>of</strong> PIK3CA mutant alleles in different BC phenotypes. Methods:<br />
DNA was obtained from 75 consecutive BC FFPE samples and was pr<strong>of</strong>iled<br />
with the OncoCarta Panel v1.0 (Sequenom). Frequencies <strong>of</strong> mutant alleles<br />
(% mutant allele, mA) <strong>of</strong> PIK3CA mutations were extracted from the<br />
MassARRAY spectrum data. The viable tumor area (TA) was scored by H&E.<br />
Pts were stratified by BC phenotypes: ER�/HER2- (LUM); HER2� (HER2);<br />
triple negative (TN). Results: 25.3% <strong>of</strong> pts had PIK3CA mutations. The<br />
mean PIK3CA mA (p�0.04) and mean TA (p�0.07) differed among<br />
phenotypes. LUM tumors demonstrated greater frequencies <strong>of</strong> PIK3CA<br />
mutation (38% vs.15%, p�0.05) and mean PIK3CA mA (31% vs.17%,<br />
p�0.01) than non-LUM.There was a significant linear correlation between<br />
mA and TAfor LUM tumors (r�0.91); when HER2 and TN tumors were also<br />
considered this relation was less pronounced (r�0.66). Overall, LUM pts,<br />
median age 55 (42-84), had significantly better clinical outcomes<br />
(p�0.00024), whilst analyses <strong>of</strong> prognosis did not differ among LUM pts<br />
(PIK3CA mutant vs. wild-type, p�0.68) nor mutant pts (LUM vs. non-<br />
LUM, p�0.36). Outcomes will be presented. To assess the intratumoral<br />
heterogeneity <strong>of</strong> PIK3CA mutations, we determined the mA/TA ratio. The<br />
mA/TA ratio should be 0.5 for a homogeneous distribution <strong>of</strong> the heterozygous<br />
PIK3CA mutation. The ratio <strong>of</strong> LUM patients was close to 0.5, while<br />
for non-LUM it was lower, suggesting a non-homogeneous distribution <strong>of</strong><br />
PIK3CA mutant alleles in non-LUM tumors. Conclusions: Our analysis<br />
indicates that LUM tumors tend to be homogeneous regarding PIK3CA<br />
mutation as compared to HER2 and TN. This suggests that PIK3CA<br />
oncogenic activation could be an early hit for tumor initiation in LUM<br />
tumors, which could be more specifically targetable; thus, pts would derive<br />
greater benefit from PI3K-inhibitors plus hormonal therapy.<br />
Visit abstract.asco.org and search by abstract for the full list <strong>of</strong> abstract authors and their disclosure information.