1239P - Clinical Genotyping and Efficacy Outcomes: Exploratory Biomarker Data from the Phase II Abigail Study of 1st-Line Bevacizumab + Chemotherapy in Non-Squamous Non-Small-Cell Lung Cancer (NS-NSCLC)

Abstract

ABSTRACT Background ABIGAIL (BO21015; NCT00700180) is a phase II, randomized, multicentre study exploring correlation between biomarkers (BMs) and best overall response (BOR) to bevacizumab with carboplatin/gemcitabine (CG) or carboplatin/paclitaxel (CP) in chemonaive patients with advanced/recurrent ns-NSCLC. ABIGAIL efficacy, safety and plasma baseline results have been reported. This abstract presents exploratory clinical genotyping data from this study. Methods 303 patients with untreated advanced/recurrent ns-NSCLC were randomized 1:1 to receive bevacizumab 7.5 mg/kg or 15 mg/kg until progression or unacceptable toxicity (with 6 cycles of CG or CP). Patients who consented provided blood and tumour samples for BM analysis. Exploratory analyses were conducted to assess whether genetic variants in the VEGFA pathway may act as biomarkers for efficacy and safety. Here we report data from DNA analysis for 12 single-nucleotide polymorphisms (SNPs) across 3 genes: VEGFA (5 SNPs), VEGFR-1 (3 SNPs) and VEGFR-2 (4 SNPs). SNPs were identified using specific individual genotyping assays. Results VEGFA: c. + 405/c.-634 (CG), VEGFA: c.-460T > C; c. -1498T > C (CT) and VEGFA: c.-2578 C > A (AC) were all associated with >50% higher odds of responding to treatment. VEGFR-1 rs9554316 (GT) was associated with >30% higher risk of progression and >40% higher risk of death. VEGF: c. + 936 C > T (CT) was associated with higher incidence of hypertension. When p-values were adjusted for treatment and prognostic factors, no SNPs were associated with significantly higher risk of hypertension. Conclusions One SNP was associated with increased risk of progression/death, while 3 others were associated with increased BOR. However, adjustment for multiple testing would no longer result in statistically significant p-values. SNPs analysed in this study have been previously reported as showing potential predictive value in other studies: VEGFA SNPs in breast cancer (E2100) and NSCLC (E4599); VEGFR1 SNP in pancreatic cancer (AVITA). More exploratory analyses from this and other trials of bevacizumab may provide further insight. Disclosure C. Pallaud: Own stock in and currently employed by F. Hoffmann-La Roche Ltd. M. Reck: Attended advisory boards for Roche, Lilly, BMS, AstraZeneca and Daiichi Sankyo. Received honoraria for lectures from Roche, Lilly, Daiichi Sankyo and AstraZeneca. B. Szima: Received funding for research. C. Yu: Attended advisory boards for Roche, AstraZeneca, Takeda and Pfizer. S.J. Scherer: Currently employed by Roche/Genentech. V. Archer: Currently employed by Roche. T.S.K. Mok: Attended advisory boards for AstraZeneca, Roche, Eli Lilly, Merck Serono, Eisai, BMS, BeiGene, AVEO, Pfizer, Taiho, BI and GSK Biologicals. On the IASLC board of directors. Received research funding from AstraZeneca. All other authors have declared no conflicts of interest.