434P Impact of molecular markers status on treatment effects comparing EGFR and VEGF monoclonal antibodies (mAbs) in untreated metastatic colorectal cancer (mCRC): Pooled individual patient data (IPD) analysis of randomized trials from the ARCAD database

Background Chemotherapy combined with either VEGF or EGFR targeting mAbs improves survival of patients (pts) with mCRC. In clinical practice, the selection of antibody may take into account multiple factors, including RAS/BRAF mutation status and tumour sidedness. Methods IPD from 2 randomized first-line trials with head-to-head comparison of chemo+EGFRmAb v chemo+VEGFmAb in mCRC were pooled. Biomarker subpopulations (see table) were analysed. Overall survival (OS) and progression-free survival (PFS) were compared between groups by Cox model, stratified by studies and adjusted by age, gender and performance status. Treatment effects were estimated by adjusted hazard ratio (HRadj) and 95% confidence interval (CI). Within each biomarker subset, mAb efficacy was explored according to type of chemo, gender, sidedness and site of metastasis. Interaction tests were performed. P-values < 0.01 were considered statistically significant to account for multiple comparisons. Results 1967 pts from 2 studies with data available for ≥ 1 biomarker were analysed. For the KRAS and NRAS MT groups, PFS favoured VEGF targeting (median 12.0 v 8.0 mos in KRAS MT, median 12.9 v 8.1 mos in NRAS MT), but OS did not differ significantly (see table). No difference in PFS or OS was observed in the BRAF MT or triple WT groups. Exploratory analyses of sidedness demonstrated inferior survival with EGFR mAbs (vs. VEGFmAb) in triple WT tumors on the right side of the colon (OS: HRadj 1.61, p=.056, pinteraction .099; PFS: HRadj 2.39, p=.0007, pinteraction .041)