Potential predictive and prognostic biomarkers identified in baseline plasma samples from the VELOUR trial.

Abstract

638 Background: Biomarkers of resistance or susceptibility to vascular endothelial growth factor (VEGF) inhibition have not been validated. Ziv-aflibercept (ZALTRAP), a fusion protein that binds and neutralizes ligands of VEGFR1 and VEGFR2 (e.g., VEGFA, PLGF, and VEGFB), is approved for the treatment of metastatic colorectal cancer (mCRC) in combination with FOLFIRI in patients who have progressed after oxaliplatin therapy. We are currently analyzing plasma samples from VELOUR, the registration trial of FOLFIRI +/- ziv-aflibercept, in an effort to identify prognostic and predictive factors. Methods: VELOUR was a prospective, multicenter, multinational, randomized (1:1), double-blind, parallel-arm phase III study conducted in 1,226 patients. Retrospective analysis of circulating proteins from 553 baseline plasma samples was performed. Samples were analyzed for levels of 98 analytes using multiplex assays and ELISA. Biomarker values were dichotomized around the median value and analyzed with respect to overall survival (OS). Results: The biomarker study sub-population was similar to the overall VELOUR population. For OS, the hazard ratio (HR) was 0.809 in the plasma biomarker population vs. HR =0.817 in the overall VELOUR population. For PFS, the HR = 0.752 in the plasma biomarker population vs. HR = 0.758 in VELOUR. Patient demographics, including ECOG status, were similar between the groups. Several biomarkers were identified as potentially predictive or prognostic (or both) of OS, with a HR<0.7 (false discovery rate of 0.05 and interaction p < 0.10). No biomarker subset corresponded to worse OS with ziv-aflibercept treatment. Conclusions: Multiple potential prognostic and predictive biomarkers were identified in VELOUR and will be presented. These results are exploratory and require prospective studies for validation. Patient subsets with elevated expression of alternative angiogenic factors or increased pro-inflammatory markers may correlate with poor outcome overall, and in some cases are not improved by addition of ziv-aflibercept treatment. Further investigation of this dataset continues, including analysis of on-treatment plasma samples in a subset of patients.