Analysis of overall survival and safety during the course of the phase III VELOUR trial comparing FOLFIRI and ziv-aflibercept or placebo in mCRC patients who progressed on prior oxaliplatin treatment.

Abstract

3574 Background: VELOUR, a large, international, randomized, placebo (pbo)-controlled study, compared efficacy and safety of FOLFIRI with ziv-aflibercept (known as aflibercept outside the United States) or with pbo in 1,226 metastatic colorectal cancer patients who received prior oxaliplatin treatment. Ziv-aflibercept demonstrated statistically significant, clinically meaningful improvements in median overall survival (OS) (13.5 vs 12.06 mos; hazard ratio [HR]=0.817, P=0.0032), progression-free survival, and response rate. This analysis estimates treatment effect and safety over time course of the study. Methods: HRs by 6-mo time periods were estimated using piecewise Cox proportional hazard model. NCI-CTCAE v3.0 was used to grade adverse event (AE) severity. Results: HR improved over time (Table), consistent with survival curves that continue to separate past the median time point, indicating that the magnitude of ziv-aflibercept treatment effect continues to increase over time. Incidence of grade 3 AEs (45.1% vs 62.0%) was higher in the ziv-aflibercept/FOLFIRI arm; incidences of grade 4 AEs were 17.4% (pbo) vs 21.4% (ziv-aflibercept). More common AEs only occurred in a small proportion of ziv-aflibercept/FOLFIRI cycles (eg, grade ≥3 hypertension and diarrhea occurred in 3.6% and 2.8% of cycles, respectively). The majority of grade 3/4 AEs occurred in early treatment (first 3-4 cycles). Most patients experienced only a single episode of grade ≥3 AEs with ziv-aflibercept/FOLFIRI. Importantly, AEs in VELOUR did not impact patients’ ability to receive chemotherapy. Conclusions: Treatment with ziv-aflibercept/FOLFIRI showed continuous, consistent improvement in OS over time. While combined grade 3/4 AEs were higher with ziv-aflibercept, AEs occurred early in treatment in a small proportion of total cycles; the majority were single-episode in nature. Clinical trial information: NCT00561470. [Table: see text]