Magnitude of progression-free survival (PFS) improvement and treatment (Tx) duration in metastatic colorectal cancer (mCRC) for bevacizumab (BV) in combination with oxaliplatin-containing regimens: An analysis of two phase III studies
4073 Background: In trials of BV with FOLFOX (fluorouracil, leucovorin, oxaliplatin) for mCRC, variability in the magnitude of PFS improvement has been reported [HR=0.61 in E3200 vs HR= 0.83 (FOLFOX or CAPOX (capecitabine and oxaliplatin)) in NO16966]. We propose that differences in rates of treatment discontinuation (D/C) for adverse events (AE) between these studies may have resulted in differences in the observed benefits associated with BV. We explored Tx duration (proportion of patients on Tx) and Tx D/C data at median PFS for the BV containing arms of each study. Methods: ECOG study E3200 randomized previously treated patients with mCRC to FOLFOX ± BV (10 mg/kg). NO16966 employed a 2x2 design that randomized previously untreated patients with mCRC to CAPOX vs FOLFOX and to BV (5 mg/kg) or placebo. In both trials, study Tx was defined as any component of the prescribed regimen. PFS was estimated from Kaplan-Meier curves, and hazard ratios (HR) for PFS were estimated by Cox regression. Results: Median PFS for the BV containing arm of the study: 30 weeks for E3200; 42 weeks for NO16966 Conclusion: These data suggest possible differences between the two studies in Tx duration and Tx D/C patterns with a greater proportion of patients on NO16966 discontinuing Tx for any AE. Duration of study Tx might have affected both the incidence of AEs and the magnitude of PFS benefit observed for the addition of bevacizumab to oxaliplatin-based chemotherapy in these studies. Attention to Tx duration and Non-PD Tx D/C in future clinical trials will be important when considering PFS as a primary efficacy endpoint. [Table: see text] No significant financial relationships to disclose.