Bevacizumab continuation versus no continuation after first-line chemo-bevacizumab therapy in patients with metastatic colorectal cancer: A randomized phase III noninferiority trial (SAKK 41/06).

3503 Background: Chemotherapy plus bevacizumab is a standard option for first-line treatment in metastatic colorectal cancer patients. We assessed whether no continuation is non-inferior to continuation of bevacizumab after stop of first-line chemotherapy. Methods: In an open-label, phase 3 multicenter study conducted in Switzerland, patients with unresectable metastatic colorectal cancer having non-progressive disease after 4-6 months of standard first-line chemotherapy plus bevacizumab were randomly assigned in a 1:1 ratio to continuing bevacizumab (7.5 mg/kg every 3 weeks) or no treatment. CT scans were done every 6 weeks between randomization and disease progression. The primary endpoint was time to progression (TTP). A non-inferiority limit for hazard ratio (HR) of 0.727 was chosen to detect a difference in TTP of 6 weeks or less, with a one-sided significant level of 10% and a statistical power of 85%. Results: The per-protocol population comprised 262 patients. Median follow-up is 28.6 months (range, 0.6-54.9 months). Median TTP was 17.9 weeks (95% CI 13.3-23.4) for bevacizumab continuation and 12.6 weeks (95% CI 12.0-16.4) for no continuation; HR 0.72 (95% CI 0.56-0.92). Median progression free-survival and overall survival, both measured from start of first-line treatment, was 9.5 months and 24.9 months for bevacizumab continuation and 8.5 months (HR 0.73 (95% CI 0.57 - 0.94)) and 22.8 months (HR 0.87 (95% CI 0.64 – 1.18)) for no continuation. Median time from randomization to second-line treatment was 5.9 months for bevacizumab and 4.8 for no continuation. Grade 3-4 adverse events in the bevacizumab continuation arm were uncommon. Conclusions: Non-inferiority could not be demonstrated. The 95% confidence intervals for the TTP HR indicate superiority of bevacizumab continuation after stop of first-line chemotherapy. The median differences in TTP and in time between randomization and start of second-line treatment were of moderate magnitude being less than 6 weeks. The results of an accompanying cost analysis will be presented at the meeting. Clinical trial information: NCT00544700.