Disease Course Patterns After Discontinuation of Bevacizumab: Pooled Analysis of Randomized Phase III Trials

Preclinical studies have suggested accelerated tumor growth, local invasion, and distant metastasis after withdrawal of treatment with some antiangiogenic agents. To investigate whether discontinuation of bevacizumab treatment is associated with accelerated disease progression or increased mortality, we retrospectively analyzed five randomized, placebo-controlled phase III studies in 4,205 patients with breast, colorectal, renal, and pancreatic cancer. Time from treatment discontinuation to progressive disease or death was analyzed in patients discontinuing bevacizumab/placebo as a result of adverse events (AEs). Mortality rates were assessed at 30, 60, 90, 120, 150, 180, and 210 days after the last bevacizumab/placebo dose in the following two groups: patients discontinuing bevacizumab/placebo as a result of AEs and patients discontinuing bevacizumab/placebo for any reason. In the same groups, time from treatment discontinuation to death was analyzed. Data on disease progression pattern were available and analyzed in four of the five studies. In the pooled analysis, median time from discontinuation as a result of AEs to progression/death was 3.0 months (95% CI, 2.6 to 3.8 months) for placebo and 4.0 months (95% CI, 3.4 to 4.6 months) for bevacizumab (hazard ratio, 0.93; 95% CI, 0.79 to 1.10). Mortality rates from 30 days to 210 days after treatment discontinuation and time from discontinuation to death were similar in bevacizumab- and placebo-treated patients. In addition, similar patterns of disease progression were seen in bevacizumab- and placebo-treated patients. This retrospective analysis of five placebo-controlled clinical trials does not support a decreased time to disease progression, increased mortality, or altered disease progression pattern after cessation of bevacizumab therapy.