Background Gastrointestinal perforation is a serious adverse event associated with bevacizumab, an inhibitor of vascular endothelial growth factor (VEGF) widely used in current cancer treatment. The association is highlighted by a black-box warning issued by the US Food and Drug Administration, recommending that bevacizumab be permanently discontinued in patients with gastrointestinal perforation. However, no significant association has yet been established between bevacizumab and gastrointestinal perforation in randomised controlled trials. We did a systematic review and meta-analysis of published randomised controlled trials to assess the overall risk of gastrointestinal perforation associated with bevacizumab treatment. Methods We searched PubMed and Web of Science for articles published between January, 1966, and July, 2008. Additionally, abstracts presented at American Society of Clinical Oncology conferences held between January, 2000, and July, 2008, were searched to identify relevant clinical trials. Eligible studies included prospective randomised controlled trials in which bevacizumab was compared with controls in combination with standard anti-neoplastic therapy. Summary incidence rates, relative risks, and 95% CIs were calculated using a fixed-effects or random-effects model, depending on the heterogeneity of the included studies. Findings 12294 patients with a variety of solid tumours from 17 randomised controlled trials were included in our analysis. The incidence of gastrointestinal perforation was 0.9% (95% Cl 0.7-1.2) among patients receiving bevacizumab, with a mortality of 21.7% (11.5-37.0). Patients treated with bevacizumab had a significantly increased risk of gastrointestinal perforation compared with patients treated with control medication, with a relative risk of 2.14 (95% Cl 1.19-3.85; p=0.011). Risk varied with bevacizumab dose and tumour type. Relative risks for patients receiving bevacizumab at 5 and 2.5 mg/kg per week were 2.67 (95% CI 1.14-6.26) and 1.61 (0.76-3.38), respectively. Higher risks were observed in patients with colorectal carcinoma (relative risk 3.10, 95% CI 1.26-7.63) and renal cell cancer (relative risk 5.67, 0.66-48.42). Interpretation The addition of bevacizumab to cancer therapy significantly increased the risk of gastrointestinal perforation compared with controls. The risk may vary with bevacizumab dose and tumour type. Further studies are recommended to investigate the use of bevacizumab in selected patients who have recovered from gastrointestinal perforation.
