Twenty-two RCTs were included in the review (14,277 participants). Eight trials were double-blinded. A random allocation sequence was generated in all trials. Jadad scores for each study were not reported.
Compared with non-bevacizumab treatment, addition of bevacizumab to cancer chemotherapy was significantly associated with an increased risk of high-grade bleeding (RR 1.60, 95% CI 1.19 to 2.15; 20 RCTs).
In patients who received high doses of bevacizumab at 5mg/kg per week (versus non-bevacizumab treatment) there was a significantly higher risk of high-grade bleeding for those with non-small-cell lung cancer (RR 3.41, 95% CI 1.68 to 6.91; three RCTs), renal cell carcinoma (RR 6.37, 95% CI 1.43 to 28.33; two RCTs) and colorectal cancer (RR 9.11, 95% CI 1.70 to 48.79; two RCTs).
In patients who received low doses of bevacizumab at 2.5 mg/kg per week (versus non-bevacizumab treatment) there was a significantly higher risk of high-grade bleeding for those with non-small cell lung cancer (RR 2.48, 95% CI 1.01 to 6.10; two RCTs) but not in patients with colorectal cancer.
No significant heterogeneity was observed on these outcomes. There was no evidence of publication bias.