Eleven RCTs were included in the review (6,439, range 115 to 1,043); two phase II trials and nine phase III trials. Randomised treatment allocation sequences were generated in all trials and six trials were of double-blind placebo controlled design. All trials were deemed to have adequate follow-up (median, 6.7 to 25.9 months).
Bevacizumab and chemotherapy was found to significantly reduce the relative risk of all-grade anaemia (RR 0.79, 95% CI 0.66 to 0.94; four RCTs) and high-grade anaemia (RR 0.72, 95% CI 0.57 to 0.90; 11 RCTs). No evidence of significant heterogeneity was found.
The relative risk of high-grade anaemia was 0.71 (95% CI 0.54 to 0.94; 10 RCTs) with bevacizumab 5mg/kg/week and 0.74 (95% CI 0.53 to 1.03; three RCTs) with bevacizumab 2.5mg/kg/week compared with chemotherapy alone; no significant difference was found between the dose levels (p=0.88). The effect bevacizumab on high-grade anaemia was not significantly affected by tumour histology (p=0.78) or chemotherapy regimen (single agent vs. doublet combination p=0.98 and platinum versus non-platinum p=0.43). No significant correlation was found between RR of high-grade anaemia and hazard ratios of progression-free survival or overall-survival.
No evidence of publication bias was found.