Comparative efficacy and safety of low-dose versus high-dose bevacizumab in ovarian cancer: An indirect treatment comparison

Objective. First-line therapy for ovarian cancer involves cytoreductive surgery and platinum-based chemotherapy, with or without bevacizumab. Bevacizumab can be administered at low (7.5 mg/kg every three weeks [Q3W]) or high dose (15 mg/kg Q3W). This study compared the efficacy and safety of these dosing strategies. Methods. Systematic literature review of Embase, MEDLINE (R), and CENTRAL (18/09/2023) identified randomized controlled trials (RCTs) evaluating bevacizumab versus any therapy or control in ovarian, fallopian tube, or primary peritoneal cancer. Indirect treatment comparisons (ITC) of response, survival, and safety outcomes were performed, including sensitivity/subgroup analyses adjusting for heterogeneity. Results. Six RCTs (sample size: 24-1528 patients) were included for ITC. Five evaluated high-dose bevacizumab with chemotherapy. The common comparator was carboplatin + paclitaxel. Trials mainly included stage III (n = 4) or stage II-III (n = 1) ovarian cancer patients; one did not report cancer stage. Primary analyses showed no significant differences between low- versus high-dose bevacizumab for partial response (risk ratio [95 % confidence interval]: 0.66 [0.42, 1.02]), complete response (1.76 [0.76, 4.11]), objective response rate (1.01 [0.63, 1.61]), progressive disease (1.08 [0.38, 3.10]), clinical benefit (0.89 [0.76, 1.03]), any grade >= 3 adverse event (1.53 [0.96, 2.44]), specific grade >= 3 adverse events, overall survival (hazard ratio: 0.93 [0.77, 1.13]), or progression-free survival (1.02 [0.86, 1.22]). Sensitivity and subgroup analyses confirmed findings. Conclusions. This ITC found no significant difference in clinical outcomes between low- and high-dose bevacizumab combination therapy. Despite limitations of small sample size and heterogeneities, findings suggest that bevacizumab dose may not significantly impact ovarian cancer outcomes. (c) 2025 The Authors. Published by Elsevier Inc. This is an open access article under the CC BY-NC license (http:// creativecommons.org/licenses/by-nc/4.0/).