Clinical efficacy and safety of aflibercept biosimilars in neovascular age-related macular degeneration: a systematic review and meta-analysis of randomised controlled trials.

Zhang, Charles; Sonik, Neel R; Ayoubi, Mohammad; AbouKasm, Georges; Zhu, Daniel; Fan, Jason C; Albini, Thomas A; Yannuzzi, Nicolas A · Br J Ophthalmol · 2026

meta_analysis · Level I

Where this comes from

Abstract

To evaluate the efficacy and safety of aflibercept biosimilars compared with reference aflibercept in the treatment of neovascular age-related macular degeneration (nAMD). A systematic review and meta-analysis were conducted following Preferred Reporting Items for Systematic Reviews and Meta-Analyses (PRISMA) guidelines and registered on PROSPERO (CRD4201069147). Phase 3 randomised controlled trials (RCTs) comparing aflibercept biosimilars to reference aflibercept in treatment-naïve nAMD were included. All included studies employed a standardised on-label aflibercept dosing regimen consisting of three initial monthly loading doses followed by injections every 8 weeks. The primary outcome was the mean difference (MD) in best-corrected visual acuity (BCVA) at week 8. Secondary outcomes included BCVA at week 52, proportion of patients gaining ≥15 letters or maintaining vision (<15-letter loss), central subfield thickness (CST) and adverse events (AEs). Six RCTs (2022-2024) were included, comprising 2029 eyes (1020 biosimilar; 1009 reference). BCVA at week 8 did not differ significantly between groups (MD -0.34 letters; 95% CI -1.12 to 0.44; I²=0%; p=0.39), as was BCVA at week 52 (MD -0.04; 95% CI -1.18 to 1.10; I²=1%; p=0.94). No differences were observed in the proportion of patients gaining ≥15 letters, maintaining vision, CST changes or AEs. Aflibercept biosimilars demonstrated no significant difference in efficacy at 8 weeks and 52 weeks as well as safety relative to reference aflibercept in eyes with treatment-naïve nAMD following an on-label aflibercept treatment protocol; however, real-world extended durability remains untested. These findings were supported by a low risk of bias and a high-certainty of evidence using the Grading of Recommendations, Assessment, Development and Evaluations tool.