To compare the efficacy and safety of bevacizumab (Avastin; F. Hoffmann-La Roche Ltd, Basel, Switzerland) versus ranibizumab (Lucentis; Novartis Pharma AG, Basel, Switzerland) for neovascular age-related macular degeneration (nAMD) after 2 years when using a treat-and-extend protocol.

Multicenter, randomized, noninferiority trial with a noninferiority limit of 5 letters.

Patients 50 years of age or older with previously untreated nAMD in 1 eye and best-corrected visual acuity 20/25 to 20/320.

Patients were assigned randomly to receive intravitreal injections with either ranibizumab 0.5 mg or bevacizumab 1.25 mg. Injections were given every 4 weeks until inactive disease was achieved. The treatment interval then was extended by 2 weeks at a time up to a maximum of 12 weeks. In the event of a recurrence, the treatment interval was shortened by 2 weeks at a time.

Mean change in visual acuity at 2 years.

Of a total of 441 randomized patients, 339 patients (79%) completed the 2-year visit. According to per-protocol analysis at 2 years, bevacizumab was equivalent to ranibizumab, with 7.4 and 6.6 letters gained, respectively (95% confidence interval [CI] of mean difference, −4.1 to 2.5; P = 0.634). Intention-to-treat analysis was concordant, with a gain of 7.8 letters for bevacizumab and 7.5 letters for ranibizumab (95% CI of mean difference, −3.2 to 2.7; P = 0.873). The 2-year results did not show any significant difference in mean central retinal thickness, with a decrease of −113 μm for bevacizumab and −122 μm for ranibizumab (95% CI of mean difference, −32 to 15; P = 0.476). There was a statistically significant difference between the drugs regarding the number of treatments given, with 18.2 injections for bevacizumab and 16.0 injections for ranibizumab (95% CI of mean difference, −3.4 to −1.0; P ≤ 0.001). The number of serious adverse events was similar between the groups over the course of the study.

At 2 years, bevacizumab and ranibizumab had an equivalent effect on visual acuity and reduction of central retinal thickness when administered according to a treat-and-extend protocol for nAMD. There was no significant difference in the number of serious adverse events between the treatment groups.