Purpose To test the hypothesis that tolerating some subretinal fluid (SRF) in patients with neovascular age-related macular degeneration (nAMD) treated with ranibizumab using a treat-and-extend (TE intensive arm: SRF intolerant) or resolution of all IRF only (relaxed arm: SRF tolerant except for SRF >200 μm at the foveal center) before extending treatment intervals. A 5-letter noninferiority margin was applied to the primary outcome. Main Outcome Measures Mean change in best-corrected VA (BCVA), and central subfield thickness and number of injections from baseline to month 24. Results Of the 349 participants randomized (intensive arm, n = 174; relaxed arm, n = 175), 279 (79.9%) completed the month 24. The mean change in BCVA from baseline to month 24 was 3.0 letters (standard deviation, 16.3 letters) in the intensive group and 2.6 letters (standard deviation, 16.3 letters) in the relaxed group, demonstrating noninferiority of the relaxed compared with the intensive treatment (P = 0.99). Similar proportions of both groups achieved 20/40 or better VA (53.5% and 56.6%, respectively; P = 0.92) and 20/200 or worse VA (8.7% and 8.1%, respectively; P = 0.52). Participants in the relaxed group received fewer ranibizumab injections over 24 months (mean, 15.8 [standard deviation, 5.9]) than those in the intensive group (mean, 17 [standard deviation, 6.5]; P = 0.001). Significantly more participants in the intensive group never extended beyond 4-week treatment intervals (13.5%) than in the relaxed group (2.8%; P = 0.003), and significantly more participants in the relaxed group extended to and maintained 12-week treatment intervals (29.6%) than the intensive group (15.0%; P = 0.005). Conclusions Patients treated with a ranibizumab T&E protocol who tolerated some SRF achieved VA that is comparable, with fewer injections, with that achieved when treatment aimed to resolve all SRF completely.
