Robyn Guymer AM1,2, Caroline Markey3, Ian McAllister4,5, Mark Gillies6,7, Alex Hunyor8,9,7, Jennifer Arnold10

Purpose: This 24-month, phase IV, randomized, single-masked, multi-centre study tested the hypoth- esis that tolerating a degree of subretinal fluid (SRF) can achieve similar BCVA using a treat-and-extend (T&E) regimen relative to SRF not being tolerated.
Method: Subjects with active CNV were random- ized to receive ranibizumab 0.5 mg monthly until complete resolution of SRF/intraretinal fluid (IRF) (intensive arm: SRF intolerant) or until resolution of IRF only (relaxed arm: SRF tolerant [<200 μm SRF]) before extending intervals. A 5-letter non-inferiority margin was applied to the primary outcome of mean change in BCVA from baseline at Month 24. Results: Of the 349 subjects randomized (intensive: 174; relaxed: 175), 279 (79.9%) completed the study. At Month 24, the mean ± SD BCVA gain was 3.2 ± 16.5 letters in the intensive group and 2.5 ± 16.6 letters in the relaxed group, suggesting non- inferiority of the relaxed to the intensive treatment (P = 0.787). There was no difference between the intensive and relaxed groups in proportion of sub- jects with ≥6/12 VA (55.1% vs 58.3%; P = 0.901) nor ≤6/60 VA (8.2% vs 8.6%; P = 0.397). Subjects in the relaxed group received fewer mean SD ranibi- zumab injections (15.8 5.9) than the intensive group (17 6.5; P = 0.001). More subjects in the relaxed than intensive group extended to 12 week intervals (28.2% vs 12.9%; P = 0.002); fewer sub- jects in the relaxed than intensive group remained on 4 weekly intervals (2.8% vs 13.5%; P = 0.003). Conclusion: Our study provides evidence that a degree of SRF can be tolerated without affecting VA when treating patients with ranibizumab T&E.