Patterns of anti-vascular endothelial growth factor discontinuation in neovascular age-related macular degeneration

Objective: To report on anti -vascular endothelial growth factor (anti-VEGF) discontinuation in neovascular age -related macular degeneration (nAMD). Design: Retrospective cohort study. Participants: Treatment -naive nAMD patients initiating anti-VEGF injections between 2015 and 2021. Methods: Demographics, treatment start and end dates, number of injections, treatment length, reason for discontinuation, and baseline and final data (i.e., age, best -corrected visual acuity, and central subfield thickness) were recorded. Statistical analyses using STATA 17.0 assessed differences between baseline and final values and between treatment -discontinuation subgroups. Results: A total of 619 eyes of 502 treatment -naive patients (9015 injections) were included (age, 81.6 +/- 8.4 years; 64.0% female). Discontinuation rate was 58.3% (361 of 619), with 310 patients discontinuing because of the lack of visual benefit (n = 152), severe comorbidity or death (n = 82), transferred (n = 33), stable off active treatment (n = 19), lack of benefit plus stable off treatment (n = 14), patient decision (n = 6), and ocular comorbidity (n = 4). Among the 309 remaining patients, 51 (16.5%) were lost to follow-up. Discontinuation occurred within the first year in 49.3% (n =178). Visual acuity was at least maintained in all groups and improved in the following groups: severe comorbidity or death ( p < 0.0001), lost to follow-up ( p = 0.0003), transferred ( p = 0.0004), and stable off treatment ( p = 0.0053). The lack of visual benefit group had no improvement in vision regardless of treatment length. Compared with other subgroups, those stable off treatment group was younger ( p = 0.0055), had better baseline vision ( p = 0.0018), received more injections ( p = 0.0437) over a longer time ( p = 0.0034), and achieved better final vision ( p < 0.0001). Conclusion: While there was a high discontinuation rate over 7.5 years, most were attributable to disease or treatment factors and nonmodifiable patient factors. Discontinuation frequently occurred within the first year.