Objective To determine visual outcome following prophylactic intravitreal bevacizumab in patients with plaque-irradiated uveal melanoma.

Design, Setting, and Participants Retrospective, nonrandomized, interventional cohort study at Wills Eye Hospital, Philadelphia, Pennsylvania. Prophylactic bevacizumab was administered between 2008 and 2018 to 1131 eyes with irradiated uveal melanoma (bevacizumab group) and compared with 117 eyes with irradiated uveal melanoma between 2007 and 2009 (no bevacizumab [historical control] group).

Interventions Prophylactic intravitreal bevacizumab was provided at the time of plaque removal as well as 6 subsequent injections at 4-month intervals over 2 years.

Main Outcomes and Measures Visual acuity.

Results The median patient age was 61 years, 1195 of 1248 patients were white (96%), and 632 of 1248 were women (51%). The median tumor thickness was 4.0 mm, and median distance to foveola was 3.0 mm. A difference was not identified (bevacizumab vs control group) in demographic features, clinical features, or radiation parameters. The mean follow-up was 40 months vs 56 months (mean difference, −18; 95% CI, −24 to −13; P < .001). By survival analysis, the bevacizumab group demonstrated less optical coherence tomography evidence of cystoid macular edema at 24 months (28% vs 37%; hazard ratio [HR], 1.5; 95% CI, 1.1-2.2; P = .02) and 36 months (44% vs 54%; HR, 1.5; 95% CI, 1.1-2.1; P = .01), less clinical evidence of radiation maculopathy at 24 months (27% vs 36%; HR, 1.5; 95% CI, 1.0-2.2; P = .03), 36 months (44% vs 55%; HR, 1.50; 95% CI, 1.1-2.0; P = .01), and 48 months (61% vs 66%; HR, 1.4; 95% CI, 1.0-1.9; P = .03), and less clinical evidence of radiation papillopathy at 18 months (6% vs 12%; HR, 2.0; 95% CI, 1.1-3.9; P = .04). Nonparametric analysis documented better visual acuity outcomes in the bevacizumab group at all points, including 12 months (median logMAR visual acuity [Snellen equivalent]: 0.30 [20/40] vs 0.48 [20/60]; mean difference, −0.28; 95% CI, −0.48 to −0.07; P = .02), 24 months (0.40 [20/50] vs 0.70 [20/100]; mean difference, −0.52; 95% CI, −0.75 to −0.29; P < .001), 36 months (0.48 [20/60] vs 1.00 [20/200]; mean difference, −0.49; 95% CI, −0.76 to −0.21; P = .003), and 48 months (0.54 [20/70] vs 2.00 [counting fingers]; mean difference, −0.71; 95% CI, −1.03 to −0.38; P < .001).

Conclusions and Relevance These findings from a retrospective cohort of plaque radiotherapy and prophylactic intravitreal bevacizumab in patients with uveal melanoma suggest better visual outcomes when compared with nonrandomized historical control individuals through 4 years.

For ocular oncologists using brachytherapy to treat patients with uveal melanoma, primary goals include achieving excellent local tumor control and preserving vision. Techniques, such as intraoperative ultrasonography, have helped to refine brachytherapy to achieve excellent local tumor control. However, as addressed by Shields et al1 in this issue of JAMA Ophthalmology, a continued challenge remains finding ways to preserve vision in the face of radiation retinopathy and optic neuropathy. One reason that many patients choose globe-sparing radiation treatment over enucleation is to preserve some useful vision. However, in the largest, randomized study of patients with uveal melanoma, the Collaborative Ocular Melanoma Study, 43% to 49% of patients treated with brachytherapy experienced substantial vision loss within 3 years of treatment.2 Studies using anti–vascular endothelial growth factor (anti-VEGF) agents have reported improved vision outcomes associated with their use, but how to optimally use these agents to try to prevent vision loss remains an important question.