Abstract
BACKGROUND: Geographic atrophy (GA) is a significant cause of vision loss in patients with age-related macular degeneration (AMD). Current treatments are limited to anti-complement drugs, which have limited efficacy to delay progression with significant risk of complications. Levodopa (L-DOPA) is a byproduct of melanin synthesis that is associated with reduced development of neovascular AMD. In this study, we determined if L-DOPA was associated with a reduced likelihood of new-onset GA. METHODS: We performed a retrospective analysis in the Vestrum Health Retina Database. We included eyes with non-neovascular AMD without GA and 1-5 years of follow-up. Eyes were divided into two groups. Exposed to L-DOPA before or on the date of non-neovascular AMD without GA diagnosis, and eyes not exposed to L-DOPA. We extracted age, sex, AREDS2 status, dry AMD stage, smoking history, and conversion rate to GA at years 1 through 5. Propensity score matching was used to match L-DOPA and control groups. Cox proportional hazard regression, adjusting for age, sex, AMD severity, AREDS2 use, smoking status, and L-DOPA use was employed to calculate hazard ratios for new-onset GA detection. RESULTS: We identified 112,089 control and 844 L-DOPA exposed eyes with non-neovascular AMD without GA. After propensity score matching, 2532 control and 844 L-DOPA exposed eyes remained that were well-matched for age, sex, AMD severity, AREDS2 use, and smoking status. We found that L-DOPA exposure was associated with a significantly reduced likelihood (HR = 0.68, 95% CI: 0.48-0.95, P = 0.025) of new-onset GA detection. CONCLUSION: L-DOPA use was associated with reduced detection of new-onset GA.