Tetracyclines and Risk of Proliferative Vitreoretinopathy after Rhegmatogenous Retinal Detachment

Objective To evaluate the association between documented systemic tetracycline administration and the risk of proliferative vitreoretinopathy (PVR)–related outcomes after rhegmatogenous retinal detachment (RRD). Design Retrospective cohort study with 1-year follow-up. Participants Adults aged 18 years or older with RRD were identified within a multicenter federated electronic health record network. After 1:1 propensity score matching, 2886 patients with documented systemic tetracycline administration were compared with 2886 matched controls. Methods Systemic tetracycline exposure was defined as a recorded prescription from 1 month before to 6 months after RRD-related cohort entry. Propensity score matching was performed on demographics, socioeconomic determinants, systemic and ophthalmic comorbidities, medications, laboratory values, and health care utilization. Time-to-event analyses were conducted using Kaplan–Meier estimation and Cox proportional hazards models. Prespecified sensitivity analyses and active comparator analyses were performed to assess robustness. Main Outcome Measures Hazard ratios (HRs) for PVR-related outcomes, defined as complex retinal detachment repair, nondiabetic proliferative retinopathy, and tractional retinal detachment. Reoperation was evaluated as a secondary outcome. Results Baseline characteristics were well balanced after matching. During comparable follow-up, systemic tetracycline administration was associated with a lower risk of complex retinal detachment repair (HR, 0.33; 95% confidence interval [CI], 0.25–0.43), nondiabetic proliferative retinopathy (HR, 0.63; 95% CI, 0.45–0.89), and tractional retinal detachment (HR, 0.57; 95% CI, 0.43–0.75). Reoperation was also less frequent among tetracycline-exposed patients (HR, 0.34; 95% CI, 0.26–0.46). Findings remained consistent across prespecified sensitivity analyses, including alternative exposure definitions, follow-up durations, doxycycline-only exposure, and active comparator analyses. E-values suggested robustness to unmeasured confounding, and positive and negative control outcomes supported the internal validity of the observed associations. Conclusions Documented systemic tetracycline administration was associated with a lower risk of PVR-related outcomes after RRD. These findings support further prospective evaluation of tetracyclines as a potential adjunctive strategy for PVR prevention. Financial Disclosure(s) The authors have no proprietary or commercial interest in any materials discussed in this article.