Association between Systemic Methotrexate Therapy and Proliferative Vitreoretinopathy.

Abstract

Objective: Proliferative vitreoretinopathy (PVR) is a major cause for failure of retinal detachment (RD) repair. We sought to determine if patients taking systemic methotrexate therapy had a lower incidence of PVR.

Design: Multicenter retrospective cohort study.

Participants: Patients aged 18 years or over who were documented in the Sight Outcomes Research Collaborative (SOURCE) repository to have received a diagnosis of RD and have undergone RD repair surgery between 2004 and 2024. We only included patients who had been in the SOURCE system for at least six months prior to the diagnosis date and had no prior record of RD repair surgery. We excluded patients with an unknown laterality of the primary RD repair, history of PVR, proliferative diabetic retinopathy, serous RD, retinal dialysis, or ocular trauma.

Methods: The exposure variable of interest was systemic methotrexate use, as documented from the medication list. The outcome of interest was the presence of new-onset PVR within 6 months of surgery. Incident PVR was modeled in log binomial regression models that included terms for systemic methotrexate use and method of primary RD repair. Regression models included inverse probability weights based on propensity scores for systemic methotrexate use. We conducted similar analyses for other antimetabolite agents (e.g., azathioprine, mycophenolate mofetil).

Main outcome measures: Cumulative incidence of PVR and adjusted risk ratio (RR) of developing PVR within 6 months of the initial RD.

Results: Of the 2674 eligible patients, 48 (1.8%) were taking systemic methotrexate at the time of RD repair. The 6-month cumulative incidence of PVR following the primary RD was 4.2% for patients taking systemic methotrexate compared with 9.2% for those not taking methotrexate (adjusted RR 0.58, 95% CI 0.47-0.71). Similar results were found for azathioprine (adjusted RR 0.28, 95% CI 0.22-0.37) but not mycophenolate mofetil (adjusted RR 0.93, 95% CI 0.77-1.11).

Conclusions: Patients taking systemic methotrexate or azathioprine were significantly less likely to develop PVR within six months of primary RD compared with those not taking methotrexate or azathioprine.