Prevalence and Risk Factors of Epiretinal membrane in the Multicenter Uveitis Steroid Treatment (MUST) Trial

Abstract

Purpose: To report baseline prevalence and associated risk factors of epiretinal membrane (ERM) in patients with severe active non-infectious intermediate, posterior or panuveitis recruited into the Multicenter Uveitis Steroid Treatment (MUST) Trial.

Methods: All participants underwent a standardized interview, systemic examination and ophthalmic examination. Baseline OCT images were graded at the Reading Center according to the presence and severity of ERM, plus any other associated ERM complications such as macular traction. Generalized estimating equations were used to fit logistic regression models to assess risk factors while accounting for between eye correlation in patients with bilateral uveitis.

Results: Of the 479 eyes with uveitis in the MUST Trial, 435 eyes (91%) from 243 individuals had OCT images that were gradable for ERM at randomization. Time from uveitis onset > 5 years, posterior synechiae, visual acuity (VA) < 20/100, prior IOP-lowering surgery, cataract, active uveitis, and any systemic disease were associated with increased risk of inability to assess ERM with OCT.<br /> Of the 435 gradeable eyes, 126 (29%) had an ERM. In a multivariable analysis, having diabetes (OR=2.21, p=0.043), age > 50 years (OR = 2.95, p < 0.001), the presence of retinal vasculitis (OR=2.68, p=0.013), macular edema (OR=1.87, p=0.013) and uveitis activity (OR = 1.83, p = 0.050) were associated with the presence of ERM. VA worse than 20/100, cataract [including prior cataract surgery], and Vogt-Koyanagi Harada disease were associated with increased ERM prevalence; however the association was abrogated by adjustment for other risk factors in the final multivariable model.

Conclusions: Detection of ERM by OCT in patients with uveitis may be limited by the optical impact of uveitic complications such as cataract and posterior synechiae. ERM is a common complication of intermediate, posterior and panuveitis, and was associated with increasing age, diabetes, macular edema and retinal vasculitis at randomization. Additional analysis of longitudinal data is needed to estimate the incidence of ERM and determine the residual impact of ERM on VA (once uveitis activity has been treated) and its outcome over time.