Skip to main content
  • By Ramana S. Moorthy, MD
    Uveitis

    This retrospective study found that epiretinal membrane (ERM) is a common complication of uveitis that is associated with older patient age, intermediate uveitis, posterior uveitis, panuveitis, male sex and previous cataract surgery. It can also contribute independently to vision loss in uveitic eyes.

    The authors reviewed the charts of patients with uveitis examined from January 2008 to December 2011 at the National Eye Institute for the presence of ERM using SD-OCT. The investigators carefully performed a masked analysis of the SD-OCT images and considered only those ERMs found in the central macular cube with evidence of contractility, including distortion, corrugation or flattening of the inner retinal surface. 

    Of the 598 patients, 248 (41 percent) had ERM in at least one eye. These patients were significantly older (52 vs. 42 years) with a longer mean duration of uveitis (8.3 vs. 5.4 years). Using the Standardization of Uveitis Nomenclature (SUN) classification system, the prevalence of ERM was 28.1% in anterior uveitis, 57% in intermediate uveitis and 43.5% in posterior uveitis and panuveitis. 

    Unilateral ERMs were found in 141 subjects. When compared to fellow eyes without ERM, eyes with ERM were significantly more likely to have undergone vitrectomy, retinal laser or intraocular injection (steroids or other medication).

    Eyes with ERM (20/76) had significantly lower mean visual acuity than those without ERM (20/60). Among patients with ERM, the presence of active vitreous inflammation, greater central subfield thickness (200 to 350 microns), anatomic classification of posterior uveitis, history of cataract surgery and longer duration of uveitis all had a significant negative effect on visual acuity.

    ERM in such a large cohort of uveitis patients using SD-OCT has not been previously studied particularly using the tools of nomenclature. The large cohort has allowed a very detailed multivariate analysis of ERMs in uveitis, which is extremely useful. 

    Most of the findings of this study confirm the association of ERM with parameters such as greater duration of disease, previous cataract or vitrectomy, previous intraocular injections, and posterior and panuveitis. The strong association of intermediate uveitis and ERM is interesting and may reflect the role of vitritis in the development of vitreoretinal interface changes. Furthermore, this study points out the increased sensitivity of SD-OCT for detecting ERMs compared to fundus photos. 

    As a retina-trained uveitis specialist, I found that this study validated my own clinical gestalt over the years. ERMs are common in uveitis patients. They are not often symptomatic and, thus, do not usually require treatment.

    If, however, pars plana vitrectomy is being performed in an eye that has ERM for other reasons (e.g., diagnostic vitrectomy, repair of rhegmatogenous retinal detachment, etc.), I will carefully evaluate the macula for vitreo-retinal interface changes preoperatively and intraoperatively, and probably would peel the ERM (including the internal limiting membrane) because leaving such changes behind will likely result in worsening of the ERM and vision after vitrectomy surgery (a risk factor in and of itself for ERM in this study).