Relation of anatomy with function following the surgical treatment of idiopathic epiretinal membrane: a multicenter retrospective study.

Purpose: To investigate the prognostic factors associated with functional and anatomical outcomes and to assess the longitudinal course of visual acuity and retinal morphology after vitreoretinal surgery for idiopathic epiretinal membrane (ERM).

Methods: This multicenter, retrospective study included a total of 634 eyes who underwent surgery for idiopathic ERM in 22 academic centers nationwide in Turkey. Data on best-corrected visual acuity (BCVA) and optical coherence tomography features (central foveal thickness (CFT), ERM and foveal contour morphology, ellipsoid zone (EZ) integrity) were collected and compared at baseline, 6-month, 12-month, and 24-month follow-ups. Prognostic factors for functional (having ≥ 20/25 Snellen BCVA) and anatomical (having normal/shallow foveal contour) recoveries after surgery were investigated by means of multivariate regression analyses. A cutoff value of preoperative BCVA optimizing functional recovery was calculated using receiver operating characteristic curve analysis.

Results: At a median follow-up of 24 months, 37.4% of the eyes achieved ≥ 20/25 BCVA and 54% regained normal or shallow foveal contour. Functional recovery was more likely in eyes with better baseline BCVA and intact EZ (R = 0.356, p < 0.001). The cutoff baseline BCVA value for good visual prognosis was 0.35 logarithm of the minimum angle of resolution (Snellen 20/44) (sensitivity 60%, specificity 85%, p < 0.001). Anatomical recovery was negatively associated with advanced age, higher baseline CFT, foveal herniation-type ERM morphology, and internal limiting membrane (ILM) peeling (R = 0.225, p < 0.001). The negative effect of ILM peeling on anatomical recovery was not significant after the first postoperative year (p = 0.05). Mean BCVA values and foveal morphology progressively improved at each visit. Cases with convex baseline foveal contour continued to change towards normal foveal depression over 24 months of follow-up, which took longer than the eyes with shallow/flat contoured cases. One-third of eyes with severe baseline EZ defects showed recovery at follow-up and achieved significantly greater visual acuity gains than the remaining eyes with persistent defects (p < 0.001).

Conclusions: Functional and anatomical restoration of the eyes appears to be a slow process after ERM surgery. This process may take much longer in eyes with worse foveal morphology at baseline. Although photoreceptor disruption may be reversible in some eyes, full functional recovery is unlikely when it persists.