Impact of internal limiting membrane peeling on long-term functional outcomes of vitrectomy for idiopathic epiretinal membrane

Actuality. Idiopathic epimacular fibrosis (EMF) is a fibrocellular formation on the inner boundary membrane of the retina and leads to a decrease in central visual acuity, metamorphosis and thickening of the macula. The standard of treatment is vitrectomy with EMF removal, however, the need for routine MIC peeling remains controversial. Peeling reduces the risk of EMF recurrence, while macular photosensitivity may decrease due to the phenomenon of dissociation of the nerve fiber layer (DONFL – Dissociated Optic Nerve Fiber Layer). Goal. To determine the effect of VPM peeling on the long-term functional outcomes of vitrectomy in patients with idiopathic EMF and to identify morphological predictors of the outcome. Materials and methods. A retrospective cohort study included 96 patients (96 eyes) with idiopathic EMF operated on in 2018-2019. All underwent 25G vitrectomy with induction of posterior vitreous detachment, removal of epimacular fibrosis, in one group an additional circular peeling of the VPM was performed after brilliant blue dual staining (n = 48), in the second group fibrosis removal was limited (n = 48). The follow-up was at least 36 months. The ophthalmological examination included maximally corrected visual acuity (MCI), central retinal thickness, microstructural features according to spectral optical coherence tomography (OCT), and average macular photosensitivity according to microperimetry. EMF staging was performed according to the Govetto classification [3], taking into account the presence of an ectopic inner foveal layer (EIFL – Ectopic Inner Foveal Layer), ellipsoid zone defects (EZ), hyperreflective foci, and stratification of the nerve fiber layer. The main endpoints are ICOSIS and average macular photosensitivity after 36 months, the secondary endpoints are the frequency of EMF recurrence. For statistical analysis, a t-test was used, either the Mann–Whitney criterion, the Fisher exact criterion and the Pearson or Spearman correlation (p<0.05). Results. The groups did not differ in age, initial MSCI, central retinal thickness, and EMF stage, however, hyperreflective foci (43.8% vs. 20.8%) and EZ defects (39.6% vs. 10.4%) were more common in the peeling group. After 36 months, both groups showed significant improvement compared to baseline, but the MCI was higher in the group without peeling (0.56±0.20 versus 0.45±0.20, p = 0.031), and the average photosensitivity was 28.0±0.97 dB versus 26.4±0.19 dB (p = 0.002). EMF recurrence occurred in 14.6% of patients without peeling and in 4.2% of patients with peeling (p<0.05). The initial ICOS was positively correlated with the final ICOS (r = 0.769) and photosensitivity (r = 0.402), the presence of an EZ defect was associated with worse ICOS (r = -0.223) and photosensitivity (r = -0.372), HPM peeling was associated with lower ICOS (r = -0.338) and photosensitivity (r = -0.883). Conclusion. The rejection of VPM peeling during idiopathic EMF surgery provided better long-term functional parameters (ICOS and photosensitivity), but was accompanied by a higher recurrence rate. The predictors of the outcome were the initial visual acuity, the stage of EMF, the preservation of the ellipsoid zone, the presence of EIFL and hyperreflective foci. The decision on HPV peeling should be made individually, taking into account the morphological status of the macula and the risk of recurrence.