ILM peeling in proliferative diabetic retinopathy: a meta-analysis
Reviewed by Sofia Rokerya

This meta-analysis highlights the efficacy of internal limiting membrane (ILM) peeling on treating posterior segment complications other than macular oedema in patients with proliferative diabetic retinopathy (PDR) treated with vitrectomy. PubMed, EMBASE and Cochrane Library were searched for databases. Eleven studies comparing vitrectomy with and without ILM peeling (ILM-P and ILM-NP groups, respectively) were included in this meta-analysis. The primary outcome was the final best-corrected visual acuity (BCVA). Secondary outcomes were final central retinal thickness (CRT) and formation of the postoperative epiretinal membrane (ERM).The postoperative BCVA at 6 months was better in those who underwent ILM peeling (95% confidence interval (CI): -0.51 to -0.15) than in those did not, whereas the final BCVA was not significantly different between the groups (95% CI: -0.30 to 0.00). The final CRT was significantly thin (95% CI: -61.82 to -35.63), with a low occurrence of postoperative ERM in the ILM peeling group (95% CI: 0.12-0.26). In the subgroup analysis based on PDR complications (vitreous haemorrhage or tractional retinal detachment (T-RD)), it was found that ILM peeling significantly improved final BCVA specifically in patients with T-RD. This benefit was not observed in patients undergoing surgery for vitreous haemorrhage. However, the risk of postoperative ERM was consistently reduced with ILM peeling across both types of PDR complications. In conclusion, the meta-analysis identified that ILM peeling effectively prevented early postoperative ERM formation in patients with PDR complications; however, the benefit of visual improvement was not evident at follow-ups longer than 6 months. Limitations: most studies included were non-randomised retrospective studies, and the cohort in each study was small. There was heterogeneity among the studies, including the follow-up periods and inclusion criteria for eligibility. The GRADE assessment indicated moderate-to-high quality evidence regarding the effect of ILM peeling on final BCVA, but low-to-moderate quality evidence on final CRT and postoperative ERM. However, despite exclusion of poor-quality studies, the final analysis did not show any substantial change. The outcomes were reported at a short postoperative period (6 months). A language bias existed because only studies written in English were included within the meta-analysis.

Efficacy of internal limiting membrane peeling in eyes with proliferative diabetic retinopathy: a meta-analysis.
Chung Y-R, Lee E, Kim H, et al
OPHTHALMOLOGICA 
2025;248(5–6):386–400.