Internal Limiting Membrane Flaps in Macular Hole Surgery: A Systematic Review and Individual Participant Data Meta-analysis

To compare anatomic and visual outcomes of internal limiting membrane (ILM) flaps versus peeling in macular hole surgery. We also assessed the impact of hole size, symptom duration, and different flap types on outcomes. The benefit of ILM flaps over standard ILM peeling in idiopathic, full-thickness macular holes (iFTMHs) remains unclear. Prospectively registered systematic review and individual participant data (IPD) meta-analysis of randomized controlled trials comparing conventional ILM peeling with ILM flaps, as well different ILM flap subtypes, in adults undergoing primary iFTMH surgery (CRD42023494971). No exclusions were made based on hole size, symptom duration, or perioperative choices. Searches were performed in MEDLINE, Embase, Cochrane Library, and trial registries (January 2000-March 2023). Critical outcomes were hole closure and postoperative visual acuity at 6 months or nearest time point. Multilevel regression models were adjusted for age, sex, hole size, lens status, and preoperative visual acuity, allowing for non-linear effects. Evidence was appraised with Cochrane Risk of Bias, GRADE, and ICEMAN tools. Subgroup analyses considered hole size, symptom duration, flap subtypes, tamponade choice, and risk-of-bias. Of 14 eligible trials, 13 provided IPD for 792 eyes. Most (68.3%) had MLD ≥500 μm, with limited representation of holes <400 μm and ≥900 μm. The adjusted odds ratio (OR) for primary closure with ILM flap versus peeling was 4.80 (95% CI, 2.77-8.30; P<0.001), with a relative risk of 1.26 (1.20-1.30) (GRADE: moderate-certainty), and a number needed to treat of 6. Compared to peeling, the ILM flap group showed better postoperative visual acuity at 3-6 months, with a mean difference of -0.14 logMAR (-0.18 to -0.09; P<0.001), about 7 letters ETDRS (GRADE: moderate-certainty). ILM flaps were likely more beneficial for holes ≥500 μm (OR for closure: 3.14 to 9.64, P<0.001; MD in vision: -0.23 to -0.13, P<0.001). Non-linear analyses suggested probable benefits across a broader range of hole sizes (ICEMAN: moderate-confidence). Results were consistent across risk-of-bias assessments, with no significant differences between ILM flap techniques. ILM flaps likely improve closure and visual recovery compared to peeling alone in iFTMH, with greater effects likely in holes >500 μm.