Microstructural and microperimetric comparison of internal limiting membrane peeling and insertion in large...

The internal limiting membrane (ILM) insertion technique was widely used to treat large macular hole (MH) for the high closure rate. However, the prognosis of closed MH after ILM insertion compared to ILM peeling remains controversial. This study aimed to compare foveal microstructure and microperimeter in large idiopathic MH surgically closed by ILM peeling and ILM insertion.

This retrospective, non-randomized, comparative study included patients with idiopathic MH (minimum diameter ≥ 650 μm) who underwent primary pars plana vitrectomy (PPV) with ILM peeling or ILM insertion. The initial closure rate was recorded. Patients with initially closed MHs were divided into two groups according to the surgery methods. The best-corrected visual acuity (BCVA), optical coherence tomography (OCT) and microperimeter-3 (MP-3) outcomes of two groups were compared at baseline, 1 and 4 months postoperatively.

For idiopathic MH (minimum diameter ≥ 650 μm), ILM insertion had a significantly higher initial closure rate than ILM peeling (71.19% vs. 97.62%, P = 0.001). Among 39 patients with initially closed MHs who were on regular follow-up, twenty-one were assigned to the ILM peeling group and 18 to the ILM insertion group. Postoperative BCVA improved significantly in both groups. The final BCVA (logMAR) (0.40 vs. 0.88, P < 0.001), macular hole sensitivity (19.66 dB vs. 14.14 dB, P < 0.001), peripheral sensitivity of macular hole (24.63 dB vs. 21.95 dB, P = 0.005), and fixation stability (FS) within 2 degrees (82.42% vs. 70.57%, P = 0.031) were significantly better and external limiting membrane (ELM) defect (330.14 μm vs. 788.28 μm, P < 0.001) and ellipsoid zone (EZ) defect (746.95 μm vs. 1105.11 μm, P = 0.010) were significantly smaller in the ILM peeling group than in the ILM insertion group.

For initially closed MHs (minimum diameter ≥ 650 μm), both ILM peeling and ILM insertion significantly improved the microstructure and microperimeter in the fovea. However, ILM insertion was less efficient at microstructural and functional recovery after surgery.