Optical Coherence Tomography Predictors of Anatomical and Functional Recovery After Macular Hole Surgery

To examine whether preoperative optical coherence tomography (OCT)-derived macular hole parameters and postoperative outer retinal restoration are associated with anatomical closure and visual recovery after macular hole surgery and to determine whether preoperative geometry and preoperative symptom duration predict different outcome domains.

This retrospective study included 48 eyes of 45 patients who underwent surgery for macular hole between January 1, 2013, and August 1, 2019, and completed at least 6 months of follow-up. Surgery consisted of 23-gauge pars plana vitrectomy (PPV) with internal limiting membrane (ILM) peeling; an inverted ILM flap was additionally used in 18 eyes (37.5%) and autologous blood in 4 eyes (8.3%). Combined phacoemulsification was performed in 12 eyes (25.0%). Best-corrected visual acuity (BCVA) was recorded preoperatively and at Postoperative Months 1, 3, and 6. Preoperative OCT measurements included minimum hole diameter, basal hole diameter, hole height, macular hole index (MHI), tractional hole index (THI), diameter hole index (DHI), and hole form factor (HFF). Between-group differences are reported with Cohen's d and 95% confidence intervals (CIs), and a multivariable linear regression model was constructed for Month-6 BCVA.

The cohort comprised 23 women (51.1%) and 22 men (48.9%) with a mean age of 68.4 ± 8.7 years. Anatomical success was achieved in 38 eyes (79.2%), and functional success in 36 eyes (75.0%). BCVA improved significantly through postoperative Month 6. Larger minimum and basal hole diameters were associated with worse postoperative BCVA, whereas higher MHI and THI values were associated with better postoperative BCVA. Eyes with preserved external limiting membrane (ELM) and/or ellipsoid zone (EZ) integrity at postoperative Months 3 and 6 had significantly better visual outcomes. Symptom duration was the dominant correlate of anatomical closure, and its association with BCVA strengthened across follow-up (r = 0.523 preoperatively to r = 0.808 at Month 6) while that of THI weakened over the same interval. HFF was the only preoperative geometric parameter that differed between closure patterns (d = 0.86; 95% CI for the mean difference, 0.024 to 0.272; p = 0.036), yet it showed no association with BCVA at any time point. In a multivariable analysis, preoperative BCVA and Month-6 EZ integrity were the only independent correlates of Month-6 BCVA (adjusted R-squared = 0.76).

Preoperative OCT morphology is clinically informative in macular hole surgery. Minimum and basal hole diameters, MHI, and THI may help estimate postoperative visual recovery, while restoration of the ELM and EZ appears closely related to better functional outcome. In this cohort, the determinants of the two outcome domains were largely dissociable: preoperative geometry chiefly tracked functional recovery, whereas symptom duration dominated anatomical closure. HFF behaved as a marker of outer retinal restoration rather than of visual acuity.