Optical coherence tomography angiography features of the chorioretinal complex and choriocapillaris perfusion before and after vitrectomy with conventional versus fovea-sparing internal limiting membrane peeling for idiopathic macular hole

Abstract

Purpose: To assess optical coherence tomography angiography (OCTA)-measured changes in the chorioretinal complex and choriocapillaris perfusion density in the macula before and after vitrectomy with fovea-sparing versus conventional internal limiting membrane (ILM) peeling for idiopathic macular hole (IMH). Material and Methods: Eyes with stage-2 to stage-4 holes as per the classification by Gass received 25-G vitrectomy with conventional or fovea-sparing ILM peeling and gas tamponade with 20% SF6 or 15% С3F8. IMH diameter, foveal avascular zone (FAZ) area in the deep retinal plexus and choriocapillaris perfusion density (CPD) were assessed before and 1 month after surgery. Results: Totally, 70 patients had an IMH surgery in 71 eyes. The mean age ± standard deviation (SD) was 65.7 ± 6.8 years, median IMH duration (interquartile range or IQR), 3.0 (1.0-6.0) months, median best-corrected visual acuity or BCVA (IQR), 0.19 (0.1-0.25), and median maximum IMH diameter (IQR), 673.5 (549.5–1010.5) µm. In eyes with IMH and fellow eyes, the median FAZ area (IQR) was 0.51 (0.15–0.53) mm2, and 0.46 (0.10–0.74) mm2, respectively (р = 0.49), and mean CPD ± SD, 0.11 ± 0.06, and 0.29 ± 0.13 (р = 0.0001), respectively. Thirty-four eyes received conventional ILM peeling and 37 eyes, fovea-sparing ILM peeling, and there was no significant intergroup difference in baseline characteristics. One month after surgery, IMH closure was achieved in 63/71 eyes (i.e., the closure rate was 88.7% for total operated eyes, and 88.2% and 89.2%, respectively, for eyes in conventional ILM peeling and fovea-sparing ILM peeling groups), and median BCVA (IQR) improved to 0.60 (0.4–0.8) (р = 0.00001). After IMH closure, in operated eyes, median FAZ area (IQR) decreased to 0.30 (0.12–0.6) mm2, but the difference was not significant, whereas mean CPD ± SD increased significantly from 0.11 ± 0.06 to 0.25 ± 0.10 (р = 0.0001). No significant difference in OCTA-based retinal microcirculation and choriocappillaris characteristics was observed between the conventional ILM peeling and fovea-sparing ILM peeling groups. Conclusion: The presence of macular hole is accompanied by abnormal perfusion in the choriocapillaris, but the CPD recovers after IMH closure. Postoperative CPD recovery is not influenced by the type (conventional or fovea-sparing) of ILM peeling.