Assessment of Macular Hole Closure Speed Using Trans-tamponade OCT and 3-Dimensional Macular Hole Segmentation in the DISCOVER Study.

Abstract

Introduction: To explore opportunities for individualized postoperative positioning duration in macular hole (MH) surgery. Methods: This post hoc analysis comprised eyes that had full-thickness MH (FTMH) repair in the prospective DISCOVER intraoperative optical coherence tomography (OCT) study. Preoperative spectral-domain OCT (SD-OCT) and postoperative day 1 (POD1) trans-tamponade OCT were analyzed. Preoperative SD-OCT macular cubes were imported and analyzed using a machine learning-enhanced segmentation platform. Nine preoperative segmentation parameters were analyzed (maximum foveal height, apex area, base area, central foveal area, maximum apex diameter, maximum base diameter, maximum depth, minimum foveal width, volume). In addition, 2 new metrics-the macular hole index and tractional hole index-were calculated. POD1 trans-tamponade OCTs were obtained and evaluated for hole closure. Results: Of the 66 eyes with an FTMH, 16 (24%) had a chronic MH and 5 (8%) were reoperations from FTMH nonclosure after previous surgery. Nine eyes (14%) had an open MH on POD1 trans-tamponade OCT (6 chronic MHs [66%]; 2 reoperations [22%]); the remaining 57 MHs (86%) were closed. Multiple segmentation parameters were significantly associated with POD1 closure. Conclusions: Volumetric MH measurements and trans-tamponade POD1 OCT closure status are important in predicting MH closure speed and the need for postoperative positioning. Individualized positioning duration in MH could lower perioperative morbidity and vastly enhance quality of life.