Abstract
Purpose :
To evaluate the predictive parameter among preoperative measurements that best predicts postoperative visual outcome in the epiretinal membrane (ERM) surgery.
Methods :
Thirty-three consecutive patients with idiopathic unilateral ERM patients between 2015 and 2018 were enrolled. Moreover, age-matched healthy eyes were selected as the control group. Based on preoperative optical coherence tomography (OCT), we further divided the patients with ERM into two groups: type 1, loosely attached ERM, and type 2, tight adherent ERM. We documented the vision and thickness of various retinal layers: nerve fiber layer, ganglion cell layer, inner plexiform layer (GCL+IPL), inner nuclear layer (INL), outer retinal layer (ORL), and retinal pigment epithelium/Bruch complex layer before and after the surgery. We analyzed the eyes that underwent combined ERM and cataract surgery or ERM surgery with subsequent cataract surgery. The association between postoperative visual acuity and these variables was analyzed using multiple linear regression analysis.
Results :
We identified 11 eyes with type 1 adhesion and 22 eyes with type 2 adhesions. Both groups demonstrated significantly thicker GCL+ IPL thicknesses than the controls. The postoperative best-corrected visual acuity (BCVA) in type II patients was worse than type I pateints (p=0.026). The preoperative GCL + IPL and INL layers were significantly thicker in type II patients (p=0.005 and p= 0.033, respectively) than in type I patients. Multiple linear regression analysis showed that GCL+IPL thickness was an independent predictor of postoperative visual acuity (VA).
Conclusions :
Idiopathic ERM demonstrated significantly thicker inner retinal layers (GCL + IPL and INL). However, the ORL thickness was similar between the normal eyes and ERM eyes. The preoperative GCL + IPL and INL layers were significantly thicker in patients with type II ERM than that in patients with type I ERM. Preoperative GCL + IPL thickness is an independent prognostic factor for the ERM surgery.
This abstract was presented at the 2022 ARVO Annual Meeting, held in Denver, CO, May 1-4, 2022, and virtually.