Related Experiment Video
Updated: Aug 5, 2026
Investigating Retinal Circuits and Molecular Localization by Pre-Embedding Immunoelectron Microscopy
Published on: July 12, 2024
Internal Limiting Membrane Loop Configurations in Epiretinal Membrane: Prevalence, Imaging Associations and Optical
Anamika Patel1, Alessandro Feo1, Néda Abraham1
1From the Retina Division (A.P., A.F., N.A., P.T., M.A., S.F., H.H.C., D.S.), Stein Eye Institute, University of California, Los Angeles, David Geffen School of Medicine at UCLA, Los Angeles, California, USA.
Purpose:
To characterize internal limiting membrane (ILM) loop configurations on optical coherence tomography (OCT) in eyes with epiretinal membrane (ERM) and evaluate their association with visual acuity (VA), ILM tear, posterior vitreous detachment (PVD), and ERM grade.
Design:
Retrospective observational cross-sectional study.
Subjects:
A total of 138 eyes from 105 patients with nonsurgical ERM at a single academic center.
Methods:
All eyes underwent spectral-domain OCT. ILM loops were identified on cross-sectional B-scans as loop-like elevations or folds of the ILM at the vitreoretinal interface. Loop number and macular distribution were recorded. PVD was graded using the Engelbert classification and ERM severity using the Govetto classification. Associations between ILM loops and VA, ILM tear, PVD grade, and ERM grade were analyzed using Fisher's exact test and the Mann-Whitney U test.
Main Outcome Measures:
Prevalence of ILM loops and their association with Snellen VA, ILM tear status, PVD grade, and ERM grade.
Results:
ILM loops were identified in 21 of 138 eyes (15.2%). Loop-positive eyes demonstrated a median VA of 0.10 logMAR (IQR, 0.00-0.30; approximately 20/25 Snellen) compared with 0.18 logMAR (IQR, 0.00-0.30; approximately 20/30 Snellen) in loop-negative eyes (P = .683). ILM loops were significantly more prevalent in eyes with ILM tear (16/32 [50.0%] vs 5/106 [4.7%]; odds ratio, 20.2; 95% CI, 6.50-62.80; P < .0001). Multiple loops were present in 15/21 eyes (71.4%) and solitary loops in 6 eyes (28.6%). Complete PVD was present in 19 of 21 loop-positive eyes (90.5%) compared with 83 of 117 loop-negative eyes (70.9%), although this did not reach significance (odds ratio, 3.89; 95% CI, 0.86-17.63; P = .065). Loop-positive eyes demonstrated higher ERM grades than loop-negative eyes (grade 3 in 7/21 [33.3%] vs 10/117 [8.5%]; P = .035).
Conclusions:
ILM loops are a distinct OCT feature in ERM and are strongly associated with ILM tears. These configurations may represent ILM retraction following ILM tear, although the cross-sectional design precludes determination of a temporal or causal relationship. These interesting lesions are a common OCT finding, and therefore their recognition and understanding will enhance and support evaluation by retinal physicians.
