Abstract
Purpose
To describe and compare the histopathological and ultrastructural findings of the internal limiting membrane (ILM) with and without epiretinal membrane (ERM), and investigate differences between idiopathic and secondary ERM, using light and transmission electron microscopy.
Design
Cross-sectional analytical study, with pathologists and image assessors masked to clinical and imaging information.
Participants
Consecutive participants undergoing pars plana vitrectomy (PPV) with peeling of ILM, ERM, or both were enrolled. One eye per participant was included.
Methods
Specimens were collected intraoperatively following a standardized protocol and analyzed for cellular morphology, extracellular matrix composition, and ultrastructural alterations. Fisher exact test and unpaired t tests were used for group comparisons. Univariate and multivariate regression models assessed associations between clinical and imaging characteristics and histopathological features, with significance set at P < 0.001.
Main Outcome Measures
Cellular and extracellular matrix ultrastructural features and their associations with clinical characteristics and imaging findings.
Results
A total of 107 eyes (90 ILM with ERM; 17 ILM without ERM) from 117 participants were analyzed. Mean age was 64.4 ± 10.9 years, and 59.8% were female. Internal limiting membrane with ERM exhibited cellular proliferation with fibrous astrocytes, fibroblasts, myofibroblasts, fibrocytes, and retinal pigment epithelium (RPE) cells, accompanied by extracellular matrix remodeling and newly formed collagen. Common ultrastructural features included ILM vacuolization, Müller cell fragments, electron-lucent bands, and disorganized collagen fibrils. Intraocular lens implantation (incidence rate ratio [IRR] = 0.333, P = 0.012) and proliferative diabetic retinopathy (IRR = 0.190, P = 0.002) were associated with reduced ILM vacuolization. Secondary ERM demonstrated significantly higher proportions of RPE cells (40.5% vs 2.8%, P < 0.001), intraretinal cysts (73.0% vs 31.8%, P < 0.001), and subfoveal ellipsoidal band loss (64.1% vs 15.9%, P < 0.001) compared to idiopathic ERM.
Conclusions
This study provides a comparative ultrastructural assessment of ILM and ERM, revealing distinct differences between idiopathic and secondary ERM. These findings suggest differing underlying biological processes that merit future mechanistic investigation. According to fibrocellular changes reported, macular disruption observed on OCT in older eyes may indicate susceptibility to secondary ERM, raising the possibility, yet to be validated, that ILM peeling during PPV for other indications could offer prophylactic benefit.
Financial Disclosure(s)
The author has no/the authors have no proprietary or commercial interest in any materials discussed in this article.
Keywords: Electron microscopy, Epiretinal membrane, Histopathology, Internal limiting membrane, Ocular pathology
Epiretinal membrane (ERM), a thin fibrocellular layer, is a common retinal condition frequently encountered in ophthalmologic practice and can significantly impair vision.1 Progressive membrane thickening may lead to metamorphopsia and visual loss if left untreated.2 The prevalence of ERM varies widely, from 1.9% to 28.9%.1,3, 4, 5, 6, 7, 8 Epiretinal membrane is generally classified into primary (idiopathic) and secondary forms. Idiopathic ERM has been associated with aging and posterior vitreous detachment (PVD).9,10 In contrast, secondary ERM often arises in the setting of underlying ocular conditions such as retinal vascular disease, intraocular inflammation, retinal breaks, or retinal detachment (RD).11,12 Systemic factors like diabetes mellitus (DM), hyperlipidemia, and hypercholesterolemia have also been implicated as risk factors.1,7,8,13 Despite these associations, the pathophysiology behind idiopathic ERM formation remains poorly understood.
Epiretinal membrane diagnosis and staging is typically assessed using OCT with aids of fundus photography. The standard treatment is pars plana vitrectomy (PPV) with membrane peeling, frequently combined with internal limiting membrane (ILM) removal to reduce recurrence and improve visual outcomes.14,15 In macular hole surgery, ILM flaps are commonly used to facilitate closure, and studies have reported that these procedures may help prevent secondary ERM formation.16,17 These postoperative observations suggest a possible association between ILM pathology and ERM development, particularly in diseases involving ILM defects such as retinal breaks or macular holes. Prior studies have suggested that the ILM may serve as a scaffold for cellular migration and inflammatory activity, contributing to ERM pathogenesis.11
Previous studies have described a variety of cellular and structural components in ERM and ILM.18,19 The presence of retinal pigment epithelium (RPE) in ERM has been attributed to migration from retinal breaks, supporting a role in ERM formation.18 Histopathological changes in the ILM have also been reported, such as vacuolization, fibrillary structures, hyaloid adhesion, and alterations in organelles and structural proteins, including newly formed collagen.18, 19, 20, 21, 22 While these pathological features have been studied using light microscopy (LM), electron microscopy (EM), and imaging modalities such as OCT, fundus photography, and OCT angiography, a comprehensive comparison between diseased and normal ILM remains limited.19,20 Comparing ILM ultrastructure in eyes with and without ERM may help identify key pathological differences, thereby addressing gaps in our understanding of the inflammatory mechanisms involved in ERM formation. Furthermore, comparing idiopathic and secondary ERM in terms of ILM and ERM morphology may help distinguish their underlying pathophysiological mechanisms. Stratifying pathological findings by baseline characteristics and clinical imaging features may yield additional insights into risk factors for idiopathic ERM.
Over the past 2 decades, the lack of integrated analyses linking ultrastructural pathology with clinical and imaging characteristics has limited understanding of ERM formation. The present study addresses this gap by providing one of the largest systematically collected ILM and ERM specimen sets, analyzed using both LM and transmission EM (TEM) under a standardized, published protocol. This study aimed to describe the pathological characteristics of ILM and ERM and correlate these findings with baseline and imaging parameters. Our primary objective was to analyze the EM morphology and pathological changes of cells and the extracellular matrix in ILM with and without ERM, and in idiopathic versus secondary ERM. Additionally, we aimed to investigate the association of these findings with clinical characteristics and imaging features, to advance the understanding of ERM pathogenesis and generate insights relevant to future ERM-preventive or antifibrotic therapeutic strategies.
Methods
Written informed consent was obtained from all participants prior to any procedures. The study was approved by the Institutional Review Board of the Faculty of Medicine, Chulalongkorn University, Bangkok, Thailand (approval no.: 0948/2023), and conducted in accordance with the Declaration of Helsinki (2013), the Belmont Report, Council for International Organizations of Medical Sciences Guidelines, and the International Conference on Harmonization Good Clinical Practice standards.
Study Design and Setting
This was a cross-sectional analytical study, with both the pathologist and image assessors masked to clinical information. The study described and compared the histopathological findings of the ILM with and without ERM, and between idiopathic and secondary ERM, using light and TEM in eyes undergoing membrane peeling. The recruitment process, sample size calculation, eligibility criteria, and masking process are detailed in our published protocol.23 We recruited consecutive participants with ERM or those requiring ILM peeling, such as for macular hole repair or prophylactic ILM peeling during RD surgery, who underwent PPV with membrane peeling. The inclusion and exclusion criteria are detailed in Supplemental Table 1, available at www.ophthalmologyscience.org. Internal limiting membrane with and without ERM were stratified using standardized OCT diagnostic criteria adapted from Govetto et al24, and the corresponding clinical diagnosis–based stratification is also presented in Table 1 of the baseline characteristics.
Table 1.
Baseline Characteristics of Included Participants
| Baseline Characteristics | Surgical Diagnosis |
OCT Diagnosis of ERM |
||||
|---|---|---|---|---|---|---|
| ILM Only n = 23 (21.5%) | ILM with ERM n = 84 (78.5%) | P | OCT Negative for ERM n = 17 (15.9%) | OCT Positive for ERM n = 90 (84.1%) | P | |
| Age | 58.9 (10.4) | 65.9 (10.1) | 0.004 | 64.7 (8.9) | 64.3 (10.8) | 0.900 |
| Female | 13 (56.5%) | 51 (60.7%) | 0.811 | 14 (82.4%) | 50 (55.6%) | 0.057 |
| Right eye | 11 (47.8%) | 38 (45.2%) | 1.000 | 8 (47.1%) | 41 (45.6%) | 1.000 |
| Lens status | ||||||
| Aphakia | 1 (4.3%) | 6 (7.1%) | 0.703 | 0 (0.0%) | 7 (7.8%) | 0.156 |
| Cataract | 13 (56.5%) | 39 (46.4%) | 12 (70.6%) | 40 (44.4%) | ||
| IOL | 9 (39.1%) | 39 (46.4%) | 5 (29.4%) | 43 (47.8%) | ||
| Vitreous status | ||||||
| Vitreous humor | 15 (65.2%) | 65 (77.4%) | 0.124 | 12 (70.6%) | 68 (75.6%) | 0.526 |
| Postvitrectomized | 8 (34.8%) | 14 (16.7%) | 5 (29.4%) | 17 (18.9%) | ||
| Silicone oil | 0 (0.0%) | 5 (6.0%) | 0 (0.0%) | 5 (5.6%) | ||
| Surgery type | ||||||
| Combined cataract and vitrectomy | 10 (43.5%) | 33 (39.3%) | 0.811 | 10 (58.8%) | 33 (36.7%) | 0.109 |
| Vitrectomy | 13 (56.5%) | 51 (60.7%) | 7 (41.2%) | 57 (63.3%) | ||
| Retinal break seen during surgery | ||||||
| Not seen | 6 (26.1%) | 56 (66.7%) | < 0.001 | 5 (29.4%) | 57 (63.3%) | 0.001 |
| Seen, involving the peeled area | 10 (43.5%) | 7 (8.3%) | 8 (47.1%) | 9 (10.0%) | ||
| Seen, not involving the peeled area | 7 (30.4%) | 21 (25.0%) | 4 (23.5%) | 24 (26.7%) | ||
| Hypertension | 12 (52.2%) | 40 (47.6%) | 0.815 | 7 (41.2%) | 45 (50.0%) | 0.601 |
| Dyslipidemia | 11 (47.8%) | 35 (41.7%) | 0.640 | 8 (47.1%) | 38 (42.2%) | 0.792 |
| DM | ||||||
| No | 10 (43.5%) | 59 (70.2%) | 0.016 | 10 (58.8%) | 59 (65.6%) | 0.346 |
| IFG | 2 (8.7%) | 2 (2.4%) | 2 (11.8%) | 2 (2.2%) | ||
| Yes | 3 (13.0%) | 14 (16.7%) | 2 (11.8%) | 15 (16.7%) | ||
| DM with ocular involvement | ||||||
| Yes + NPDR | 2 (8.7%) | 1 (1.2%) | 0 (0.0%) | 3 (3.3%) | ||
| Yes + PDR | 6 (26.1%) | 8 (9.5%) | 3 (17.6%) | 11 (12.2%) | ||
| Ocular diseases | ||||||
| Uveitis | 0 (0.0%) | 6 (7.1%) | 0.337 | 2 (11.8%) | 4 (4.4%) | 0.242 |
| Vascular occlusive diseases | 0 (0.0%) | 3 (3.6%) | 1.000 | 0 (0.0%) | 3 (3.3%) | 1.000 |
| Glaucoma | 0 (0.0%) | 8 (9.5%) | 0.197 | 1 (5.9%) | 7 (7.8%) | 1.000 |
| Corneal diseases | 0 (0.0%) | 2 (2.4%) | 1.000 | 0 (0.0%) | 2 (2.2%) | 1.000 |
| Retinal detachment | 0 (0.0%) | 3 (3.6%) | 1.000 | 0 (0.0%) | 3 (3.3%) | 1.000 |
| Retinal degenerative diseases | 1 (4.3%) | 1 (1.2%) | 0.385 | 0 (0.0%) | 2 (2.2%) | 1.000 |
DM = diabetes mellitus; ERM = epiretinal membrane; IFG = impaired fasting glucose; ILM = internal limiting membrane; IOL = intraocular lens; NPDR = nonproliferative diabetic retinopathy; PDR = proliferative retinopathy.
All tests were done using Fisher exact test (%) excepted for age (mean [standard deviation]).
Specimen Harvesting, Processing, and Examination
The detailed specimen harvesting and processing can be found in our published protocol.23 Only specimens from the foveal and parafoveal areas, with or without the perifoveal areas, were collected and immediately immersed in a cryovial for fixation. Specimen size was not restricted in our study, as tissue was harvested according to surgical requirements and safety considerations. The specimens were fixed in 3% glutaraldehyde. The thin sections were stained with uranyl acetate and lead citrate for TEM examination. For LM examination, toluidine blue–stained thick sections were examined at 400× magnification, with nuclei counted in high-power fields, depending on specimen size. Counts were expressed as cells per mm2, and notable histopathological findings were recorded. Ultrastructural examination was performed on thin sections using the JEOL-JEM 1400 Plus TEM. Scanning focused on the grid containing the highest specimen yield. Detailed examination parameters are listed in Supplemental Table 2, available at www.ophthalmologyscience.org.
Outcomes
The primary outcome was the morphology and pathological changes of cells and extracellular matrix in the ILM. The secondary outcome was the predominant cell types and their associated pathological features in the ILM and ERM, stratified by idiopathic and secondary ERM. All collected variables are summarized in Supplemental Table 3, available at www.ophthalmologyscience.org.
Statistical Analyses
Statistical analyses were performed using STATA SE 18.5 (StataCorp LLC), with a significance level of 0.05. Descriptive statistics were used to characterize the cellular and extracellular matrix morphology and pathological changes in ILM with and without ERM. Fisher exact test was applied to categorical variables, while unpaired t tests assuming equal variances were used for continuous variables. Due to multiple comparisons, only results with a P value < 0.001 were considered statistically significant.
Multivariate analyses included both known ERM-associated factors and variables with P values < 0.1 from univariate analyses. For count data, distribution assumptions were checked to determine suitability for Poisson or negative binomial regression models, while binomial regression was used for binary outcomes. Where appropriate, post hoc grouping of factors was considered to enhance clinical interpretability. Sensitivity analyses were conducted by excluding poor-quality specimens from the TEM assessment.
Post hoc grouping was performed by combining PVD stages adapted from Johnson et al25: stages 1 and 2 were classified as no foveal detachment, while stages 3 and 4, as well as cases with foveal detachment and undetermined optic nerve status, were classified as foveal detachment. Additionally, ERM stages 1–3, based on the classification adapted from Govetto et al24, were grouped together for analysis.
Results
Demographic Data
A total of 120 eyes from 117 participants were recruited for this study from 21st August 2023 to 13th May 2024. However, only 107 eyes from 107 participants, 90 from ILM with ERM, and 17 from ILM without ERM group, were included in the final analysis. Thirteen eyes were excluded due to failure to obtain interpretable histopathological results. Among these, three eyes belonged to participants for whom the fellow eye was still eligible for analysis. The mean age of participants was 64.4 ± 10.9 years, and 64 (59.8%) were female. Only ERM without accompanying ILM was observed in 27 of the 84 ILM with ERM specimens, while cellular membrane structures resembling ERM were identified in 5 ILM without ERM specimens, as illustrated in Figure 1.
Figure 1.
Study flow. Red: cellular structures resembling ERM or only ERM were observed in those specimens. ERM = epiretinal membrane.
Discrepancies between surgical and OCT diagnoses are summarized in Supplemental Table 4, available at www.ophthalmologyscience.org. Seventeen eyes were allocated to the ILM without ERM group. Surgical diagnoses were detailed in Supplemental Appendix 1, available at www.ophthalmologyscience.org. Baseline characteristics of the 107 included eyes, based on both surgical and OCT diagnoses, are presented in Table 1. The proportion of participants with retinal breaks including macular hole observed during surgery was significantly higher in the ILM without ERM group compared to the ILM with ERM group (P < 0.001 for surgical diagnosis; P = 0.001 for OCT diagnosis). Clinical imaging findings for the analyzed eyes are shown in Table 2. Surgical outcome was described in Supplemental Appendix 2, available at www.ophthalmologyscience.org. The visual acuity of included participants was displayed in Supplemental Table 5, available at www.ophthalmologyscience.org.
Table 2.
Clinical Imaging Characteristics of Included Participants
| Baseline Characteristics | OCT for ERM |
||
|---|---|---|---|
| OCT Negative for ERM n = 17 (15.9%) | OCT Positive for ERM n = 90 (84.1%) | P | |
| Fundus photos | |||
| Acquired | 11 (100.0%) | 67 (87.0%) | 0.350 |
| Poor quality | 0 (0.0%) | 10 (13.0%) | |
| Retinal break∗ | |||
| Not seen | 5 (45.5%) | 52 (67.5%) | 0.111 |
| Seen, involving the peeled area | 4 (36.4%) | 10 (13.0%) | |
| Seen, not involving the peeled area | 2 (18.2%) | 14 (18.2%) | |
| Vitreous hemorrhage∗ | 2 (18.2%) | 7 (9.1%) | 0.318 |
| Retinal Detachment∗ | |||
| Not seen | 7 (63.6%) | 59 (76.6%) | 0.377 |
| Seen, involving the peeled area | 3 (27.3%) | 13 (16.9%) | |
| Seen, not involving the peeled area | 1 (9.1%) | 4 (5.2%) | |
| Myopic status∗ | |||
| High myopic fundus | 1 (9.1%) | 13 (16.9%) | 0.578 |
| Normal | 6 (54.5%) | 53 (68.8%) | |
| Posterior staphyloma | 1 (9.1%) | 3 (3.9%) | |
| AMD∗ | |||
| No AMD | 11 (100.0%) | 68 (88.3%) | 1.000 |
| Early AMD | 0 (0.0%) | 3 (3.9%) | |
| Intermediate AMD | 0 (0.0%) | 1 (1.3%) | |
| Advance dry AMD (GA) | 0 (0.0%) | 1 (1.3%) | |
| Depigmentation∗ | |||
| Not seen | 9 (81.8%) | 54 (70.1%) | 0.677 |
| Seen, involving the peeled area | 0 (0.0%) | 8 (10.4%) | |
| Seen, not involving the peeled area | 1 (9.1%) | 6 (7.8%) | |
| Increase pigment∗ | |||
| Not seen | 9 (81.8%) | 58 (75.3%) | 1.000 |
| Seen, involving the peeled area | 0 (0.0%) | 5 (6.5%) | |
| Seen, not involving the peeled area | 1 (9.1%) | 5 (6.5%) | |
| Any pigmentary abnormalities∗ | |||
| Not seen | 9 (81.8%) | 60 (77.9%) | 0.505 |
| Seen, involving the peeled area | 0 (0.0%) | 5 (6.5%) | |
| Seen, not involving the peeled area | 1 (9.1%) | 3 (3.9%) | |
| OCT | 17 (100%) | 90 (100%) | |
| CST | 357 (221) | 359 (108) | 0.967 |
| PVD∗ | |||
| Stage 1 | 4 (28.6%) | 9 (10.8%) | 0.364 |
| Stage 2 | 0 (0.0%) | 1 (1.2%) | |
| Stage 3 | 1 (7.1%) | 4 (4.8%) | |
| Stage 4 | 6 (42.9%) | 43 (51.8%) | |
| Foveal detachment with undetermined optic nerve | 3 (21.4%) | 26 (31.3%) | |
| VMT∗ | |||
| None | 10 (71.4%) | 74 (89.2%) | 0.102 |
| VMT focal | 4 (28.6%) | 7 (8.4%) | |
| VMT broad | 0 (0.0%) | 2 (2.4%) | |
| Hole∗ | |||
| No hole | 5 (35.7%) | 62 (73.8%) | 0.011 |
| Lamellar hole | 4 (28.6%) | 11 (13.1%) | |
| FTMH | 5 (35.7%) | 11 (13.1%) | |
| Hole size | 858.3 (211.1) | 725.3 (732.3) | 0.735 |
| DRIL∗ | 6 (42.9%) | 35 (42.2%) | 1.000 |
| Retinoschisis∗ | 5 (35.7%) | 19 (23.8%) | 0.338 |
| Intraretinal cyst∗ | 13 (86.7%) | 41 (50.6%) | 0.011 |
| Loss of subfoveal ellipsoidal band∗ | 10 (66.7%) | 32 (38.6%) | 0.052 |
| ERM staging∗ | |||
| Not seen | 17 (100.0%) | 0 (0.0%) | <0.001 |
| Cannot stage | 0 (0.0%) | 20 (24.1%) | |
| Stage 1 | 0 (0.0%) | 12 (14.5%) | |
| Stage 2 | 0 (0.0%) | 11 (13.3%) | |
| Stage 3 | 0 (0.0%) | 22 (26.5%) | |
| Stage 4 | 0 (0.0%) | 18 (21.7%) | |
| OCTA | 0 (0.0%) | 5 (5.6%) | 1.000 |
| Increase FAZ∗ | 0 (0.0%) | 3 (60.0%) | - |
AMD = age-related macular degeneration; CST = central subfield thickness; DRIL = disorganization of retinal inner layers; ERM = epiretinal membrane; FAZ = foveal avascular zone; FTMH = full-thickness macular hole; GA = geographic atrophy; OCTA = OCT angiography; PVD = perifoveal vitreous detachment; VMT = vitreomacular traction.
All tests were done using Fisher exact test (%) excepted for CST and hole size (mean [standard deviation]).
Some images were of poor quality, and certain factors could not be evaluated.
Histopathological Findings of ILM with and without ERM
Histopathological examination was performed on 108 specimens, with successful TEM analysis in 107 cases. The key histopathological findings of ILM with and without ERM are summarized in Table 3. Thick-section images are visually illustrated in Figure 2, while thin-section images, along with corresponding clinical imaging, are presented in Figure 3. Notable pathological features of the ILM are depicted in Figure 4, and the presence of ERM-associated cells within the specimens is highlighted in Figure 5. Epiretinal membrane specimens demonstrated cellular proliferation involving fibrous astrocytes, myofibroblasts, fibroblasts, fibrocytes, and RPE, along with extracellular matrix deposition of newly formed collagen, indicating active fibrocellular remodeling. In case without retinal break seen in fundus photo or during surgery, lamellar hole, or full-thickness macular holes (FTMH), RPE was not present. Long-spacing collagen was observed exclusively in ILM specimens with ERM, although this finding was not statistically significant. Notably, some ILM without ERM specimens also showed single up to a few cellular structures resembling ERM or early-stage membrane formation.
Table 3.
Histopathological Findings of ILM with and without ERM
| Histopathological Findings | OCT Diagnosis of ERM |
||
|---|---|---|---|
| OCT Negative for ERM n = 17 (15.9%) | OCT Positive for ERM n = 90 (84.1%) | P | |
| Histopathology | |||
| Total nuclei | 33.353 (73.6) | 33.356 (54.8) | 1.000 |
| Nuclei/HPF | 30.419 (54.0) | 42.866 (61.1) | 0.435 |
| TEM examination | |||
| ILM | 12 (70.6%) | 63 (70.0%) | 1.000 |
| Vacuolization of ILM | 10 (83.3%) | 41 (65.1%) | 0.317 |
| Vacuole count | 2.011 (3.2) | 3.795 (8.1) | 0.454 |
| ILM consistency | |||
| Compact | 10 (83.3%) | 44 (69.8%) | 0.491 |
| Loose | 2 (16.7%) | 19 (30.2%) | |
| Electronlucent band | 1 (5.9%) | 11 (12.2%) | 0.686 |
| Cellular structures resembling ERM or ERM | 12 (70.6%) | 71 (78.9%) | 0.527 |
| Fibrous astrocyte | 1 (8.3%) | 7 (9.9%) | 1.000 |
| Myofibroblast | 4 (33.3%) | 37 (52.1%) | 0.350 |
| Fibroblast | 1 (8.3%) | 11 (15.5%) | 1.000 |
| Fibrocyte | 1 (8.3%) | 0 (0.0%) | 0.145 |
| Retinal pigment epithelium | 4 (33.3%) | 16 (22.5%) | 0.471 |
| Cell fragments | 12 (100%) | 60 (84.5%) | 0.352 |
| Vitreous matrix | 8 (61.5%) | 54 (75.0%) | 0.325 |
| Newly formed collagen | 2 (16.7%) | 8 (11.3%) | 0.633 |
| Long-spacing collagen | 0 (0%) | 13 (14.4%) | 0.123 |
ERM = epiretinal membrane; HPF = high-power field; ILM = internal limiting membrane; TEM = transmission EM.
All tests were done using Fisher exact test (%) excepted for total nuclei, nuclei per HPF and vacuole count (mean [standard deviation]).
Figure 2.
Histopathological findings of included specimen. Thick section of the specimens on the toluidine blue stained 1-micron thick-section. A: Epiretinal membrane without definite ILM under LM. B: Folding ILM without definite ERM under LM. C: A case with definite ILM (arrow) under LM and multilayered ERM containing scattered pigmented cells (circle). ERM = epiretinal membrane; ILM = internal limiting membrane; LM = light microscopy.
Figure 3.
Fundus photo, OCT, and TEM results of ILM with and without ERM. Case 1: A 57-year-old female with senile incipient cataract, proliferative diabetic retinopathy, superior tractional retinal detachment, and localized rhegmatogenous retinal detachment in the left eye. No ERM was detected on OCT or intraoperatively. Case 2: A 67-year-old female with a history of BRVO and primary open-angle glaucoma in the left eye, diagnosed with senile incipient cataract and ERM secondary to BRVO. Epiretinal membrane stage 4 was detected on OCT and seen intraoperatively. (A) OCT of the left eye of case 1 with ILM without ERM. (B) Fundus photograph of the left eye of case 1 with ILM without ERM. (C) Transmission EM ultrastructure of case 1. The ILM with various thickness is observed without ERM. The blue box indicated 10 boxes for ultrastructure examination. (D) OCT of the left eye of case 2 with ILM with secondary ERM due to a history of BRVO in the left eye. (E) Fundus photograph of the left eye of case 2 with ILM with secondary ERM. (F) Transmission EM ultrastructure of case 2, showing ERM is seen with ILM. Arrow indicates the cellular components of the ERM. BRVO = branch retinal vein occlusion; ERM = epiretinal membrane; ILM = internal limiting membrane; TEM = transmission EM.
Figure 4.
Transmission EM findings of pathological-suspected features of the ILM. (A) Internal limiting membrane with loose consistency was observed with no significant differences between groups: 2 ILM without ERM versus 19 ILM with ERM (P = 0.491), and 9 idiopathic versus 10 secondary ERM eyes (P = 0.788). (B) Internal limiting membrane with compact consistency was identified in 10 ILM without ERM and 41 ILM with ERM (P = 0.491), and in 23 idiopathic versus 21 secondary ERM eyes (P = 0.788). Vacuolization (arrows) and ERM tissue on the vitreous surface were observed in 10 ILM without ERM and 41 ILM with ERM (P = 0.317), and in 26 idiopathic versus 15 secondary ERM eyes (P = 0.008). (C) Long-spacing collagen (white arrows) was detected in 0 ILM without ERM and 13 ILM with ERM (P = 0.123), and in 10 idiopathic versus 3 secondary ERM eyes (P = 0.034). Vacuolated ILM was also present. (D) The ERM is composed of multiple layers of cells overlying collagen fibers on the vitreous aspect of the ILM. Muller’s foot processes (circular structures) are observed on the retinal side. A region of altered collagen is present (A). Arrows indicate cellular components of the ERM, and stars denote the collagen layer. ERM = epiretinal membrane; ILM = internal limiting membrane.
Figure 5.
Transmission EM findings of cells and extracellular matrix of included specimen. (A) Retinal pigment epithelium shows pigment globules of uniform size (arrow) and a prominent nucleolus. Retinal pigment epithelium was identified in 4 ILM without ERM and 16 ILM with ERM (P = 0.471), and differed significantly between 1 idiopathic and 15 secondary ERM eyes (P < 0.001). (B) A myofibroblast (arrow) with cell processes and subplasmalemmal aggregates of cytoplasmic filaments (square) is observed. Myofibroblasts were found in 4 ILM without ERM and 37 ILM with ERM (P = 0.350), and in 16 idiopathic versus 21 secondary ERM eyes (P = 0.346). (C) Internal limiting membrane with ERM showing a single Ce and Co on the vitreous side. Müller cell foot processes are present at the retinal surface of the ILM (arrow). (D) Thick, newly formed collagen fibers (arrow) are seen within the ERM. Newly formed collagen was present in 2 ILM without ERM and 8 ILM with ERM (P = 0.633), and in 1 idiopathic versus 7 secondary ERM eyes (P = 0.055). Ce = cell layer; Co = collagen layers; ERM = epiretinal membrane; ILM = internal limiting membrane.
To further investigate cellular differences, we performed univariate analyses with significant factors entered multivariate analysis. For nuclei count, negative binomial regression was used, with vitreous hemorrhage, PVD, foveal hole, disorganization of retinal inner layers, and ERM grade included as candidate predictors. Vacuole count was also analyzed using negative binomial regression, with potential predictors including lens status, vitreous status, intraoperative retinal breaks, diabetes, RD, pigmentary changes, vitreomacular traction status, disorganization of retinal inner layers, ERM grade, and increased foveal avascular zone. Finally, the presence of ERM cells was assessed using logistic regression, with vitreous status, retinal breaks on fundus photography, PVD status, and ERM grade considered as predictors.
Selected factors with a P value < 0.1 from univariate analysis were included in the multivariate model. None of these factors had previously established associations or hypothesized links to cellular proliferation or migration, or to vacuole formation. Negative binomial regression models were employed to assess differences in nuclei count in Supplemental Figure 1, available at www.ophthalmologyscience.org and vacuole count in Supplemental Figure 2, available at www.ophthalmologyscience.org. For vacuole count, intraocular lens (IOL) was found to have an incidence rate ratio of 0.333 95% confidence interval (CI) (0.142–0.784) P = 0.012 and DM with proliferative diabetic retinopathy (PDR) with an incidence rate ratio of 0.190 95% CI (0.065–0.555) P = 0.002. Other factors were excluded from nuclei and vacuole count analysis due to insufficient data. Additionally, logistic regression analysis was conducted to evaluate the presence of ERM cells in Supplemental Figure 3, available at www.ophthalmologyscience.org. After multivariate analysis, none of the evaluated variables including nuclei count, vacuole count, or presence of ERM cells remained statistically significant.
Histopathological Findings of Idiopathic and Secondary ERM
Among the analyzed specimens, 90 eyes were diagnosed as ILM with ERM from OCT. A total of 45 eyes were categorized as secondary ERM, with categorization detailed in Supplemental Table 6, available at www.ophthalmologyscience.org. The baseline characteristics of these cases are summarized in Supplemental Table 7, available at www.ophthalmologyscience.org, while their clinical imaging findings are detailed in Supplemental Table 8, available at www.ophthalmologyscience.org. Vitreous status differed significantly between groups. In idiopathic ERM, 95.6% of eyes were intact vitreous, with the remainder being postvitrectomized or filled with silicone oil. In contrast, in secondary ERM, only 55.6% had intact vitreous, while the rest were postvitrectomized or silicone oil–filled. Vitreous status, retinal breaks, including FTMH, RD, and DM with ocular involvement, were used as a criterion for defining secondary ERM.
The proportion of eyes without a macular hole was higher in the idiopathic ERM group compared to the secondary ERM group (79.5% vs 67.5%), with FTMH classified in the secondary group (27.5% vs 0.0%). The overall difference across macular hole types was statistically significant (P < 0.001). Similarly, intraretinal cysts were significantly more prevalent in secondary ERM (73.0% vs 31.8%, P < 0.001), as was the loss of the subfoveal ellipsoidal band (64.1% vs 15.9%, P < 0.001).
In the TEM ultrastructural analysis, RPE cells were significantly more prevalent in secondary ERM cases (40.5% vs 2.8%, P < 0.001). Idiopathic ERM tended to show more frequent ILM vacuolization (81.2% vs 48.4%, P = 0.008) and long-spacing collagen (23.3% vs 6.4%, P = 0.034), features previously associated with vitreous remodeling.26 However, these findings are not statistically significant. No significant difference in vitreous matrix presence was observed between the groups (P = 1.000). The key histopathological findings are summarized in Table 4, while thin-section images, along with corresponding clinical imaging, are presented in Figure 6.
Table 4.
Histopathological Findings of Idiopathic and Secondary ERM
| Histopathological Findings | ERM Category |
||
|---|---|---|---|
| Idiopathic n = 43 (47.8%) | Secondary n = 47 (52.2%) | P | |
| Histopathology | |||
| Total nuclei | 27.2 (52.8) | 39.0 (56.5) | 0.308 |
| Nuclei/HPF | 32.9 (51.6) | 52.0 (67.9) | 0.140 |
| TEM Examination | |||
| ILM | 32 (74.4%) | 31 (66.0%) | 0.491 |
| Vacuolization of ILM | 26 (81.2%) | 15 (48.4%) | 0.008 |
| Vacuole count | 4.3 (7.3) | 3.2 (8.9) | 0.600 |
| ILM consistency | |||
| Compact | 23 (71.9%) | 21 (67.7%) | 0.788 |
| Loose | 9 (28.1%) | 10 (32.3%) | |
| Electronlucent band | 7 (16.3%) | 4 (8.5%) | 0.340 |
| ERM | 35 (81.4%) | 36 (76.6%) | 0.615 |
| Fibrous astrocyte | 3 (8.6%) | 4 (11.1%) | 1.000 |
| Myofibroblast | 16 (45.7%) | 21 (58.3%) | 0.346 |
| Fibroblast | 6 (17.1%) | 5 (13.9%) | 0.753 |
| Fibrocyte | 0 (0.0%) | 0 (0.0%) | . |
| Retinal pigment epithelium | 1 (2.9%) | 15 (41.7%) | <0.001 |
| Cell fragments | 31 (88.6%) | 29 (80.6%) | 0.514 |
| Vitreous matrix | 27 (75.0%) | 27 (75.0%) | 1.000 |
| Newly formed collagen | 1 (2.9%) | 7 (19.4%) | 0.055 |
| Long-spacing collagen | 10 (23.3%) | 3 (6.4%) | 0.034 |
ERM = epiretinal membrane; HPF = high-power field; ILM = internal limiting membrane; TEM = transmission electron microscope.
All tests were conducted using Fisher exact test for categorical variables (reported as %) and unpaired t test total nuclei, nuclei per HPF, and vacuole count (reported as mean [standard deviation]). A P value < 0.001 was considered statistically significant.
Figure 6.
Fundus photo, OCT, and TEM results of idiopathic and secondary ERM. Case 3: An 81-year-old male with a history of pathologic myopia in the left eye, diagnosed with senile incipient cataract and idiopathic ERM. Epiretinal membrane stage 2 was detected on OCT and seen intraoperatively. Case 4: A 70-year-old male with a history of quiet BARN and ocular OIS in the left eye, diagnosed with senile incipient cataract and ERM secondary to BARN and OIS. Epiretinal membrane stage 2 was detected on OCT and seen intraoperatively. (A) OCT of the left eye of case 3 with idiopathic ERM. (B) Fundus photograph of the left eye of case 3 with idiopathic ERM. (C) Transmission EM ultrastructure of case 3. The ERM is observed on the vitreous side of the ILM and is composed of collagen layers and a few layers of cells. Arrows indicate the cellular components of the ERM, while stars mark the collagen layer. The ILM exhibits vacuolization (arrowheads). (D) OCT of the left eye of case 4 with secondary ERM due to a history of BARN and OIS in the left eye. (E) Fundus photograph of the left eye of case 4 with secondary ERM. (F) TEM ultrastructure of case 4, showing closely packed membranes consisting of the ILM and ERM. Arrows indicate the cellular components of the ERM, while stars mark the collagen layer. BARN = bilateral acute retinal necrosis; ERM = epiretinal membrane; ILM = internal limiting membrane; OIS = ischemic syndrome; TEM = transmission EM.
Sensitivity Analysis
The sensitivity analysis of histopathological findings for ILM with and without ERM is summarized in Supplemental Table 9, available at www.ophthalmologyscience.org compared with Table 3, while the analysis for idiopathic and secondary ERM is presented in Supplemental Table 10, available at www.ophthalmologyscience.org compared with Table 4. Retinal pigment epithelium cells remained significantly more prevalent in secondary ERM cases (43.8% vs 3.0%, P < 0.001).
Discussion
This study incorporated a systematic pathological assessment of intraocular specimens, along with detailed clinical imaging and patient characteristics, to establish potential associations between structural alterations and ERM development. Compared to previous studies that utilized only LM or immunohistochemical staining, our TEM analysis provides a more detailed visualization of ultrastructural components.19 In line with previous findings, we observed that ILM specimens associated with ERM, especially secondary ERM, exhibited cellular proliferation and extracellular matrix deposition of collagen, reinforcing the hypothesis that fibrocellular proliferation is a key feature of secondary ERM pathology.16,18,20 Since some ILM without ERM specimens demonstrated few cellular layers resembling ERM, this may represent an early-stage inflammatory response, variance of ERM cellular proliferation, or reflect the limited sensitivity of OCT in detecting ERM, particularly in eyes with underlying conditions such as RD that affect image quality.19 Rather than implying clinical benefit from routine pathological assessment, our findings highlight a diagnostic gap that emerging high-resolution or multimodal OCT technologies may help bridge. Future imaging innovations could enable earlier detection and monitoring of ERM development by visualizing subclinical membrane changes that currently fall below OCT axial resolution.
Consistent with previous reports, we identified basement membrane fragments on the ILM, suggesting Müller cell involvement in ILM remodeling.27,28 However, unlike some prior studies that reported ILM thickening in patients with ERM, our findings showed variability in ILM consistency, likely due to differences in tissue processing and sectioning angles.29 Potential slant cuts during ultramicrotomy may have contributed to discrepancies in measured thickness, highlighting the challenge of obtaining uniform ILM specimens for quantitative assessment. Furthermore, we also observed an electronlucent band and vacuolization, which we hypothesized to be a reaction to the inflammatory response or the degeneration of the ILM, although artefactual change of compact collagen fiber cannot be entirely excluded, and we were unable to identify its significance in this study as well as in previous reports.20
Correlating with baseline characteristics and clinical imaging, our analysis revealed significantly reduced ILM vacuole counts in eyes with IOL and in those with PDR. In pseudophakic eyes, this reduction may reflect structural alterations in the ILM following cataract surgery, possibly due to changes in vitreous dynamics or mechanical stress from prior interventions.30,31 In PDR, reduced vacuolization may be driven by chronic ischemia–induced Müller cell activation, resulting in persistent gliosis and deposition of dense glial and basement membrane–like material along the vitreoretinal interface.32 In our study, ILM vacuolization was observed more frequently in idiopathic than in secondary ERM, although the difference did not reach statistical significance and an area requiring further investigation in future studies. Unlike previous report, this trend may reflect the classification of PDR within the secondary ERM group, where ischemia-driven gliotic and fibrovascular changes lead to less vacuolization in ILM ultrastructure.20 These observations highlight the influence of surgical history and retinal pathology on ILM ultrastructure, warranting further research into the clinical relevance of vacuolization.
A standardized specimen collection protocol, adapted from previous studies, was developed, formally registered, and published to ensure reproducibility.18, 19, 20,23,33,34 While concerns exist regarding the potential effects of Brilliant Blue G staining on pathological results, we believe that Brilliant Blue G is largely dissolved in the 3% glutaraldehyde in a 0.1 M phosphate buffer (pH 7.3) solution, minimizing its impact on staining outcomes, as previously reported.20 For this reason, we did not analyze the relationship between Brilliant Blue G staining time and the pathological findings. Additionally, variations in specimen harvesting time may introduce slight differences in staining duration, but our methods did not result in increased surgical complications.35,36 The specimen loss in our study did not correlate with baseline characteristics or clinical imaging findings, suggesting that other factors, such as specimen size or random loss during processing, may have contributed to this issue.
Comparing idiopathic and secondary ERM, ILM vacuolization was more frequent, though not statistically significant, in idiopathic ERM cases. While previous reports have described this pathological finding in patients with DM, our study suggests that ILM alterations in idiopathic ERM may be more closely associated with degenerative processes rather than reactive fibrosis.20 This distinction further differentiates idiopathic ERM from secondary ERM in terms of pathophysiology, highlighting potential differences in disease mechanisms and cellular interactions with the ILM. We hypothesize that multiple factors may contribute to the formation of vacuolization within the ILM. Furthermore, RPE cells were significantly more prevalent in secondary ERM cases suggesting a stronger role for retinal breaks, which are frequently identified intraoperatively, and subsequent RPE migration in the pathogenesis of secondary ERM.37 Newly formed collagen and an increased nuclei count per high-power field after sensitivity analysis were also more common, though not statistically significant, in secondary ERM, indicating a higher degree of extracellular matrix remodeling. These findings support the hypothesis that secondary ERM arises from a more pronounced reparative response, likely driven by retinal insults such as inflammation, trauma, or RD.12
Baseline characteristic analysis showed a significant age difference between ILM specimens with and without ERM, with older patients more likely to have ERM diagnosed intraoperatively. This age association was not observed in OCT-based diagnoses, likely because OCT's superior sensitivity enables detection of subclinical ERM in younger patients that would otherwise go undiagnosed with conventional clinical examination alone. These younger patients, who would traditionally be classified as not having the ERM based on clinical assessment, were reclassified into the ERM group when OCT revealed previously undetected ERMs. The proportion of eyes with retinal breaks observed during surgery was significantly higher in the ILM without ERM group, likely reflecting the high proportion of cases undergoing FTMH repair or surgery for RD. Conversely, prior studies have reported an association between retinal breaks and ERM formation, which we also identified as a contributing factor in secondary ERM.12 In clinical imaging, the proportion of participants with lamellar holes, FTMH, or intraretinal cysts observed in OCT was higher, although not statistically significant, in the ILM without ERM, group likely due to these eyes being primarily operated for macular hole or RD repair.
Secondary ERM cases exhibited a significantly higher prevalence of intraretinal cysts and subfoveal ellipsoidal band loss. Disorganization of retinal inner layers was also more frequent in secondary ERM cases, though not statistically significant. These findings indicate more severe retinal structural disruption compared to idiopathic ERM, suggesting that secondary ERM is associated with more chronic and severe inner retinal architectural changes, possibly due to underlying retinal pathologies such as PDR or vascular occlusions.9,12,38 Epiretinal membrane staging also showed trends, with secondary ERM cases more frequently classified as stage 4, reinforcing the idea that secondary ERM is associated with more aggressive fibrocellular proliferation and structural distortion.11 Within the ILM with ERM group, FTMH was present exclusively in secondary ERM cases, consistent with our definition of secondary ERM as arising from identifiable causes such as macular hole, RD, or inflammation. This finding supports the notion that secondary ERM is associated with greater macular structural disruption. Unlike prior reports, we did not observe a significant association between FTMH and increased central subfield thickness.9,39 Also, participants with idiopathic ERM were older, though not statistically significant, than those with secondary ERM, which may suggest an age-related degenerative component in idiopathic cases, as described in previous studies.11 Given the fibrocellular changes observed in our study, macular structural disruption in older eyes on OCT may reflect increased susceptibility to secondary ERM formation. This raises the question of whether ILM peeling during PPV for other indications could play a prophylactic role, although this remains speculative and warrants investigation in future studies. However, interpretation of findings in secondary ERM is limited by heterogeneity in its underlying causes, which may dilute group differences and complicate comparative analyses.
A major strength of this study is the combined use of LM and TEM, which allows for a more comprehensive analysis of ILM and ERM pathology compared to studies relying on single-modality microscopic techniques. Additionally, our methodology incorporates detailed preoperative imaging, clinical imaging, and surgical outcomes, facilitating robust correlation analyses. The standardized specimen collection and masked evaluation further enhance the reliability of our findings. The use of TEM may provide superior confirmation of ultrastructural cellular morphology compared to immunohistochemistry, offering more detailed visualization at the subcellular level.
Despite these strengths, several limitations must be acknowledged. This cross-sectional study with a short 3-month follow-up limits the ability to assess ILM regeneration, ERM recurrence, and long-term visual outcomes. A longitudinal design is needed to correlate ultrastructural findings with disease progression. The sample size, based on surgical volume, may affect generalizability, and validation in larger, multicenter studies is warranted. Despite standardized processing, potential artifacts—such as sectioning angle variability—could affect ultrastructural analysis and limit consistency with prior studies. Our novel specimen collection method, while replicable, may have overlooked some pathological features. Cellular components, including inflammatory cells, were also present in ILM without ERM, suggesting nonspecificity. Although membranes were obtained from specific locations, the specimens may represent only partial segments of the entire ILM or ERM. We excluded patients with prior cataract surgery due to uncertain timing as a causative factor, and no established cutoff exists. In our multivariate regression analysis, none of the predictors remained significant, which may be due to the analysis being a secondary outcome and therefore potentially underpowered. Additionally, multicollinearity or overfitting may have influenced the validity of the model. We also acknowledge the possibility of misclassification arising from cell identification based solely on TEM morphology. Incorporating immunohistochemical validation of specific cell types could offer future benefits, particularly in settings without access to EM. In centers equipped with EM, the use of immunogold labeling may further enhance cellular characterization. Lastly, the single-center design may limit broader applicability to diverse populations or surgical settings.
Our study provides novel insights into ILM with and without ERM ultrastructure and distinguishes key difference of idiopathic from secondary ERM. We identified distinct cellular proliferation and extracellular matrix deposition in ILM with ERM, along with ultrastructural changes such as vacuolization of the ILM, Müller cell fragments on retinal side, an electronlucent band on the vitreous side, and variations in ILM consistency. Our standardized specimen collection achieved a high success rate without increased postoperative complications and may be generalizable to other future intraocular TEM studies. We hypothesized that idiopathic ERM likely arises from degenerative processes, while secondary ERM appears inflammation-driven, as supported by the presence of RPE cells which were exclusively observed in eyes with retinal break. OCT features such as FTMH, intraretinal cysts, and loss of the subfoveal ellipsoidal band may indicate secondary ERM, where early membrane peeling could be beneficial. Longitudinal studies are needed to track ILM changes and optimize ERM management strategies.
Acknowledgments
The authors would like to extend their gratitude to Sirinapa Srikam and Wilawan Ji-au, EM laboratory specialist of the Department of Pathology, Faculty of Medicine, Chulalongkorn University and King Chulalongkorn Memorial Hospital, Thai Red Cross Society, Bangkok, Thailand, Stephen Kerr, PhD, Senior statistician and the director of the Biostatistics Excellence Center, Faculty of Medicine, Chulalongkorn University and King Chulalongkorn Memorial Hospital, Thai Red Cross Society, Bangkok, Thailand, as well as Dr Buravej Assavapongpaiboon of the ocular pathology unit, Department of Ophthalmology; Dr Wajamon Supawatjariyakul of the uveitis unit; and Dr Chanida Saree Khome, Dr Nathapon Treewipanon, Dr Patthicha Pinyosawadsakul, and Dr Pimpisa Vudhichaiphun of the retina unit, Center of Excellence in Retina, Faculty of Medicine, Chulalongkorn University and King Chulalongkorn Memorial Hospital, Thai Red Cross Society, Bangkok, Thailand, for their invaluable contributions in implementing and facilitating the methodology outlined in this study.
Manuscript no. XOPS-D-25-00792.
Footnotes
Disclosure(s):
All authors have completed and submitted the ICMJE disclosures form.
The authors have no proprietary or commercial interest in any materials discussed in this article.
This study granted the Ratchadapiseksomphot Research Funds Type I (grant no.: GA67/038), the 90th Anniversary of Chulalongkorn University Ratchadapiseksomphot Research Funds (grant no.: GCUGR1125671149M), and the grants for research of the Center of Excellence in Retina, Faculty of Medicine, Chulalongkorn University and King Chulalongkorn Memorial Hospital, Thai Red Cross Society (grant no.: 25010119). The funding organization had no role in the design or conduct of this research.
Support for Open Access publication was provided by Chulalongkorn University and the Faculty of Medicine, Chulalongkorn University.
HUMAN SUBJECTS: Human subjects were included in this study. Written informed consent was obtained from all participants prior to any procedures. The study was approved by the Institutional Review Board of the Faculty of Medicine, Chulalongkorn University, Bangkok, Thailand (approval no.: 0948/2023), and conducted in accordance with the Declaration of Helsinki (2013), the Belmont Report, CIOMS Guidelines, and the International Conference on Harmonization Good Clinical Practice standards.
No animal subjects were used in this study.
Author Contributions:
Conception and design: Pothikamjorn, Somkijrungroj, Prasanpanich, Surawatsatien, Tulvatana
Data collection: Pothikamjorn, Somkijrungroj, Prasanpanich, Surawatsatien, Tulvatana
Analysis and interpretation: Pothikamjorn, Somkijrungroj, Prasanpanich, Surawatsatien, Tulvatana
Obtained funding: Pothikamjorn, Somkijrungroj, Tulvatana
Overall responsibility: Pothikamjorn, Somkijrungroj, Prasanpanich, Surawatsatien, Tulvatana
Supplemental material available at www.ophthalmologyscience.org.
Supplementary Data
References
- 1.Mitchell P., Smith W., Chey T., et al. Prevalence and associations of epiretinal membranes. The Blue Mountains eye study, Australia. Ophthalmology. 1997;104:1033–1040. doi: 10.1016/s0161-6420(97)30190-0. [DOI] [PubMed] [Google Scholar]
- 2.Sidd R.J., Fine S.L., Owens S.L., Patz A. Idiopathic preretinal gliosis. Am J Ophthalmol. 1982;94:44–48. doi: 10.1016/0002-9394(82)90189-1. [DOI] [PubMed] [Google Scholar]
- 3.Duan X.R., Liang Y.B., Friedman D.S., et al. Prevalence and associations of epiretinal membranes in a rural Chinese adult population: the Handan eye study. Invest Ophthalmol Vis Sci. 2009;50:2018–2023. doi: 10.1167/iovs.08-2624. [DOI] [PubMed] [Google Scholar]
- 4.You Q., Xu L., Jonas J.B. Prevalence and associations of epiretinal membranes in adult Chinese: the Beijing eye study. Eye (Lond) 2008;22:874–879. doi: 10.1038/sj.eye.6702786. [DOI] [PubMed] [Google Scholar]
- 5.Miyazaki M., Nakamura H., Kubo M., et al. Prevalence and risk factors for epiretinal membranes in a Japanese population: the Hisayama study. Graefes Arch Clin Exp Ophthalmol. 2003;241:642–646. doi: 10.1007/s00417-003-0723-8. [DOI] [PubMed] [Google Scholar]
- 6.Kawasaki R., Wang J.J., Sato H., et al. Prevalence and associations of epiretinal membranes in an adult Japanese population: the Funagata study. Eye (Lond) 2009;23:1045–1051. doi: 10.1038/eye.2008.238. [DOI] [PubMed] [Google Scholar]
- 7.Ng C.H., Cheung N., Wang J.J., et al. Prevalence and risk factors for epiretinal membranes in a multi-ethnic United States population. Ophthalmology. 2011;118:694–699. doi: 10.1016/j.ophtha.2010.08.009. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 8.Kawasaki R., Wang J.J., Mitchell P., et al. Racial difference in the prevalence of epiretinal membrane between Caucasians and Asians. Br J Ophthalmol. 2008;92:1320–1324. doi: 10.1136/bjo.2008.144626. [DOI] [PubMed] [Google Scholar]
- 9.Stevenson W., Prospero Ponce C.M., Agarwal D.R., et al. Epiretinal membrane: optical coherence tomography-based diagnosis and classification. Clin Ophthalmol. 2016;10:527–534. doi: 10.2147/OPTH.S97722. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 10.Wiznia R.A. Posterior vitreous detachment and idiopathic preretinal macular gliosis. Am J Ophthalmol. 1986;102:196–198. doi: 10.1016/0002-9394(86)90144-3. [DOI] [PubMed] [Google Scholar]
- 11.Lee G.W., Lee S.E., Han S.H., et al. Characteristics of secondary epiretinal membrane due to peripheral break. Sci Rep. 2020;10 doi: 10.1038/s41598-020-78093-9. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 12.Fung A.T., Galvin J., Tran T. Epiretinal membrane: a review. Clin Exp Ophthalmol. 2021;49:289–308. doi: 10.1111/ceo.13914. [DOI] [PubMed] [Google Scholar]
- 13.Xiao W., Chen X., Yan W., et al. Prevalence and risk factors of epiretinal membranes: a systematic review and meta-analysis of population-based studies. BMJ Open. 2017;7 doi: 10.1136/bmjopen-2016-014644. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 14.Guber J., Pereni I., Scholl H.P.N., et al. Outcomes after epiretinal membrane surgery with or without internal limiting membrane peeling. Ophthalmol Ther. 2019;8:297–303. doi: 10.1007/s40123-019-0185-7. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 15.Zhang B., Dong X., Sun Y. Effect of internal limiting membrane peeling for idiopathic epiretinal membrane. Heliyon. 2023;9 doi: 10.1016/j.heliyon.2023.e14079. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 16.Schumann R.G., Wolf A., Mayer W.J., et al. Pathology of internal limiting membrane specimens following intravitreal injection of ocriplasmin. Am J Ophthalmol. 2015;160:767–778. doi: 10.1016/j.ajo.2015.06.020. [DOI] [PubMed] [Google Scholar]
- 17.Pradhan D., Agarwal L., Joshi I., et al. Internal limiting membrane peeling in macular hole surgery. Ger Med Sci. 2022;20 doi: 10.3205/000309. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 18.Smiddy W.E., Maguire A.M., Green W.R., et al. Idiopathic epiretinal membranes. Ultrastructural characteristics and clinicopathologic correlation. Ophthalmology. 1989;96:811–820. doi: 10.1016/s0161-6420(89)32811-9. [DOI] [PubMed] [Google Scholar]
- 19.Gandorfer A., Rohleder M., Kampik A. Epiretinal pathology of vitreomacular traction syndrome. Br J Ophthalmol. 2002;86:902–909. doi: 10.1136/bjo.86.8.902. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 20.Beyazyildiz Ö., Tirhiş M.H., Hekimoğlu E.R., et al. Histopathological analysis of internal limiting membrane surgically peeled from eyes with epiretinal membrane. Curr Eye Res. 2016;41:258–265. doi: 10.3109/02713683.2015.1008642. [DOI] [PubMed] [Google Scholar]
- 21.Kritzenberger M., Junglas B., Framme C., et al. Different collagen types define two types of idiopathic epiretinal membranes. Histopathology. 2011;58:953–965. doi: 10.1111/j.1365-2559.2011.03820.x. [DOI] [PubMed] [Google Scholar]
- 22.Regoli M., Tosi G.M., Neri G., et al. The peculiar pattern of type IV collagen deposition in epiretinal membranes. J Histochem Cytochem. 2020;68:149–162. doi: 10.1369/0022155419897258. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 23.Pothikamjorn T., Somkijrungroj T., Prasanpanich M., et al. Comparative electron microscopy analysis of internal limiting membrane and epiretinal membrane ultrastructure from vitrectomy surgery: a study protocol. PLoS One. 2025;20 doi: 10.1371/journal.pone.0331248. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 24.Govetto A., Lalane R.A., 3rd, Sarraf D., et al. Insights into epiretinal membranes: presence of ectopic inner foveal layers and a new optical coherence tomography staging scheme. Am J Ophthalmol. 2017;175:99–113. doi: 10.1016/j.ajo.2016.12.006. [DOI] [PubMed] [Google Scholar]
- 25.Johnson M.W. Perifoveal vitreous detachment and its macular complications. Trans Am Ophthalmol Soc. 2005;103:537–567. [PMC free article] [PubMed] [Google Scholar]
- 26.Vogt D., Deiters V., Zaytseva Y., et al. Vitreoretinal interface features in epiretinal membrane associated with pseudoexfoliation syndrome. Graefes Arch Clin Exp Ophthalmol. 2025;263:2227–2236. doi: 10.1007/s00417-025-06854-1. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 27.Semeraro F., Morescalchi F., Duse S., et al. Current trends about inner limiting membrane peeling in surgery for epiretinal membranes. J Ophthalmol. 2015;2015 doi: 10.1155/2015/671905. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 28.Yang Y.S., Lee J.S., Son G., Sohn J. Epiretinal proliferation associated with lamellar hole or macular hole: origin and surgical prognosis. Korean J Ophthalmol. 2019;33:142–149. doi: 10.3341/kjo.2018.0070. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 29.Haritoglou C., Gandorfer A., Gass C.A., et al. The effect of indocyanine-green on functional outcome of macular pucker surgery. Am J Ophthalmol. 2003;135:328–337. doi: 10.1016/s0002-9394(02)01969-4. [DOI] [PubMed] [Google Scholar]
- 30.Neal R.E., Bettelheim F.A., Lin C., et al. Alterations in human vitreous humour following cataract extraction. Exp Eye Res. 2005;80:337–347. doi: 10.1016/j.exer.2004.09.015. [DOI] [PubMed] [Google Scholar]
- 31.Scarfone H.A., Rodriguez E.C., Rufiner M.G., et al. Vitreous-lens interface changes after cataract surgery using active fluidics and active sentry with high and low infusion pressure settings. J Cataract Refract Surg. 2024;50:333–338. doi: 10.1097/j.jcrs.0000000000001359. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 32.Wu R.H., Xu M.N., Lin K., et al. Inner limiting membrane peeling prevents secondary epiretinal membrane after vitrectomy for proliferative diabetic retinopathy. Int J Ophthalmol. 2022;15:1496–1501. doi: 10.18240/ijo.2022.09.13. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 33.Pothikamjorn T., Prasanpanich M., Somkijrungroj T. Comparative evaluation of anterior lens capsule electron microscopic pathology in a case undergoing simultaneous bilateral cataract surgery: a study of CAPSULaser and continuous curvilinear capsulorhexis. Am J Ophthalmol Case Rep. 2025;39 doi: 10.1016/j.ajoc.2025.102400. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 34.Frisina R., Tessarolo F., Marchesoni I., et al. Microscopic observation of proliferative membranes in fibrocontractive retinal disorders. J Ophthalmol. 2019;2019 doi: 10.1155/2019/9647947. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 35.Alam M.R., Arcinue C.A., Mendoza N.B., Freeman W.R. Recalcitrant cystoid macular edema after pars plana vitrectomy. Retina. 2016;36:1244–1251. doi: 10.1097/IAE.0000000000000892. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 36.Dave V.P., Pathengay A., Schwartz S.G., Flynn H.W., Jr. Endophthalmitis following pars plana vitrectomy: a literature review of incidence, causative organisms, and treatment outcomes. Clin Ophthalmol. 2014;8:2183–2188. doi: 10.2147/OPTH.S71293. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 37.Kampik A., Kenyon K.R., Michels R.G., et al. Epiretinal and vitreous membranes. Comparative study of 56 cases. Arch Ophthalmol. 1981;99:1445–1454. doi: 10.1001/archopht.1981.03930020319025. [DOI] [PubMed] [Google Scholar]
- 38.Joltikov K.A., Sesi C.A., de Castro V.M., et al. Disorganization of retinal inner layers (DRIL) and neuroretinal dysfunction in early diabetic retinopathy. Invest Ophthalmol Vis Sci. 2018;59:5481–5486. doi: 10.1167/iovs.18-24955. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 39.Yazici A.T., Alagöz N., Celik H.U., et al. Idiopathic and secondary epiretinal membranes: do they differ in terms of morphology? An optical coherence tomography-based study. Retina. 2011;31:779–784. doi: 10.1097/IAE.0b013e3181ef8786. [DOI] [PubMed] [Google Scholar]
Associated Data
This section collects any data citations, data availability statements, or supplementary materials included in this article.