An epiretinal membrane (ERM) is a relatively common macular disease characterized by fibrocellular proliferation above the internal limiting membrane (ILM), which can lead to decreased vision, metamorphopsia, and / or macropsia.
Vitrectomy is the first-line therapy to remove an ERM. Although improvements in surgical technology and instrumentation have made early intervention more common, there is no consensus regarding the correct timing for surgical intervention to repair an ERM. Regarding visual prognosis after ERM surgery, preoperative visual acuity (VA) and the duration of symptoms have been known as the most important factors since the pre-optical coherence tomography (OCT) era.1 Recent OCT studies have demonstrated an association between visual function and the macular retinal structures of both the inner2–4 and outer5–9 layers in eyes with ERMs.
An ectopic inner foveal layer (EIFL) is a new OCT-based clinical finding in eyes with ERMs proposed by Govetto et al.,10 which may partially overlap with an abnormally thick inner retinal layer (IRL) in the foveal center or foveal central thick IRL.2,11 Govetto et al. demonstrated that the presence of a continuous EIFL in eyes with ERMs was associated with significant vision loss.10 Joe et al. also concluded that the IRL thickness of the fovea was the major determinant of VA in eyes with an idiopathic ERM.2 In addition, Yang et al. reported that in eyes with an ERM and abnormally thick IRL in the foveal center, postoperative visual outcomes were correlated with preoperative central IRL thicknesses.11 However, there have been no available data on visual prognosis after vitrectomy in ERM eyes with and without EIFLs.
Thus, the purpose of this study was to compare surgical outcomes of idiopathic ERM treatment between eyes with and without EIFLs.
Patients and Methods
This was a retrospective, nonrandomized, comparative study. The institutional review board of Hayashi Eye Hospital approved this study, which adhered to the tenets of the Declaration of Helsinki. All patients provided written informed consent after receiving an explanation of the nature and possible consequences of the study.
Patients who underwent vitrectomy to repair an idiopathic ERM at Hayashi Eye Hospital, Fukuoka, Japan, between June 2015 and May 2016 were investigated retrospectively. Vitrectomies were performed by three experienced surgeons (KY, AH, and SM).
Exclusion criteria included preexisting macular diseases (eg, age-related macular degeneration, vascular occlusive disease, and / or diabetic retinopathy) and non-idiopathic conditions such as uveitis, trauma, and / or post-vitrectomy status. Eyes that required a second surgery or had less than a 12-month follow-up period were also excluded from this study.
Best-corrected VA (BCVA) and OCT examinations, including EIFL, ellipsoid zone (EZ), and cotton ball sign evaluations and central macular thickness (CMT) measurement, were performed before and at 6 and 12 months after vitrectomy. In addition, the presence or absence of a dissociated optic nerve fiber layer (DONFL) was evaluated 6 and 12 months postoperatively.
A 25- or 23-gauge vitrectomy was performed using the Constellation Vision System (Alcon Japan, Tokyo, Japan) under local anesthesia. Phacoemulsification and intraocular lens implantation were also performed in phakic cases aged 50 years or older.12,13
After core vitrectomy following the creation or confirmation of posterior vitreous detachment, the ERM and ILM within the vascular arcade were peeled with the assistance of 0.025% brilliant blue G dye.14 If a leak was noted, the sclerotomy site was sutured at the end of surgery to prevent transient postoperative hypotony.15
Optical Coherence Tomography
Spectral-domain OCT examinations were performed using Cirrus HD-OCT (Carl Zeiss Meditec, Dublin, CA). Scans with a signal strength of 6 or greater were considered appropriate. According to previous reports,10,11 the steepest part of the foveal excavation was set as the foveal center if a foveal pit was present. However, if a foveal pit was absent, the foveal center was identified as the absent point of the EIFL, the point of greatest outer nuclear layer thickness, and the intersection of the fixation point.
The average CMT — that is, the distance from the ILM to the retinal pigment epithelium in the central subfield as defined by the Early Treatment Diabetic Retinopathy Study chart (central 1-mm ring) — was automatically measured in the 200 × 200 macular cube scan mode.
As described in a previous study,10 an EIFL was defined as present on OCT if continuous hypo- and hyperreflective bands extending from the inner nuclear layer and inner plexiform layer / ganglion cell layer complex traversed the foveal center (Figure 1). An intact EZ was defined as a continuous hyperreflective line. The cotton ball sign was defined as a roundish or diffuse highly reflective region observed between the EZ and interdigitation zone at the foveal center.16 OCT images were evaluated by two observers blinded to the clinical data (RM and ST) based on 6-mm, high-definition five-line raster scans with an interval of 0.125 mm. When there was disagreement, a third investigator (TS) was consulted to reach a final decision.
(a–c) Optical coherence tomography (OCT) images of a 65-year-old woman before (a) and at 6 (b) and 12 (c) months after vitrectomy for the repair of an epiretinal membrane (ERM) without an ectopic inner foveal layer (EIFL). Best-corrected visual acuities (BCVAs) in logarithm of the minimum angle of resolution (logMAR) units before and at 6 and 12 months after vitrectomy were 0.30, −0.08, and −0.08, respectively. Her central macular thickness (CMT) measurements were 354 µm, 354 µm, and 358 µm before and at 6 and 12 months after vitrectomy, respectively. (d–f) OCT images of a 54-year-old woman before (d) and at 6 (e) and 12 (f) months after vitrectomy to repair an ERM with an EIFL. The EIFL resolved postoperatively (e and f). BCVAs in logMAR units before and at 6 and 12 months after vitrectomy were 0.40, 0.10, and 0.10, respectively; CMTs measured 524 µm, 367 µm, and 360 µm, respectively.
A DONFL was defined as being present if arcuate-shaped thinning of the retina was detected on en face OCT images and corresponding retinal thinning was detected on cross-sectional OCT images.17,18 OCT images were evaluated by two observers blinded to the clinical data (RM and ST); if there was disagreement, a third investigator (TS) was consulted to reach a final decision.
Data are presented as the mean ± standard deviation. BCVA was measured using a Landolt C acuity chart, and the decimal BCVA was converted to the logarithm of the minimum angle of resolution (log-MAR) units for statistical analyses.
The significance of differences between cases with and without EIFLs was determined using the Mann-Whitney rank-sum test because the data were not normally distributed. The significance of differences in the ratios between cases with and without EIFLs was determined using the Chi-square or Fisher's exact test. The Friedman repeated measures analysis of variance on ranks was performed to compare the BCVA and CMT within the same subjects at different perioperative periods. This was followed by the Tukey test to detect significant differences between each value.
All statistical analyses were performed using SigmaPlot version 12.0 for Windows (Systat Software, San Jose, CA). A P value less than .05 was considered statistically significant.
Seventy-seven eyes with idiopathic ERMs in 76 consecutive patients with an average age of 68.9 years ± 10.4 years were investigated retrospectively (Table 1). An EIFL was detected in 23 out of 77 eyes (29.9%) with an ERM preoperatively. There were no significant differences in the patients' ages or sexes, preoperative lens status, axial length, vitrectomy gauge, or preoperative appearance of the cotton ball sign between ERM cases with and without an EIFL. The EZ was disrupted in a total of seven eyes (9.1%). The ratio of eyes with an ERM and disrupted EZ was significantly (P = .002) greater in eyes with an EIFL than in those without an EIFL.
Patient and Ocular Demographics in ERM Cases With and Without EIFL
BCVAs before and at 6 and 12 months after vitrectomy were 0.28 ± 0.18, 0.10 ± 0.13, and 0.09 ± 0.13, respectively, in eyes without EIFLs; and 0.37 ± 0.21, 0.17 ± 0.12, and 0.15 ± 0.13, respectively, in eyes with EIFLs (Table 1, Figure 2). In both groups, there were significant (P ≤ .001) differences in BCVAs among the three time points, and the BCVAs at 6 and 12 months after vitrectomy were significantly better (P < .05) than those before vitrectomy. There was no significant difference in preoperative BCVA between the two groups, whereas BCVAs at 6 and 12 months after vitrectomy were significantly better (P = .014 and P = .019, respectively) in eyes without EIFLs than in those with EIFLs.
The best-corrected visual acuity (BCVA) time course in logarithm of the minimum angle of resolution (logMAR) units is shown. The abscissa represents perioperative time points and the ordinate represents the BCVA in logMAR units. Statistical analyses were performed using the Friedman repeated measures analysis of variance on ranks followed by the Tukey test (*P < .05).
CMTs before and at 6 and 12 months after vitrectomy were 372.9 µm ± 65.7 µm, 362.6 µm ± 35.8 µm, and 353.4 µm ± 32.5 µm in eyes without an EIFL; and 479.0 µm ± 89.4 µm, 365.5 µm ± 51.6 µm, and 359.4 µm ± 47.0 µm in eyes with an EIFL, respectively (Table 1, Figure 3). In both groups, there were significant (P ≤ .001) differences in CMT among the three time points, whereas CMTs after vitrectomy were significantly thinner (P < .05) compared to those before vitrectomy only in eyes with an EIFL. Preoperative CMT was significantly (P ≤ .001) thicker in eyes with EIFLs than in those without EIFLs, whereas postoperative CMTs were not significantly different between the two groups.
The central macular thickness (CMT) time course is shown. The abscissa represents perioperative time points and the ordinate represents the CMT. Statistical analyses were performed using the Friedman repeated measures analysis of variance on ranks followed by the Tukey test (*P < .05).
A DONFL was detected in five out of 54 eyes (9.3%) without EIFLs 6 and 12 months postoperatively, whereas it was detected in one (4.3%) and two (8.7%) out of 23 eyes with EIFLs 6 and 12 months postoperatively, respectively. There was no significant difference in the ratio of the presence of a DONFL postoperatively between the two groups.
An EIFL was not detected 12 months after vitrectomy in 18 (78.3%) out of 23 eyes in this study (Figure 1, Table 2). There were no significant differences in patient or ocular demographics between eyes with and without a persistent EIFL except for preoperative CMT, which was significantly thicker in eyes without EIFLs than in those with EIFLs (P = .019).
Patient and Ocular Demographics in ERM Cases With and Without Persistent EIFL
We compared the surgical outcomes of idiopathic ERM treatment in eyes with and without EIFLs. The EIFL prevalence was 29.9% in this study. The prevalence of an EIFL was reported to range from 32.5% to 36.6%,2,10,19 which was in agreement with the findings of this study.
The ratio of a disrupted EZ was significantly higher in eyes with EIFLs than in eyes without EIFLs preoperatively. In addition, a trend of worse BCVA in eyes with EIFLs was found preoperatively but without any significant difference when compared with those without EIFLs, which showed the same trend as in previous studies.2,10 Joe et al. demonstrated that the IRL (ie, EIFL) thickness of the fovea was the major determinant of visual acuity in eyes with ERMs.2 Govetto et al. also reported that the presence of an EIFL was an independent risk factor for lower VA.10
We compared visual outcomes after ERM vitrectomy in eyes with and without EIFLs. In both groups, BCVAs at 6 and 12 months after surgery were significantly better than those before vitrectomy. However, postoperative BCVAs were significantly better in eyes without EIFLs compared with eyes with EIFLs at both time points. These results suggest that the presence of an EIFL may be a predictor of a worse visual prognosis after ERM surgery, although vitrectomy was effective in improving VA in both eyes with and without EIFLs.
In this study, preoperative CMT was significantly thicker in eyes with EIFLs than in eyes without EIFLs; however, there was no statistically significant difference in preoperative BCVAs between the two groups. Postoperative CMTs were significantly reduced compared to previtrectomy values only in eyes with EIFLs, resulting in no significant differences in CMTs at 6 or 12 months after vitrectomy between eyes with and without EIFLs. However, postoperative BCVAs were significantly better in eyes without EIFLs than in those with EIFLs. There is controversy regarding the correlation between CMT and VA in eyes with ERMs. Cho et al. demonstrated that CMTs had significant correlations with BCVAs pre- and postoperatively.4 Suh et al. also mentioned that preoperative CMT was correlated with pre-and postoperative BCVAs.5 In addition, Kim et al. demonstrated that preoperative CMT was significantly correlated with final BCVA.7 On the other hand, others demonstrated that central foveal thicknesses were not significantly correlated with BCVAs pre- or postoperatively.8,9 In our current study, the preoperative CMT did not have any correlation with pre- or postoperative BCVAs. In addition, there were no significant correlations between CMTs and BCVAs at any time point (data not shown). Thus, CMT may not be a good indicator of VA in ERM cases.20
Recent OCT studies have demonstrated associations between visual function and the macular retinal structures of both the inner2–4 and outer5–9 layers in eyes with ERMs. However, most OCT parameters were measured manually by using software equipped with an OCT instrument or measurement software such as ImageJ (available at https://imagej.nih.gov/ij/). It seems unrealistic to use these structural parameters as an indication for ERM surgery in a daily clinical setting. In addition, it is sometimes difficult to make an accurate measurement of the retinal segment thickness and/or make an accurate evaluation of photoreceptor integrity in ERM cases complicated by substantial retinal thickening and remarkable anatomic disruption of the macula because retinal layers are substantially distorted and not clearly identified on OCT.7,10 Thus, an EIFL may be a new and reliable predictor of visual prognosis after ERM surgery.
Regarding the mechanism by which an EIFL develops, Yang et al. demonstrated that an inward tractional force caused by an ERM induced thickening of foveal structures and that chronic traction led to apposition of the adjacent inner retina — that is, the EIFL — and loss of normal cell alignment in the fovea.11 Thus, the EIFL would reopen after the release of inward traction by ERM removal in cases with a relatively short duration of traction, while a long-standing EIFL would prevent postoperative foveal restoration because of the irreversible anatomical change. Among cases with EIFLs preoperatively, the EIFL was not detected in 18 out of 23 eyes (78.3%) at 12 months after vitrectomy in this study, although there were no significant differences in BCVAs between cases with and without a persistent EIFL at any time point. Govetto et al.10 proposed an ERM staging scheme based on OCT findings and demonstrated anatomic progression to more advanced stages in 23 out of 131 ERMs (17.5%) during the follow-up period (range: 3 months to 84 months). Thus far, we cannot predict exactly when an EIFL develops. Therefore, these results suggest that early intervention before EIFL development may be desirable to obtain a better visual prognosis after ERM surgery, which has been facilitated by recent improvements in surgical technology and instrumentation.
There are some limitations to this study. First, this was a single-center, retrospective comparative study in which vitrectomies were performed by three surgeons. Nevertheless, the data obtained from our relatively large number of consecutive ERM cases that underwent vitrectomy provides new information about outcomes in ERM surgery. Second, combined cataract surgery performed in patients aged 50 years or older can affect visual prognosis. However, patients with a severe cataract were not included in this study to minimize the effect of the cataract on the studied surgical outcome. Patients whose OCT images showed a signal strength of 5 or less were also excluded from this study. Lastly, we evaluated only VA as a functional parameter of the surgical outcome; however, meta-morphopsia and macropsia are also predominant symptoms of ERM in addition to decreased vision. Further studies are needed to investigate associations between EIFLs and these symptoms in eyes with ERMs.
In conclusion, we compared surgical outcomes of ERM treatment in eyes with and without an EIFL and demonstrated that postoperative BCVA was significantly better in eyes with an ERM and without an EIFL than in those with an EIFL, although there were significant improvements in post-BCVA values compared to those obtained previtrectomy in both groups. In addition, among cases with a preoperative EIFL, there were no significant differences in BCVAs between cases with and without a persistent EIFL pre- or postoperatively. These results suggest that vitrectomy may be effective in improving visual acuity in ERM eyes both with and without an EIFL. However, ERM surgery may be desirable before EIFL development to obtain a better prognosis.
- Pesin SR, Olk RJ, Grand MG, et al. Vitrectomy for premacular fibroplasia. Prognostic factors, long-term follow-up, and time course of visual improvement. Ophthalmology. 1991;98(7):1109–1114. doi:10.1016/S0161-6420(91)32169-9 [CrossRef]
- Joe SG, Lee KS, Lee JY, Hwang JU, Kim JG, Yoon YH. Inner retinal layer thickness is the major determinant of visual acuity in patients with idiopathic epiretinal membrane. Acta Ophthalmol. 2013;91(3):e242–243. doi:10.1111/aos.12017 [CrossRef]
- Okamoto F, Sugiura Y, Okamoto Y, Hiraoka T, Oshika T. Time course of changes in aniseikonia and foveal microstructure after vitrectomy for epiretinal membrane. Ophthalmology. 2014;121(11):2255–2260. doi:10.1016/j.ophtha.2014.05.016 [CrossRef]
- Cho KH, Park SJ, Cho JH, Woo SJ, Park KH. Inner-retinal irregularity index predicts postoperative visual prognosis in idiopathic epiretinal membrane. Am J Ophthalmol. 2016;168:139–149. doi:10.1016/j.ajo.2016.05.011 [CrossRef]
- Suh MH, Seo JM, Park KH, Yu HG. Associations between macular findings by optical coherence tomography and visual outcomes after epiretinal membrane removal. Am J Ophthalmol. 2009;147(3):473–480.e3. doi:10.1016/j.ajo.2008.09.020 [CrossRef]
- Inoue M, Morita S, Watanabe Y, et al. Inner segment/outer segment junction assessed by spectral-domain optical coherence tomography in patients with idiopathic epiretinal membrane. Am J Ophthalmol. 2010;150(6):834–839. doi:10.1016/j.ajo.2010.06.006 [CrossRef]
- Kim JH, Kim YM, Chung EJ, Lee SY, Koh HJ. Structural and functional predictors of visual outcomes of epiretinal membrane surgery. Am J Ophthalmol. 2012;153(1):103–110.e1. doi:10.1016/j.ajo.2011.06.021 [CrossRef]
- Shimozono M, Oishi A, Hata M, et al. The significance of cone outer segment tips as a prognostic factor in epiretinal membrane surgery. Am J Ophthalmol. 2012;153(4):698–704, 704.e1. doi:10.1016/j.ajo.2011.09.011 [CrossRef]
- Shiono A, Kogo J, Klose G, et al. Photoreceptor outer segment length: A prognostic factor for idiopathic epiretinal membrane surgery. Ophthalmology. 2013;120(4):788–794. doi:10.1016/j.ophtha.2012.09.044 [CrossRef]
- Govetto A, Lalane RA 3rd, Sarraf D, Figueroa MS, Hubschman JP. 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 [CrossRef]
- Yang HS, Kim JT, Joe SG, Lee JY, Yoon YH. Postoperative restoration of foveal inner retinal configuration in patients with epiretinal membrane and abnormally thick inner retina. Retina. 2015;35(1):111–119. doi:10.1097/IAE.0000000000000276 [CrossRef]
- Ogura Y, Takanashi T, Ishigooka H, Ogino N. Quantitative analysis of lens changes after vitrectomy by fluorophotometry. Am J Ophthalmol. 1991;111(2):179–183. doi:10.1016/S0002-9394(14)72256-1 [CrossRef]
- Melberg NS, Thomas MA. Nuclear sclerotic cataract after vitrectomy in patients younger than 50 years of age. Ophthalmology. 1995;102(10):1466–1471. doi:10.1016/S0161-6420(95)30844-5 [CrossRef]
- Enaida H, Hisatomi T, Hata Y, et al. Brilliant blue G selectively stains the internal limiting membrane/brilliant blue G-assisted membrane peeling. Retina. 2006;26(6):631–636.
- Bamonte G, Mura M, Stevie Tan H. Hypotony after 25-gauge vitrectomy. Am J Ophthalmol. 2011;151(1):156–160. doi:10.1016/j.ajo.2010.06.042 [CrossRef]
- Tsunoda K, Watanabe K, Akiyama K, Usui T, Noda T. Highly reflective foveal region in optical coherence tomography in eyes with vitreomacular traction or epiretinal membrane. Ophthalmology. 2012;119(3):581–587. doi:10.1016/j.ophtha.2011.08.026 [CrossRef]
- Tadayoni R, Paques M, Massin P, Mouki-Benani S, Mikol J, Gaudric A. Dissociated optic nerve fiber layer appearance of the fundus after idiopathic epiretinal membrane removal. Ophthalmology. 2001;108(12):2279–2283. doi:10.1016/S0161-6420(01)00856-9 [CrossRef]
- Hisatomi T, Tachibana T, Notomi S, et al. Internal limiting membrane peeling-dependent retinal structural changes after vitrectomy in rhegmatogenous retinal detachment. Retina. 2018;38(3):471–479. doi:10.1097/IAE.0000000000001558 [CrossRef]
- Govetto A, Su D, Farajzadeh M, et al. Microcystoid macular changes in association with idiopathic epiretinal membranes in eyes with and without glaucoma: Clinical insights. Am J Ophthalmol. 2017;181:156–165. doi:10.1016/j.ajo.2017.06.023 [CrossRef]
- Scheerlinck LM, van der Valk R, van Leeuwen R. Predictive factors for postoperative visual acuity in idiopathic epiretinal membrane: A systematic review. Acta Ophthalmol. 2015;93(3):203–212. doi:10.1111/aos.12537 [CrossRef]
Patient and Ocular Demographics in ERM Cases With and Without EIFL
|EIFL (−)||EIFL (+)||P Value|
|Eyes / Patients (n)||54/53||23/23|
|Age (Years, Mean ± SDs)||69.1 ± 10.7||68.4 ± 9.8||.428|
|Sex (Female : Male)||30:23||16:7||.420|
|Preoperative Lens Status (Phakia : IOL)||35:19||13:10||.667|
|Axial Length (mm, Mean ± SDs)||24.75 ± 1.50||24.47 ± 1.73||.624|
|Vitrectomy Gauge Size Performed (25 : 23)||41:13||14:9||.288|
|Preoperative EZ Status (Intact : Disrupted or Unidentified)||53:1||17:6||.002|
|Preoperative Cotton Ball Sign (Positive : Negative or Unidentified)||13:41||7:16||.765|
|BCVA (logMAR, Mean ± SDs)|
| Pre||0.28 ± 0.18||0.37 ± 0.21||.061|
| 6M||0.10 ± 0.13||0.17 ± 0.12||.014|
| 12M||0.09 ± 0.13||0.15 ± 0.13||.019|
|CMT (µm, Mean ± SDs)|
| Pre||372.9 ± 65.7||479.0 ± 89.4||< .001|
| 6M||362.6 ± 35.8||365.5 ± 51.6||.555|
| 12M||353.4 ± 32.5||359.4 ± 47.0||.501|
Patient and Ocular Demographics in ERM Cases With and Without Persistent EIFL
|Persistent EIFL (−)||Persistent EIFL (+)||P Value|
|Eye / Patients (n)||18/18||5/5|
|Age (Years, Mean ± SDs)||68.2 ± 10.5||69.2 ± 7.3||.848|
|Sex (Female : Male)||12:6||4:1||1.000|
|Preoperative Lens Status (Phakia: IOL)||10:8||3:2||1.000|
|Axial Length (mm, Mean ± SDs)||24.73 ± 1.68||24.15 ± 2.34||.541|
|Vitrectomy Gauge Size Performed (25:23)||11:7||3:2||1.000|
|Preoperative EZ Status (Intact : Disrupted or Unidentified)||12:6||5:0||.272|
|Preoperative Cotton Ball Sign (Positive : Negative or Unidentified)||6:12||1:4||1.000|
|BCVA (logMAR, Mean ± SDs)|
| Pre||0.36 ± 0.23||0.41 ± 0.11||.447|
| 6M||0.16 ± 0.13||0.19 ± 0.08||.491|
| 12M||0.15 ± 0.14||0.15 ± 0.09||.970|
|CMT (µm, mean ± SDs)|
| Pre||497.9 ± 89.5||411.0 ± 50.2||.019|
| 6M||364.2 ± 54.6||370.2 ± 43.9||.825|
| 12M||353.9 ± 51.4||379.0 ± 17.5||.301|