Inner retinal layer thickness is the major determinant of visual acuity in patients with idiopathic epiretinal membrane

Editor, Idiopathic epiretinal membrane (ERM) may be asymptomatic until the tangential tractional forces on the retina cause significant thickening and intraretinal structural distortion at the macula (Pournaras et al. 2000). In clinical practice, we often witness asymptomatic patients with prominent ERM. It is unclear whether central subfield thickness (CST), inner retinal layer thickness (IRLT), outer retinal layer thickness (ORLT) or photoreceptor inner segment/outer segment (IS/OS) disruption is the major contributing factor to reduced visual acuity in ERM (Michalewski et al. 2007; Arichika et al. 2010). Medical records for 123 patients (134 eyes: 107 phakic eyes and 27 pseudophakic eyes) with idiopathic ERM were reviewed. Cases of ERM secondary to other diseases or accompanied by lens opacity of N3, C3, P3 or greater (LOCS III) (Chylack et al. 1993) were excluded. Fluorescein angiography was performed in all eyes to exclude associated vascular occlusive disease. Best-corrected visual acuity (BCVA) was converted to the logMAR scale. SD-OCT examinations were performed using a Cirrus™ SD-OCT system (Carl Zeiss Meditec Inc., Dublin, CA, USA) with software version 4.0. CST was measured automatically. Inner retinal layer thickness (length from the vitreoretinal interface to the outer border of the inner nuclear layer) and ORLT (length from the inner border of the outer plexiform layer to the photoreceptor IS/OS junction) were measured manually on the horizontal raster image passing through the foveal centre. The mean age was 65.1 ± 8.5 years. The CST values ranged from 189 to 661 μm, and BCVA in logMAR ranged from 0.00 to 1.30. ERM was diffusely attached in 105 eyes (fovea-attached type; type I) and spared the fovea in 29 eyes (pseudohole type; type II). While BCVA in type II eyes was uniform (0.0957 ± 0.11), it varied widely in type I eyes. BCVA correlated with CST, but showed a stronger correlation with IRLT (p < 0.001). Interestingly, BCVA decreased only when IRLT was thicker than normal (>75 μm), regardless of CST thickening. Photoreceptor layer damage, defined as the disruption of the IS/OS junction, was found in only two eyes. Based on these results, eyes with type I ERM were classified into three groups: preserved foveal depression (type 1a), elevated fovea but no IRL thickening (type Ib), and elevated fovea and abnormally thick IRL (type Ic). Eyes with no increase in CST (type Ia and II) maintained good visual acuity (Fig. 1). While both type Ib and type Ic eyes showed significant thickening of the CST, their BCVA profiles were markedly different. Whereas the BCVA of most type Ib eyes was 0.8 or better, that of type Ic eyes varied depending on the IRLT (p < 0.001), but not on the ORLT, at the fovea (p = 0.09). Correlations between best corrected visual acuity and foveal layers' thickness. The first column shows data from all patients. The second, third, fourth and fifth columns show the data for patients in Type Ia (27 eyes), Type Ib (29 eyes), Type Ic (49 eyes) and Type II (29 eyes), respectively. The third row shows the correlations between foveal inner retinal layer thickness (IRLT) and visual acuity. The bottom row shows the correlations between central subfield thickness (CST) and visual acuity. In general visual acuity shows a stronger correlation with IRLT than CST (correlation coefficients: 0.74 and 0.62, respectively), in group analysis, however only Type Ic showed a significant correlation between IRLT and VA. White arrow indicates IRLT and black arrow outer retinal layer thickness. We have shown that IRLT at the foveal centre affects vision more significantly than CST in patients with idiopathic ERM. Patients with a thickened macula do not necessarily experience reduced vision. Inner retinal layer thickness at the central fovea lacks several layers, measuring only 50–80 μm by high-resolution OCT (Koo et al. 2012). The cut-off value in the present study was 75 μm. The pathogenesis of the thickened IRL observed in type Ic eyes is unknown. We speculate that when the tangential traction becomes intense, the inward peak of the outer nuclear layer becomes exaggerated, resulting in the attachment of adjacent parafoveal inner retinal layers. Once attached, the cells in each retinal layer lose polarity and normal neural transmission fails. This may explain why VA is severely affected by a thickened IRL. Unlike previous reports, photoreceptor IS/OS disruption was rarely observed in our patient population. Previous studies may have included cases of ERM secondary to other retinal diseases such as macular branch vein occlusion or cases of long-standing advanced stage (Inoue et al. 2011). In conclusion, the present findings suggest that IRLT of the fovea is the major determinant of visual acuity in patients with idiopathic ERM and may provide important information on the optimal timing of surgical intervention in these patients. The authors indicate no financial support or financial conflicts of interest. SGJ and YHY made design and conduct of study; SGJ, KSL, JYL, JH, JK and YHY performed collection, analysis and interpretation of data; SGJ and YHY carried out literature search; SGJ and YHY carried out preparation, review and writing; and SGJ, KSL, JYL, JH, JK and YHY were involved in approval of the manuscript. The study was carried out with approval from the Institutional Research Board at the Asan Medical Centre and adhered to the tenets set out in the Declaration of Helsinki.