Associations between Metamorphopsia and Foveal Microstructure in Patients with Epiretinal Membrane

Abstract

Purpose.: To investigate the relationship between the severity of metamorphopsia and the foveal microstructure measured with spectral-domain optical coherence tomography (SD-OCT) in patients with epiretinal membrane (ERM).

Methods.: This study included 48 eyes of 48 patients with idiopathic ERM and 18 age-matched normal controls. We examined the logarithm of the minimum angle of resolution best-corrected visual acuity (logMAR BCVA) and the severity of metamorphopsia using M-CHARTS. Central foveal thickness (CFT), central retinal thickness at the fovea (CRT-1mm) and parafovea (CRT-3mm), and macular volume (MV) were measured with SD-OCT software. Based on the obtained OCT image, we divided the 1.0 × 1.0-mm area centered on the fovea into nine sections and quantified the following parameters using an image-processing program: mean thickness of the ganglion cell layer (GCL), inner nuclear layer (INL), and outer retinal layer (ONL+OPL: outer nuclear layer and outer plexiform layer), the degree of the photoreceptor inner and outer segment junction (IS/OS), and external limiting membrane (ELM) disruption.

Results.: CFT, CRT-1mm, CRT-3mm, MV, mean GCL, INL, and ONL+OPL thickness were significantly larger in patients with ERM than in normal controls. Multiple regression analysis revealed that the severity of metamorphopsia was significantly related to the mean INL thickness (P < 0.0001). LogMAR BCVA had a significant correlation with the degree of IS/OS disruption (P < 0.05), whereas other parameters were not relevant.

Conclusions.: In patients with idiopathic ERM, the degree of metamorphopsia is associated with INL thickness, and IS/OS status influences visual acuity.

Introduction

Epiretinal membrane (ERM) is a nonvascular fibrocellular proliferation that develops on the surface of the internal limiting membrane, resulting in retinal wrinkling and distortion. Metamorphopsia is one of the most common symptoms in patients with ERM, with 80 to 85% of patients complaining of moderate to severe distortion. ^1,2 Visual acuity improves in many patients after successful removal of ERM, whereas it has been reported that metamorphopsia can still be present even after successful surgery. ¹ In addition, a study using the 25-item National Eye Institute Visual Function Questionnaire demonstrated that the severity of metamorphopsia strongly influenced vision-related quality of life in patients with ERM. ^3,4  

Recent advancement of optical coherence tomography (OCT) technologies has provided critical insights into various retinal conditions, including ERM. ^5,6 Although conventional time-domain OCT had limitations in detecting subtle pathologic changes, the new generation of spectral-domain OCT (SD-OCT) with its high resolution and high scanning speed has allowed layer-by-layer evaluation of the retina. ^7,8 SD-OCT enhances the resolution with which the intraretinal architectural morphology, such as ganglion cell layer (GCL), inner nuclear layer (INL), outer plexiform layer (OPL), outer nuclear layer (ONL), photoreceptor inner, and outer segment junction (IS/OS), and external limiting membrane (ELM), which can be clearly seen. ⁷ Studies with SD-OCT revealed that the visual acuity in ERM patients deteriorated due to disruption of the IS/OS, ^9–13 thickening of the central macula, ^11–15 and thickening of the outer retina. ^14,15 It has also been reported that preoperative central retinal thickness and integrity of the IS/OS junction may be prognostic factors of visual acuity after ERM surgery. ^9,10,16  

However, few studies have investigated the relation between metamorphopsia and OCT findings. Watanabe et al. ¹⁴ reported that INL was thicker in ERM patients with broad metamorphopsia than that in ERM patients without metamorphopsia. This study used the Amsler charts, which have been widely used to detect and evaluate metamorphopsia in patients with macular diseases. ^17,18 With the Amsler charts, however, it is difficult to quantify the severity of metamorphopsia because the patients have to self-describe the degree of image distortion. In contrast, M-CHARTS (Inami Co., Tokyo, Japan) is an instrument that can easily and quantitatively evaluate the degree of metamorphopsia associated with macular diseases, with the patients only to answer whether the line is either distorted or not distorted. ^19,20  

The purpose of the present study was to quantify metamorphopsia using M-CHARTS in ERM patients and to investigate the relationship between the severity of metamorphopsia and the anatomic features of the macular region assessed with SD-OCT. 

Methods

Patients

We included 48 eyes of 48 consecutive patients with idiopathic ERM, who were referred to the Tsukuba University Hospital between January 7, 2010 and October 24, 2011. There were 29 males and 19 females, averaging 64.5 ± 10.7 years of age (mean ± SD). Eighteen age-matched subjects served as normal controls. This study was approved by the Institutional Review Board at the Tsukuba University Hospital and was in adherence to the tenets of the Declaration of Helsinki. Signed informed consent was obtained from all study subjects. 

ERM was defined as a translucent or semitranslucent membrane with macular thickening involving the center of the macula, with or without distortion and wrinkling of the inner retinal surface on biomicroscopy and SD-OCT. Exclusion criteria included patients with a previous history of vitreoretinal surgery and ophthalmic disorders except mild refractive errors and mild cataract, and best-corrected visual acuity (BCVA) of >0.7 (logMAR). Eyes with secondary ERM due to retinal vascular disease, uveitis, trauma, and retinal breaks were also excluded from the study. 

BCVA was measured using the Landolt Chart, and the degree of metamorphopsia was quantified using M-CHARTS, which consist of 19 dotted lines, with dot intervals ranging from 0.2 to 2.0° of visual angle. If the straight line is substituted with a dotted line and the dot interval is changed from fine to coarse, the distortion of the line decreases with the increasing dot interval, until the dotted line appears straight. ^19,20 First, a vertical straight line (0°) was shown to the patient. If the patient recognized the straight line as straight, the metamorphopsia score was 0. If the patient recognized the straight line as irregular or curved, then subsequent pages of M-CHARTS, in which the dot intervals of the dotted line change from fine to coarse, were shown one after another. When the patient recognized a dotted line as being straight, the visual angle that separated the dots was considered to represent his/her metamorphopsia score for vertical lines. Then the M-CHARTS were rotated 90° and the same test was performed using horizontal lines. The examinations were repeated three times, and their mean values were used for data analyses. Each examination was performed at 30 cm, with the refraction of the eye exactly corrected for this distance. 

Measurement of Retinal Microstructure with SD-OCT

Retinal images were obtained using SD-OCT (Cirrus high-definition OCT; Carl Zeiss, Dublin, CA). Macular cube 512 × 128 scans and 5-line Raster scans were performed for each eye using a commercial analytic software package (Cirrus analysis software, version 3.0; Carl Zeiss). Scans with signal strength of more than 8/10 were considered to be appropriate, and a representative image was selected. Based on the obtained OCT image, the following nine parameters were measured: central foveal thickness (CFT), central retinal thickness at the fovea (CRT-1mm: within a circle of diameter of 1 mm), central retinal thickness at the parafovea (CRT-3mm: within a circle of diameter of 3 mm), macular volume (MV: volume of the 5 × 5-mm retinal area centered on the fovea), thickness of the ganglion cell layer (GCL), inner nuclear layer (INL), and outer retinal layer (ONL+OPL; outer nuclear layer and outer plexiform layer), degree of the photoreceptor inner and outer segment junction (IS/OS) disruption, and degree of the external limiting membrane (ELM) disruption. 

The high-definition OCT device (Cirrus; Carl Zeiss) automatically calculates CRT-1mm, CRT-3mm, and MV. Based on images obtained by 5-line Raster scans, we divided the 1.0 × 1.0-mm area centered on the fovea into nine sections at 0.25-mm intervals, measuring the thickness of the GCL, INL, and ONL+OPL at each section to use the mean scores of nine sections in subsequent analyses (Fig. 1). A freely available image-processing program (ImageJ software, developed by Wayne Rasband, National Institutes of Health, Bethesda, MD; available at http://rsbweb.nih.gov/ij/index.html) was used to quantify the thickness of each retinal layer. To simply quantify the degree of IS/OS disruption, we assessed in each of the above nine sections whether IS/OS was disrupted, and defined the number of disrupted sections as the degree of IS/OS disruption. Diagnosis of a disrupted IS/OS was made based on loss and irregularity of the hyperreflexive line corresponding to the IS/OS junction. We also assessed the degree of ELM disruption in a similar fashion. Two graders (YS, YO) measured the retinal layer thickness and assessed the status of the IS/OS and ELM line. Both graders were masked to the clinical findings of the patients, including their visual acuity and metamorphopsia scores. 

Figure 1. 

 

This schema indicates the measurement position of the thickness of each retinal layer and the integrity of IS/OS and ELM. Five horizontal broken lines represent scanning lines of 5-line Raster scans with SD-OCT. The 1.0 × 1.0-mm area centered on the fovea was divided into 9 sections at 0.25-mm intervals. Each retinal layer and integrity of IS/OS and ELM were measured at each section.

Statistical Analysis

The mean scores were compared and SD values were calculated for each parameter of visual function and OCT measurements. BCVA, measured using the Landolt Chart, was expressed as the logarithm of minimal angle of resolution (logMAR). The Mann-Whitney U test was performed to compare age and each OCT parameter between ERM patients and normal controls. The associations between the severity of metamorphopsia, logMAR BCVA, and OCT parameters were examined by the Spearman rank correlation test. Multiple regression analysis was performed to investigate the relationship between various explanatory variables, severity of metamorphopsia, and logMAR BCVA. Variables tested were CFT, CRT-1mm, CRT-3mm, MV, mean thickness of GCL, INL, ONL+OPL, degree of IS/OS disruption, and degree of ELM disruption. All tests of associations were considered statistically significant if P < 0.05. The analyses were carried out using a commercial software package (StatView software, version 5.0; SAS, Inc., Cary, NC). 

Results

Table 1 lists visual functions and OCT parameters of patients with ERM and normal controls. OCT parameters such as CFT, CRT-1mm, CRT-3mm, MV, mean GCL thickness, mean INL thickness, and mean ONL+OPL thickness were significantly larger in patients with ERM than those in normal controls. 

Table 1. 

 

Visual Function and Optical Coherence Tomography Parameters of Patients with Epiretinal Membrane and Normal Controls

	Patients with Epiretinal Membrane	Normal Controls	P Value
No. of eyes	48	18
Male/female	29/19	12/6	—
Age (y)	64.5 ± 10.7 (26−84)	62.6 ± 13.2 (26–84)	0.561
LogMAR BCVA	0.24 ± 0.16 (−0.08–0.52)	−0.03 ± 0.09 (−0.18–0.15)	<0.0001*
Metamorphopsia score
Vertical	0.63 ± 0.58 (0.0–2.0)	0.02 ± 0.05 (0.0–0.2)	<0.0001*
Horizontal	0.82 ± 0.66 (0.0–2.0)	0.00 ± 0.00 (0.0–0.0)	<0.0001*
Average	0.72 ± 0.57 (0.0–2.0)	0.01 ± 0.03 (0.0–0.1)	<0.0001*
CFT (μm)	380 ± 148 (84–658)	194 ± 30 (145–258)	<0.0001*
CRT-1mm (μm)	429 ± 97 (218–614)	258 ± 20 (215–293)	<0.0001*
CRT-3mm (μm)	411 ± 60 (327–538)	314 ± 23 (269–341)	<0.0001*
MV (mm3)	11.9 ± 1.3 (8.7–15.5)	10.1 ± 1.0 (8.1–11.6)	<0.0001*
Mean GCL thickness (μm)	125 ± 44 (45–260)	90 ± 9 (74–105)	0.001*
Mean INL thickness (μm)	87 ± 33 (37–175)	38 ± 3 (31–43)	<0.0001*
Mean ONL+OPL thickness (μm)	195 ± 26 (143–254)	151 ± 16 (125–173)	<0.0001*
Degree of IS/OS disruption (No.)	1.0 ± 1.7 (0–7)	0.0 ± 0.0 (0–0)	<0.05*
Degree of ELM disruption (No.)	2.0 ± 2.4 (0–9)	0.0 ± 0.0 (0–0)	<0.005*

 Values are presented as the mean ± SD (range).

*  Significant at P < 0.05.

In simple regression analysis, the average metamorphopsia score exhibited a significant correlation with CFT (r = 0.294, P < 0.05), CRT-1mm (r = 0.312, P < 0.05), mean GCL thickness (r = 0.415, P < 0.005), and mean INL thickness (r = 0.711, P < 0.0001; Fig. 2), but not with CRT-3mm (r = 0.238, P = 0.115), MV (r = 0.146, P = 0.359), mean ONL+OPL thickness (r = −0.130, P = 0.383), degree of IS/OS disruption (r = −0.079, P = 0.599), or degree of ELM disruption (r = 0.034, P = 0.820). Vertical metamorphopsia scores exhibited a significant correlation with mean INL thickness in the vertical cross-section (r = 0.641, P < 0.0001) and mean INL thickness in the horizontal cross-section (r = 0.384, P < 0.001). Horizontal metamorphopsia scores showed a significant correlation with mean INL thickness in the vertical cross-section (r = 0.565, P < 0.0001) and mean INL thickness in the horizontal cross-section (r = 0.671, P < 0.0001). 

Figure 2. 

 

Average metamorphopsia score versus mean INL thickness in patients with epiretinal membrane.

LogMAR BCVA showed a significant correlation with CFT (r = 0.440, P < 0.005), CRT-1mm (r = 0.368, P < 0.05), mean GCL thickness (r = 0.363, P < 0.05), mean INL thickness (r = 0.328, P < 0.05), degree of IS/OS disruption (r = 0.575, P < 0.0001), and degree of ELM disruption (r = 0.528, P < 0.0001; Fig. 3), but not with CRT-3mm (r = 0.284, P = 0.059), MV (r = 0.044, P = 0.772), or mean ONL+OPL thickness (r = −0.243, P = 0.100). 

Figure 3. 

 

The logMAR BCVA versus the degree of photoreceptor IS/OS disruption in patients with epiretinal membrane.

Table 2 summarizes the results of multiple regression analysis on the relation between metamorphopsia and OCT parameters. The average metamorphopsia score was significantly related to the mean INL thickness (P < 0.0001), but not to other parameters. Multiple regression revealed that logMAR BCVA had a significant correlation with the degree of IS/OS disruption (P < 0.05), whereas other parameters were not relevant (Table 3). 

Table 2. 

 

Multiple Regression Analysis of the Average Metamorphopsia Score and Optical Coherence Tomography Parameters in Patients with Epiretinal Membrane

	β	SE	P Value
CFT (μm)	0.103	0.001	0.703
CRT-1mm (μm)	–0.055	0.002	0.893
CRT-3mm (μm)	–0.372	0.004	0.417
MV (mm3)	0.037	0.114	0.888
Mean GCL thickness (μm)	–0.207	0.003	0.376
Mean INL thickness (μm)	1.080	0.003	<0.0001*
Mean ONL+OPL thickness (μm)	–0.135	0.003	0.451
Degree of IS/OS disruption (No.)	0.003	0.044	0.998
Degree of ELM disruption (No.)	–0.100	0.035	0.532

 β, standard regression coefficient; SE, standard error.

*  Significant at P < 0.05.

Table 3. 

 

Multiple Regression Analysis of the LogMAR BCVA and Optical Coherence Tomography Parameters in Patients with Epiretinal Membrane

	β	SE	P Value
CFT (μm)	0.391	0.002	0.100
CRT-1mm (μm)	–0.480	0.001	0.244
CRT-3mm (μm)	0.274	0.001	0.547
MV (mm3)	–0.268	0.034	0.316
Mean GCL thickness (μm)	–0.229	0.001	0.256
Mean INL thickness (μm)	0.208	0.001	0.266
Mean ONL+OPL thickness (μm)	–0.173	0.001	0.404
Degree of IS/OS disruption (No.)	0.473	0.013	<0.05*
Degree of ELM disruption (No.)	0.181	0.010	0.262

*  Significant at P < 0.05.

We present two representative cases: Case 1 (Figs. 4A, 4B) and Case 2 (Figs. 4C, 4D). The logMAR BCVA of Case 1 (0.40) is similar to that of Case 2 (0.50), whereas the average metamorphopsia score of Case 2 (0.1) was significantly smaller than that of Case 1 (1.15). In both cases, IS/OS and ELM were intact. The ONL was slightly thicker in Case 1 than that in Case 2, whereas the INL of Case 1 (112 μm) was significantly thicker than that of Case 2 (52 μm). 

Figure 4. 

 

( A, B) Image of the right eye of a 60-year-old woman with ERM. LogMAR BCVA was 0.40, whereas M-CHARTS metamorphopsia scores were 0.5 for vertical lines and 1.8 for horizontal lines. (A) Fundus photograph. (B) SD-OCT image: INL (arrows) was thick, and the mean INL thickness was 112 μm. (C, D) Image of the right eye of a 70-year-old woman with ERM. LogMAR BCVA was 0.50, whereas M-CHARTS metamorphopsia scores were 0 for vertical lines and 0.2 for horizontal lines. (C) Fundus photograph. (D) SD-OCT image: INL (arrows) was thin, and the mean INL thickness was 52 μm.

Discussion

This study presents the relationship between metamorphopsia and foveal microstructure in 48 patients with idiopathic ERM. Based on the quantification of metamorphopsia with M-CHARTS and evaluation of foveal microstructure by nine parameters with an image-processing program, we were able to examine the association between metamorphopsia and OCT parameters in detail. We found that metamorphopsia was affected by multiple factors, of which INL thickness had the largest impact on the severity of metamorphopsia as shown by the multiple regression analysis. 

ERM affected thickness of all retinal layers compared with normal controls. It has been known that ERM affects the underlying retinal structures. These results are consistent with the finding of a previous report. ¹⁵  

No report so far has discussed the association between metamorphopsia and total retinal thickness. Meanwhile, several studies reported that central retinal thickness and visual acuity were negatively correlated. ^{6,11–13,15,21–23} In our study, both metamorphopsia and visual acuity were significantly correlated with CFT and CRT-1mm. Our results suggest that retinal thickening itself causes deterioration not only in visual acuity but also in metamorphopsia. 

The severity of metamorphopsia was associated with the GCL and INL thickness, but not with thickening of the outer retina by the simple regression analysis. In particular, a multiple regression analysis revealed that among nine parameters of foveal microstructure, INL thickness was the only single factor that showed a significant association. Watanabe et al. ¹⁴ classified ERM patients into three groups according to the extent of metamorphopsia using the Amsler charts, and reported that patients with broad metamorphopsia exhibited thicker INL than patients without metamorphopsia. Since the severity of metamorphopsia was evaluated based on the number of distorted squares on the Amsler charts, the result mainly indicates the extent of central metamorphopsia. In contrast, M-CHARTS have been used to quantify the severity of metamorphopsia in patients with ERM and macular hole. ^19,20,24 M-CHARTS can evaluate frequency components of distortion in metamorphopsia. The high-frequency component of metamorphopsia, which is usually observed in mild ERM, is detected by fine dotted lines, but not by coarse dotted lines. In severe ERM, on the other hand, the large amplitude and low-frequency components of metamorphopsia increase, so it is easy to detect them by all kinds of lines, including coarse dotted lines. Thus, the inner foveal retina seems to be associated with the frequency components of distortion in metamorphopsia, but not with the extent of central metamorphopsia. 

Arimura et al. ²⁰ investigated the relationship between the degree of retinal contraction and metamorphopsia using M-CHARTS in patients with ERM, and found that the vertical metamorphopsia score correlated with the degree of horizontal contraction, and the horizontal metamorphopsia score correlated with the degree of vertical contraction. In our study, vertical metamorphopsia correlated with INL thickness in the vertical and horizontal cross-sections, respectively. Horizontal metamorphopsia also showed a correlation with INL thickness in both sections. The difference in these results might have arisen from different sizes of the evaluated areas. Whereas Arimura et al. ²⁰ observed the central 20° macula area, we performed retinal layer measurement in a 0.5 × 0.5-mm small area centered on the fovea. We thus might not have found any difference between the vertical and horizontal sections. 

Previous studies that investigated if, in eyes with ERM, B-wave and oscillatory potentials were more reduced than A-wave using focal macular electroretinography found that there was a significant correlation between relative B-wave amplitude (affected/normal contralateral eye) and visual acuity. ^25,26 The authors suggested that ERM damaged the neurons in the inner retinal layers. Visual acuity, as well as metamorphopsia, was significantly correlated with mean GCL and INL thickness, but not with thickening of outer retina in our results. Abnormality of the inner retinal layers can influence not only visual acuity but also metamorphopsia. It is still unclear why damages of the inner retina augment metamorphopsia. Our speculation is as follows: When INL is stretched or becomes edematous due to vitreous traction or membrane contraction, the structures of horizontal, bipolar, amacrine, and Müller cell bodies, which comprise INL, may change. This inhibits the normal function of synaptic junctions and lowers photoreceptor sensitivity, causing metamorphopsia. In addition, thickening of GCL and INL may induce aberration in the retina itself and thus quality of vision may be deteriorated. 

In our results, metamorphopsia was not associated with the degree of IS/OS or ELM disruption. Ooto et al. ²⁷ demonstrated no significant association between disruption of the IS/OS on SD-OCT and severity of metamorphopsia measured using M-CHARTS in eyes with ERM. This is consistent with our findings. In contrast, visual acuity was significantly correlated with the degree of IS/OS and ELM disruption in our study. In particular, the multiple regression analysis revealed that among nine parameters of foveal microstructure, IS/OS disruption was the only single factor that showed a significant association. The relationship between disruption of the IS/OS and visual acuity in eyes with ERM remains in dispute. Michalewski et al. ¹² and Mitamura et al. ¹³ found that visual acuity was worse in eyes with a more severe degree of IS/OS disruption, but Suh et al., ¹¹ Arichika et al., ¹⁵ and Ooto et al. ²⁷ did not find a significant difference in visual acuity between eyes with disrupted IS/OS and eyes with intact IS/OS. Few studies have investigated the association between the status of ELM and visual function. Theodossiadis et al. ²⁸ demonstrated that the preoperative condition of ELM and IS/OS is the most important and significant factor affecting postoperative visual acuity after secondary ERM peeling. Wakabayashi et al. ²⁹ reported that the integrity of IS/OS junction and ELM signals may account for visual restoration in patients with preoperative macula-off rhegmatogenous retinal detachment. Based on these findings, it is suggested that the integrity of IS/OS is strongly related not to metamorphopsia but to visual acuity. 

The limitations of this study include a relatively small sample size, measurements based on only five horizontal B-scan cross-sections, and imaging inaccuracy in each of the retinal layers of ERM eyes by manual segmentation. Future studies with a larger sample size and improved OCT technologies will be needed. 

In conclusion, our results suggest that metamorphopsia was strongly associated with thickness of INL, whereas visual acuity was associated with the integrity of IS/OS in patients with ERM. These results emphasize that in patients with ERM, visual acuity and metamorphopsia, two components of visual function, are associated with different retinal layers. 

Footnotes

 The authors have no individual or family investments, stock, or business ownership exceeding 1% of a company's worth, consulting, retainers, patents, or other commercial interests in the product or company described in the current article. There is no involvement in the marketing of any product, drug, instrument, or piece of equipment discussed in the manuscript that could cause or be perceived to be a conflict of interest.

Footnotes

 Disclosure: F. Okamoto, None; Y. Sugiura, None; Y. Okamoto, None; T. Hiraoka, None; T. Oshika, None