Influence of Retinal Vessel Printings on Metamorphopsia and Retinal Architectural Abnormalities in Eyes With Idiopathic Macular Epiretinal Membrane

Abstract

Purpose.: To investigate whether the presence of retinal vessel printings (RVPs) in eyes having idiopathic epiretinal membrane (ERM) is associated with a higher degree of metamorphopsia or with more prominent abnormalities in the retinal architecture compared with eyes not having RVPs.

Methods.: A cross-sectional study of 36 eyes in 36 patients was conducted. The patients were divided into two groups (18 eyes per group) on the basis of the presence or absence of RVPs on fundus autofluorescence (FAF) imaging. Metamorphopsia was assessed using M-CHARTS. The optical coherence tomography (OCT) and FAF images were recorded using the Spectralis HRA+OCT. The macula was divided into 15 squares, and the areas between the displaced retinal vessels and the corresponding RVPs were calculated.

Results.: Vertical, horizontal, and average metamorphopsia scores did not differ between the two groups. However, only two patients in the group with RVPs had an average metamorphopsia score of 0.6 or less versus nine patients in the group without RVPs (P = 0.029). There was no correlation between the area of displacement and vertical (P = 0.44), horizontal (P = 0.8), or average (P = 0.38) metamorphopsia scores. The eyes with RVPs manifested a higher degree of irregularity of the external limiting membrane (ELM) and the photoreceptor inner segment/outer segment (IS/OS) lines (P = 0.042 and P = 0.014, respectively). The presence of RVPs was the only independent variable associated with an average metamorphopsia score of 0.6 or higher.

Conclusions.: The presence of RVPs in eyes with idiopathic macular ERM is usually associated with an average metamorphopsia score of 0.6 or higher using M-CHARTS and with a higher degree of irregularity of the ELM and IS/OS lines at the fovea.

Introduction

Epiretinal membrane (ERM) is a frequently diagnosed pathologic condition in the elderly. It is a fibrocellular proliferation that develops on the surface of the internal limiting membrane. 

The traction forces exerted on the retina during ERM formation can be either anteroposterior or tangential. Anteroposterior forces produce vertical traction and consequently an increased thickness of the retina, whereas tangential forces drag the superficial retinal layers away from their original location and cause a straightening or curling of the superficial retinal vessels. ¹ Both retinal thickness and retinal distortion may increase over time because ERM is a dynamic condition. Until recently, there was no way of determining the degree of displacement of the superficial retinal vessels from their original position caused by an ERM. 

In 2010, Shiragami et al. ² reported that fundus autofluorescence (FAF) ^3,4 can show lines of increased autofluorescence similar in caliber and running parallel to retinal vessels in eyes operated on for retinal detachment (RD). These hyperautofluorescent lines would indicate the original location of the retinal vessels, which have been displaced together with the retina after retinal reattachment. We called these lines “retinal vessel printings” (RVPs) in a previous article. ⁵ The RVPs are also visible in some eyes of patients with ERM. ⁶  

Although RVPs associated with RD ^2,7 and RVPs associated with ERMs ⁶ may appear similar based on a cursory examination (both indicate the displacement of retinal vessels from their original position), they differ from each other in two fundamental aspects. First, for a given eye RVPs in the macular region associated with RD lie exclusively either above (more often, a sign of inferior retinal displacement) or below (rarely, a sign of superior retinal displacement) the retinal vessels they originate from (Fig. 1). Conversely, in the presence of macular ERM, provided that the traction focus is located between the arcades, RVPs running above and below the retinal vessels may be seen in the same eye (Fig. 1). This is because, in the case of RD involving the macula, the full-thickness macula (with its vessels) is separated from the RPE, so it moves en bloc and may reattach above or below its original location. Conversely, in the case of ERM, the retinal displacement almost exclusively involves the superficial layers that are dragged toward the focus of traction. In other words, in the case of ERM, the dislocation of the superficial layers (and their vessels) is not the result of a true retinal translocation, as happens in the case of RD, but is rather the result of superficial retinal dragging. The second aspect, which is a consequence of the first, is that after RD repair the distance between retinal vessels and corresponding RVPs remains mostly stable or changes slowly during the follow-up period. On the other hand, after ERM peeling the retinal vessels tend to progressively approximate and eventually cover up the corresponding RVPs (Fig. 2). This process occurs because the release of traction after ERM peeling may cause the displaced inner layers of the retina (and the dragged superficial vessels) to return to their original locations. Conversely, a retina displaced following RD surgery should be able to slide en bloc on the RPE to return to its original position. 

Figure 1

 

Differences between RVPs associated with RD and RVPs associated with ERM in fundus autofluorescence imaging. (A, B) In the macular region, RVPs (arrows) associated with macula-off RD are located exclusively above (as in the case presented here) or exclusively below the retinal vessels that they are related to, depending on the direction of retinal displacement (downward or upward, respectively). (C, D) In the case of ERM, the position of RVPs (arrows) relative to the corresponding retinal vessels depends on the location of the traction focus of the membrane. If the traction focus is located between the vascular arcades, RVPs running above and below retinal vessels may be seen in the same eye.

Figure 2

 

Follow-up observation of RVPs associated with RD and ERM. (A) Preoperative view showing RD involving the macula. No RVPs are visible at this stage. (B) At 1 month after operation, as a result of downward displacement of the retina, there are RVPs (arrows) located above the retinal vessels. (C) At 6 months after operation, the positions of RVPs (arrows) are unchanged. (D) Preoperative view showing an ERM exerting tangential traction with a consequent upward displacement of the inferior arcade. The corresponding RVP (arrows) is visible inferiorly below the retinal arcade. (E) At 1 month after operation, the RVP has approximated to the corresponding retinal vessel. (F) At 6 months after operation, the RVP is even closer to the corresponding retinal vessel.

At present, there are few data in the literature comparing morphologic and functional features of eyes affected by macular ERM with and without RVPs. ⁶ The presence of RVPs and their distance from retinal vessels could in theory provide information about the severity of the tangential traction caused by the membrane and might possibly relate to the severity of optical coherence tomography (OCT) features and some symptoms, especially metamorphopsia, associated with ERM. The objective of the present study was to investigate whether the presence of RVPs in eyes having idiopathic ERM is associated with a higher degree of metamorphopsia or with more prominent abnormalities in the retinal architecture compared with eyes not having RVPs. 

Methods

Study Participants and Clinical Examinations

The study was a cross-sectional case series assessing data from patients with idiopathic macular ERM. All patients were referred to the University of Molise, Campobasso, Italy, between September 7, 2011, and October 30, 2012. The study was approved by the institutional review board at the University of Molise and adhered to the tenets of the Declaration of Helsinki. Signed informed consent was obtained from all study participants. 

We included 18 eyes of 18 patients having ERM with RVPs and 18 eyes of 18 patients having ERM without RVPs on FAF imaging; this latter group served as the control subjects. The controls were chosen such that their central retinal thicknesses (CRTs), a sign of severity of anteroposterior traction, matched those of the eyes with RVPs to better compare the impact of tangential traction on metamorphopsia and OCT features. In fact, previous evidence has shown that retinal thickening itself causes deterioration not only in visual acuity (VA) but also in metamorphopsia. ⁸  

The center of the fovea (umbo) was determined in the OCT images by identifying the bulge-like structure of the photoreceptor inner segment/outer segment (IS/OS) junction line at the fovea. ⁹ Then, the central foveal thickness, defined as the distance between the inner retinal surface and the inner border of the RPE, and the CRT-1 mm at the fovea, defined as the average thickness within a circle of diameter of 1 mm centered on the umbo, were measured with the built-in scale of the OCT system. We included eyes with CRT-1 mm ranging from 350 to 600 μm. 

Excluded from the study were patients with a history of ophthalmic surgery (except for patients with noncomplicated phacoemulsification), patients with secondary causes of ERM (e.g., diabetic retinopathy, retinal vein occlusion, peripheral retinal tear, and uveitis), and patients with vitreomacular traction and lamellar holes. We also excluded eyes with reduced vision attributed to other retinal pathologie conditions such as glaucoma and age-related macular degeneration. Finally, because opacification of the lens may cause variations in the assessment of metamorphopsia, we excluded patients with visually significant cataracts. 

All patients underwent a comprehensive ophthalmic examination. Before dilating the pupil, each patient's best-corrected VA (BCVA), using Early Treatment Diabetic Retinopathy Study charts at 4 m, and metamorphopsia, using M-CHARTS (Inami Co., Tokyo, Japan), were assessed. 

Metamorphopsia Assessment

Introduced in 1999 by Shinoda et al., ¹⁰ M-CHARTS (Inami Co.) represent a simple and reliable method for quantifying the degree of metamorphopsia. Previous studies ^11,12 demonstrated that these charts are useful in measuring metamorphopsia in patients with ERM and that they correlated well with the severity of membrane proliferation and contraction. 

M-CHARTS (Inami Co.) consist of 19 dotted lines, with dot intervals ranging from 0.2 to 2.0° in visual angle. If a 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 increasing dot interval until the dotted line appears straight. 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 the 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 or her metamorphopsia score for vertical lines. Then, the 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 the data analyses. Each examination was performed at 30 cm, with the refraction of the eye exactly corrected for this distance. 

Average metamorphopsia scores were calculated by summing the horizontal and vertical scores and dividing the results by two. Because according to the literature ¹³ average metamorphopsia scores of less than 0.5 measured using M-CHARTS (Inami Co.) are usually not subjectively disturbing to the patients, we applied a cutoff at 0.6 in the logistic regression analysis model. 

Optical Coherence Tomography

The OCT and FAF images (excitation wavelength of 488 nm and barrier filter of 500 nm) and infrared pictures (50 and 35°) were acquired using an OCT system (Spectralis HRA+OCT; Heidelberg Engineering, Heidelberg, Germany) after pupil dilation. This OCT system records up to 40,000 A-scans per second with an axial digital resolution of approximately 4 μm and a transversal digital resolution of approximately 15 μm in the high-resolution mode. The patients were asked to focus on a target, and line scans were performed. The OCT recording protocol consisted of a sequence of 97 horizontal sections, covering an area of 30° horizontally and 20° vertically recorded in the high-resolution mode (1536 and 1024 A-scans, respectively), with a distance of approximately 60 μm between individual sections. 

To evaluate the irregularity (disruption or loss) of the external limiting membrane (ELM) and the IS/OS line at the fovea, we analyzed the central, 1-mm part of five horizontal scans (the central scan centered on the umbo and the two scans above and below the central scan, respectively) (Fig. 3). A score ranging from 0 to 5 (0 indicates no scan showing irregularity of the lines, and 5 indicates all scans showing irregularity of the lines) was assigned to each analyzed area to quantify the degree of irregularity of the ELM and IS/OS lines. 

Figure 3

 

Protocol used to evaluate the irregularity of the ELM and the IS/OS lines at the fovea. The central, 1-mm portion of five horizontal scans (the central scan centered on the umbo and the two scans above and below the central scan), spaced 60 μm from each other, was assessed.

Two graders (RdO and MF) assessed the status of the ELM and IS/OS lines. Both graders were unaware of the FAF imaging findings in the patients and their VA and metamorphopsia scores. When there was disagreement between the two graders, a third grader (CC) decided which grader's judgment should be accepted. 

FAF Imaging

After acquisition, the FAF images were divided for analysis using a reference grid. The reference grid was made according to the following technique. First, a layer of 992 × 992 pixels was created. Then, horizontal and vertical lines were positioned to subdivide the layer into 100 squares (each measuring 9.92 pixels squared). The FAF images (image resolution of 496 pixels squared) were imported into Adobe Photoshop (CS5 version 12.0; Adobe Systems Incorporated, San Jose, CA), and the grid was superimposed on the images. The grid was centered on the fovea (whose position was detected on the corresponding OCT image, as already described), and the central 15 squares (with the fovea centered on square number eight) were used to calculate the displacement of the retinal vessels (Fig. 4). 

Figure 4

 

(A) The grid used to calculate the areas of displacement. It was centered on the fovea, and 15 squares (fovea centered on square number eight) were used to calculate the displacement of the retinal vessels. (B) The areas between the displaced retinal vessels and the corresponding RVPs were manually drawn in each of the 15 squares. The value of each area was automatically calculated by the software. A total score was obtained for each eye by summing the values of the areas of displacement found in each of the 15 squares. For representative purposes, the areas of displacement in square 12 are highlighted in yellow.

To measure the displacement, the images were imported into freeware image manipulation software (ImageJ, version 46r; provided in the public domain by Wayne Rasband, National Institutes of Health [http://imageJ.nih.gov/ij]). In these images, 10 pixels correspond to 0.2 mm. Using the “polygon selection” function, we manually drew the areas between the displaced retinal vessels and the corresponding RVPs in each of the 15 squares. The value of each area was automatically calculated by the software. A total score was obtained for each eye by summing the values of the areas of displacement found in each of the 15 squares (Fig. 4). 

Statistical Analysis

The mean scores were compared, and SDs were calculated for each parameter of visual function and the OCT measurements. Mann-Whitney U test and Fisher exact test were performed to compare clinical and OCT parameters between the two groups. Spearman rank correlation test was used to evaluate the correlation between the area of retinal displacement and the metamorphopsia scores. Logistic regression analysis was performed to determine whether there was a significant association between the independent variables (logarithm of the minimum angle of resolution [logMAR] BCVA, lens status, RVPs, CRT-1 mm, ELM, and IS/OS line irregularity) and the dependent variable (average metamorphopsia score). P < 0.05 was considered statistically significant. The analyses were performed using MedCalc version 11.5.1 (MedCalc software bvba, Mariakerke, Belgium). 

Results

The visual function and clinical and OCT parameters of the patients are given in Table 1. Both groups were comparable in terms of age, sex, and logMAR BCVA. 

Table 1

 

Clinical Parameters, Visual Function, Metamorphopsia Scores, and OCT Characteristics of the Sample*

Variable	ERM With RVPs	ERM Without RVPs	P Value
Eyes, n	18	18	…
Ratio of men to women	11:7	10:8	1.0†
Age, y	68 (5.9) [59–78]	72 (7.4) [55–83]	0.061
Ratio of right eyes to left eyes	10:8	7:11	0.5†
Ratio of phakic to IOL	12:6	13:5	1.0†
logMAR BCVA	0.33 (0.20)	0.23 (0.19)	0.165
Metamorphopsia score
Vertical	0. 71 (0.45)	0.62 (0.44)	0.484
Horizontal	1.01 (0.58)	0.88 (0.67)	0.391
Average	0.86 (0.45)	0.75 (0.52)	0.303
Eyes with average metamorphopsia score ≤0.6, n	2	9	0.027†
Central foveal thickness, μm	473.50 (79.40) [352–615]	460.94 (102.18) [308–613]	0.937
CRT-1 mm, μm	481.17 (67.54) [377–590]	478.50 (72.17) [354–594]	0.937
ELM line irregularity score	2.61 (1.75)	1.44 (1.09)	0.042
IS/OS line irregularity score	2. 44 (1.88)	0.94 (1.21)	0.014
Ratio of presence to absence of intraretinal cysts	4:14	4:14	1.0*
Ratio of presence to absence of foveal detachment	3:15	2:16	1.0*

*  Values are expressed as the mean (SD) [range] unless otherwise indicated. Boldface indicates statistical significance.

†  Fisher exact test. Other P values are by Mann-Whitney U test.

Vertical, horizontal, and average metamorphopsia scores did not differ between the two groups. However, among the patients in the group with RVPs, only two had an average metamorphopsia score of 0.6 or less versus nine among the patients in the group without RVPs. This difference was statistically significant (P = 0.029). 

The mean (SD) area of displacement calculated on the FAF images in the group with RVPs was 1.73 (1.22) mm² (range, 0.38–4.37 mm²). There was no correlation between the area of displacement and the vertical (P = 0.44), horizontal (P = 0.8), or average (P = 0.38) metamorphopsia scores. 

Regarding the OCT parameters, there were two eyes that did not show irregularity in the ELM line and four eyes that did not show irregularity in the IS/OS line in the group with RVPs (P = 1.0). In the group without RVPs, there were three eyes that did not show irregularity in the ELM line and seven eyes that did not show irregularity in the IS/OS line (P = 0.11). 

The eyes with RVPs manifested a higher degree of irregularity of the ELM and the photoreceptor IS/OS lines versus the control group (Fig. 5), with a mean (SD) score of 2.61 (1.75) vs. 1.44 (1.09) for the ELM line and a mean (SD) score of 2.44 (1.88) vs. 0.94 (1.21) for the IS/OS line; the differences were statistically significant (P = 0.042 and P = 0.014, respectively). For the groups with and without RVPs, VA was not significantly different in eyes with and without irregularities in the IS/OS line (P = 0.87 and P = 0.40, respectively); similarly, the presence of irregularity in the ELM did not affect VA in either group (P = 0.47 and P = 0.76, respectively). 

Figure 5

 

Representative scans showing the relationship between the presence of RVPs on FAF imaging and the irregularity of the ELM and the IS/OS junction of photoreceptor lines at the fovea visualized in OCT scans of eyes with epiretinal membranes. (A) No RVPs are visible on FAF imaging. (B) No irregularities of the ELM and IS/OS lines are detected in the OCT scan. (C) The RVP (arrow) at the inferior aspect of the macula. (D) The ELM appears intact, but there is an interruption of the IS/OS line at the fovea (arrow). (E) The RVP in the FAF image (arrow) associated with the disruption of both the ELM and the IS/OS line at the fovea in OCT scans ([F], arrow).

Central foveal thickness, CRT-1 mm, and the number of eyes with intraretinal foveal cysts and foveal detachment did not differ between the two groups. A correlation between decreased VA and increased CRT-1 mm was found only in eyes having ERMs not associated with RVPs (P = 0.0021). No correlation was found in either group between VA and disruption scores of the ELM and IS/OS lines, VA, and vertical, horizontal, and average metamorphopsia scores. 

Logistic regression analysis was used to investigate the possible variables related to average metamorphopsia scores of 0.6 or higher. The following variables were included in the model: logMAR BCVA, lens status, RVPs, CRT-1 mm, and irregularity of the ELM and IS/OS lines. The only independent variable associated with an average metamorphopsia score of 0.6 or higher was the presence of RVPs (Table 2). 

Table 2

 

Multiple Logistic Regression Model of Variables Associated With Metamorphopsia in Patients With Idiopathic ERM*

Variable	Odds Ratio (95% Confidence Interval)	P Value
logMAR BCVA	0.0384 (0.0001–19.4934)	0.3050
Lens status	0.2046 (0.0104–4.0144)	0.2961
RVPs	0.0570 (0.0035–0.9334)	0.0446
CRT-1 mm, μm	0.9829 (0.9597–1.0067)	0.1574
ELM line irregularity	0.7642 (0.2779–2.1014)	0.6023
IS/OS line irregularity	1.5893 (0.6235–4.0513)	0.3319

*  Zero indicates an average metamorphopsia score of 0.6 or higher; 1 indicates an average metamorphopsia score of 0.5 or less. Boldface indicates statistical significance.

Discussion

Although modern OCT scanners and built-in software permit the study of the retina in great detail and the comparison of changing features based on follow-up observation, FAF imaging offers a unique opportunity to visualize the displacement of superficial retinal vessels that occurs in eyes with ERM. The RVPs are the mark of such displacement, and their visualization in eyes with ERM may provide immediate and easily interpretable information about the severity and direction of tangential traction. 

In this study, we used FAF imaging to investigate the differences in terms of the degree of metamorphopsia and architectural retinal changes in eyes having ERM with and without RVPs. We also explored the relationship between the amount of tangential displacement revealed by FAF imaging and metamorphopsia scores. 

We found that 16 of 18 patients with RVPs had an average metamorphopsia score of 0.6 or higher versus only nine of 18 patients in the control group. Furthermore, the presence of RVPs was associated with a higher degree of irregularity of the ELM and IS/OS lines at the fovea. 

To date, few studies have attempted to assess the degree of retinal contraction and to quantitatively relate the degree of retinal contraction with the degree of metamorphopsia in eyes with idiopathic ERM. Arimura et al. ¹² investigated the relationship between retinal contraction using confocal scanning laser ophthalmoscope (cSLO) images obtained with an argon blue laser beam and metamorphopsia using M-CHARTS (Inami Co.) over a period of 3 years. The authors found that the vertical metamorphopsia score correlated with the degree of horizontal contraction, whereas the horizontal metamorphopsia score correlated with the degree of vertical contraction. 

Kofod and la Cour ¹⁴ quantified the amount of displacement of superficial vessels during follow-up observation. They found that the retinal tangential movements were significantly greater in patients who had subjectively reported worsening of metamorphopsia compared with patients who had unchanged metamorphopsia. 

The results of the present study confirm, but only in part, these previous observations. In line with previous studies, we found that the presence of RVPs, a sign of tangential traction, is normally associated with severe metamorphopsia. 

In fact, we found that most eyes with RVPs had an average metamorphopsia score of 0.6 or higher. Furthermore, among the variables with the potential to influence the severity of metamorphopsia (i.e., VA, lens status, presence of RVPs, foveal thickness, and irregularity of the ELM and IS/OS lines at the fovea), the presence of RVPs was the only factor associated with an average metamorphopsia score of 0.6 or higher. 

However, when we compared the metamorphopsia scores (vertical, horizontal, and average) between the two groups, we found no significant differences. Furthermore, in eyes with RVPs, the severity of metamorphopsia did not correlate with the extent of the areas of retinal displacement. So, the expected relationship between the severity of the retinal tangential distortion/contraction and the severity of metamorphopsia was only in part confirmed. Several reasons may explain our results. First, previous studies ^12,14 focused on the changes in retinal contraction (and the corresponding metamorphopsia) followed up over time in the same eyes, whereas in the present study we pooled together and compared data collected from different patients at a single point in the evolution of the disease. Because metamorphopsia severity depends on multiple factors, including complex processes occurring at the level of the visual cortex, it is probably more likely to identify an association between increasing retinal distortion and increasing metamorphopsia in the same patient followed up over time rather than between a certain amount of retinal displacement and a certain degree of metamorphopsia in different patients. This is true primarily because different patients may have different perceptions of metamorphopsia. ¹⁵  

Second, we used a different method to evaluate the retinal displacement. In previous studies, ^12,14 the progressive movement of the retinal vessels was calculated on the basis of measurements taken at specific points of the vessels, by recording the horizontal and vertical components of the vectors. Conversely, in our model we examined the dislocation of every single point of the vessels by calculating the total area of the displacement (i.e., the area between the dislocated retinal vessels and the corresponding RVPs). We thought that such an approach would be more accurate and lead to more reliable results because we consider the tangential displacement along the entire course of the vessels. In fact, the displacement of the vessels may be quite complex, with some portions being stretched or shifted more than others. ⁵ As a consequence, measurements taken at single points may not be sufficiently representative of the global impact of tangential traction on a certain area. 

Third, RVPs might be more evident in some eyes than in others (depending on eye characteristics and/or on the equipment used to record the images). This may have introduced a bias in the selection of the patients for each group. 

An interesting aspect of this study is that, although the two groups did not differ with regard to macular thickness and the presence of intraretinal cysts or foveal detachment, they differed with regard to the degree of irregularity of the ELM and IS/OS lines. The higher degree of irregularity found in the eyes with RVPs could be secondary to a more pronounced cellular damage/loss or to a more conspicuous misalignment of the retinal layers (because of tangential traction) compared with eyes without RVPs. In fact, the reflectivity of the retinal structures depends, among other factors, on the angular incidence of the OCT beam. If the ELM and the IS/OS lines are not perpendicular to the direction of the incoming OCT beam (because of distortion caused by an ERM), this situation might produce weaker backscattering that can be interpreted as an irregularity of the lines. 

This explains why some data reported in the literature regarding the relationship between the irregularity of the retinal lines and VA and metamorphopsia in eyes with ERM may appear contradictory. In fact, it has been reported that IS/OS disruption may or may not be associated with VA loss ^16–23 and that foveal thickening may or may not correlate with VA. ^16–18  

In the present work, VA was not significantly different between eyes with and without RVPs, and we found a significant correlation between poor VA and increased foveal thickness only in the group without RVPs. Furthermore, we observed no difference in VA between eyes with irregular IS/OS and eyes with intact IS/OS in either group. In addition, in line with previous studies, ^8,13,23 we did not find a correlation between the irregularity score of the ELM and IS/OS lines and metamorphopsia scores. 

The pathogenesis of RVPs and the reason why RVPs are detectable only in some eyes affected by ERMs ⁶ remain unknown. Shiragami et al. ² and Nitta et al. ⁶ interpreted the lines as the result of increased metabolic activity from portions of RPE previously shaded by the retinal vessels and then acutely exposed to light irradiation because of retinal vessel displacement. However, there is no scientific evidence proving that exposure to light is able to cause an increased autofluorescence signal in the short term. We detected RVPs in eyes tamponaded with silicone oil as soon as 5 days after operation (dell'Omo R, dell'Omo E, unpublished data, 2012), and Shiragami et al. ² observed RVPs 10 days after operation. If an increased metabolic activity induced dramatic changes in the previously shielded RPE–photoreceptor cells and this caused increased autofluorescence, one would expect that these acute changes would suddenly manifest and then soon disappear. In contrast, follow-up observation shows that RVPs are still visible months and even years after the displacement has occurred. ²  

The persistence of RVPs suggests that profound differences in terms of composition and characteristics of fluorophores (responsible for the autofluorescence signal) exist between the RPE cells belonging to RVPs and the neighboring cells. These characteristics may include the presence of melanin and lipofuscin granules with different fluorescent spectral properties compared with the cells exposed to light. Indeed, A2E, the major blue-absorbing fluorophore, can undergo a variety of changes, including some induced by light such as photoisomerization and oxidation; this process can produce various A2E derivatives with different fluorescent spectral properties. ²⁴ In addition, other by-products of the visual cycle could account for the observed differences. Therefore, it is possible that RVPs are a sign of these long-standing differences and are not induced but simply unveiled by the displacement. In other words, RVPs could be present under normal circumstances but not visible because they are masked by the overlying retinal vessels. Under normal circumstances, the retinal vessels often show a parallel narrow band of higher autofluorescence (lining only one side of the retinal vessels). These bands have been interpreted in the past as a refraction effect at the vessel walls, ⁴ but if our hypothesis is true, they could actually represent a clue to the existence of RVPs (before a displacement occurs). In fact, it is conceivable that, depending on the axial orientation of the laser scanner (when using a cSLO) and on the position of the eye when images are recorded, some retinal vessels may not superimpose perfectly on the corresponding RVPs, rendering narrow parts of RVPs visible. 

This hypothesis could also explain why RVPs are visible only in some eyes with ERM. If the displacement progresses very slowly, a substantial change in the aforementioned characteristics could occur over time in the previously shielded RPE cells, making them similar to the adjacent cells that are exposed to light and therefore leaving no clue of retinal vessel displacement on FAF imaging. Conversely, the presence of RVPs might indicate that strong tangential traction forces have caused significant displacements in the short term. 

The technical characteristics of the device used to record the FAF signal (the fundus camera or the cSLO) might also affect the sensitivity for detection of RVPs. The cSLO system is based on excitation of a blue laser beam (λ = 488 nm) with emission detected using a cutoff filter (λ = 500 nm), whereas for the modified fundus camera two bandwidth filters are applied operating in a longer wavelength range (excitation wavelength of 535–580 nm and barrier filter of 615–715 nm). 

The FAF signal is made up of several fluorophores with different absorption and emission spectra, as well as different amounts of fluorescence intensity. Therefore, devices using different filters may be more or less sensitive to detect the fluorescence originating from RVPs. 

Fundus camera systems are inherently more affected by scattering than are cSLO methods of imaging. On the other hand, the gradations of grayscale values in fundus camera images appear more smooth and capable of rendering subtle changes than a commercially available cSLO. Further studies are warranted to clarify the relevance, if any, of these technical differences to the ability to detect RVPs. 

The limitations of this study include a small sample size, measurements based on only five horizontal central sections, imaging inaccuracy due to manual segmentation of the areas of displacement, and a subjective evaluation of the degree of irregularity of the outer retinal lines on OCT. However, to the best of our knowledge, this study is the first to compare eyes with macular ERM with versus without RVPs. Our study is also the first to date to investigate the relationship between the amount of retinal displacement caused by ERM on the basis of FAF imaging and quantitatively measured metamorphopsia. Finally, an alternative hypothesis for the pathogenesis of RVPs is proposed. 

In conclusion, our results suggest that the presence of RVPs on FAF images recorded using cSLO in eyes with idiopathic macular ERMs is usually associated with an average metamorphopsia score of 0.6 or higher using M-CHARTS (Inami Co.) and with a higher degree of irregularity of the ELM and IS/OS lines at the fovea compared with eyes without RVPs. Further prospective studies are warranted to investigate if RVPs are a useful prognostic factor for estimating the speed of progression of architectural changes in the outer retina and metamorphopsia scores in eyes with ERM. 

Acknowledgments

The authors thank Giovanni Staurenghi and Andrea Giani for sharing their knowledge and expertise of the subjects discussed. 

A portion of this work has been published previously as an ARVO abstract: dell'Omo R, et al. IOVS 2013: ARVO E-Abstract 3313. 

Disclosure: R. dell'Omo, None; F. Cifariello, None; E. dell'Omo, None; A. De Lena, None; R. Di Iorio, None; M. Filippelli, None; C. Costagliola, None