Quantification of Metamorphopsia in Patients with Macular Hole

Abstract

purpose. To describe a novel method for the evaluation of metamorphopsia within the central 10° of the visual field in 55 patients with idiopathic macular hole. The test evaluates the interocular disparity between the two eyes in metamorphopsia.

methods. Semicircular test and reference stimuli of variable diameters were used in a binocular test that measured interocular size disparity between perceptually iseikonic stimuli in subjects with a unilateral macular hole. A group of 11 healthy subjects was used as the reference.

results. In 55 patients with a macular hole, interocular disparity demonstrated a mean value of 0.71° with stimuli in the range 1.0° to 2.5° in diameter. This number declined to 0.41° with stimuli in the range of 9.0° to 10.0° in diameter. Both hole diameter and eccentricity had a significant effect on mean disparity (P < 0.001).

conclusions. The level of metamorphopsia declined as a function of eccentricity and affected the central 10° of visual field. Macular hole size had an independent effect on interocular disparity. These results confirm reports that visuospatial distortion in the presence of macular hole is primarily the result of radial displacement of photoreceptors. (ClinicalTrials.gov number, NCT00302328.)

Seeing distorted faces and broken lines, metamorphopsia, is a common complaint in patients with retinal diseases such as age-related macular degeneration, central serous chorioretinopathy, vitreomacular traction, reattached retinal detachment, epiretinal fibrosis, ^{1 2 3} and macular hole. ⁴ In these diseases, the essential pathogenic event is likely to be either cell death or merely dislocation of photoreceptors that change their projection in physical space, or local sign. ⁵ In addition, metamorphopsia can be caused by anterior segment irregularities that cause distortion in visual perception secondary to changes in the optical signal itself before it reaches the retina. Finally, damage to the optic pathways or the visual cortex itself can also result in metamorphopsia. ^{6 7 8}  

The formation of a macular hole is associated with crowding of photoreceptors and other retinal structures around the hole and possibly also some degree of eversion of the photoreceptor layer in the vicinity of the hole. It is unknown how far this abnormality extends into the surrounding retina. The prominent central metamorphopsia in patients with macular hole suggests that it may be confined to the macula. 

Since it was first described in 1953, the Amsler grid has been used as a simple test for detection of metamorphopsia, but it is less useful for providing an accurate measure of the degree of metamorphopsia. ⁹ Objective measurement of metamorphopsia in eyes with a macular hole has been limited to mapping a fixed area of the visual field. ^{10 11} The new method presented herein provides measurements of metamorphopsia within the central 10° of visual field (5° eccentricity). The method relies on the finding that the displacement of photoreceptors around a macular hole is, presumably, more or less uniformly radial, following a meridional direction away from the center of the fovea and being of roughly equal magnitude in all directions. ^{10 12} It follows from this mechanism, that a circular or semicircular stimulus has the ideal characteristic of covering all directions simultaneously and reducing the description of the visual disorder to a simple linear function. In the present study, we applied semicircular test and reference stimuli of various diameters in a binocular test that measured the spatial discrepancy between perceptually iseikonic stimuli in subjects with a unilateral macular hole and normal binocular vision before the onset of this condition. 

Methods

Subjects

Study participants were recruited from the Copenhagen Macular Hole (COMAH) study, a randomized controlled clinical trial comparing different methods of surgical treatment, and from a pool of healthy volunteers. Inclusion criteria were idiopathic macular hole of Gass stage 2 or 3, duration of symptoms less than 12 months, and best-corrected visual acuity (BCVA) better than or equal to 20/200 (1.0 logMAR on the Early Treatment of Diabetic Retinopathy Study [EDTRS] chart). Acuity measures were recorded as logMAR and transposed to Snellen notation. Exclusion criteria were previous macular hole in either eye or macular hole in both eyes, epiretinal fibrosis, prior intraocular surgery except cataract extraction, any disease affecting or potentially affecting retinal function including diabetic retinopathy, a history of glaucoma or glaucoma diagnosed at the screening visit, anisometropia greater than 3 D, amblyopia, nystagmus, and strabismus. 

All subjects underwent a comprehensive ophthalmic examination including, manual undilated refraction, assessment of BCVA at a 4-m distance, slit lamp biomicroscopy, mydriatic funduscopy, fundus photography, optical coherence tomography (Stratus OCT3; Carl Zeiss Meditec, Dublin, CA). All patients who were not pseudophakic at the time of inclusion underwent phacoemulsification with pseudophakic lens implantation one month before baseline. The minimum inner diameter of the hole (minimum diameter) and the hole diameter at the level of the retinal pigment epithelium (maximum diameter) were measured with the built-in calipers of the system software. ¹³  

The study included 55 patients with a unilateral stage 2 (n = 19) or stage 3 (n = 36) macular hole and a reference population (not age-matched) of 11 healthy subjects with normal binocular vision. No subject had a refractive error greater than ±5 D or anisometropia greater than 3 D. All healthy subjects had BCVA 20/20 or better in both eyes, no metamorphopsia, and normal ophthalmoscopy. 

All participants gave their written informed consent before inclusion. The study was approved by the ethics committee of Copenhagen County (KA04144) and was conducted according to the Declaration of Helsinki. 

Quantification of Metamorphopsia

The metamorphopsia test was conducted with the patient seated, facing a computer screen (cathode ray tube monitor, SyncMaster 959nf; Samsung, Seoul, South Korea; 118-Hz frame rate, 17-in., 1024 horizontal, 768 vertical pixels) at 50-cm distance with the head resting on a chin and forehead support. 

Test spectacle frames mounted with appropriate lenses were used to correct for refraction and distance. In addition, a green filter was mounted in front of the test eye and a red filter in front of the fellow eye, to enable dichoptic stimulation using red and green test stimuli and binocular stimuli using black. The intersection of two black lines, one horizontal and one vertical, at the center of the screen served as the patient’s fixation target throughout the procedure. This task is manageable for patients with central scotomas as the components of the fixation cross fall on the peripheral retina. Test stimuli were introduced in the form of two high-luminance, high-contrast dichoptically presented half-disks of different colors (red and green; Fig. 1 ). The room illumination was kept dim enough to provide desirable stimulus contrast on the video display. When objectively equally sized, they would form a single round disk, divided vertically into two halves of equal size but different color. A patient with unilateral macular hole, however, sees two half-disks of uneven size, if the rim of the half-disk seen by the diseased eye falls within the region of the visual field affected by the metamorphopsia. In a patient with micropsia, the test stimulus must be larger than the reference stimulus. The patient is then presented a series of combinations of unevenly and evenly sized half-disks and prompted to answer whether the disk are of equal size or not (two-alternative, forced-choice paradigm). 

The test determines the eccentricity necessary for a test stimulus (presented to the diseased eye) to match perceptually the reference stimulus (presented to the healthy eye). This difference in eccentricity is referred to as interocular disparity and will be used synonymously with metamorphopsia throughout this article, although, strictly speaking, metamorphopsia can be both a monocular and binocular phenomenon. Interocular disparity is taken as a measure of the angular displacement of receptors that were originally placed closer to the center of the foveola. The reference eccentricity is defined as the angular distance from the point of fixation to the rim of the reference stimulus. The maximum interocular disparity expresses the metamorphopsia function. 

Reference stimuli ranged in diameter from 1° to 10° in increments of 0.5°. One degree of visual angle is equal to 288 μm on the retina in the emmetropic eye. ¹⁴ The test stimuli ranged from being identical with the reference to an extra width of up to 1.5°. The total number of combinations was 78 pairs of red and green semicircular disks, which were presented in random order at two trials, with a short break between each trial. The viewing time for each stimulus pair was approximately 1 second. 

Reproducibility was assessed by examining patients twice at baseline and using the Mann–Whitney rank sum test for repeated measurements. Disparity data were ranked in five groups of reference eccentricity ranges each containing the mean value of the individual reference eccentricities within that range. Nonparametric statistics were applied where key outcomes were found not to demonstrate a normal distribution. The five groups of reference eccentricities were ranked in three classes according to minimum and maximum macular hole diameter. Two-way analysis of variance was used to assess the effect on interocular disparity of reference eccentricity and hole dimensions. Comparisons between hole dimensions, visual acuity, and maximum interocular disparity were made using the Spearman’s rank correlation analysis 

Results

The study comprised 55 patients with a unilateral macular hole and 11 healthy subjects (Table 1) . The measure of interocular disparity did not show any systematic or statistically significant difference between first and second baseline examination. Consequently, the average of first and second examination was used for analysis. 

All healthy subjects were able to see the entire range of reference and test stimuli. Interocular disparity between a subject’s two eyes ranged from −0.25° to 0.50° of visual angle in healthy subjects within all reference eccentricities. Median disparity was 0.0° for all but one stimulus size (Fig. 2) . The variation displayed an insignificant increase with eccentricity. 

In the patient group, 48 reference stimuli, ranging in size from 1.0° (n = 27) to 3.0° (n = 1), could not be matched by a test stimulus due to the effect of a central scotoma. Interocular disparity demonstrated a mean value of 0.71° in the 1.0° to 2.5° reference stimuli range declining to 0.58° in the 3.0° to 4.5° range. The interocular disparity function demonstrated a general effect of eccentricity (P < 0.001), the highest degrees of disparity being found closest to the rim of the central scotoma and reaching a plateau value of approximately 0.4° in the range from 6° to 10° of reference stimulus eccentricity (Fig. 2) . 

Stratification of the disparity function versus eccentricity function by the greatest diameter of the macular hole (Fig. 3) , found at its bottom, and by its smallest diameter (Fig. 4) , usually found near the top of the hole, demonstrated that higher degrees of disparity were associated with larger holes (P < 0.001). The effect of different levels of hole diameter did not show statistically significant dependence on eccentricity (P = 0.466). 

Maximum interocular disparity did not correlate with visual acuity (n = 55, r = 0.147, P > 0.286; Fig. 5 ) but correlated significantly with minimum (n = 55, r = 0.478, P < 0.001) and maximum (n = 55, r = 0.414, P < 0.002; Fig. 6 ) hole diameter. However, a step-wise regression analysis showed that the prime determinant of maximum interocular disparity was minimum hole diameter (P < 0.001) and that maximum diameter did not significantly add to the ability to predict maximum disparity. Maximum interocular disparity was within the limits of the size of the minimum hole diameter in all but two patients. 

Discussion

We mapped retinal interocular disparity associated with a macular hole as a function of visual field eccentricity, using a simple stimulus designed specifically to suit the visual distortion induced by a macular hole. In addition to confirming previous work that demonstrated radial photoreceptor displacement away from the center of the retina in eyes with a macular hole, ^{10 12} this study adds to the understanding of the nature of the visual distortion. It demonstrates that interocular size disparity is not constant over the visual field but decreases with the distance from the macular hole and that evidence of displacement can be found to at least 5° eccentricity, far beyond the limits of the hole itself (Fig. 7) . 

Our findings may explain why patients describe a pincushion distortion (i.e., an image deformation that is most prominent around the point of fixation and gradually decreases toward the periphery). ^{15 16 17} Our results are supported by the detailed psychophysical observations of Burke, ¹⁶ a psychophysicist who had metamorphopsia due to a macular hole and who found that a thin annulus with an inner diameter of 1.5° appeared as a small circle of approximately 0.6° diameter with a hole, when seen by an eye with a macular hole. This corresponds to the 0.75° median interocular size disparity for a 1.5° reference stimulus found in our study. Burke also found gradually decreasing disparity for larger size stimuli. Conversely, he found that in his own case of macular hole, stimuli greater than 4.0° were not distorted. Even though all our patients demonstrated interocular disparity beyond 4°, one should bear in mind that our study presents data from 55 patients with macular holes of various diameters. 

Our finding that metamorphopsia could be detected to at least 5° eccentricity supports the idea that photoreceptors are displaced laterally. Tangential movement of photoreceptors has been suggested by Gass ^{18 19} and others, ^{10 12} and it is believed that macular hole is a result of a dehiscence and subsequent radial displacement of photoreceptors. This notion leads us to suggest that tangential traction imposed on the retina causes movements of photoreceptors that follow a continuum from radial displacement of foveal cells to merely angular misalignment of more peripherally located cells. This theory could explain the gradual tapering off of metamorphopsia found in our study. 

Our data show that interocular disparity correlated with both the inner (minimal) and the outer (maximal) hole diameters of a macular hole, meaning that larger holes gave rise to the largest degree of metamorphopsia. From this perspective it should be possible to predict the magnitude of distortion by determining, from optical coherence tomography imaging, the photoreceptor border along the margin of the hole. The OCT instrument used in this study did not enable delineation of this tissue layer, and it is methodologically problematic, because a change in orientation of the photoreceptors is likely to change their reflectivity, as seen from the pupil. On the other hand, the minimum diameter of the hole should set the maximum extent of photoreceptor movement and hence should also set the upper limit for visuospatial distortion, as is generally confirmed by our findings (Fig. 6) . 

In a study using M-CHARTS (Inami Co., Tokyo, Japan), which are based on hyperacuity or vernier acuity, Arimura et al. ¹¹ did not find a correlation between hole size and metamorphopsia measured using the M-CHARTS. However, there was a significant correlation between metamorphopsia scores and fluid cuff size if the macular hole size was <0.5 mm. It is likely that the measurements with M-CHARTS, which assess metamorphopsia in a fixed area of visual field, are affected by the central scotoma in larger size holes. There are a couple of methodological problems intrinsic to this type of test that also apply to the Amsler Grid. There is no restriction on display time, and as a result, the individual may scan the image, thereby facilitating the filling-in process. In addition, both tests present bottom-up information in a way that can be anticipated by the patient and therefore will be affected by a top-down expectation. 

Besides the apparent correlation to hole morphology, perceptual filling-in may be involved in visuospatial distortion. A common example of this phenomenon is the filling-in of the receptor-free area of the retina (physiologic blind spot) with part of the image from the surrounding area. ²⁰ The perceptual phenomenon of filling-in has been observed with both real and artificial scotomas, ^{17 21 22 23 24 25} and the perceived image may, among other image properties, reflect the texture and color of the area surrounding the scotoma. ^{21 25}  

Gilbert and Wiesel ²² demonstrated that scotomas induced by localized binocular retinal lesions result in expansion of the receptive field of cortical neurons corresponding to the rim of the resulting scotomas in cats and monkeys. In humans, perceptual completion of paracentral scotomas has been reported after ischemic brain insults, ⁷ and recently this observation was supported by objective signs of cortical reorganization in a study using functional magnetic resonance imaging (fMRI). ⁸ Burke ¹⁶ concluded that the degree of metamorphopsia in macular hole could not be explained by photoreceptor displacement alone, but should also be attributed to cortical filling-in. It has been described that objects positioned near a scotoma are perceptually deflected toward the scotoma and that distortion decreases with the distance of the stimulus from the scotoma, ^{8 16} the latter being confirmed in our study. Filling-in across a macular scotoma has also been described when viewing an Amsler grid—for some patients the perception within the retinal area of the scotoma being that of a distorted Amsler grid. ⁹  

In patients with macular hole, part of an external object is projected on a retinal area devoid of photoreceptors, and consequently part of that object is not detected. Any object falling on the rim of the macular hole is distorted, because it is detected by photoreceptors that have changed their physical projection in space. Apparently, the missing part of the external object is perceptually filled in so that rather than detecting a gap in the external world, the patient perceives a continuous world, although partly distorted. It seems to hold that the phenomenon of perceptual filling-in coexists with metamorphopsia as a symptom. 

It could be hypothesized that the degree of metamorphopsia may reflect the amount of retinal damage present before surgery and thus would be linked to visual acuity. However, metamorphopsia function showed no correlation to preoperative visual acuity, and Wittich et al. ^{17 26} have demonstrated this finding after surgery. These results are interesting, as interocular disparity may be used independently in assessing visual function. Conversely, this finding is challenged by a study ¹² based on the technique described by Jensen and Larsen, stating that mean preoperative and postoperative visual acuities were significantly better in patients with photoreceptor displacement. In the latter study, however, the failure to detect photoreceptor displacement could reflect a larger central scotoma, thus accounting for the different finding. Further studies will test whether this newly developed disparity test can be used to predict the anatomic and functional outcome of vitrectomy for macular hole. 

One weakness in our direct-comparison test is that, given enough time, the patient may roam around the two images to scan for contour differences. On the other hand, a certain amount of time is necessary for elderly patients to make a judgment about the stimuli. Patients were encouraged to make the semicircular stimuli fuse to the shape of a circle, and viewing time was limited to approximately 1 second, in that it is well known that saccadic eye movements occur at ∼120 ms. If scanning of the test images and fixation instability have affected the test, interocular disparity would presumably be underestimated, because the patient would tend to look for contour differences and thus be able to determine the objective relative size of the stimulus pair. Although, some participants had symptoms that could be ascribed to binocular rivalry, no person reported that only one stimulus was visible for a prolonged period, which suggests that ocular dominance did not severely affect the test. Another shortcoming of the test is the requirement of a fellow eye without metamorphopsia for comparison, but most of the patients with macular hole have this on one eye only. The physiologic limitation of Panum’s area limits the possibility to assess visuospatial distortion in more eccentric areas. 

Interocular disparity in macular hole can be mapped as a function of visual field eccentricity. Interocular disparity, which is taken as a measure of metamorphopsia, was not constant over the central 10° of visual field and extended beyond the area of clear photoreceptor–retinal pigment epithelial separation marked by the maximum hole diameter. Furthermore, interocular disparity was significantly dependent on both minimum and maximum hole diameter. Finding that maximum disparity generally did not exceed minimum hole diameter, we suggest that the visuospatial distortion in macular hole, is primarily the result of radial displacement of photoreceptors, although perceptual filling-in could also contribute to these findings. 

 Supported by the Danish Eye Health Society (Værn om Synet), the Danish Medical Research Council, Juvenile Diabetes Research Foundation Grant 8-2002-130, the John and Birthe Meyer Foundation, and the Carl and Nicoline Larsens Foundation.

 Submitted for publication November 12, 2007; revised March 23 and April 16, 2008; accepted June 20, 2008.

 Disclosure: K. Krøyer, None; U. Christensen, None; M. Larsen, None; M. la Cour, None

 The publication costs of this article were defrayed in part by page charge payment. This article must therefore be marked “advertisement” in accordance with 18 U.S.C. §1734 solely to indicate this fact.

 

Figure 1.

 

Left–right stimulus pair. The test used dichoptic stimulus pairs consisting of red and green semicircular disks. A black cross visible to both eyes was used for fixation. In this example, the left semicircular green disk (depicted in light gray) served as the reference stimulus and the right semicircular red disk (depicted in dark gray) served as the test stimulus. In this example, a normal observer perceived the right stimulus to be larger than the left stimulus. A patient with a macular hole, however, judged the two stimuli to be of equal diameter.

Table 1.

 

Baseline Characteristics of Patients and Control Subjects

	Patients	Controls
Age (y)
Range	54–78	29–60
Mean ± SD	67 ± 6	47 ± 11
Sex
Male	11	5
Female	44	6
Spherical equivalent in study eye
Range	−4.6 to −1.8	—
Mean ± SD	−1.0 ± 1.1	—
Lens status
Phakic	0	—
Pseudophakic	55	—
Visual acuity study eye (logMAR); (Snellen)
Range	1.0–0.4; 20/200–20/50	—
Mean ± SD	0.71 ± 0.13; 20/100 ± (1 line)	—
Visual acuity control eye (logMAR); (Snellen)
Range	0.3–−0.2; 20/40–20/12.5	—
Mean ± SD	0.02 ± 0.10; 20/20 ± (1 line)	—
Stage of macular hole
II	19	—
III	36	—
Minimum size of macular hole (μm)
Range	199–748	—
Mean ± SD	442 ± 125	—
Maximum size of macular hole (μm)
Range	486–1459	—
Mean ± SD	916 ± 216	—

Figure 2.

 

Median interocular disparity in 55 patients with a uniocular stage 2 or 3 macular hole (○) and in a reference population of 11 healthy control subjects (•). Disparity data of the patients have been grouped in five ranges of reference eccentricity in the bar graph (inset). Error bars, SD.

Figure 3.

 

Mean interocular disparity as a function of reference stimulus diameter for 55 patients with a uniocular stage 2 or 3 macular hole. Data have been divided into three groups according to maximum hole diameter and significant main effect of mean disparity. There is a significant effect on mean disparity of eccentricity (P < 0.001) and hole diameter (P < 0.001). There is not a statistically significant interaction between maximum hole diameter and eccentricity (P = 0.688). Error bars, SEM.

Figure 4.

 

Mean interocular disparity as a function of reference stimulus diameter for 55 patients with a uniocular stage 2 or 3 macular hole. Data have been divided into three groups according to minimum hole diameter and significant main effect of mean disparity. There is a significant effect on mean disparity of eccentricity (P < 0.001) and hole diameter (P < 0.001). There is not a statistically significant interaction between minimum hole diameter and eccentricity (P = 0.466). Error bars, SEM.

Figure 5.

 

Visual acuity as a function of maximum interocular disparity. Maximum level of interocular disparity showed no correlation to visual acuity (r = 0.147, P > 0.286), meaning that interocular disparity may be used independently in assessing visual function.

Figure 6.

 

Metamorphopsia in relation to hole diameter in eyes with macular hole. Correlation between maximum interocular disparity and diameter of the hole measured at its minimum diameter near the opening of the hole (•) and at its maximum diameter at the bottom of the hole (○). Converting macular hole minimum and maximum diameters to degrees of visual angle demonstrated that maximum interocular disparity was generally within the limits predicted by the smallest hole diameter. Maximum interocular disparity correlated significantly with minimum (r = 0.478, P < 0.001) and maximum (r = 0.414, P < 0.002) hole diameter. A step-wise regression analysis showed that the prime determinant of maximum interocular disparity was minimum hole diameter (P < 0.001) and that maximum diameter did not add significantly to the ability to predict maximum disparity. Inset: distribution of minimum (▪) and maximum (□) hole diameters. Solid line: line of equality.

Figure 7.

 

Fundus photograph of macular hole with superimposed 6° and 10° visual field (inner and outer white circles, respectively), corresponding to 3° and 5° reference eccentricity.