Effect of Amblyopia Treatment on Macular Thickness in Eyes With Myopic Anisometropic Amblyopia

Abstract

Purpose.: To determine whether abnormal macular thickness in myopic anisometropic amblyopia differed after amblyopia treatment. Furthermore, to investigate whether effect of treatment on macular thickness was associated with subject age or improvement in stereoacuity.

Methods.: Seventeen children (mean age: 9.0 [±3.0] years, ranging from 5.7–13.9 years) with myopic anisometropic amblyopia (visual acuity [VA] in amblyopic eyes: 20/80–20/400) were recruited and treated with 16-week refractive correction, followed by an additional 16-week refractive correction and patching. Macular thickness, best-corrected VA, and stereoacuity were measured both before and after amblyopia treatment. Factorial repeated-measures analysis of variance was performed to determine whether macular thickness in amblyopic eyes changed after amblyopia treatment.

Results.: Mean baseline VA in the amblyopic eye was 1.0 ± 0.3 logMAR and improved to 0.7 ± 0.3 after amblyopia treatment (P < 0.0001). The interaction between eye and amblyopia treatment was statistically significant for average foveal thickness (P = 0.040). There was no treatment effect on fellow eyes (P = 0.245); however, the average foveal thickness in the amblyopic eye was significantly reduced after amblyopia treatment (P = 0.049). No statistically significant interactions were found for the other macular thickness parameters (P > 0.05).

Conclusions.: Abnormal central macula associated with myopic anisometropic amblyopia tended to be thinner following amblyopia treatment with no significant changes in peripheral macular thickness.

Amblyopia is the most common reason for monocular visual impairment in children and young- and middle-aged adults.^1,2 Anisometropia as the most frequent cause of amblyopia has been evaluated in numerous studies.^3–5 However, anisometropia associated with high myopia was often excluded from those studies. Pang et al.⁶ were among the first to address high myopia and reported that both refractive correction and patching significantly improved the visual acuity (VA) of eyes with amblyopia associated with myopic anisometropia. Kutschke et al.⁷ reported that amblyopia due to myopic anisometropia had a poorer visual outcome with full-time patching than amblyopia caused by hyperopic anisometropia. Underlying structural abnormalities and macular hypoplasia have been proposed as causes of the poor treatment outcomes in myopic anisometropia.^8–10 Knowledge of the macular structure and associated factors in amblyopia associated with myopic anisometropia will further our understanding of this type of amblyopia and may assist clinicians in predicting and understanding treatment outcomes. 

Optical coherence tomography (OCT) is a technology with good reproducibility for measuring the retina in vivo.^11,12 The application of OCT in young children has been demonstrated.^13–15 Macular thickness measured with OCT in different types of amblyopia has been reported with inconsistent results. Most of the recent studies show that amblyopic eyes have anomalous maculae.^10,16–20 Pang et al.¹⁰ reported that amblyopic children with unilateral high myopia tend to have a thicker central macula and thinner peripheral macula in the amblyopic eye compared to the fellow eye. Huynh et al.¹⁸ reported a thicker foveola and a thinner inner macular ring in amblyopic eyes with strabismus or hyperopic anisometropia. Furthermore, they reported that abnormal central macular thickness only occurred in untreated amblyopic eyes, not in treated amblyopic eyes. Their cross-sectional study results suggest that macular thickness of the amblyopic eyes might be altered by amblyopia treatment. To the best of our knowledge, there has been no longitudinal study to measure the effect of amblyopia treatment on macular morphology in children with amblyopia or in the type of amblyopia associated with myopic anisometropia. 

The purpose of this study was to determine whether abnormal macular thickness (thicker central macula and thinner peripheral macula) in myopic anisometropic amblyopia changed following amblyopia treatment. Furthermore, we investigated whether the effect of amblyopia treatment on macular thickness was associated with subject age or improvement in stereoacuity. 

Methods

Study Population and Data Collection

Both the study protocol and informed consent forms were approved by the Institutional Review Board of the Illinois College of Optometry. In accordance with the tenets of the Declaration of Helsinki, written informed consent was obtained from the parent or legal guardian of each child, and each subject gave verbal assent to participate in the study. 

A total of 31 children diagnosed with myopic anisometropic amblyopia were identified at the Illinois Eye Institute, an urban ambulatory eye clinic, with 20 subjects being qualified for amblyopia treatment. These 20 subjects were enrolled in the study (eligibility criteria for amblyopia treatment are listed in Table 1). Seventeen children completed the study with three subjects lost to follow-up. At the beginning of the study, all subjects underwent a comprehensive eye examination including monocular distance VA, cover test at distance and near, stereoacuity with both the Randot Preschool Stereoacuity Test and Stereo Fly (Stereo Optical Company, Inc., Chicago, IL, USA), manifest refraction, dilated fundus examination with indirect ophthalmoscopy, cycloplegic retinoscopy, and A-scan ultrasound biometry. Monocular distance VA was tested at 3 m (10 feet) using a computer-based electronic VA tester,^21,22 which is commonly used in amblyopia studies.^23,24 Single character optotypes with surrounding bars were used according to the Amblyopia Treatment Study VA protocol.²⁵ For children aged <7 years, HOTV characters were used, and for children aged ≥7 years, Early Treatment Diabetic Retinopathy study (ETDRS) characters were presented. Cycloplegia was induced with 1% cyclopentolate, 1% tropicamide, and 2.5% phenylephrine (one drop each). Other than refractive error and amblyopia, the subjects had no concurrent ocular disease. 

Table 1

 

Eligibility and Exclusion Criteria

 

All subjects underwent treatment initially with their refractive correction alone for 16 weeks and with refractive correction and patching treatment for an additional 16 weeks. Detailed information on refractive correction and patching treatment, including patching compliance, was described previously.⁷ Measurements of OCT (Stratus OCT 3; Carl Zeiss Meditec, Inc., Dublin, CA, USA) were performed at the enrollment visit and at the visit immediately following the 16 weeks of patching treatment. A single, skilled technician who was masked to the diagnoses of the subjects performed all OCT measurements after subjects' pupils were dilated to at least 5 mm in diameter. Scans were performed using the fast macular thickness protocol in which six 6-mm lines in a radial spoke-like pattern are obtained in a continuous automated sequence. All scans had signal strength of at least 6.^18,26–30 Although the Stratus OCT operation manual²⁶ recommends a minimum signal strength of 5 (as was selected as the minimal inclusion criterion in various studies^18,27,31), a signal strength of 6, used in previous OCT studies,^29,30 was chosen as the enrollment criterion for this study. In addition, all the macular images were collected using the following protocol to ensure that the foveal depression was in the center of the scan. Alignment mode was used by the technician to ensure each 6-mm line scan was through the center of the fovea. When a scan was observed to be off the center of the fovea even if the subject's fixation was accurate, the procedure of adjusting scan placement was performed to move the scan location to the center of the fovea. The following 10 parameters were measured for each subject: foveal minimum thickness; average foveal thickness; inner nasal, superior, temporal, and inferior macular thickness; and outer nasal, superior, temporal, and inferior macular thickness. The foveal area is defined as the central circle of 1-mm diameter. Inner macula refers to the ring around the foveal area of 3-mm diameter, and outer macula refers to the ring around the inner macula of 6-mm diameter. 

Data Analysis

To analyze refractive error based on cycloplegic retinoscopy, spherical equivalent was defined as the spherical power plus half of the minus cylinder power. Visual acuity was converted to logMAR for analysis. Interocular difference (IOD) in macular thickness was calculated by subtracting macular thickness of the fellow eye from that of the amblyopic eye. A 2 (before treatment, after treatment) × 2 (amblyopic eye, fellow eye) factorial repeated-measures ANOVA was used to determine whether macular thickness in amblyopic eyes differed before and after amblyopia treatment. Independent t-test also was performed to compare change of IOD in subjects with stereoacuity improvement to those without improvement. The association between subject age and change of macular thickness was tested using partial correlation while controlling for VA improvement in amblyopic eyes. All data were analyzed using statistical software (SPSS version 21.0; SPSS, Inc., Chicago, IL, USA). A value of P < 0.05 was considered statistically significant. 

Results

The mean age (±SD) of our subjects was 9.0 ± 3.0 years, ranging from 5.7 to 13.9 years. Table 2 shows the clinical profiles of the children with myopic anisometropic amblyopia. The average logMAR VA at enrollment was 1.0 ± 0.3 (Snellen: 20/200), ranging from 0.6 to 1.3 (Snellen: 20/80–20/400). The mean spherical equivalent refractive error in the amblyopic eyes was −10.65 ± 3.28 diopters (D; range, −5.00 to −15.75 D), while the mean spherical equivalent in the fellow eyes was −1.28 ± 2.60 D. The mean magnitude of anisometropia (spherical equivalent) was −9.37 ± 3.57 D (range, −3.63 to −14.88 D). Mean axial length was 26.6 ± 1.5 mm in the amblyopic eyes and 23.2 ± 1.3 mm in the fellow eyes. The average logMAR VA improvement in the amblyopic eyes was 0.3 ± 0.1 with amblyopia treatment (P < 0.0001).⁶ The mean spherical equivalent refractive error in the amblyopic eyes was −10.88 ± 3.15 D after amblyopia treatment with no significant difference (P = 0.052) compared with that before amblyopia treatment, while the mean spherical equivalent in the fellow eyes after treatment was −1.39 ± 2.40 D (P = 0.069), comparable with that before treatment. Table 3 shows mean macular thickness measurements of amblyopic and fellow eyes before and after amblyopia treatment. Before amblyopia treatment, central macula (both minimum and average foveal thickness) was statistically significantly thicker in amblyopic eyes than in fellow eyes (t[16] = 2.456, P = 0.027 and t[16] = 2.254, P = 0.040, respectively). Inner and outer maculae were statistically significantly thinner in amblyopic eyes than in fellow eyes (all P values < 0.01) before amblyopia treatment (Table 3). Our macular thickness findings before amblyopia treatment were consistent with our previous report.¹⁰ Factorial repeated-measures ANOVA showed the interaction between eye and amblyopia treatment was statistically significant for average foveal thickness (F[1, 16] = 5.047, P = 0.040). There was no treatment effect on fellow eyes (t[16] = −1.209, P = 0.245), but the average foveal thickness in the amblyopic eyes showed a statistically significant reduction after amblyopia treatment (t[16] = 2.139, P = 0.049) even though some subjects showed an increase in central macular thickness after treatment (Table 2). No statistically significant interaction between eye and amblyopia treatment was found in the other macular thickness parameters (all P values > 0.05). 

Table 2

 

Clinical Profiles of the Children With Myopic Anisometropic Amblyopia

 

Table 3

 

Macular Thickness (μm) of Amblyopic and Fellow Eyes in Children With Myopic Anisometropia Before and After Amblyopia Treatment (n = 17)

 

With improvement of VA in amblyopic eyes being controlled, subject age was not correlated with the change of macular thickness in amblyopic eyes (all P values > 0.05). The association of stereoacuity with macular thickness change was also evaluated. Nine subjects had no stereoacuity improvement with either Randot Preschool Stereoacuity Test or Stereo Fly. The remaining eight subjects had stereoacuity improvement with one or both stereo tests. The average change in IOD of foveal minimum thickness was 30.5 ± 43.5 μm in subjects with stereoacuity improvement and 2.33 ± 22.4 μm in subjects without stereoacuity improvement without statistical significance (P = 0.11). The average change of IOD in average foveal thickness was 17.3 μm in subjects with stereoacuity improvement and 5.2 μm in subjects without improvement. Again, this change was not statistically significant (P = 0.23). No significant difference was identified in inner and outer macular thickness change between the two groups. 

Discussion

This study found that central macula (average foveal thickness) was thicker in amblyopic eyes compared with fellow eyes before amblyopia treatment, and the central macular thickness of the amblyopic eyes tended to decrease to the fellow eye level after amblyopia treatment (refractive correction and patching). Minimal changes in inner and outer macular thickness were found following amblyopia treatment. 

Although the destructive effect of amblyopia on both the lateral geniculate nucleus and cortical neurons is well known,^32–34 possible involvement of the retina in amblyopia has been disputed.^35–40 While several electrophysiological studies have shown that pattern electroretinograms and electro-oculograms in amblyopic eyes were reduced compared with normal eyes,^39,40 other studies have not concurred.^37,38 With optical coherence tomography, the retinal structure can be measured reliably.^11,12 Huynh et al.¹⁸ tested 53 amblyopes with either strabismus or hyperopic anisometropia and reported that amblyopic eyes had greater foveal minimum thickness and a thinner inner macular ring than normal eyes. We studied 31 amblyopic children with high myopia and have reported that macular thickness in amblyopic eyes was significantly different from that of the normal fellow eyes.¹⁰ Both foveal minimum thickness and average foveal thickness were greater in the amblyopic eyes with high myopia than in the normal fellow eyes.¹⁰ In contrast, both inner and outer macular thickness were thinner in the amblyopic eyes compared with that of the fellow eyes.¹⁰ Our previous findings are consistent with the study by Huynh et al.,¹⁸ even though different types of amblyopia were evaluated in the two studies. Huynh et al.¹⁸ have proposed that arrest of normal postnatal changes in amblyopic eyes could affect the normal maturation of the macula, including movement of Henle's fibers away from the foveola, and result in an increased foveal thickness measured by OCT. One may expect that amblyopia treatment with both refractive correction and patching could have the potential to reverse the anomalies in the macula. 

A few studies have compared macular thickness in treated amblyopic eyes to that in untreated or residual amblyopic eyes.^18,41,42 Tugcu et al.⁴² studied macular thickness in the amblyopic eyes of 14 children with persistent amblyopia (VA 20/25 or worse) and 18 with resolved amblyopia (VA better than 20/25) and reported no difference in macular thickness between the two groups. However, the children with persistent amblyopia started with worse logMAR VA (0.5 vs. 0.3) and achieved the same degree of VA improvement (3 logMAR lines) as children with resolved amblyopia.⁴² These circumstances might contribute to their finding no difference between the two groups. In their cross-sectional study, Huynh et al.¹⁸ investigated 33 treated and 12 untreated unilateral amblyopes with either strabismus or hyperopic anisometropia and found that IOD in foveal minimum thickness was 1.9 μm in treated versus 12.7 μm in untreated amblyopes (P > 0.05). They also reported that IOD in central macular thickness (average foveal thickness) was −0.3 μm in treated versus 9.2 μm in untreated amblyopes (P < 0.05). In contrast, they found minimal difference in IOD in inner and outer macula thickness between treated and untreated amblyopes.¹⁸ Interestingly, our study revealed findings similar to those of Huynh et al.,¹⁸ although with different types of amblyopia. Conversely, Chen et al.⁴¹ compared 26 treated bilateral amblyopes to 53 untreated bilateral amblyopes and reported no significant thickness difference in any region of the macula between the two groups. It is well established that the deleterious effect on the lateral geniculate nucleus and cortical neurons is more robust in unilateral amblyopia than in bilateral amblyopia due to binocular competition^43–47 between the amblyopic eye and the normal fellow eye in unilateral but not in bilateral amblyopia. Perhaps the competition theory also plays a role in the anomalies of retinal morphology associated with amblyopia and results in more dramatic changes in unilateral amblyopic eyes than in bilateral amblyopic eyes. This theory may explain the contradictory findings by Huynh et al.¹⁸ and Chen et al.⁴¹ Compared to the previous studies, a longitudinal study in which macular thickness is measured both before and after amblyopia treatment provides stronger evidence. A thorough literature search reveals that ours is the first longitudinal study to determine the effect of amblyopia treatment on macular thickness. 

Along with Huynh et al.,^10,18 we have proposed that arrest of normal postnatal changes in amblyopic eyes results in an increased foveal thickness measured by OCT. Furthermore, based on the findings in the current study, we propose that amblyopia treatment has the potential to partially reverse the anomalies in the central macula associated with amblyopia. This proposal is based on our finding that the concave shape of the fovea was partially normalized following amblyopia treatment with no change in the peripheral macula (Fig.). For children with myopic anisometropic amblyopia, the amblyopic eyes are exposed to blurred images. With amblyopia treatment including 16 weeks of both refractive correction and patching of the nonamblyopic eye, we propose that the clear image received by the amblyopic eyes may facilitate the development of the macula and concave shape of the fovea. During macular development, the lateral migration of inner retinal elements that might scatter light has been conventionally considered to assist in achieving high-level acuity. In addition, Snyder and Miller⁴⁸ have suggested that refraction by the foveal pit may serve to magnify the image and increase its resolving power. Conversely, Williams⁴⁹ has proposed that the profile of the foveal pit has little obvious optical benefit. More recently, Provis et al.⁵⁰ have suggested that there are few advantages to having a pit in the central retina. In fact, a critically important factor in high-level acuity is foveal cone density.^51,52 In this study, we found the foveal concavity was deeper following amblyopia treatment; however, no association was found between VA improvement and change of central macular thickness in the amblyopic eyes. Our findings are consistent with the proposals from Williams⁴⁹ and Provis et al.⁵⁰ 

Figure

 

(A) Schematic diagram of macular thickness of a normal eye (dashed line) versus that of an amblyopic eye with unilateral high myopia before amblyopia treatment (solid red line). (B) Schematic diagram of macular thickness of a normal eye (dashed line) versus that of an amblyopic eye following amblyopia treatment (solid red line), showing increased concavity in the fovea with no change in the peripheral macula.

The following may be reasons why we found only partial reversal of macular anomalies associated with myopic anisometropic amblyopia. One could be that the critical period for macular development is short.^53,54 In order to achieve a full recovery, earlier amblyopia treatment may be needed than that in our study. Second, our subjects were only treated with 16 weeks of refractive correction followed by the addition of 16 weeks of patching. Thus, it is possible that more macular change might be revealed with a longer period of treatment. Leone et al.⁵⁵ proposed that increased macular thickness found in several studies may be due to inadvertent measurement of a parafoveal eccentric point in amblyopia. To address this issue, central fixation was confirmed in each subject in our study by ensuring the location of the foveal depression in the inner circle of the macular scan. The new generation OCT, spectral-domain OCT, provides more accurate measurement of macular thickness compared to the time-domain OCT. We are conducting a cross-sectional study to investigate the macular thickness of amblyopic eyes using the Cirrus high-definition OCT. One may question whether macular structure might have changed over the course of the study even without amblyopia treatment. This question could be addressed with a control group of age-matched children with bilateral myopia magnitude matching that of our amblyopic subjects (mean refractive error = −10.79 D) but without amblyopia (VA of 20/25 or better). We have attempted to enroll such a control group. However, within a 2-year period, only one control subject was identified. The challenge we faced when identifying control subjects was that bilateral amblyopia was present in individuals with a magnitude of myopia matching our myopic anisometropic amblyopia subjects. Thus, without a control group, the effect of normal aging on macular thickness during the study period cannot be ruled out. Because central macular thickness did not change in the fellow eye during the course of the study, it is unlikely that factors other than amblyopia treatment contributed to the change we found in the macula of the amblyopic eyes. 

The association between myopia and macular thickness has been reported with inconsistent results.^56–61 Earlier studies reported myopia had no effect on macular thickness.⁵⁶ More recent studies have found that myopia, especially high myopia (<−5.00 D), was associated with abnormal macular thickness.^57–61 Luo et al.⁵⁹ reported that outer macular thickness was thinner in moderate myopia compared to both low myopia and nonmyopia; however, they did not find any difference in either the foveal or inner macular thickness among no, low, and moderate myopia. Lam et al.⁶¹ found a thicker fovea and thinner outer macula associated with myopia but no change in inner macular thickness comparing myopia (including low, moderate, and high myopia) to nonmyopia.⁶¹ Wu et al.⁵⁷ reported a thicker fovea and thinner inner and outer macula in high myopia than nonmyopia. If myopia were associated with a shallow fovea, foveal morphology of myopes would be expected to be different from that of emmetropes. However, Dubis et al.⁶² studied foveal pit morphology in myopes and emmetropes and found no difference in foveal depth, diameter, and slope between those two groups. Odell et al.⁶³ studied 113 normal adult subjects and 43 subjects with retinal pathologies and reported that ocular magnification affected the accuracy of macular thickness maps. Furthermore, they found that as the deviation from average axial length increased, the OCT error in ETDRS thickness plots became greater.⁶³ Thus, Odell et al.⁶³ suggested that macular thickness be adjusted for axial length. Without measuring axial length after amblyopia treatment, macular thickness cannot be adjusted accordingly. In the current study, macular thickness before amblyopia treatment was compared with that after amblyopia treatment. With no significant change in myopia in the amblyopic eyes before and after amblyopia treatment in our subjects, we expect that axial length should not differ significantly after amblyopia treatment and myopia should not be the underlying reason for the change we found in the central macula of the amblyopic eyes. It is our assumption that if myopia were the major reason for the differences in macular thickness, one would expect that the findings in amblyopes associated with hyperopic anisometropia should be different from those associated with myopic anisometropia. However, our macular thickness findings are in agreement with those in the Sydney Childhood Eye Study in which the refractive error of amblyopic eyes was hyperopic (approximately +2.00 D).¹⁸ 

It is well established that central maculae in blacks are thinner than those in Caucasians and Hispanics.^64–67 Although the majority of our subjects were black children, macular thickness before treatment was compared with that after treatment in this study. Thus, the racial difference in macular thickness should not affect our study results. 

To the best of our knowledge, this is the first longitudinal study to determine whether macular thickness can change with amblyopia treatment in amblyopic eyes associated with myopic anisometropia. There are several limitations to our study. First, there were a small number of subjects enrolled, which might explain our finding a trend toward foveal minimum thickness change without statistical significance. An effort has been made to enroll more subjects. However, due to frequent OCT upgrading and the inability to compare macular thickness data obtained with time-domain OCT and spectral-domain OCT,^68,69 it is difficult to conduct a longitudinal study using OCT over a long period of time. Second, an attempt was made to enroll control children with normal vision and the same degree of myopia as the amblyopic eyes, and within a 2-year period, only one control subject was identified. Finally, our subjects had myopic anisometropic amblyopia. Caution is needed in generalizing these findings to other types of amblyopia. 

In summary, amblyopic eyes in children with myopic anisometropia had thicker foveae and thinner peripheral maculae compared to the fellow eyes before amblyopia treatment. Central maculae tended to be thinner following amblyopia treatment with no changes in peripheral maculae. 

Acknowledgments

The authors thank Rebecca Tudor for her assistance in performing OCT on all subjects in this study. They also thank Li Deng, PhD, at the New England College of Optometry, for her statistical support and help. 

Presented in part at the annual meeting of the Association for Research in Vision and Ophthalmology, Seattle, Washington, United States, May 2013. 

Supported by the Illinois Society for the Prevention of Blindness and the Illinois College of Optometry Research Fund. 

Disclosure: Y. Pang, None; K.A. Frantz, None; S. Block, None; G.W. Goodfellow, None; C. Allison, None