This prospective study enrolled 31 patients, aged from 5 to 12 years, with hyperopic anisometropic amblyopia. We examined outpatients who met the following inclusion criteria of hyperopic anisometropic amblyopia: no history or evidence of intraocular surgery, neurologic or retinal disease, glaucoma, nystagmus, and strabismus. The VA difference between the amblyopic and normal eyes was at least 2 lines of Snellen acuity. All patients underwent a detailed eye examination including best corrected VA (BCVA) determination using Snellen chart from 6m distance, manifest refraction and cycloplegic refraction after pupillary dilation with 1% cyclopentolate hydrochloride and 1% tropicamide, alternative cover test, duction and version testing, intraocular pressure (IOP) measurement, slit-lamp biomicroscopy and fundus examination.

In this study, anisometropia was defined as a cycloplegic, spherical equivalent difference greater than 2.00 diopter (D) between fellow eyes. To calculate the mean VA, Snellen VA was converted to logMAR scale, and the mean logMAR VA was reconverted to the Snellen VA. IOP, anterior segment and fundus examination were normal in all eyes.

The macular and RNFL thicknesses were measured by OCT (OCT 3000, version A 3.0; Carl Zeiss-Humphrey system, Dublin, CA, USA). Macular scans, consisting of six radial scans, each scan is 30° apart, around the fovea, were performed to evaluate macular thickness, obtain the retinal thickness from 10 points including the fovea, and construct the 6 mm diameter map (Fig. 1). The mean of these values was used to calculate the average macular thickness. To measure the peripapillary thickness of the nerve fiber layer, circular scans were performed around the optic disc (Figs. 2, 3). Internal fixation was used for macular scanning, and external fixation for optic disc scanning. Multiple images were taken from each eye by an experienced operator.

Retinal thickness was measured as the distance between vitreoretinal interface and the layer corresponding to the pigment epithelium and choriocapillaries, while the foveal thickness was defined as the minimal value located at the image center. The scans were subjected to analysis with standard software provided with the apparatus. Student's two-tail t test was used for data analysis, and a P value of less than 0.05 was considered statistically significant.

The study included 31 unilateral anisometropic amblyopia patients without any other ocular or neurologic disease. There were 16 boys and 15 girls, of mean age 7.7 years (5~12 years), with hyperopic anisometropia (+2.00 D to +6.50 D). Mean corrected VA was 0.29 (0.2 to 0.6) and 1.0 (0.9 to 1.0), and mean refractive error (spherical equivalent of the cycloplegic refraction) was +3.71 D (+2.00 D to +6.50 D) and +1.00 D (emmetropic to +3.00 D) in the amblyopic and normal eyes, respectively (Table 1). Examination of the anterior segment and fundus revealed no abnormalities and IOP was within the normal range.

The macular thickness ranged from 237 to 285 µm (252.5±13.7 µm) in the amblyopic eyes and 230 to 283 µm (249.7±13.3 µm) in the normal eyes. There was no statistically significant difference between the two (P>0.05). The thickness of the peripapillary RNFL ranged from 101 to 142 µm (115.2±9.7 µm) in the amblyopic eyes and 96 to 130 µm (109.6±8.4 µm) in the normal eyes. The former layer was significantly thicker than the latter (P=0.019) (Table 2).

The amblyopic process may have an effect on various levels of the visual pathway. Atrophy involving the cells in the lateral geniculate nucleus that receive input from the amblyopic eye16 has been reported. However, retinal involvement in the amblyopic eye is controversial.17-19 Several experiments have demonstrated that retinal ganglion cells can suffer modifications with light deprivation from birth, including cell loss,9 mean nucleolar volume diminution in ganglion cell cytoplasm and internal plexiform layer thinning in rats and cats,10 and reduction in optic nerve size area in mice.19 Arden and Wooding22 reported that electroretinograms (ERG) elicited by patterned stimuli in humans with various types of amblyopia were significantly reduced. These results suggest that in humans, amblyopia may be associated with changes in retinal function at the level of production of the pattern ERG (PERG), which is presumed to be preganglionic. Other investigators,23,24 on the contrary, have not observed a PERG deficit when optical focus, fixation alignment, and stability were individually optimized.

Recently, several in vivo, structural study techniques such as scanning laser polarimeter (GDx) and OCT have been described and used to evaluate the RNFL and macular thicknesses.

Using a third generation nerve fiber analyzer (GDx; Laser Diagnostic Technologies, San Diego, CA), Colen et al25 measured RNFL thickness in strabismic amblyopia and found no statistically significant difference between the strabismic amblyopic eyes and normal eyes. In the study of Bozkurt et al,26 GDx was performed on 18 anisometropic, 2 strabismic and 4 combined amblyopic eyes and there were no significant differences in the retardation measurements of the nerve fiber layer between amblyopic and normal eyes. However, in a study of using OCT, Yen et al27 reported a thicker RNFL in refractive amblyopia and a significantly different RNFL thickness between the amblyopic and normal eyes in refractive amblyopia patients. They suggested that refractive amblyopia affects the process of postnatal reduction of ganglion cells and that RNFL thickness may be thicker than the normal eye.

In the present study, the peripapillary RNFL and macular thicknesses were measured by OCT and the results showed that the macular thickness was not significantly different between the anisometropic amblyopic eyes and the normal eyes, although the peripapillary RNFL was significantly thicker in the amblyopic eyes. Our results corroborate those of a previous OCT study27 that suggested RNFL is thicker in refractive amblyopia and that the amblyopic process may involve the peripapillary RNFL.

Mrugacz et al28 showed that foveal retinal and RNFL thicknesses were significantly decreased, especially in high myopia. In our study, normal eyes were emmetropic or weakly hyperopic. The average refractive error was +1.00 D, which ensured that the control eye selection was reasonable as the comparison was not with thinner controls.

The average foveal thickness in our study was 183.2 µm in the amblyopic eyes and 178.7 µm in the normal eyes. This result is very similar to that of many other studies using OCT, including Hee et al29 and Kang et al.30 There was no statistically significant difference in average foveal thickness between the two eyes (P>0.05).

In conclusion, peripapillary RNFL was significantly thicker in hyperopic anisometropic amblyopia, whereas the macular thickness was not significantly different between amblyopic and normal eyes. This is the first study to evaluate the macular and foveal thicknesses in anisometropic amblyopic eyes. Our results suggest that although the amblyopic process may not have any significant effect on the macula, it may exert a significant effect on peripapillary RNFL. Further studies, including postmortem, would be instrumental in examining the retinal, histologic and structural differences between amblyopic and normal eyes.