Assessment of macular microvascular changes in children following treatment of anisometropic myopic amblyopia using optical coherence tomography angiography

The Avanti RTVue-XR system with optical software (Angio Vue version 2017.1.0.155; Optovue, Inc.) was utilized for OCTA imaging, which uses a split-spectrum amplitude-decorrelation angiography (SSADA) approach to retrieve OCT angiography information.

The Angio Retina protocol was used to acquire OCTA images, using a scan area of 6 × 6 mm².

Automated segmentation was used for the definition of different vascular plexuses: the superficial capillary plexus (SCP) is identified between 3 μm below the internal limiting membrane (ILM) and 15 μm below the inner plexiform layer (IPL), the deep capillary plexus (DCP) is located between 15 and 70 μm below the IPL, and the choriocapillaris is found between 30 and 60 μm below the RPE reference.

AngioAnalytics are measurement tools that allow for the assessment of vascular area density (VD) and non-flow area in order to determine the size of the FAZ. The vessel density at the level of the superficial and deep retinal capillaries, as well as the choriocapillaris, was assessed and analyzed using a vessel density map, and defined as the relative density of flow as a proportion of the total area. After image acquisition, the FAZ area was measured in square millimeters (mm²) at the level of the superficial and deep retinal plexuses using the non-flow feature of the OCTA software. The area of the FAZ is automatically estimated by the program (RTVue-XR version: 2017.1.0.151) when the operator taps on the center of the FAZ.

In amblyopic eyes, measured VD values and FAZ areas were corrected by the magnification factor of the image according to the following formula: (where D_t is the true size of the measured retinal feature, D_m is the measured size on the OCTA image, and AL is axial length) [19].

The best-corrected visual acuity, foveal and parafoveal thickness, vessel density (VD), and FAZ area were compared using OCTA at three levels (superficial, deep, and choriocapillaris) between the sound and amblyopic eyes.

Also, a comparison was made between the amblyopic eyes before and after treatment with either glasses or glasses with patching.

The data was coded, tabulated, and statistically analyzed using the Statistical Package for Social Sciences (SPSS) program (software version 25; SPSS Inc., IBM Corp., New York, USA, 2017).

Descriptive statistics were done for parametric (normally distributed) quantitative data by mean, standard deviation (SD), and minimum and maximum of range and for qualitative data by frequency and percentage.

Analyses were done between the two groups for parametric quantitative data using the independent samples t test, and for non-parametric quantitative data using the Mann–Whitney U test, while analyses between the two times within each group for parametric quantitative data were done using the paired samples t test.

Analyses were done for qualitative data using the chi square test.

Pearson’s correlation was done between continuous quantitative variables.

Amblyopic eyes showed a significant increase in parafoveal thickness (p < 0.001) compared with their sound eyes, without a significant difference in foveal thickness between the two groups. The latter finding was consistent with the findings of Singh et al. and Taskiran Comez et al., who reported a nonsignificant difference in central macular thickness (CMT) between better and worse eyes with anisomyopia [21, 22]. In contrast to our findings, Rajavi and his co-workers reported that amblyopic eyes had increased macular retinal thickness at the foveal area (1-mm ring) than fellow eyes and controls, and no significant difference was found at the 3 mm ring (parafoveal) [23], this could be attributed to various inclusion criteria for their research participants, as they included patients with varying degrees of amblyopia (mild, moderate, and severe), whereas the current study included children with moderate and severe amblyopia.

Before treatment, there was a statistically significant decrease in vascular density in amblyopic eyes at the level of the superficial retinal plexuses (whole image and parafoveal) compared with fellow eyes. This finding was consistent with the findings of Hamurcu and colleagues, who found a statistically significant decrease in whole and parafoveal superficial capillary plexus (SCP) vessel density in amblyopic eyes compared to healthy controls [24].

Although the choroid primarily nourishes the external retina, neurovascular coupling has demonstrated how a light stimulus can modify the vascular parameters of the superficial plexuses that nourish the internal retina [25], suggesting that the lower VD in the superficial capillary plexus in amblyopic eyes could be due to improper development due to a lack of normal visual experience [8].

Previous studies found a significantly lower vessel density in the deep retinal plexus as well as the superficial retinal plexus in amblyopic eyes, and this difference could be explained by comparing the amblyopic eyes to healthy controls rather than their fellow eyes [8]. Furthermore, amblyopic eyes in their studies had various etiologies (ametropic, strabismic, anisometropic, and meridional amblyopia) [7], and strabismic amblyopia in Yilmaz et al. [9].

In the current study, there was a significant decrease in vessel density at the choriocapillaris level (whole image and foveal) in the amblyopic anisomyopic eyes compared with the sound fellow eye. This finding is consistent with previous observations obtained by Liu et al., who concluded that when anisometropia exceeded 1.50 D, the choriocapillaris vascular density in the more myopic eyes was considerably lower than in the less myopic eyes [26].

In contrast to our findings, Borrelli and colleagues discovered significantly increased choriocapillaris vessel density in amblyopic eyes (anisometropic and strabismic) when compared to control eyes, but not in the same patient’s sound eye. Furthermore, their patients were hypermetropic (mean refractive error was 4.3 ± 6.2), with a thicker retina and increased vessel density as a result [27].

There was a wide variation in OCTA findings in amblyopia with diverse aetiologies. Therefore, we chose one form of amblyopia (anisomyopic amblyopia) to be specific in our results regarding this type and avoid any variation with different types of amblyopia.

To the best of our knowledge, this is the first study to compare the effects of effective amblyopia therapy on OCTA measures such as FAZ area alterations and vessel density changes at the level of the retinal plexus and choriocapillaris in anisomyopic patients to the same amblyopic eyes before treatment.

After proper and successful treatment of amblyopia with glasses alone or glasses and patching, a significant improvement was reported in amblyopic eyes in terms of best-corrected visual acuity (p < 0.001), with higher VD values in superficial and deep capillary plexuses (whole, foveal, and parafoveal, p < 0.001 for all), and choriocapillaris (foveal, p = 0.030).

Similar findings were reported by Gunzenhauser and colleagues, who found a significant increase in VD values in some regions of their study (whole image and superior hemisphere) at the level of deep capillary plexus and superficial foveal vessel density after treatment using 3 × 3-mm OCTA scans, despite the fact that their study was conducted on patients with strabismic amblyopia, and they suggested that the increase in retinal vessel density could be attributed to strabismus surgery in strabismic patients [11]. Our patients were anisometropic, treated with glasses only or with patching, so the possibility of improved retinal vessel density due to surgery was excluded from our explanations.

Also, Salerni et al. found a higher macular vascular density at the level of superficial plexuses in successfully treated amblyopic patients than in unresponsive amblyopic and normal eyes, as their study only investigated vessel density at the superficial retinal plexuses, and participants had different types of amblyopia (strabismus, anisometropia, and meridional) [20].

Zhang and his associates examined the effect of treatment on the retinal vessel density in anisometropic amblyopia by comparing newly diagnosed amblyopes with treated amblyopes and controls and found a significant reduction in superficial and deep vessel density compared with controls and a significantly lower superficial vessel density between amblyopes and treated amblyopes, but they did not compare pre- and post-treatment vessel density [12].

As the optical treatment of amblyopia should be the first line of treatment in patients with refractive error [28], glasses were prescribed for all our patients. Fifty-three (53.12%) of the patients had best-corrected visual acuity 0–0.2 log MAR with glasses after 12–16 weeks. Forty-seven (46.88%) patients needed occlusion therapy with their glasses, and improvement in the mean best-corrected visual acuity reached 0.1 ± 0.1 log MAR after a mean follow-up period of 25.4 ± 3.31 weeks. This finding is in line with that of Cotter and colleagues, who discovered that by 9 weeks, about half of the children had reached their best acuity [29]. However, the length of time depends on the patients' compliance with glasses wear [30].

To our knowledge, this is the first research to identify the effect of optical correction alone on OCTA parameters, while the few earlier studies only identified the effect of patching treatment with glasses if required [11, 12, 20].

Amblyopic eyes were divided into two subgroups according to treatment (glasses and glasses with patching) and compared before treatment as regards best-corrected visual acuity, refractions, and OCTA parameters. Amblyopic eyes treated with glasses had a significantly lower refraction and better best-corrected visual acuity with smaller superficial and deep FAZ areas when compared with amblyopic eyes treated with glasses and patching. Furthermore, the VD values in the superficial (foveal and parafoveal) and deep (whole image, foveal, and parafoveal) retinal plexuses were significantly higher, which could be the major reason for these eyes’ rapid recovery with only optical treatment, as Huang and his co-workers suggested that the clearer visual stimulation caused by optical therapy may promote the process of macular vascular pruning in the amblyopic eyes, thereby improving macular perfusion and increasing FAZ circularity [31]. Wang and Xia also discovered that a decrease in vessel density could contribute to poor best-corrected visual acuity [16].

On the other hand, there was higher vessel density in the whole image and foveal choriocapillaris in patients treated with glasses with patching (Table 3).

When the difference between pre-treatment and post-treatment parameters in two subgroups was compared, amblyopic eyes treated with glasses and patching showed greater improvement in vessel density in superficial capillary plexuses (foveal and parafoveal) and deep capillary plexuses (whole image, foveal, and parafoveal), and this finding may be attributed to a longer time of treatment in the patching subgroup with additional stimulation to photoreceptors and, as a result, increased metabolic and oxygen requirements for this layer, which is supplied in part by the deep capillary plexus (10–15%) [32].

Amblyopic eyes treated with glasses alone improved more in choriocapillaris foveal VD (Table 5).

In amblyopic eyes, there was a positive correlation between refraction and post-treatment superficial foveal VD, deep whole-image VD, and deep foveal and parafoveal VD.

In the current study, all amblyopic eyes were myopic. Myopia may lead to the narrowing and straightening of blood vessels [33]. Treatment for amblyopia in myopic eyes may stimulate the retinal layers’ need for more oxygen, resulting in increased blood circulation.

On the other hand, there was a moderate negative correlation between refraction and post-treatment superficial and deep FAZ areas, which means that amblyopic eyes with higher degrees of myopia in our study had larger FAZ areas. This might be explained by the fact that a thicker retina has a greater metabolic need, which is correlated with a decrease in the FAZ area and vice versa [34]. Myopic eyes already have a thinner retina and, as a result, a greater FAZ.

Before treatment, amblyopic eyes’ log MAR best-corrected visual acuity was negatively correlated with post-treatment improvement in superficial foveal VD, deep capillary plexus VD (whole image, foveal, and parafoveal), (higher VD after treatment with lower log MAR values, or better visual acuity). As a result, severe amblyopia results in less recovery of vessel density following successful anisomyopic amblyopia treatment.

Although the purpose of this study was to evaluate microvascular changes after amblyopia treatment in eyes with anisomyopic amblyopia to avoid variable factors and detect specific results, it has several limitations, including a small number of patients, and participants who are younger and have a better response to amblyopia treatment cannot be included as they are uncooperative. Further research with a larger number of patients and diverse age groups is required in each type of amblyopia separately, as well as in anisometropic amblyopia due to different refractive errors.

In conclusion, amblyopic eyes had higher parafoveal thickness and lower superficial and choriocapillaris VD than sound eyes before treatment, which was significantly improved after visual acuity improvement in both glasses and glasses with patching subgroups. Following successful anisomyopic amblyopia treatment, severe amblyopia results in less recovery of the retinal vessel density.