Choroidal vessel density in unilateral hyperopic amblyopia using en-face optical coherence tomography

Amblyopia is defined as a disorder with dysfunctional processing of visual information, such as reduced recognition visual acuity (VA), resulting from either disuse due to the absence of a clear image on the retina or misuse due to abnormal binocular interactions [1]. Traditionally, morphological abnormalities have not been considered present in the ocular structure of amblyopic eyes. In contrast, recent studies using optical coherence tomography (OCT) have shown structural changes to the choroid such as thickening in hyperanisometropic amblyopic eyes compared with that of fellow or healthy eyes, independent of axial length (AL) [2‐5]. In chorioretinal diseases such as central serous chorioretinopathy [6] and Vogt-Koyanagi-Harada disease [7], the pathogenesis and clinical significance of choroidal structural changes is relatively well understood, and OCT findings are utilized for diagnosis and follow-up. In contrast, no consensus has been reached regarding the pathogenesis or clinical significance of choroidal structural changes in patients with amblyopia.

Recently, a new method other than thickness has been developed to evaluate choroidal structures, focusing on choroidal vasculature using binarization analyses [8, 9]. Some studies have suggested changes in choroidal vasculature in amblyopic eyes, but insufficient knowledge has been accumulated [10‐13]. The purpose of the present study was thus to investigate the clinical significance of evaluating choroidal vascular density (CVD) in patients with amblyopia.

All investigative procedures adhered to the tenets of the Declaration of Helsinki. This study was approved by the Institutional Review Board Committee at Kawasaki Medical School (registration number: 3473). This study was designed as a retrospective observational case series and was conducted from November 2013 to December 2018 in the Department of Ophthalmology at Kawasaki Medical School Hospital. This study included minors under 16 years old among the patient sample. Informed consent for all examinations was obtained from one parent of each patient.

The current findings showed that the CDV of Haller’s layer, as analyzed by en-face OCT, did not differ significantly among amblyopic, fellow and normal control eyes. Also, no significant correlation was apparent between CVD and BCVA in amblyopic eyes.

The choroidal structure has been implicated in the pathogenesis of various ocular diseases. Agrawal et al. recently reported that choroidal vascularity index (CVI: the ratio of luminal area to total choroid area, equivalent to CVD) is a robust marker for studying pathophysiologies of the choroid, because CVI is less variable than SFCT and is not associated with most physiological variables [17]. In fact, studies in patients with Stargardt disease [18] and retinopathy of prematurity [19] have found that CVI is a more sensitive biomarker than SFCT and that a decrease in CVI is associated with a decrease in BCVA. On the other hand, the present study observed no specific changes in the CVD of amblyopic eyes, and no correlation with BCVA, suggesting that CVD may not be an adequate biomarker for evaluating visual acuity loss in amblyopic patients. However, visual dysfunction in amblyopia can be evaluated using many parameters other than visual acuity. For example, amblyopic eyes show poorer accommodative performance than fellow or normal eyes [20]. Accommodation has been reported directly related to choroidal structure [21]. Therefore, novel findings may be obtained by using indices other than BCVA (e.g., accommodation) in the investigation of the relationship between amblyopia and CVD in the future.

Several reports have examined choroidal vascular structures in patients with amblyopia [10‐13]. In reports using analyses by binarization of a single-section OCT image, Nishi et al. found that the luminal area was significantly larger and the stromal area was significantly smaller in amblyopic eyes than in control hyperopic eyes [10]. Another study showed a reduction in the luminal area and widened stromal area after treatment only in amblyopic eyes [13]. Beak et al. reported that the ratio of choroidal luminal area to total choroidal area (i.e., CVD) of amblyopic eyes showed different characteristics from normal control eyes, such as amblyopic eyes showing higher CVD that did not increase in direct proportion to the increased choroidal thickness [12]. In contrast to those previous reports [10, 12, 13], the present study did not find any difference in CVD between amblyopic patients and normal healthy subjects. Our study extracted en-face images at a fixed distance from Burch’s membrane and analyzed the CVD. On the other hand, CVD analysis using B-scan images, as used in previous studies [10, 12, 13], is able to analyze only one cross-section image, while allowing analysis of the structure of the inner and outer layers of choroid simultaneously. Differences in results between these studies may thus be due to differences in the methods used for CVD analysis.

On the basis of analyzing binarized en-face OCT images, Terada et al. reported that the outer choroidal vascular areas (i.e., CVD of Haller’s layer) were larger in both amblyopic (61.49 ± 4.95%) and fellow eyes (61.48 ± 3.73%) than in healthy eyes (55.69 ± 1.83%) [11]. They also indicated that the cutoff for distinguishing between amblyopic patients and controls was 59%. In our current study, the CVD in amblyopic and fellow eyes exceeded 59% (59.11 ± 0.66% and 59.23 ± 0.81%, respectively), consistent with a report by Terada et al. However, the CVD in normal control eyes (59.29 ± 0.74%) also exceeded 59%, resulting in comparable CVDs among different eye conditions. In normal eyes, CVD has been reported to correlate positively with SFCT [9]. However, SFCT was slightly thicker in Terada’s control group (351.9 ± 60.7 μm) than in our control group (323.8 ± 69.1 μm). The higher CVD in the normal control eye is thus not attributable to differences in SFCT. Substantial individual differences might be present in CVD. We should therefore be cautious of using CVD analysis of en-face OCT images to evaluate amblyopia.

The present study showed that SFCT was significantly thicker in the amblyopic eye than in fellow and normal control eyes, while CVD was comparable among all eyes. The results indicate that choroidal blood flow with respect to total choroidal volume may be increased in amblyopic eyes than in fellow and normal control eyes. A possible future direction of this work would be to investigate the ratio of choroidal vessels to total volume of the choroid in amblyopic eyes.

Several limitations should be considered when interpreting the findings of the present study. First, the sample size was small. A further study of a larger number of patients with detailed classifications of the degree and type of amblyopia will need to be undertaken to validate our findings. Second, CVD could only be assessed in a 3 × 3-mm localized area centered on the fovea. This was due to the large difference in thickness between the center and periphery of the choroid, making it difficult to capture the same depth of vascular structure in the center and periphery if the analyzed area of the en-face image flattened with Bruch’s membrane is expanded. In the future, new tools, such as high-penetration doppler optical coherence angiography [22] and hybrid three-dimensional models of the choroidal vasculature [23], are expected to provide a more detailed assessment of choroidal vascular structures.

In conclusion, the CVD of Haller’s layer did not differ significantly between amblyopic eyes and fellow or normal control eyes. Evaluation of CVD using en-face OCT may have little clinical significance in elucidating the mechanisms underlying amblyopia.