Visual Acuity-Related Outer Retinal Structural Parameters on Swept Source Optical Coherence Tomography and Angiography in XLRS Patients and Carriers

Abstract

Purpose: To assess the quantitative differences in vessel density and retinal thickness of X-linked retinoschisis (XLRS) patients and RS1 mutation carriers, and the correlation with best-corrected visual acuity (BCVA) with swept source optical coherence tomography (SS-OCT) and OCT angiography (OCTA).

Methods: We analyzed the correlation between the BCVA of XLRS patients and the SS-OCT and OCTA findings including the detailed structural characteristics of XLRS patients.

Results: Besides the schitic changes in various retinal layers, the structural disturbance of outer retina was universally found. In 29 eyes included in the quantitative analysis, XLRS patients showed lower vessel density of the superficial capillary plexus, deep capillary plexus and lower thickness of the outer nuclear layer. BCVA was correlated with the thickness of the outer plexiform layer and outer nuclear layer and the thickness from the outer limiting membrane to the retinal pigment epithelium. Carriers showed higher thickness of outer plexiform layer and smaller foveal avascular zone area.

Conclusions: SS-OCT and OCTA could identify the pathological alterations of the individual retinal layers and capillaries, which could pinpoint the exact location of the damages related to visual impairment. In the carriers, the subtle alterations that can be detected with SS-OCT, despite their normal visual acuity, may be caused by the lyonization.

Translational Relevance: Swept source optical coherence tomography can be used as an efficient technique to expose the retinal damage related to visual impairment for prognosis and follow-up.

Introduction

X-linked retinoschisis (XLRS) is a fairly common vitreoretinal degeneration disease that affects male early in their life with the prevalence ranging from 1 in 5000 to 1 in 20,000.¹ The mutation of retinoschisin gene (RS1) located on Xp22 is responsible for XLRS.² Retinoschisin, a 24-kDa discoidin domain-containing protein secreted by photoreceptors, can bind tightly to the surface of bipolar cells, photoreceptors, ganglion cells, and amacrine cells, which is involved in the maintenance of retinal cellular adhesion via subunit oligomerization.^3–5 Typical clinical features of XLRS include bilateral spoke-wheel foveal schisis and peripheral retinoschisis, most commonly in the inferotemporal quadrant.⁶ The severity of the visual distortion and the progression of the disease are highly variable.⁷ Additionally, although female carriers are generally asymptomatic with normal fundus appearance, multifocal electroretinogram (ERG) study reported minor dysfunction in the carriers.⁸ 

The fast-evolving technology in retinal imaging using optical coherence tomography (OCT) leads to detailed visualization of the retinal microstructure on a histopathological scale offering fast and noninvasive approach for the diagnosis and follow-up for XLRS.³ Several studies focusing on the determination of the correlations between the structural changes and the visual acuity showed that the decreased vision is associated with thinning of the outer segment.^9,10 These findings are consistent with the theory that the best-corrected visual acuity (BCVA) decreases when the retina become atrophic as the result of the disease progression. In the recent years, the swept source OCT (SS-OCT) and SS-OCT angiography (SS-OCTA) equipped with tunable laser provides quantitative measurements of the retina for a better understanding of the pathophysiology of retinal diseases.¹¹ 

The purpose of our study was to identify the possible correlations between the visual acuity and the alterations of the retinal structures and vascular density in XLRS patients and carriers on SS-OCT and OCTA, which could be used as indicators for visual prognosis. In addition, detailed image analysis of the structural alterations may shed some light on the pathogenic process of XLRS. 

Methods

This observational cross-sectional study included XLRS patients and carriers of RS1 gene mutations. Patients and carriers were enrolled at the Department of Ophthalmology, West China Hospital, Sichuan University, between July 2019 and July 2021, with written consent from the informed the parents or guardians of each patient. 

Patients with XLRS were diagnosed based on clinical manifestations characterized of foveal spoke-wheel patterns on fundus examination and foveal retinoschisis with or without the involvement of the peripheral retina on OCT examinations. They were classified based on recent OCT classifications of XLRS that includes four distinct clinical phenotypes: type 1, foveal; type 2, foveolamellar; type 3, complex; and type 4, foveoperipheral.⁶ Patients with a previous ocular surgical history, dioptricmedia opacity, or poor vision interfering with the acquisition of accurate images were excluded. Carriers were identified by having male family members suffered from XLRS, and carrying a variant of RS1 through whole exome sequencing. Age- and sex-matched healthy controls with no history of any ocular diseases were recruited as the control groups for patients and the carriers, respectively. Thorough interviews with the candidates were conducted for a complete family history for any unknown poor vision before the enrollment of the control group. Ophthalmologic examinations including BCVA, slit lamp examination, fundus photography, and SS-OCT/OCTA were performed in the outpatient clinic for all the involving subjects. The visual acuity test was performed by experienced optometrists using Snellen visual acuity chart and presented as logarithm of the minimal angle of resolution (logMAR) for analysis, with 2.6 logMAR for counting fingers, 2.7 logMAR for hand movements, 2.8 logMAR for light perception, and 2.9 logMAR for no light perception.¹² The clinical spectrum and clinical phenotypes were investigated. The genetic correlation of the disease was also analyzed. 

SS-OCT and OCTA images were acquired by VG200 SS-OCTA (SVision Imaging, Ltd., Luoyang, China) with a light source of 1050 nm and a scan rate of 200,000 A-scans per second.¹³ The automatic software program segments were corrected manually by our researchers. Images with a signal strength index of more than 7 (full scale, 10) were considered for analysis as recommended by the manufacture. The retinal thickness of the RGI layer (retinal nerve fiber layer, ganglion cell layer, and inner plexiform layer), inner nuclear layer (INL), outer plexiform layer (OPL), and outer nuclear layer (ONL) in the 5-mm diameter areas and OEIR layer (from the outer limiting membrane [OLM] to the apical retinal pigment epithelium/retinal pigment epithelium [RPE], border including interdigitation zone and ellipsoid zone) at fovea in the 1-mm diameter areas were also automatically obtained with the software (Fig. 1A). Vessel density (VD) was defined as the percentage of the area occupied by large vessels and microvasculature in the analyzed region, and the full OCT spectrum was split into several narrower bands. Inter–B-scan decorrelation was computed separately and then averaged for better flow detection and connectivity of microvascular network.¹⁴ For VD analysis, the central retina was divided into the 1.5-mm and 5.0-mm diameter central foveal region (Fig. 1B). Superficial capillary plexus (SCP) slabs were defined from the inner limiting membrane to the boundary of the inner plexiform layer and the INL (Fig. 1C), and deep capillary plexus (DCP) slabs were defined from the boundary of the inner plexiform layer and INL to that of the OPL and ONL (INL and OPL) (Fig. 1D). 

Figure 1.

 

OCT and OCTA XLRS patients. (A) SS-OCT image showing the individual retinal layers with labels. Interdigitation zone defect and structural disturbance of the ellipsoid zone were shown with a red box. Disappearance of normal morphology of myoid zone was also presented (red triangle). Peripapillary retinoschisis was also presented in the image. (B) VD analysis of central retina was divided into a 1.5-mm diameter region (inner circle) and a 5-mm diameter region (outer circle). (C) The SCP ranged from the ILM to the boundary of the inner plexiform layer (IPL) and INL (within the blue lines). (D) The DCP ranged from the boundary of the IPL and INL to that of the OPL and ONL (within the blue lines).

Statistical Analysis

The analysis was performed by the SPSS Statistics Version 24.0 software package (IBM, Armonk, NY). The Shapiro–Wilk test was used to test the normality of the data. Differences between patients and controls were calculated by means of Student's t-test for continuous variables with normally distributed data and by the Mann–Whitney test for the nonparametric variables. Correlations between BCVA and VD and retinal thickness were performed for the XLRS patients and controls with the Pearson's correlation test for parametric data and Spearman's rank correlation test for nonparametric data. All tests were two-sided with P values of 0.05 or less considered statistically significant. 

Results

A total of 34 XLRS patients (aged 9.17 ± 5.46 years old; range, 3.60–30.67 years) underwent an SS-OCT examination; 12 patients were excluded because of retinal detachment, vitreous hemorrhage, or poor cooperation. The OCT and OCTA images were obtained in 43 eyes, in which 14 eyes were excluded from quantitative analysis because the signal strength indexes of their dataset was less than 7. Finally, the OCT datasets of 29 eyes of 22 XLRS patients, 42 eyes of 24 carriers, 33 eyes of 17 patient controls, and 30 eyes of 15 carrier controls were included for quantitative retinal thickness and angiographic analyses. The average age of these 22 patients was 10.17 ± 4.20 years (range, 6.15–21.63 years), and that of the carriers was 34.57 ± 6.57 years old (range, 25.52–45.87 years). The average age of the patient controls was 10.50 ± 4.22 years old (range, 5.11–22.04 years) and that of the carriers was 32.81 ± 6.15 (range, 24.68–47.19 years) (Supplementary Table S1). The BCVA of eyes received quantitative analysis was 0.56 ± 0.19 (Supplementary Table S1). Clinical phenotype analysis and the genetic correlation were presented in the supplementary material, and thicker INL and worse BCVA were presented in type III patients (Supplementary Table S2 and Supplementary Table S4). 

OCT Structural Scan of the Macular in XLRS

High-resolution horizontal line scans including macula and optic disc of 43 eyes of XLRS patients were qualitatively analyzed for the distribution of schitic spaces and other structural malformation. Schitic spaces were seen on SS-OCT in the INL layer in all affected eyes, and the INL persists at the macular fovea (Supplementary Fig. S1A). Twenty-one eyes had schitic spaces in the OPL layer (21/43 [48.83%]) (Supplementary Fig. S1B), 18 eyes had schitic spaces in the ganglion cell layer (18/43 [41.86%]) (Supplementary Fig. S1C), 10 eyes had schitic spaces in the ONL layer (10/43 [23.26%]) (Supplementary Fig. S1D), and none of them showed schisis in the retinal nerve fiber layer (Supplementary Table S3). At the fovea of the XLRS, defects of the interdigitation zone, blurring of the OLM, the structural disturbance of the ellipsoid zone, the disappearance of normal morphology of outer segment of the photoreceptors and myoid zone were universally detected. Peripapillary retinoschisis was shown in 15 patients (15/43 [34.88%]) (Fig. 1A). 

Retinal Thickness, VD, and Foveal Avascular Zone Area Analysis in XLRS Patients

In 29 eyes of 22 XLRS patients with sufficient data for OCT analysis, XLRS patients showed higher thickness of INL (119.16 ± 76.30 µm vs 40.34 ± 2.63 µm; P < 0.000) and OPL (49.94 ± 29.15 µm vs 23.64 ± 2.62; P < 0.000). The thickness of ONL was significantly lower in the XLRS patients than that of the controls (42.15 ± 8.72 µm vs 61.27 ± 4.87 µm; P = 0.000). The thickness of OEIR at the fovea, which implied the thickness of the outer retina at the fovea, was significantly thinner in XLRS patients (55.79 ± 10.95 µm vs 89.41 ± 3.11µm; P < 0.000). There was no difference in the retinal thickness of the RGI layer between patients and controls (Supplementary Fig. S5; Table 1). 

Table 1.

 

Retinal Thickness and VD Measurement in XLRS Patients Compared With Controls

 

The foveal VD of the SCP and DCP were significantly decreased in XLRS patients (Figs. 2A and E) compared with that of the controls (Figs. 2B and F) in the 1.5-mm (22.92 ± 7.14% vs 36.96 ± 4.04%; P = 0.000; 12.48 ± 9.86 vs 28.87 ± 5.90; P = 0.000) and 5-mm (51.24 ± 5.71% vs 57.34 ± 3.09%; P = 0.000; 25.46 ± 9.03 vs 35.56 ± 4.72; P = 0.000) diameter areas. The margin of the foveal avascular zones (FAZs) in these patients were irregular, and the FAZ area of the XLRS patients was significantly larger than that of controls (0.399 ± 0.209 mm² vs 0.219 ± 0.091 mm²; P < 0.000) (Supplementary Figs. S2A–F; Table 1). 

Figure 2.

 

Representative SS-OCTA images of SCP and DCP in different subjects. (A and E) OCTA image of SCP and DCP in XLRS patients showed vessels with irregularities and reduced VD. (B and F) OCTA image of SCP and DCP in patient controls. (C and G) OCTA image of SCP and DCP in RS1 carriers. (D and H) OCTA image of SCP and DCP in carrier controls.

Correlation Analysis Between Retinal Thickness, VD, and BCVA

BCVA was positively correlated with the thickness of the INL (rs = 0.452; P = 0.014) and OPL (rs = 0.643; P < 0.000), and negatively correlated with VD of SCP and DCP in the central foveal region of 1.5 mm diameter (rs = −0.748; P = 0.000; rs = −0.654; P = 0.000) and 5-mm diameter (rs = −0.427; P = 0.001; rs = −0.497; P = 0.001), the thickness of ONL (rs = −0.793; P < 0.000) and OEIR (rs = −0.391; P = 0.04) at the fovea. In addition, BCVA was not correlated with the thickness of the RGI layer (rs = 0.192; P = 0.339) (Table 2). The correlation between retinal thickness and VD was presented in supplemental material (Supplementary Table S5). 

Table 2.

 

Correlation Between Retinal Thickness, VD, and BCVA in XLRS Patients

 

Retinal Thickness, VD, and FAZ Area Analysis in RS1 Mutation Carriers

In 42 eyes of carriers included in the quantitative analysis, the BCVA was −0.01 ± 0.02. The average retinal thickness of OPL was significantly thicker in the carriers than that of the controls (23.30 ± 2.42 µm vs 20.15 ± 3.20 µm; P < 0.000). The thickness of the RGI layer, INL, ONL, and OEIR at the fovea showed no significant difference between the carriers and the controls. The vessel densities in the carriers were not significantly different with the controls (Figs. 2C, G, D, and H; Supplementary Figs. S3A–F; Supplementary Figs. A4A–F). The FAZ area of the carriers was significantly smaller than that of the controls (0.35 ± 0.08 mm² vs 0.42 ± 0.11 mm²; P = 0.005) (Supplementary Fig. 5; Table 3). 

Table 3.

 

Retinal Thickness and VD in the Carriers Compared With the Controls

 

Discussion

The thinning of the ONL and OEIR and decreased VD of SCP and DCP found in our study could be reliable indicators for monitoring and prediction of visual function in XLRS owing to their significant correlations with vision. The universal findings of persistent INL, structural disturbance of the OLM, ellipsoid zone and myoid zone, the disappearance of the interdigitation zone, and the abnormal morphology of the outer segment of the photoreceptors in the fovea could be responsible for the functional abnormality in XLRS instead of the schisis in INL. 

OCT measurements of the macular structure have been used as markers for monitoring the progress of the disease and the treatment effects of XLRS patients using carbonic anhydrase inhibitors. Previous studies indicated that the overall retinal thickness and the severity of the retinal schisis may not be directly influential for visual function, but the remaining viable retinal neurons are.^9,15–17 XLRS patients showed a decrease in the mean kinetic and static perimetry results and decreased b-wave amplitude on electroretinography.¹⁸ We had applied 2% dorzolamide to diminish the schisis in XLRS, and the report showed a significant decrease in foveal thickness with concomitant visual improvement,¹⁵ although other investigators have published differing results.^19,20 The limited value in predicting visual function could be caused by the large variations of the schitic INL and OPL that conceal the correlation between the outer retina and vision. Therefore, high-resolution SS-OCT provides the opportunity to analyze subtle changes in the outer retinal layers that are less affected by schisis. and their correlation with visual damage could be found. Our results are consistent with the report of Bennet et al.⁹ that the outer segment thickness between the ellipsoid zone and the RPE was strongly correlated with BCVA. In another study, Padrón-Pérez et al.²¹ found that the atrophic OLM, ellipsoid zone, and interdigitation zone defects were significantly correlated with worse BCVA through SS-OCT in nine patients. These findings suggested a direct correlation between the outer retinal thickness and visual acuity, which is in line with our findings that the thinning of the ONL and OPL as well as the OEIR were significantly correlated with decreased visual acuity. The disturbance of the ellipsoid zone and myoid zone, the defects of the interdigitation zone, blurring of the OLM, and the disappearance of the normal morphology of the outer segment of the photoreceptors in the fovea were found universally in this study. The interdigitation zone represents the outer segments of photoreceptors and the apical processes of the RPE.²² Because they are the main structural components forming the OEIR, we suggested that these pathological changes could be important indicators for the malfunctioning of the photoreceptors that results in decreased visual acuity.^9,10,17 

Based on the observation in macular OCT images of XLRS patients, we speculated that the deficiency of retinoschisin in the photoreceptors is the primary cause of abnormal development of the retina in XLRS. Before birth, the INL layer is thick with the presence of retinal vessels, whereas the ONL and outer segment are thin because the cones are not developed fully during the fetal state.²³ In the first few postnatal weeks, the cones are significantly elongated, forming the mature fovea with a thicker outer retina and INL migration.²⁴ We also speculated that the deficiency of retinoschisin in the photoreceptors induces not only poorly elongated cones and the malformation of the ONL, but also hindering of the migration of INL, which results in the subsequent INL persistence, thus eventually leads to an abnormally developed visual acuity. These pathological changes are prior to the development of retinoschisis, rather than secondary damage. This notion is also supported by the previous reports of AAV8-RS1gene therapy for RS1 knockout mice. It has been accepted that photoreceptors are the predominant producers of retinoschisin both in humans and in mice.²⁵ Intravitreal injection of AAV-RS1 vectors had no curative effect if it did not diffuse to the outer retina and photoreceptors.²⁶ Previous studies showed that the AAV-RS1 vector specifically targeting on rods had the most prominent rescue effects on retinal structure and ERG results in RS1 knockout mice indicating that the photoreceptor is the key factor in the pathogenesis of the disease.²⁷ Meanwhile, RS1 knockout mice developed coalescence of the borders of the ellipsoid zone and interdigitation zone, which can be salvaged by AAV8-RS1 therapy.²⁸ This finding supports our theory that the observed of interdigitation zone defects could be the result of coalescence of the borders of the ellipsoid zone and interdigitation zone, which is the reflex of malfunction of the cones and rods on OCT images. The malfunction of cones and rods is also consistent with the previous conclusion that the effect of XLRS on rod photoreceptors cannot be ignored through ERG.²⁹ Taken together, RS1 deficiency leads to the malformation of photoreceptors and subsequent persistence of INL. These underlying cellular alterations present as thinning of ONL and distortion of the outer segment structure on SS-OCT, eventually induce decreased visual acuity. 

In contrast with the enlarged FAZ area in XLRS patients, a smaller FAZ area and thickening of the OPL were observed in RS1 carrier compared with their age-matched healthy controls. It has been shown that decreased multifocal ERG and mild structural changes could occur in the carriers.^8,30 We suspect that the retinoschisin could be produced abnormally owing to the lyonization in a certain portion of the carriers. In these individuals, the malformation of the retinoschisin results in a certain degree of malfunction in the bipolar cells that stimulates the growth of the OPL and foveal capillaries owing to the compensatory reaction, which leads to the higher VD in SCP, diminution of FAZ, and the thickening of the OPL in carriers. 

In the present study, the VD of the SCP and DCP were decreased significantly in XLRS patients, which corelated with poorer visual acuity, possibly owing to physical disruption and damage to the enlarged schitic cavity in both the inner and outer retinas.³¹ A previous study using spectral domain OCTA showed decreased VD and FAZ enlargement only in the DCP with a decrease in BCVA.³² The discrepancy could be due to the difference in the ethnicity and the sample size between the two studies.^33,34 Mastropasqua et al.³⁵ found that the VD of SCP and DCP was reduced in XLRS patients, and no difference was shown in carriers in a three-generation family by spectral domain OCT. Their OCTA study showed the interruption of the integrity of the perifoveal anastomotic arcades in SCP and vessel rarefaction in DCP and enlargement of the FAZ in SCP,³⁵ which was consistent with our findings. 

The current study has several limitations, including the limited size of the cohorts and absence of a longitudinal data to monitor the corresponding structural changes with their correlation with visual function, and more detailed structure–function assessments should be performed in further research. There may be a bias because our clinic is a referral center in south west China, and it is possible that patients enrolled in our clinic tend to have a more severe condition. In addition, although there is no statistical significance, it seems that a thicker INL and worse BCVA were present in type III patients according to the OCT classification of XLRS; a significant difference may arise with a larger sample size. The expansion of the cohort and the follow-up data will provide further information regarding the OCT imaging and their functional interpretation. This information may be of value in the treatment, management, and genetic consulting in XLRS patients and RS1 carriers. 

Conclusions

SS-OCT provides detailed imaging for the pathological alteration of the individual retinal layers and capillaries, which could pinpoint the exact location of the damage related to visual impairment for more accurate prognosis and follow-up. The thickness of OEIR representing the outer retina structural alterations are significantly correlated with the visual acuity, which could be used as an objective indicator for monitoring the effectiveness of genetic therapy. 

Acknowledgments

Funding provided by Beijing Bethune Charitable Foundation (HX-H1703040, Science and Technology Support Projects from Sichuan Province of China (Project No.2021YFS0210, 23NSFSC4121, and 2023NSFSC1667), and Tianjin Municipal Education Commission Scientific Funding- 2017KJ214. 

The procedures in our study were approved by the ethics committee of West China Hospital, Sichuan University (2017-472). 

Disclosure: Z. Tao, None; S. Bu, None; L. Liang, None; Y. Yang, None; K. She, None; F. Lu, None