Characterizing Retinal Sensitivity and Structure in Congenital Stationary Night Blindness: A Combined Microperimetry and OCT Study

Abstract

Purpose: To investigate the characteristics of microperimetry and optical coherence tomography (OCT) in congenital stationary night blindness (CSNB), as well as their structure–function association.

Methods: This cross-sectional study included 32 eyes from 32 participants with CSNB, comprising 18 with complete CSNB and 14 with incomplete CSNB, along with 36 eyes from 36 CSNB-unaffected controls matched for age, sex, and spherical equivalent. Using MP-3 microperimetry, central retinal sensitivity was assessed within a 20° field, distributed across six concentric rings (0°, 2°, 4°, 6°, 8°, and 10°). OCT was used to analyze retinal and choroidal thickness. The study aimed to assess the overall and ring-wise retinal sensitivity, as well as choroidal and retinal thickness in CSNB and CSNB-unaffected controls, with a secondary focus on the relationship between retinal sensitivity and microstructural features on OCT.

Results: In comparison with CSNB-unaffected subjects, the overall and ring-wise retinal sensitivity as well as choroidal thickness were reduced in patients with CSNB (P < 0.001). Moreover, the central sensitivity in incomplete CSNB group was lower than in complete CSNB group (25.72 ± 3.93 dB vs. 21.92 ± 4.10 dB; P < 0.001). The retinal thickness in the CSNB group was thinner outside the fovea compared with the CSNB-unaffected group. Multiple mixed regression analyses revealed that point-to-point retinal sensitivity was significantly correlated with BCVA (P = 0.002) and the corresponding retinal thickness (P = 0.004).

Conclusions: Examination of retinal sensitivity and OCT revealed different spatial distribution profiles in CSNB and its subtypes. In CSNB eyes, retinal sensitivity on microperimetry was associated with retinal thickness on OCT.

Congenital stationary night blindness (CSNB) is a nonprogressive retinal disorder characterized by night blindness, variably decreased visual acuity, refractive error (usually myopia, and occasionally hyperopia), strabismus, nystagmus, and photophobia.^1–4 Initially, Schubert and Bornschein diagnosed CSNB through the identification of its distinctive negative ERG pattern.⁵ Subsequently, Miyake et al.⁶ further subclassified Schubert–Bornschein CSNB into two main subtypes—complete CSNB (cCSNB) and incomplete CSNB (iCSNB)—based on psychophysical and ERG characteristics. Patients with cCSNB exhibit no residual rods’ responses, whereas those with iCSNB demonstrate decreased but recordable rod responses. 

Historically, ERG has been instrumental in differentiating these two CSNB subtypes, predominantly reflecting the overall retinal function. However, there is a gap in the literature regarding the detailed spatial distribution of retinal function, which is crucial for understanding the severity and progression of the disease and cannot be reflected fully by best-corrected visual acuity alone. In recent years, microperimetry has emerged as an innovative methodology for investigating the spatial retinal sensitivity in retinal disorders.^7–9 Previous investigations using the MP-1 microperimetry have indicated reduced retinal sensitivity in CSNB,¹⁰ albeit within a limited stimulus strength range (1–20 dB), prone to floor and ceiling effects, restricted in-depth disease analysis. In contrast, newer devices like MP-3, offer a broader range of visual sensitivity testing (0–34 dB),^9,11 allowing for more precise assessment of retinal impairments. This technique, using fundus tracking, enables precise mapping of retinal function, correlating it with structural characteristics obtained via optical coherence tomography (OCT), thereby providing deeper insights into the disease.¹² However, there is a relative scarcity in current research regarding the distinct microperimetric profiles of cCSNB and iCSNB subtypes and their correlation with OCT structural findings. Therefore, this study aimed to evaluate and compare macular sensitivity in patient with iCSNB and patient with cCSNB with CSNB-unaffected participants and to explore its correlation with OCT structures, thereby offering more comprehensive quantitative data to support the clinical diagnosis and treatment of CSNB. 

Methods

Study Population

This cross-sectional study research was conducted at the Refraction and Juvenile Myopia Prevention Clinic of Zhongshan Ophthalmic Center. From February 2023 to January 2024, patients diagnosed with CSNB through full-field ERG and molecular testing were recruited consecutively. Participants unable to cooperate with the procedures of microperimetry, OCT, and fundus examination were excluded from the study. Additionally, 36 CSNB-unaffected subjects matched for age, sex, and spherical equivalent were selected from the same center as a control group. The inclusion criteria for the control group were a spherical equivalent refractive error ranging from +2.00 D to −16.00 D and a corrected distance visual acuity of 0.1 logMAR (20/32) or better. Subjects with a history of ocular trauma or disease, previous ocular or refractive surgery, diabetes, or any systemic disease that could potentially affect the eyes were excluded. This research was approved by the Ethics Committee of Zhongshan Ophthalmic Center, Sun Yat-Sen University and conducted in accordance with the Declaration of Helsinki. Informed consent was obtained from all participants. In the case of minors, informed consent was obtained from their guardians. 

Clinical Examination

All patients underwent a comprehensive ophthalmic examination, including slit-lamp examination, refraction, best-corrected visual acuity, and ocular biometric parameters. Cycloplegic refraction was performed for pediatric patients, and subjective manifest refraction was used for adults. Refractive errors were documented conventionally and then transformed into spherical equivalents, calculated as sphere +1/2 cylinder. Ocular biometric parameters, including axial length were obtained using the IOLMaster700 (Carl Zeiss Meditec, Dublin, CA). 

Microperimetric Examination

All patients underwent retinal sensitivity analysis using microperimetry (MP-3, Nidek, Hiroshima, Japan). Examinations were conducted in darkness (ambient illuminance, 1 lux) without the use of mydriatic agents. Before the assessment, all patients underwent a fast test to familiarize themselves with the equipment, thereby minimizing the learning curve effect. An assessment of retinal sensitivity thresholds was conducted within a 20° diameter around the macular region, covering the central retina (1° = 300 micrometers; hence, 10° = 3000 micrometers, 20° = 6000 micrometers). The standard 4–2 staircase strategy was used, with a dynamic range of 0 to 34 decibels(dB) for each stimulus. A custom concentric pattern consisting of 33 test locations was evaluated, arranged at 2° intervals in a circular area centered on the macular fovea, specifically at 0°, 2°, 4°, 6°, 8°, and 10° from the central point. The stimuli were presented as white test lights (stimulus size Goldmann III, duration 200 ms), projected onto the corresponding locations on the retina against a dark white background with a luminance of 31.4 asb. After the examination, fundus photography was overlaid with infrared images for precise alignment. The average ring-wise sensitivity was calculated by averaging the total projected stimuli at distances of 0°, 2°, 4°, 6°, 8°, and 10° from the foveal target point (Fig. 1). 

Figure 1.

 

Retinal sensitivity distribution in the CSNB and control groups. Representative mean retinal sensitivity (dB) mapping on fundus photographs of eyes from the CSNB-unaffected control, cCSNB, and iCSNB groups. The mean retinal sensitivity of the overall, 0°, 2°, 4°, 6°, 8°, and 10° apart from the macular center of eye is attached. SE, spherical equivalent (diopter).

In a single session, the right eyes of participants with CSNB underwent two consecutive examinations. If the right eye was not eligible, the left eye was then examined. All participants had several minutes of rest between each test session to minimize the potential impact of fatigue on the assessments. To investigate the structural-functional association in the macular region, point-wise sensitivity (PWS) analysis was measured at specific retinal points corresponding to the designated locations on OCT images. Along the horizontal meridian, PWS was assessed at 2°, 6°, and 10° (1° = 300 µm) on both nasal and temporal sides. Vertically, PWS was measured at 2°, 6°, and 10° both superiorly and inferiorly. This allowed for a direct comparison of functional sensitivity data with the underlying structural OCT data at matching retinal coordinates. 

The centrality and stability of fixation were assessed using fixation indices and the bivariate contour ellipse area (BCEA). The fixation indices for 2° and 4° represent the percentage of fixation points within a circle with diameters of 2° and 4°, respectively. BCEA68, which accounts for 68% of fixation points and is automatically calculated by the MP3 microperimetry, was used for a more precise evaluation of the fixation pattern. 

OCT Scan

Choroidal thickness measurements were performed using a swept-source OCT (VG200, SVision Imaging, Ltd, Henan, China). Centered on the fovea, line scans were performed along horizontal and vertical directions. Retinal thickness and choroidal thickness measurements were conducted at positions corresponding to the central retina (0°) and at nasal, temporal, superior, and inferior locations relative to the center. These measurements were taken at the distance of 600 µm, 1800 µm, and 3000 µm, corresponding with the microperimetric assessment points at 2°, 6°, and 10° (with 1° equating to 300 µm) (Fig. 2). A built-in caliper in the instrument facilitated the measurement of retinal and choroidal thickness. Retinal thickness was measured manually as the vertical distance between the inner limiting membrane and the outer boundary of the highly reflective line of the retinal pigment epithelium, and choroidal thickness was measured manually as the vertical distance between the outer boundary of the highly reflective line of the retinal pigmental epithelium and the choroid–sclera junction. 

Figure 2.

 

Protocol of microperimetry and OCT pattern. (A–D) Infrared reflectance (IR) image and SS-OCT of horizontal and vertical directions were acquired during OCT imaging. Yellow solid triangles in C and D represent manual measurements at 14 points: HSF (horizontal central subfovea), N1 (600 µm nasal), N2 (1800 µm nasal), N3 (3000 µm nasal), T1 (600 µm nasal), T2 (1800 µm nasal), T3 (3000 µm nasal), VSF (vertical central subfovea), S1 (600 µm superior), S2 (1800 µm superior), S3 (3000 µm superior), I1(600 µm inferior), I2(1800 µm inferior), and I3(3000 µm inferior). (E) Microperimetry image containing 33 regional points, arranged at 2° intervals in a circular area centered on the macular fovea, specifically at 0°, 2°, 4°, 6°, 8°, and 10° from the central point (1° = 300 µm; 10° = 3000 µm).

Statistical Analysis

Continuous variables were first tested for normality using the Shapiro–Wilk normality test. Bland–Altman analysis was used to evaluate test-retest repeatability of overall mean macular sensitivity, calculating the 95% repeatability coefficient. The intraclass correlation coefficient was used to assess the consistency of the measurements between tests. 

The between-group comparisons were conducted between the CSNB group and CSNB-unaffected controls, with subset comparisons within the CSNB group being performed using t tests if the data followed a normal distribution; if not, the Mann–Whitney U test was used. Additionally, to identify factors associated with retinal sensitivity, univariate and multivariable mixed model analyses were carried out. Only variables with a P value of <0.10 in univariate analysis were included in the multivariable analysis. All statistical analyses were performed using SPSS Statistics version 22.0 (IBM, Chicago, IL), and a P value of <0.05 was considered statistically significant. 

Results

Demographics

The study included 32 patients with CSNB (26 male, 6 female) and 36 CSNB-unaffected controls matched for age, sex, and spherical equivalent. The demographic and ocular characteristics of the patients were delineated in Table 1. Genetic defects were identified in all individuals with CSNB, with variants found in NYX (10 cases), GRM6 (5 cases), TRPM1 (2 cases), LRIT3 (1 case), and CACNA1F (14 cases). Individual gene variants were given in Supplementary Table S1. Based on ERG examinations, 18 patients were categorized as cCSNB and 14 as iCSNB (Fig. 3). In the cCSNB, scotopic ERG to dim flashes was undetectable, with electronegative waveforms to bright flashes. Photopic ERG displayed normal amplitude. In the iCSNB, scotopic dim flash ERG showed subnormal amplitude and an electronegative waveform to bright flashes. Photopic responses were affected more severely, with a markedly subnormal and delayed ERG. Compared with the CSNB-unaffected control group, the CSNB group exhibited significantly decreased visual acuity, with the iCSNB subgroup presenting worse visual acuity than the cCSNB subgroup. Regarding refractive error, myopia was observed universally in the cCSNB subgroup, whereas the iCSNB subgroup exhibited a mix of myopia and hyperopia. Additionally, the spherical equivalent values were markedly more negative in the cCSNB subgroup than in the iCSNB subgroup (Table 2) (cCSNB group, −9.89 ± 2.64D; iCSNB, −4.28 ± 3.91D; P < 0.0001). 

Table 1.

 

Patients’ Demographics and Mean Retinal Sensitivity, Fixation Stability and OCT Parameters in Patients With CSNB and CSNB-Unaffected Subjects

 

Figure 3

 

Representative full-field electroretinography waveforms of CSNB-unaffected, cCSNB, and iCSNB patients.

Table 2.

 

Patient Demographics and Mean Retinal Sensitivity and OCT Parameters in Patients With cCSNB and iCSNB

 

Microperimetry

Among the 32 CSNB participants, 20 (62.5%) underwent the testing twice at baseline. The Bland–Altman analysis was used to calculate the 95% repeatability coefficient by comparing the mean overall retinal sensitivity obtained from the first and second tests (test 1 and test 2), with the value of 1.89 (95% CI, −2.1 to 1.7) dB. The Bland-Altman plot indicated no significant proportional biases (Supplementary Fig. S1). And the intraclass correlation coefficient was 0.925 (95% confidence interval [CI], 0.82 to 0.969), suggesting high consistency between the two tests. In the CSNB-unaffected group, the 95% repeatability coefficient for the mean overall retinal sensitivity was 0.73 dB (95% CI, −0.95 to 0.51), with an intraclass correlation coefficient of 0.947 (95% CI, 0.812–0.986). 

Microperimetric assessment of retinal sensitivity within the central 6-mm area showed that patients with CSNB exhibited significantly lower retinal sensitivity in each concentric zone compared with CSNB-unaffected subjects (Table 1). Additionally, upon analyzing the sensitivity distribution across concentric rings, a marked disparity was noted between the cCSNB and iCSNB groups, predominantly in the macular central area (0°) (25.72 ± 3.93 dB vs. 21.92 ± 4.10 dB; P < 0.001) (Table 2, Fig. 1). In contrast, the remaining ring areas did not exhibit significant variances in sensitivity (Table 2). The spatial distribution of PWS also reflected similar results (Fig. 4A). In addition, the fixation was more unstable in the CSNB group than that in the control group. 

Figure 4.

 

(A) PWS measured at seven points along the horizontal meridian (N1, N2, N3, HSF, T1, T2, and T3) and seven points along the vertical meridian (S1, S2, S3, VSF, I1, I2, and I3) within the microperimetry. (B, C) Point-wise retinal and choroidal thickness measurements at specific locations corresponding to the microperimetry points.

OCT Measurement

Overall, the retinal thickness in the CSNB group was thinner than that in CSNB-unaffected group outside the fovea (Table 1, Fig. 4B). There was almost no difference in retinal thickness between the cCSNB and iCSNB groups (Table 2). The choroid thickness at all measurement points were thinner in CSNB group compared with the CSNB-unaffected group (Fig. 4C). Horizontally, a gradual thinning trend of the choroid thickness from the temporal to the nasal side was displayed in the CSNB-unaffected, cCSNB, and iCSNB groups. Vertically, a diminishing trend in choroidal thickness from superior to inferior was seen across all groups. 

Association of OCT and Microperimetry

The multivariable mixed regression analysis was used to explore factors influencing PWS. As shown in Table 3, it was found that the PWS was correlated with best-corrected visual acuity (coefficient, −8.171; 95% CI, −13.022 to −3.319; P = 0.002) and the retinal thickness (coefficient, 0.016; 95% CI, 0.005–0.027; P = 0.004). 

Table 3.

 

Mixed Linear Regression Analyses to Determine Association of Retinal Sensitivity With Clinical and Microstructural Features on OCT

 

Discussion

To the best of our knowledge, our study was the first to apply OCT combined with microperimetry for concurrent assessment of the structural and functional features in patients with CSNB, providing a novel insight into this field. In this research, we included patients with both complete and incomplete types of CSNB, as well as CSNB-unaffected controls. A key discovery of our research was the distinctive spatial distribution pattern of retinal sensitivity observed in patients with CSNB. Compared with the CSNB-unaffected group, there was a general decrease in retinal sensitivity in patients with CSNB, affecting both overall and ring-wise sensitivity. Additionally, central (0°) retinal sensitivity in the iCSNB subtype was found to be lower than in the cCSNB subtype by approximately 3.8 dB (25.72 ± 3.93 dB vs. 21.92 ± 4.10 dB; P < 0.001). Thinning of the retinal and choroidal thickness was observed in the CSNB group compared with CSNB unaffected group. Point-to-point association on OCT and microperimetry suggested that microstructural changes correspond with functional alterations. 

In our study, patients with CSNB demonstrated significantly reduced visual acuity compared with CSNB-unaffected controls (P < 0.001, Table 1). Notably, visual acuity in patients with cCSNB was better than in patients with iCSNB, consistent with findings from the previous literature.^1,3,13 This discrepancy in visual acuity might be attributed to genetic differences associated with the two subtypes. Since 1999, several genes linked to CSNB have been identified, with cCSNB being associated with five genes including NYX, TRPM1, LRIT3, GRM6, and GPR179,^3,13–15 whereas iCSNB is related to CACNA1F and CABP4.^4,15 The study by Bijveld et al.¹ further clarified that the distinctions between cCSNB and iCSNB arose from the varying functions and cellular localizations of the affected proteins. In cCSNB, gene mutations impact the function of proteins located on the dendrites of ON bipolar cells,^16–18 whereas the OFF bipolar cell pathway remains intact. Although the proteins encoded by the iCSNB genes are located at the synaptic terminals of photoreceptors, they are involved in sustained calcium-dependent neurotransmitter release.¹⁹ Thus, both rod and cone cell signal transmissions, as well as the ON and OFF pathways, are affected in iCSNB. In humans, rod cells are primarily connected to ON bipolar cells, whereas cone cells connect with both ON and OFF bipolar cells. Therefore, patients with cCSNB primarily exhibit issues related to rod cells, whereas patients with iCSNB experience impairments in both rod and cone cells. 

The differences in the causative genes of CSNB subtypes not only influence visual acuity, but also manifest in microperimetry outcomes. Microperimetry, as a tool for measuring visual function, is applied extensively and exhibits good –-retest reliability in retinal diseases.²⁰ Previous studies have indicated that the repeatability coefficients are approximately 1.2 dB in healthy controls and range from 1.4 to 2.0 dB in patients with macular diseases,^20,21 respectively. Our study corroborates a similar overall test–retest repeatability for CSNB genotypes, with repeatability coefficients of 1.89 dB for patients with CSNB and 0.73 dB for healthy subjects. Concurrently, microperimetry assessments facilitate fundus imaging while projecting light stimuli onto test points. The precise spatial characterization of visual function achieved through fundus-controlled perimetry offers valuable insights into disease features, severity, and progression, which could not be reflected by best-corrected visual acuity. A prior study using MP-1 microperimetry to evaluate the retinal sensitivity of 11 patients with CSNB found a significant decrease in average retinal sensitivity compared with healthy subjects across all macula, especially in the outermost ring (3–6 mm).¹⁰ However, owing to the limited sample size, comparisons between cCSNB and iCSNB subtypes were not conducted. In our study, we used the more advanced MP-3 microperimetry device, which offers a dynamic testing range from 0 to 34 dB, surpassing the 0- to 20-dB range of the earlier MP-1 microperimetry. Our findings indicated that ring-wise sensitivity in patients with CSNB was markedly lower in comparison with those unaffected by CSNB. Interestingly, in the iCSNB subgroup, we noted that the central foveal sensitivity was decreased compared with the cCSNB subgroup (25.72 ± 3.93 dB vs. 21.92 ± 4.10 dB; P < 0.001), although no notable differences were observed in other concentric regions. This phenomenon could be attributed to the differential involvement of the visual systems in the two subtypes. It is well-documented that cCSNB typically manifests nearly normal photopic responses in ERG, whereas iCSNB exhibits a substantial decrease in amplitude, indicating compromised photopic responses associated with cone system functionality.⁵ Bijveld et al.¹ have delineated the distinctions between these subgroups, noting that in the patients with iCSNB, both rod and cone pathways exhibit dysfunction, whereas in the patients with cCSNB, only the rod pathway is compromised. Furthermore, anatomically, the fovea, characterized by the highest density of cone cells, may exhibit decreased visual sensitivity in the central foveal region in cases of iCSNB, where the cone system is affected. 

OCT scans of the macular region in patients with CSNB, in both horizontal and vertical orientations, predominantly revealed myopic changes. These changes were characterized by a decrease in choroidal thickness and the development of posterior staphyloma in some patients. Notably, in both iCSNB and cCSNB groups, the retinal and choroidal thicknesses were found to be thinner compared with the CSNB-unaffected group (Fig. 4). Although age and refractive error were matched between the control and CSNB groups, the thinning of the retinal and choroidal structures in the CSNB group suggested that these changes may be attributed to the disease itself. In human subjects with cCSNB as well as animal models, retinal morphology is generally considered preserved and stable. However, subtle cellular changes have been observed in mice lacking Lrit3 and Trpm1.^14,22,23 Several studies have documented the thinning of the inner nuclear layer in cases of cCSNB, as measured by spectral-domain OCT.^24,25 This thinning extends to the inner plexiform layer, ganglion cell layer and nerve fiber layer, as observed in Lrit3^−/− mice model.²⁶ Notably, the pattern of this thinning in cCSNB differs from the changes seen in progressive rod–cone dystrophies.²⁷ Unlike progressive rod–cone dystrophies with notable retinal degradation, patients with CSNB generally maintain intact retinal structures, indicative of the condition's nondegenerative nature. Future studies should concentrate on detailed quantitative segmentation of each retinal layer to elucidate the impact of different genotypes on the structure of the retina. 

Previous assessments of structure–function correlations have been conducted in various retinal diseases through microperimetry sensitivity maps and point-to-point matching with OCT imaging,^12,28,29 which could enhance our understanding of underlying pathophysiology. This study evaluated the correlation between OCT microstructural changes and retinal sensitivity in patients with CSNB. OCT B-scans provided detailed insights into retinal microstructure. Patients with CSNB typically maintain intact retinal structure in previous studies, especially when compared with patients with progressive cone-rod dystrophy. Point-to-point matching analysis revealed higher PWS was associated with better best-corrected visual acuity and a thicker retinal thickness. Future advancements in wide-angle retinal imaging devices will aid in a deeper understanding of the pathogenic mechanisms. 

There are two primary limitations to this study. First, owing to the cross-sectional study design, we cannot provide prognostic information about the progression of CSNB. Further longitudinal studies will be crucial to assessing changes of structure and function in patients with CSNB. Second, the relatively small sample size may encumber the generalizability and statistical robustness of our findings. Within the rare disease paradigm, however, the sample size in this study stands as comparatively large, urging subsequent research endeavors to amplify sample dimensions for bolstering result reliability and authentication. 

In conclusion, this study provides information on spatial distribution of retinal sensitivity in cCSNB and iCSNB, with a significant reduction noted at the fovea in iCSNB. By elucidating the relationship between structure and function, robust structural biomarkers will be identified to determine disease severity. Future investigations could benefit from incorporating diverse microperimetry techniques, such as dark adaptation scotopic microperimetry and color stimulus microperimetry, to enhance the comprehensive understanding of the underlying pathophysiology of CSNB. 

Acknowledgments

Supported in part by grants from the National Natural Science Foundation of China [grant numbers 30309010012226, 82371089]. 

Disclosure: M. Yu, None; W. Hao, None; M. Wang, None; Z. Ruan, None; Z. Li, None; C. Xiang, None; L. Wang, None; Y. Hu, None; X. Yang, None