A Classification System Using Optical Coherence Tomography for X-linked Retinoschisis and Its Association with Genetic Variants

Abstract Background To develop and apply a novel optical coherence tomography (OCT)-based classification system for X-linked retinoschisis (XLRS) and evaluate its association with best-corrected visual acuity (BCVA) and RS1 genotypes. Methods In this retrospective cohort study, 72 patients (139 eyes) with molecularly confirmed XLRS from a tertiary referral center were analyzed. Spectral-domain OCT images were classified into five phenotypic types (1A, 1B, 2, 3, 4, and 5A, 5B) based on foveoschisis configuration and outer retinal layer integrity. Associations between OCT types and BCVA, age, and genotype (truncating vs. missense) were evaluated. Results A total of 139 eyes were analyzed (mean age 15.5 ± 9.3 years). BCVA ranged from 2.0 to 0.18 logMAR (mean 0.68 ± 0.39). Type 1A had the best BCVA, significantly outperforming Types 2, 3, 4, and 5B (p<0.01). Type 5A showed comparable BCVA to 1A and outperformed all other types except 1A and 1B. Types 3, 4, and 5B exhibited worse BCVA, with complete outer retinal layer loss. No significant BCVA or age differences were observed between truncating and missense variants (p=0.692, p=0.559). However, truncating variants were associated with older age within Type 1 (OR = 1.13, 95% CI: 1.02–1.24, p=0.016), suggesting a slower but progressive visual decline compared to missense variants. Conclusions This OCT-based classification reflects structural severity and correlates with visual function in XLRS. It supports phenotype-based risk stratification and may guide patient selection and monitoring in gene therapy trials for XLRS.