X-linked Retinoschisis

PURPOSE: To describe the natural course, phenotype and genotype of patients with X-linked retinoschisis (XLRS).

DESIGN: Retrospective cohort study.

PARTICIPANTS: Three hundred forty patients with XLRS from 178 presumably unrelated families.

METHODS: This multicenter, retrospective cohort study reviewed medical records of XLRS patients for medical history, symptoms, visual acuity (VA), ophthalmoscopy, full-field electroretinography and retinal imaging (fundus photography, spectral-domain optical coherence tomography (SD-OCT), fundus autofluorescence).

MAIN OUTCOME MEASURES: Age at onset, age of diagnosis, severity of visual impairment, annual visual decline, electroretinogram and imaging findings.

RESULTS: In total, 340 patients were included with a mean follow-up time of 13.2 years (range, 0-50.1 years). The median age to reach mild visual impairment and low vision was 12 and 25 years, respectively. Severe visual impairment and blindness were predominantly observed in patients above 40 years old, with a predicted prevalence of 35% and 25% at the age of 60, respectively. The VA increased slightly in the first two decades of life, and subsequently transitioned in to an average annual decline of 0.44% (P < 0.001). There was no significant difference in decline of VA between variants that were predicted to be severe and mild (P = 0.239). The integrity of the ellipsoid zone (EZ) as well as the photoreceptor outer segment (PROS) length in the fovea on SD-OCT significantly correlated with visual acuity (Spearman’s ρ = -0.604, P < 0.001 and Spearman’s ρ = -0.759, P < 0.001; respectively). Fifty-three different RS1 variants were found. The most common variants were the founder variant c.214G>A (p.(Glu72Lys) (101 subjects, 29.7%) and a deletion of exon 3 (38 subjects, 11.2%).

CONCLUSION: A large variability in phenotype and natural course of XLRS was seen in this study. In most patients, XLRS showed a slow deterioration starting in the second decade of life, suggesting an optimal window of opportunity for treatment within the first three decades of life. The integrity of EZ as well as the PROS length on SD-OCT may be important in choosing optimal candidates for treatment, and as potential structural endpoints in future therapeutic studies. No clear genotype-phenotype correlation was found.