Genotype-Phenotype Correlations of COL2A1 and COL11A1 Patients

To describe and compare the ophthalmologic and extraophthalmologic features of patients with Stickler syndrome because of pathogenic variants in COL2A1 and COL11A1.

Retrospective cross-sectional study nested in a multicentric cohort study.

Records of patients with a confirmed molecular diagnosis of Stickler syndrome followed up in the ophthalmology department at Necker-Enfants Malades and Cochin University hospitals (Paris) between 2016 and 2024 were retrospectively reviewed. Demographic data, clinical findings from ophthalmologic examination, and extraophthalmologic features were recorded. Patients with an incomplete file, lack of genetic evidence despite a compatible clinical phenotype, and those presenting with rare variants were excluded.

Among 110 patients with confirmed Stickler syndrome, 90 (82%) had a COL2A1 variant and 20 (18%) a COL11A1 variant. The median age at last follow-up was 24.4 years (IQR 0.9-77.6), and the median follow-up duration was 10.8 years (IQR 4.2-27.3). Retinal detachment occurred in 50% of patients with COL2A1 variants (45/90) and 45% of those with COL11A1 variants (9/20), with no statistically significant difference between groups (P = .81). Twenty-four patients (22%) had a bilateral retinal detachment with a median time for a retinal detachment of the fellow eye of 3.0 (IQR 0-25.2) years. The patients with COL11A1 variants were significantly different from those with COL2A1 variants in terms of deafness frequency (50% [10/20] vs 13% [12/90]; P = .005; 95% CI 13.7%, 59.7%), axial lengths (28.9 ± 3.2 mm vs 26.3 ± 2.3 mm; P < .001; 95% CI +1.20, +4.02), age at retinal detachment onset (9.8 [IQR 5.3-19.7] years vs 13.3 [IQR 0.5-66] years; P = .006; 95% CI -12.44, -2.19), and median time to retinal detachment in the fellow eye (2.0 [IQR 0.2-3] years vs 4.5 [IQR 0-25.2] years; P = .009; 95% CI -7.99, -1.30).

The study highlights phenotypic difference between COL2A1- and COL11A1-related Stickler syndrome, with COL11A1 variants potentially associated with more severe ocular phenotype. Such genotype-phenotype correlations may contribute to refining patient management and guiding prophylactic interventions. These findings could support individualized follow-up strategies; however, confirmation in larger cohorts is warranted.