Abstract
Purpose :
To optimize variant classification for the EYS gene, population studies were performed using the PRO-EYS natural history study. Gene and variant models were generated to demonstrate disease-enriched alleles and genotype-phenotype correlations.
Methods :
Ninety-six unique variants were reported in clinical reports of 104 probands with EYS-related degeneration. Variants were evaluated by standard variant interpretation criteria and compared to public submissions in ClinVar. Internal allele frequencies (AF) calculated for the PRO-EYS cohort were compared with a general population database (gnomAD) by Fisher exact test with Bonferroni correction. Variant characteristics were analyzed against the ophthalmological imaging data including the total visual field hill of vision (VTOT).
Results :
Predicted loss of function (pLoF) variants represented 80% of the unique variants, including 15 exonic del/dup and 62 nonsense/frameshift/canonical splicing variants. The c.403_423delinsCTTTT:p.(Thr135LeufsTer26) variant was most prevalent with 9% cohort AF. 44% of variants were found to be statistically enriched compared with the general population, improving classification of 11 nonsense/frameshift variants in the terminal exon to pathogenic. Among the 11 missense/inframe indel variants, four were reclassified as pathogenic or likely pathogenic by AF enrichment. For PRO-EYS participant genotypes, two pLoF alleles were found in 82%. Median VTOT at baseline visit for those with 2 pLOF variants tended to be lower than those with one or no pLoF allele (Wilcoxon test; p=0.07).
Conclusions :
The PRO-EYS cohort included a larger proportion of pLoF variants and patients with biallelic pLoF alleles than previously reported. The application of AF enrichment in the PRO-EYS cohort was essential for reclassifying missense variants and truncating variants predicted to produce functional protein. Patient subgroup analyses stratified by genotypes provided novel insights on patient visual function. Additional cohort-level analyses of gene-specific disorders will further improve clinical variant interpretation for retinal degenerations.
This abstract was presented at the 2023 ARVO Annual Meeting, held in New Orleans, LA, April 23-27, 2023.