Determinants of diagnostic yield in a multi-ethnic Asian inherited retinal disease cohort

Abstract

As the discovery of new genes causing inherited retinal disease (IRD) has plateaued, we look to other factors which could be used to maximize diagnostic yield. We analyzed whole-exome sequencing (WES) data from 506 IRD probands, focusing on the interplay between diagnostic yield, age of symptom onset or diagnosis, family history, and initial clinical diagnosis. The cohort's overall diagnostic yield was 49.2%. Diagnostic yield was negatively correlated with the age of symptom onset and positively correlated with the number of affected family members. Diseases with distinctive clinical presentations such as Bietti crystalline dystrophy (BCD) or Leber congenital amaurosis (LCA) were more reliably diagnosed than more common and heterogeneous diseases like retinitis pigmentosa (RP) and cone-rod dystrophy (CRD). Recurrent genes and variants in this Chinese-majority cohort resemble those found in Chinese cohort studies but differ from populations of European descent, with implications for the design and prioritization of gene therapies. These insights may help optimize the diagnostic utility of genetic testing for IRDs, enhance the delivery of genetic counseling for patients, and guide the development of more inclusive targeted therapies.

Fig. 1. Diagnostic yield plotted against age of diagnosis and number of affected family members.

A distribution of age of diagnosis/onset in patients who received molecular diagnosis (positive) and not (negative). Median age of diagnosis of the population is represented by horizontal line at 31. P-value is calculated using the Wilcoxon rank-sum test. B The diagnostic yield of the IRD cohort grouped by age of diagnosis. Numbers within each bar represent the total number of patients within the group. C Diagnostic yield of patients grouped by the self-reported number of family members with similar phenotypes. Numbers within boxes are the number of patients within each group.

Fig. 2. Diagnostic yield and age of diagnosis in patients grouped by their clinical phenotype.

A Overall diagnosis yield based on phenotype. The same abbreviations are used throughout the other figures in this paper. BBS Bardet-Biedl syndrome; BCD - Bietti crystalline corneoretinal dystrophy; LCA Leber congenital amaurosis; RS retinoschisis; STGD Stargardt and Stargardt-like retinal dystrophy, including ABCA4-related phenotypes; VD Vitelliform dystrophy; USH Usher syndrome; MD macular dystrophy; CRD cone or cone-rod dystrophy; RP retinitis pigmentosa. Other diagnoses with less than 5 patients are grouped together as ‘Others’ for clarity. B Age of diagnosis distribution based on molecular diagnosis.

Fig. 3. Variant frequency in the five most common pathogenic genes and others of note.

Singular variants are only observed in one patient.

Fig. 4. Possible refinement of initial clinical diagnosis based on genotyping information.

From left to right: initial clinical diagnosis, molecular diagnosis, and proposed refined diagnosis.