Diagnostic yield of panel-based genetic testing in syndromic inherited retinal disease

Abstract

Thirty percent of all inherited retinal disease (IRD) is accounted for by conditions with extra-ocular features. This study aimed to establish the genetic diagnostic pick-up rate for IRD patients with one or more extra-ocular features undergoing panel-based screening in a clinical setting. One hundred and six participants, tested on a gene panel which contained both isolated and syndromic IRD genes, were retrospectively ascertained from the Manchester Genomic Diagnostics Laboratory database spanning 6 years (2012-2017). Phenotypic features were extracted from the clinical notes and classified according to Human Phenotype Ontology; all identified genetic variants were interpreted in accordance to the American College of Medical Genetics and Genomics guidelines. Overall, 49% (n = 52) of patients received a probable genetic diagnosis. A further 6% (n = 6) had a single disease-associated variant in an autosomal recessive disease-relevant gene. Fifty-two percent (n = 55) of patients had a clinical diagnosis at the time of testing. Of these, 71% (n = 39) received a probable genetic diagnosis. By contrast, for those without a provisional clinical diagnosis (n = 51), only 25% (n = 13) received a probable genetic diagnosis. The clinical diagnosis of Usher (n = 33) and Bardet-Biedl syndrome (n = 10) was confirmed in 67% (n = 22) and 80% (n = 8), respectively. The testing diagnostic rate in patients with clinically diagnosed multisystemic IRD conditions was significantly higher than those without one (71% versus 25%; p value < 0.001). The lower pick-up rate in patients without a clinical diagnosis suggests that panel-based approaches are unlikely to be the most effective means of achieving a molecular diagnosis for this group. Here, we suggest that genome-wide approaches (whole exome or genome) are more appropriate.

Fig. 1. Diagnostic pick-up rate by IRD NGS panel testing in patients with potential syndromic IRD.

Of the 55/106 (52%) patients who were referred with a provisional clinical diagnosis, 71% received a probable genetic diagnosis compared with 25% of those patients with no known clinical diagnosis (51/106).

Fig. 2. Clinical features in HPO terms found in 33 patients referred with a provisional clinical diagnosis of Usher syndrome.

Features in the orange and yellow boxes are the two main features of Usher syndrome that were used as the diagnostic criteria for Usher syndrome in this study. Types and age of onset of hearing impairment in the patients are shown as well as the type of retinal dystrophy and any unusual features found in these patients. The patients are divided by the genetic diagnosis received from the IRD NGS panel test: probable-, possible- and no genetic diagnosis.

Fig. 3. Phenotype–genotype correlation in BBS.

BBS was diagnosed in 8/10 patients who fulfilled the diagnostic criteria or with high index of suspicion. In addition, the NGS panel was able to pick-up two early cases with only two features. BBS1 was causative in half of the diagnosed cases.

Fig. 4

Case study describing the pedigree and clinical features of two unrelated patients with different phenotypes but were found to have the same disease-associated homozygous variant in CEP290 gene causative of their conditions.