RNA-guided clarity: The potential for resolving variant uncertainty in clinical exome sequencing

Although RNA analysis holds potential to increase the diagnostic yield of exome sequencing (ES) for rare disease, its practical utility and barriers to its large-scale implementation have not been well studied. To quantify the impact of RNA analysis in clinical ES, we retrospectively assessed variants reported over a nearly 5-year period at a single clinical laboratory.

Reported variants and corresponding genes were assessed based on 4 eligibility criteria for RNA analysis at our laboratory: (1) suspected spliceogenic impact, (2) sufficient gene expression in blood, (3) characterized gene with at least moderate gene-disease validity, and (4) mechanism of disease is loss of function (LOF).

A total of 1987 unique variants were reported on ES during the study period. In total, 11.4% (226/1987) were putative splicing variants, 68.2% (1355/1987) had sufficient gene expression in blood, 83.0% (1650/1987) were in a gene with a LOF mechanism, and 98.9% (1965/1987) occurred in characterized genes. When assessed on the 4 criteria combined, 7.6% (152/1987) of variants were eligible for RNA analysis. Of those variants, which were identified in 153 individuals, 35.5% (54/152) were variant of uncertain significance (VUS), representing 5.2% (54/1035) of all VUS reported. Nine cases were successfully recruited for retrospective RNA analyses, 7 of whom had VUS and 2 who had a likely pathogenic variant. Of the VUS, 71.4% (5/7) were upgraded, whereas 28.6% (2/7) remained a VUS. The 2 likely pathogenic variants remained likely pathogenic.

These findings indicate that putative splicing variants represent a substantial portion of VUS in this clinical ES cohort, and most were identified in characterized LOF genes with RNA expression in blood. Integrating RNA analysis with clinical ES could clarify over 5% of reported VUS, even when accounting for known limitations in a rare disease setting, making it a viable strategy to increase diagnostic accuracy and the likelihood that a patient will receive a clinically meaningful report.