Abstracts

Rationale: Broad-based genetic testing methodologies like exome sequencing (ES) allow for identification of rare variants, including copy number variants (CNVs). Professional guidelines and consensus statements support ES as a first line test (J Genet Couns 2023 32(2):266-280). The diagnostic utility of ES has been shown to be significantly greater than chromosomal microarray (CMA). However, for some phenotypes CMA is often ordered prior to or in parallel with ES, due to perceived limitations of CNV calling by ES. Given that up to 15% of patients with epilepsy have been reported to have disease-associated CNVs, we queried a large genetic testing laboratory dataset to assess concordance of CNV detection in individuals receiving both CMA and ES.

Methods:
Under an approved IRB research protocol using deidentified data, we evaluated concordance of CNV detection between ES and CMA, as well as detection of the ClinGen recurrent dosage-dependent microdeletions and duplications approved list (April 2024). We identified a cohort of 11,930 individuals tested from June 2022 to October 2023 who had both CMA and ES. Variants were classified according to ACMG criteria, and clinical notes were assessed for phenotypic characterization.









Results:
For the overall cohort, 591/621 (95.2%) positive CMA findings were also detected by ES. Additionally, ~1% of negative CMA cases had a CNV detected by ES. Overall, 11,770/11,930 (98.7%) cases had equivalent detection or additional CNV findings on ES.




An indication of “seizure” or “epilepsy” was present in 2,043/11,930 (17%) individuals. In this sub-cohort, 111/116 (95.7%) positive CMA findings were also detected by ES. Again, ~1% of negative CMA cases had a CNV detected by ES. For the seizure sub-cohort, 2,015/2,043 (98.6%) cases had equivalent detection or additional CNV findings on ES.






Categories of CNVs more challenging to detect via ES were mosaic, impacted three or fewer exons, and/or fell within the pseudo-autosomal region or on the sex chromosomes in general. ES has sufficient coverage for all 39 ClinGen approved recurrent dosage-sensitive microdeletions and microduplications, and 34/39 were observed in patients.




Conclusions: ES has been shown to have a nearly 3x greater diagnostic yield than CMA. In both a general cohort and seizure-specific sub-cohort, these data suggest highly concordant detection of CNVs in ES vs CMA, including ES detection of CNVs in challenging categories and detection of common microdeletions/duplications. Given that all genetic tests have technical limitations, ES could be considered first due to its ability to identify CNVs at high concordance with CMA, CNVs below the reporting threshold of CMA, plus a host of other variant types (e.g., SNVs, UPD, indels, MEIs).

Funding: None