O1012: Variant Reclassification Among 308 Children Undergoing Genetic Testing in a Large Tertiary Center for Heritable Arrhythmia Syndromes

Background: Recent expansion of genetic testing in cardiac heritable arrhythmia syndromes (HAS) has resulted in an increasing number of variants detected. Results can be unclear for clinicians trying to make definitive diagnoses of whether a child actually carries a disease or not. While these variants are subject to change over time and these changes not only alter diagnosis and management but have psychological impact, variant reclassification is not routinely performed. We aimed to identify the yield of initial genetic testing and determine how reclassification altered diagnostic yield and its longitudinal impact on diagnosis.

Methods: Single-center retrospective study of pediatric probands < 18y who underwent HAS genetic testing between 2007-2018. A variant was considered diagnostic if classified as pathogenic or likely pathogenic (P/LP) by the testing company, while non-diagnostic variants were benign/likely benign (B/LB). Variants of unknown significance (VUS) indicate an unclear disease association. To account for varying reasons for genetic testing (e.g. cardiac arrest with QTc of 550msec vs. borderline QTc on sports screening), probands were stratified into high likelihood (HL) or low likelihood (LL) of disease based on cardiac evaluation and clinical suspicion. Diagnostic yield by likelihood was statistically compared. Multivariable logistic regression was performed to evaluate the association between likelihood and harboring a diagnostic variant while controlling for year. Initial variants were reclassified in March 2023 using American College of Medical Genetics criteria and compared to initial testing results. Variants downgraded if reclassifications were reclassified as less diagnostic (P/LP to VUS, VUS to B/LB) or upgraded if the opposite was observed.

Results: Among 308 probands (51% male), initial testing was diagnostic in 24%. Diagnostic yield was 43% (for HL) and 10% (for LL; p< 0.0001), with HL having 6-times the odds of having diagnostic results compared to LL (OR 6.0, 95% CI 3.1-11.6, p< 0.0001). Probands with LL were almost twice as likely to have a completely negative genetic test result (no variants identified; 42% vs 22%, p< 0.0001). On reclassification, classification 10% (7/71) of HL probands with diagnostic variant(s) downgraded. Among probands with a VUS, 21% upgraded while 37% downgraded (Figure). No variant initially classified as B/LB upgraded.

Conclusion: Diagnostic yield of HAS genetic testing was 24% but was significantly higher in those with high clinical suspicion (43%). Reclassification demonstrated that 10% and 37% of probands with an initially diagnostic or VUS variant (respectively) downgraded, suggesting that variant(s) initially detected may not be solely responsible for their disease based on updated data. Results demonstrate that genetic variant reclassification is important for patients with VUS and even those with P/LP variants thought to be diagnostic. Since classification can change over time and may affect clinical status, our study suggests reclassification in clinical practice should be uniformly performed for this population.