Abstract
Objective: To determine whether there is a correlation of genetic diagnosis/result with long-term seizure freedom in pediatric epilepsy patients. Methods: This was a prospective and retrospective cohort study of children with epilepsy referred for genetic testing at a single center. The primary outcomes were presence and type of genetic diagnosis (pathogenic, benign, or variant of uncertain significance) and patient epilepsy status (seizure free, treatment failure, uncertain). Epilepsy gene panels were the primary method of genetic testing. Results: The prospective cohort had 22 patients followed for >11 years and for whom genetic testing was then performed; the retrospective cohort had 78 patients with previous genetic testing followed for >8 years. In the prospective cohort, one patient each of the seizure free or treatment failure groups had a pathogenic genetic variant; mean Combined Annotation Dependent Depletion (CADD) scores 22 and 24, respectively (P = .62). In the retrospective cohort, there was no difference in the number of variants (P = .97), the variant interpretations (P = .29 ClinVar, P = .39 lab interpretation) or mean CADD scores (P = .29) between the seizure-free, treatment failure, and uncertain epilepsy patients. Whole exome and genome sequencing identified pathogenic variants in 70% of patients with treatment failure but were not performed in seizure-free patients. Significance: Our findings show no correlation of the presence or type of epilepsy gene panel result with long-term seizure freedom in pediatric patients. The yield and specificity of pathogenic variants may be higher using whole exome and whole genome sequencing in patients with treatment-resistant epilepsy. Whole exome and whole genome sequencing, or more targeted understanding of specific variants, may be needed to improve the utility of pediatric epilepsy genetic testing.