Objective

To evaluate the clinical utility of next-generation sequencing (NGS) in unexplained pediatric epilepsy, and to identify the potential predictors associated with Mendelian genetic causes.

Methods

Two hundred and ten children with unexplained epilepsy, who underwent NGS test were included. We analyzed the demographic, clinical and genetic characteristics, and executed a Logistic regression analysis for identifying predictors for Mendelian genetic causes. Patients were classified as either with isolated epilepsy or syndromic epilepsy with concurrent neurodevelopmental phenotypes.

Results

The overall diagnostic yield was 29.0% (61/210). A total of 68 variants spanning 39 genes were identified in 58 patients (27.6%, 58/210) from exome sequencing based testing. Of the 68 variants, 33 were novel ones. Besides, STAR and CNTN2 were identified to be a candidate gene for epilepsy. Patients with syndromic epilepsy had a much higher diagnostic yield than that of isolated epilepsy (53/135, 39.3% vs. 8/75, 10.7%, p = 0.000). The odds ratio of detecting genetic cause was 3.939 (95% CI 1.369–11.332) for syndromic epilepsy without epileptic encephalopathy (EE), 5.814 (95% CI 2.208–15.306) for EE, 2.958 (95% CI 1.093–8.000) for patients with seizure onset <12 months, and 2.932 (95%CI 1.414–6.080) for female. Of the 210 patients, 78.4% of patients (145/185) had at least a 50% reduction in seizure frequency and 58.9% (109/185) reached seizure freedom. There was no difference between seizure prognosis and diagnostic outcomes.

Significance

NGS is effective for Mendelian genetic etiological diagnosis for unexplained pediatric epilepsy. Female patients with syndromic epilepsy with onset within the first year of life are most likely to yield a positive test result.