Familial aggregation of seizure outcomes in four familial epilepsy cohorts

Colin A. Ellis,Danni Tu,Karen L. Oliver,Heather C. Mefford,W. Allen Hauser,Jeffrey Buchhalter,Michael P. Epstein,Quy Cao,Samuel F. Berkovic,Ruth Ottman,EPGP Consortium,Epi4K Consortium

DOI: https://doi.org/10.1111/epi.18004

IF: 6.74

2024-05-15

Epilepsia

Abstract:Objective To assess the possible effects of genetics on seizure outcome by estimating the familial aggregation of three outcome measures: seizure remission, history of ≥4 tonic–clonic seizures, and seizure control for individuals taking antiseizure medication. Methods We analyzed families containing multiple persons with epilepsy in four previously collected retrospective cohorts. Seizure remission was defined as being 5 and 10 years seizure‐free at last observation. Total number of tonic–clonic seizures was dichotomized at <4 and ≥4 seizures. Seizure control in patients taking antiseizure medication was defined as no seizures for 1, 2, and 3 years. We used Bayesian generalized linear mixed‐effects model (GLMM) to estimate the intraclass correlation coefficient (ICC) of the family‐specific random effect, controlling for epilepsy type, age at epilepsy onset, and age at last data collection as fixed effects. We analyzed each cohort separately and performed meta‐analysis using GLMMs. Results The combined cohorts included 3644 individuals with epilepsy from 1463 families. A history of ≥4 tonic–clonic seizures showed strong familial aggregation in three separate cohorts and meta‐analysis (ICC .28, 95% confidence interval [CI] .21–.35, Bayes factor 8 × 1016). Meta‐analyses did not reveal significant familial aggregation of seizure remission (ICC .08, 95% CI .01–.17, Bayes factor 1.46) or seizure control for individuals taking antiseizure medication (ICC .13, 95% CI 0–.35, Bayes factor 0.94), with heterogeneity among cohorts. Significance A history of ≥4 tonic–clonic seizures aggregated strongly in families, suggesting a genetic influence, whereas seizure remission and seizure control for individuals taking antiseizure medications did not aggregate consistently in families. Different seizure outcomes may have different underlying biology and risk factors. These findings should inform the future molecular genetic studies of seizure outcomes.

clinical neurology