Abstracts

Rationale:
Status epilepticus (SE) is a common neurological emergency with significant morbidity and mortality. Identification of any genetic predisposition to SE will facilitate the understanding of mechanisms and effective treatments. We previously found familial aggregation of SE using data from the Epilepsy Phenome/Genome Project (EPGP) obtained through a medical record abstraction (MRA). However, obtaining accurate seizure histories is a challenge that may require consolidation of multiple sources. Here, we examined whether SE showed familial aggregation in the EPGP dataset using data derived from a structured diagnostic interview (DI). We also compared the results obtained from a DI and MRA and evaluated the agreement between the sources.  
Method:
We used data from 2197 individuals with epilepsy belonging to 1043 families collected by the EPGP between 2007 and 2014. We identified participants as having traditionally defined SE (TSE, seizure ≥ 30 min) or operationally defined status epilepticus (OSE, seizure ≥ 10 min) via DI or MRA. We examined familial aggregation of TSE and OSE using generalized estimating equations (GEE), including sex, age at epilepsy onset, seizure localization, number of antiseizure medications, epilepsy duration, and history of epilepsy surgery as covariates. Odds ratios and 95% confidence intervals were computed using a binomial distribution, logit link function, and an exchangeable correlation structure. To assess concordance between MRA and DI results, we calculated the kappa statistic.
Results:
The presence of TSE or OSE was correlated with a lower age at epilepsy onset and localization-related seizures. The kappa statistic for agreement between DI and MRA was 0.49 for TSE and 0.40 for OSE. Although the methods showed moderate concordance for identification of SE, the phenotypic characteristics of the participant groups showed no significant differences. We found evidence of familial aggregation for OSE when OSE was indicated by DI only (OR 3.81, 95% CI 2.59-5.60), both DI and MRA (OR 4.55, 95% CI 2.47-8.38), and at least one source (OR 3.54, 95% CI 2.50-5.01). For TSE, we found evidence of aggregation in families when indicated by both DI and MRA (OR 4.05, 95% CI 1.18-13.91) and at least one source (OR 2.24, 95% CI 1.28-3.91). We did not find significant TSE aggregation using DI data only (OR 1.79, 95% CI 0.76-4.21). The results held after controlling for epilepsy surgery and number of antiseizure medications as markers of epilepsy severity.
Conclusion:
We found moderate agreement between data sources in identifying TSE and OSE and saw evidence of familial aggregation when restricting the sample to individuals who were identified by both DI and MRA. However, when using DI data alone, we found significant aggregation for OSE but not for TSE. The findings support a genetic contribution to OSE and a possible genetic contribution to TSE. They also underscore the challenge in identifying TSE and OSE retrospectively.
Funding:
:NIH (U01NS053998), St. Louis Children's Hospital Foundation