Abstracts

Rationale: Relatively little is known about the pathophysiology of status epilepticus (SE) despite the fact that it is the most common neurological emergency in children and a common one in adults. One approach to elucidating the underlying mechanisms is to identify predisposing genes. We have used the EPGP study population to test the hypothesis that genetic mechanisms contribute to SE. Methods: EPGP is a multi-center collaborative effort to collect detailed phenotypic and genetic data on a large number of epilepsy patients. EPGP has enrolled approximately 2,000 sibling pairs and parent-child pairs with idiopathic generalized epilepsy (IGE) or localization-related epilepsy (LRE). An analysis was performed on all IGE and LRE families with complete data in the EPGP database as of 12/31/2011. We assessed familial concordance of SE (≥30 min) or prolonged seizures (≥10 min) within family units based on both the medical record abstraction (MRA), completed via formal chart review, and a detailed diagnostic interview (DI) administered to the participant or their parent at enrollment. The observed number of concordant pairs was compared to the expected number based on a binomial distribution using a chi square test. A mixed effect logistic regression analysis was performed to adjust for age, gender and broad epilepsy type. Results: There were 1497 participants in 626 families that had 2 or more participants with completed data collection. 764 participants had IGE and 505 participants had LRE. The average length of follow-up from time of epilepsy diagnosis was 11.3 years. The data from the MRA and DI did not show strong agreement. For instance, on the MRA 94 participants had a history of SE versus 86 on the DI, of which only 37 were identified by both. Assuming the MRA is gold standard, the DI has a sensitivity of 39%, a specificity of 97%, a positive predictive value of PPV of 43%, and a negative predictive value of 97%. Similar discrepancies occurred when looking at the concordance for SE within families. For instance, the MRA identified 10 families in which both participants were concordant for SE, and the DI identified 5 such families. Only 2 of those families were identified by both. Therefore, we analyzed the MRA and DI data independently for familial concordance for the occurrence of SE and prolonged seizures. For both the MRA and the DI, the concordance rate for SE (p<0.0001) and prolonged seizures (p<0.0001) was significantly greater than expected. To ensure that we were not assessing concordance of complex febrile seizures, an analysis was performed excluding those participants with a history of a febrile seizure ≥15 min on the MRA (24 participants). The familial concordance for SE (p=0.005) and prolonged seizures (p=0.003) remained greater than expected. Adjusting for age, gender and epilepsy type did not eliminate the effect. Conclusions: The EPGP database shows greater than expected concordance for prolonged seizures and SE within families. These findings suggest that genetic mechanisms contribute to SE.