Abstracts

Rationale: Evidence suggests that epilepsy involves susceptibility genes that influence overall risk for epilepsy, as well as modifier genes that influence the phenotypic features of epilepsy. Understanding the inheritance of individual components of the epilepsy phenotype is a key step to understanding heterogeneity in the human epilepsies, and to identifying novel disease genes in the presence of such heterogeneity. The objective of this study was to determine familial correlations of individual phenotypic features of epilepsy among affected family members, as a measure of heritability and a method of selecting phenotypes for genetic analysis. Methods: Epilepsy families with two or more affected individuals were ascertained from patients attending the Cleveland Clinic Epilepsy Center. Informed consent was obtained from all participating individuals. Seizures were classified using the Cleveland Clinic Semiological Classification System. Multivariate familial correlation coefficients with their asymptotic standard errors were calculated for phenotypic features, for parent-offspring and sibling pairs, using the FCOR program in S.A.G.E. This method has the advantage of not assuming that the traits analyzed follow a normal distribution. The phenotypic variables analyzed included epilepsy type, seizure type, age of onset, seizure frequency, response to medications, history of febrile seizures, and seizure severity. Results: The study population consisted of 58 individuals from 26 independent pedigrees, comprising 33 parent-offspring and 14 sibling pairs. Baseline characteristics included a mean age of epilepsy onset of 12 years. The predominant epilepsy type was focal (64%); 56% of the study population had refractory epilepsy. Correlation coefficients for both parent-offspring and sibling pairs were high for epilepsy type and seizure type (see Table). For siblings, correlations were also high for epilepsy age of onset and seizure severity. Conclusions: These results indicate varying degrees of familial correlation, a measure of heritability, for individual phenotypic components of epilepsy. In particular, seizure type was highly correlated among affected-relative pairs, suggesting that seizure type may be more useful than epilepsy syndrome type in genetic analysis. Additional work determining correlations between phenotypic variables is important for interpreting relative-pair correlations, and for deriving the ultimate phenotypes to be used for genetic analysis.