Abstracts

Rationale: Children with epilepsy have a higher incidence of atypical language characterized by right or bilateral hemisphere language patterns. Factors that have been correlated with atypical language patterns in children with epilepsy include left handedness, focal seizures or lesions within the left hemisphere, early age of epilepsy onset and longer epilepsy duration. Atypical patterns are correlated with poor language performance. Children with medically intractable epilepsy that are evaluated for surgical resection undergo detailed investigation to determine the localization of the epileptic focus and language function. Magnetoencephalography (MEG) may be used in combination with video-electroencephalography (VEEG) to determine the location of the epileptogenic focus. Functional MRI (fMRI) is used to identify hemisphere language patterns. The objective of this study is to investigate the relationship between the location of the epileptic focus, as determined by VEEG and MEG, and the development of atypical language patterns in children with epilepsy. Methods: This is a retrospective cohort study of 68 children with epilepsy that underwent fMRI for language evaluation and MEG from January 2005 - December 2008. Reliable data was available for both procedures in 43 of 68 children. Typical language pattern was defined as left hemisphere fMRI activations on a standardized battery of pediatric fMRI language paradigms. Atypical language pattern was defined as right hemisphere or bilateral fMRI activations. The precise location of the epileptic focus was defined by a cluster (>6) of MEG spike sources in agreement with findings from ictal VEEG. The parameter distribution was tested by goodness of fit using Pearson's chi-square (χ2) test. Results: Twenty-five children demonstrated typical language patterns and 18 demonstrated atypical patterns. Atypical language was significantly associated with non-right handedness and female gender. However, atypical language patterns were not associated with side of epileptic focus. Twenty-one patients had a right hemisphere epilepsy focus, 7 (33%) had atypical language dominance, all with epileptogenic foci in mesiotemoral or the rolandic regions. The 14 children with right hemisphere onset epilepsy and typical language patterns had predominantly anterior frontal epileptic foci. Conclusions: Hemisphere of seizure onset did not predict the presence of atypical language patterns in this cohort of children with epilepsy. Moreover, children with right rolandic and mesiotemporal epileptic foci were more likely to demonstrate atypical language patterns. This unexpected finding suggests that atypical language patterns may be present in children with epilepsy regardless of the side of the epileptic focus and that the location of the epileptic focus may be related to the development of atypical language. Future studies will be directed at the role of the location of the epileptic focus in the prediction of language patterns in children with epilepsy.