Abstracts

Rationale: Perturbations of neural integrity often result in reorganization of cortical functioning, particularly in epileptic patients. In the presence of cortical pathology within the dominant hemisphere, language networks may relocate either within the same hemisphere or to contralateral homologues. However, little is known about the rates of atypical language organization in developmental and acquired pathologies of pediatric epilepsy. The present study compared differences in language organization in focal cortical dysplasia (FCD), a prototypical representation of developmental disorders of neuronal migration and differentiation, and hippocampal sclerosis (HS), a postnatally acquired brain lesion. We also evaluated relationships between histopathological characteristics, lesion location, and neuropsychological performance. Methods: 30 right-handed subjects 9-20 years of age were evaluated using functional MRI (fMRI) language paradigms and neuropsychological testing prior to epilepsy surgery. Subjects were classified histopathologically as having FCD (n=20) or HS (n=9). All underwent comprehensive neuropsychological evaluation and fMRI with receptive and expressive language paradigms. fMRI findings included either cluster activation within expected areas of the left hemisphere or atypical organization. Results: Eight subjects evidenced atypical language network activation. Twenty-three (77%) had expected receptive language activation and 7 (23%) showed atypical organization. Twenty-five subjects (83%) had classic activation for expressive language tasks and three (10%) demonstrated atypical organization. One subject failed to show any expressive BOLD signal. Histopathology did not influence language organization. However, analysis of activation by lesion location revealed atypical receptive language organization in 60% of subjects with dominant temporal pathology whereas dominant frontal lesions had no influence on activation patterns. Furthermore, atypical expressive language organization occurred in 22% of cases with left temporal lesions and 33% with left frontal pathology. Neuropsychological scores did not differ significantly with type or location of anatomical substrate, but trends were noted in select domains. Atypical organization of both temporal and frontal language areas correlated inversely with receptive language performance. Conclusions: The data suggest that prenatal and early postnatal pathologic lesions involving receptive language circuits in the dominant temporal lobe both strongly promote atypical language organization and diminished receptive language performance. In contrast, the contribution of specific histopathological substrate to language representation is minimal. These findings corroborate those of previous studies implicating proximity of pathology to eloquent cortex in the dominant hemisphere as the primary determinant of functional reorganization. However, receptive language networks demonstrate a greater propensity toward atypical activation patterns than expressive networks. These patterns are unrelated to age at seizure onset or duration of epilepsy.