Abstracts

Rationale: Children seem to be at less risk than adults for measurable and lasting language decline after epilepsy surgery. This may be due to increased plasticity of the young brain, relatively shorter duration of epilepsy, improved accuracy of pre-surgical diagnostics, and a variety of other contributing factors. Magnetoencephalography/magnetic source imaging (MEG/MSI) is one such pre-surgical diagnostic technique that facilitates prediction of post-surgical language outcome. This study describes patterns of relevant clinical features among patients with non-left language dominance on MSI, and reports post-surgical language outcome in relation to MSI findings. Methods: Data from 34 pediatric epilepsy patients (ages 3-17 years) was collected via retrospective record review. All patients were being considered for epilepsy surgery, due to focal intractable seizure disorder. All patients produced valid language lateralization data during the receptive language (CRM) task of the MSI protocol. Twenty patients had primarily left-hemisphere seizure focus; 14 patients had primarily right-hemisphere seizure focus. A subset of 20 patients from this sample underwent focal resective epilepsy surgery (10 left, 10 right) and post-surgical neuropsychological evaluation. Patients who underwent CC or hemispherectomy were excluded from the post-surgical outcome portion of the analysis. Results: MSI data demonstrated left language dominance for 16 patients (47%), right language dominance for 6 patients (18%), and bilateral language for 12 patients (35%). Of the 18 patients who showed atypical (non-left) language dominance, 15 (83%) had seizure onset < 5 years of age; 14 (78%) had duration of epilepsy > 4 years; 11 (61%) had primarily left hemisphere seizure focus. Among the subset of 20 patients who underwent focal epilepsy surgery, only 3 demonstrated > 15 point decline in Verbal Comprehension Index (VCI) after surgery. Given significant confounding factors (multiple surgeries for 2 and psychosis for 1), these 3 cases were excluded from the subsequent analyses. Duration of epilepsy and age of seizure onset were significantly correlated with pre and post-surgical VCI. 12 of the remaining 17 patients were classified as bilateral or right hemisphere dominant for language on MSI. 4 of the 17 had focal resection in the MSI-classified language-dominant hemisphere and none of these patients demonstrated language decline after surgery. Conclusions: Preliminary data analysis implicates younger age at seizure onset, longer duration of epilepsy, and left hemisphere seizure focus as prominent characteristics in the majority of patients with bilateral or right-dominant language representation on MSI. Analysis of post-surgical outcome measures confirms previously documented findings to suggest that only a relatively low proportion of a pediatric sample demonstrates significant language decline after focal resective epilepsy surgery, regardless of side of surgery, and in spite of pre-surgical diagnostic findings to suggest at least some language representation in the to-be-resected hemisphere.