Abstracts

Rationale: Historically, language lateralization in preschool aged children has been very difficult to obtain. In epilepsy patients who are surgical candidates, this has been traditionally performed using standard electro-cortical stimulation via implanted electrode arrays. Alternatively, the intra-carotid amobarbital (Wada) procedure can be used, but is also technically difficult and unreliable in this age group. Furthermore, presurgical language mapping not only determines lateralization, but can assist with intracranial electrode placement. Language lateralization has not been well established in children under age five, but there is a growing body of evidence showing that it occurs earlier than classically perceived. We present two cases where preschool children with temporal lobe epilepsy underwent language mapping using magnetoencephalography. Methods: Patient A was a 3-year-old right-handed female who began having seizures at approximately 18 months of age. Video EEG monitoring and MEG indicated a left temporal lobe seizure focus. Her MR imaging was grossly normal. Patient B was a 2-year-old right-handed female who began having seizures at 21 months of age. Video EEG monitoring and MEG indicated a right temporal lobe seizure origin. In addition, a low-grade tumor was identified in her right temporal lobe. The procedure for receptive language mapping involved having the participant passively listen to a series of words. Both patients were sedated during their language mapping procedures. There were 180 words, presented at a rate one word every 1.5 s. The responses were filtered, averaged, and localized using a single equivalent current dipole model over a time period a 300 - 800 ms. Hemispheric dominance for language was determined as a ration of language sources localized to the left relative to the right temporal lobe. Results: The results of the language mapping in Patient A showed clear right hemisphere dominance for receptive language. This was confirmed via stimulation mapping using intra-cortical electrodes. This patient underwent a left temporal lobectomy without any language deficits. For patient B, the results of the language mapping showed a clear left temporal lobe dominance for receptive language. This patient underwent a right temporal lobectomy with no language deficit post-surgically. Both patients remain seizure-free at this date. Conclusions: The discussion presents the case for the advantages of MEG for non-invasively localizing language function in preschool children. Language mapping can be done safely and quickly, even in young children, which in turn can improve surgical outcomes for preschool children.