Abstracts

Rationale: Seizures of childhood absence epilepsy (CAE) are seen as 3-4 Hz spike-and-wave discharges (SWD) on electroencephalogram and manifest as brief epochs of impaired consciousness. While absence seizures are brief, they can occur up to hundreds of times per day and greatly impact quality of life. Absence seizures are historically considered a generalized form of epilepsy. However, recent imaging work has revealed focal involvement of areas of the attention network in the brain, including the orbital/medial frontal cortex, medial/lateral parietal cortex and thalamus. Previous work has demonstrated variability in ictal blood oxygen level-dependent (BOLD) signal changes and behavioral performance. Our central hypothesis is that the selective disruption of attentional and primary information processing networks is related to the degree of ictal behavioral impairment.Methods: We performed simultaneous 32-lead electroencephalography and functional magnetic resonance imaging (EEG-fMRI) in 21 pediatric CAE patients as they performed a continuous performance task (CPT) or an easier repetitive tapping task (RTT). EEG and fMRI signals for 221 seizures were analyzed and correlated to performance on CPT or RTT. Time-frequency analysis was performed for ictal EEG at discrete frequency bands. Time course and amplitude of ictal fMRI changes were also determined for seizure and attention network areas.Results: Ictal task performance was more impaired relative to interictal task performance. Performance was relatively more impaired for the more difficult CPT compared to RTT, though variability existed from seizure to seizure and from patient to patient. EEG analysis showed an association between seizure duration and degree of impairment, though omissions occurred even in brief absence seizures. Time-frequency analysis of EEG signals revealed variable spatial distribution of power changes. Time course of ictal fMRI changes were seen in orbital frontal, parietal, thalamus and other cortical areas, though changes were highly variable across seizures. In one subject with a large number of seizures captured in the scanner, poor task performance was correlated with greater fMRI decreases in the medial and lateral frontal cortices. Conclusions: EEG and fMRI signal variability arising from seizure and attention networks may lead to variability in ictal performance on task. Determining brain-behavior relationships using EEG and fMRI may provide insight on the mechanism by which CAE causes impaired consciousness during seizures and could lead to the development of targeted therapies.