Abstracts

Rationale: Absence seizures manifest as brief impairment of consciousness marked by 3-4 Hz spike and wave discharge on electroencephalography (EEG). Prior neuroimaging studies of childhood absence seizures have demonstrated involvement of both cortical and subcortical networks. However, the relationship between these changes and ictal behavior is unknown. Limited studies have shown variable behavioral performance during childhood absence seizures, though the mechanism underlying this variability is unknown. Methods: We performed two studies to elucidate the neural mechanism for behavioral variability during seizures of childhood absence epilepsy. In study 1, we measured simultaneous behavioral, 32-lead EEG, and fMRI signals in 36 patients. In study 2, we measured simultaneous high-quality 256-lead EEG recordings with behavioral in 15 patients. Two tasks requiring varying levels of attention were used for behavioral testing: a repetitive tapping task (RTT) and a more difficult continuous performance task (CPT). Quantitative analyses were performed on both EEG and fMRI data and correlated with ictal behavioral performance. Results: In study 1, we recorded 579 seizures and in study 2, we recorded 150 seizures. Seizures with both good (<25% omissions) and poor (>75% omissions) were captured for both tasks. Task difficulty and longer seizure durations were predictive of more impaired ictal performance. For both RTT and CPT, patients demonstrated a bimodal distribution of behavior in which most seizures were associated with either fully spared or impaired performance. Data-driven clustering analysis of fMRI showed involvement of three networks including the default mode network, the task-positive network, and the primary sensorimotor and thalamic network. The three networks were sequentially activated in seizures and showed greater amplitude changes for seizures with worse performance. EEG also showed a correlation between signal amplitude in the 2.5-4 Hz seizure frequencies and worse ictal performance on both tasks. Interestingly, EEG and fMRI signal amplitude was greater for seizures with poor performance at or prior to seizure onset. Conclusions: Seizure amplitude as measured by EEG and fMRI is related to poor ictal behavioral performance. This correlation at seizure onset suggests the pre-seizures state may determine seizure severity and degree of behavioral impairment during absence seizures. Childhood absence epilepsy has significant effects on the quality of life of patients. Understanding how absence seizures impair consciousness may lead to improved treatments for this disorder.