Abstracts

Rationale: Absence epilepsy raises significant challenges for learning and quality of life in children. Current pharmacological therapies frequently fail to provide satisfactory relief from symptoms or cause intolerable side-effects. To develop better drugs, it is important to investigate the underlying neural mechanisms with valid animal models. Previous work has focused on sedated or anesthetized animals. However, these studies showed cortical fMRI increases, whereas human absence fMRI is dominated by cortical decreases. Here we investigate whether seizure hemodynamics in awake, drug-free animals are more consistent with those in human seizures. Methods: We used head-fixed, awake, drug-free functional magnetic resonance imaging (fMRI) with a high field magnet (9.4 tesla) and simultaneous electroencephalography (EEG) recordings of Genetic Absence Epilepsy Rats from Strasbourg (GAERS), an established absence epilepsy model. Animals were incrementally acclimated to the MRI environment, including restraints and noises over 3 weeks. Once acclimated, 18 animals were scanned in a total of 117 sessions, during which we recorded 1719 seizures. Echo-planar imaging (EPI) was used to acquire fMRI. Both T1- and T2-weighted anatomical images were also obtained for each recording session. Preprocessing spanned multiple software suites and included detection and removal of motion and artifact epochs, realignment, functional to structural registration, and spatial smoothing. We then conducted both a general linear model (GLM) and region of interest (ROI) analysis of group data. Results: Our first analysis was a voxel-wise GLM comparing seizure to non-seizure periods. Seizures were associated with a decrease in blood oxygen level dependent (BOLD) signal in multiple cortical regions and an increase in thalamic nuclei. Second, we performed an ROI-based BOLD signal time-course analysis. Similar to our findings in the first analysis, time-courses showed differences between the cortex and thalamus both during and after seizures. This resembles the cortical decreases and thalamic increases seen in human fMRI during absence seizures. Conclusions: We conclude that the hemodynamics of seizures in awake GAERS are like those in humans with absence epilepsy, supporting the translatability of electrophysiological and behavioral findings from awake GAERS. We also conclude that like in human patients, the BOLD signal in awake GAERS seizures has significant differences between cortex and thalamus. This suggests the importance of any future mechanistic investigations using awake animal models to ensure translational validity and facilitate development of new therapies. Funding: NIH R37NS100901