Abstracts

Rationale: Studies of absence seizures (AS) using EEG with functional MRI (EEG-fMRI) show a consistent network with prominent thalamic activation and a variety of cortical changes. Despite evidence suggesting a role of frontal cortex in seizure generation, group studies have not detected consistent AS-related changes in this region. We hypothesized that this may be due to intra-group variability in frontal cortical activation.Methods: We studied 13 subjects with AS during EEG-fMRI to classify the different individual patterns of frontal cortical activation associated with AS. We used visual inspection to classify patients according to changes in frontal lobe blood oxygen level dependent (BOLD) signal change. We performed event related time-course analysis in selected areas of interest.Results: Based upon visual inspection of surface-rendered activation maps we identified two sub-groups that could be distinguished by the activation in the dorso-lateral pre-frontal cortex (DLPFC). One group of patients (n=7) showed a primarily positive signal change in the DLPFC (DLPFC-POS), whereas the other group (n=6) showed a primarily negative signal change (DLFPC-NEG). A random effects group analysis showed signficant differences in BOLD activation between the groups (p < 0.05 corrected for multiple comparisons). Analysis of the timing of signal changes within the DLPFC, and in the thalamus, caudate nucleus and parietal cortex demonstrated that the DLPFC-POS group showed a more prominent BOLD increase associated with AS onset and a greater negative deflection following this. Furthermore analysisis of clinical features suggested that these patterns of BOLD activation may have a clinical correlate.Conclusions: In this study we identified two major patterns of frontal cortical BOLD change in individuals with absence seizures; a predominantly positive cortical BOLD signal change in the dorso-lateral pre-frontal cortex or a negative BOLD signal change. Comparison of phenotypic and electrographic differences between groups suggested potential clinical and EEG correlates of these patterns. Event related time course analysis showed distinct differences in the pattern of BOLD signal change between these groups. Our results suggest that the frontal cortical BOLD response may differ between individuals which may affect the way seizures are generated or propagated. These observed differences may have phenotypic implications.