Abstracts

Rationale: On positron emission tomography (PET), patients with temporal lobe epilepsy (TLE) have reduced GABA-A receptor binding and glucose metabolism (RMRGlu) in seizure foci. Previous studies have focused on measuring independent neurotransmitter system abnormalities rather than considering the relation between GABA receptor binding and RMRglu. PET studies in healthy controls showed strong negative coupling between GABA-A BP and RMRglu in insular and mesial temporal lobe regions. We predicted decoupling between GABA-A binding and regional glucose metabolic rate (RMRGlu) in temporal lobe epilepsy patients. Methods: We studied 9 patients and 10 healthy volunteers who underwent both [11C]Flumazenil and [18F] FDG PET on an ECAT high resolution research tomograph (full width half maximal resolution 3 millimeters). [11C]Flumazenil binding potential (BP) was calculated using the time-activity curve from the pons as the input function. Skull-stripped MRI images (3DSkullStrip, AFNI) were segmented into grey matter, white matter, and CSF binary masks (FSL FAST) and ROIs were applied to MRI images in Talairach space (infralimbic cortex, left and right interior insula, anterior hippocampus, posterior hippocampus, amygdala, anterior cingulate, posterior cingulate, superior temporal gyrus, parietal-occipital cortex). Images were converted back into MRI space, multiplied by the gray matter mask, and applied to partial volume corrected RMRGlu and GABA-A BP images. Mean RMRGlu and BP values within each region were compared between groups with independent samples T-tests. Correlations between glucose metabolism and GABA-A binding in each region were determined for both patients and controls. Results: We found significant correlations between GABA-A BP and RMRGlu in healthy controls in ipsilateral superior temporal gyrus, as well as both ipsi- and contralateral subgenual prefrontal cortex. In contrast, significant relations were not found between RMRglu and GABA-A BP in TLE patients. One region (posterior hippocampus), showed significant correlation values in TLE patients but not healthy controls (Figs. 1 & 2). Conclusions: These results support our previous pilot study showing high correlations between GABA-A BP and RMRGlu within areas of increased GABA transmission. This relationship was not found in TLE patients, suggesting potential decoupling of these two systems.