Epilepsy is one of the most common neurological disorders, and patients whose seizures are not controlled suffer from many adverse effects. The goal of this study is to investigate the effects of epilepsy on brain structure and function, and to test innovative approaches to treatment when seizures cannot be controlled by currently available approaches. Methods: Patients undergo video-EEG monitoring to determine seizure type and focus localization. Positron emission tomography (PET) and magnetic resonance imaging (MRI) are used to study cerebral metabolism, blood flow, and structure. Antiepileptic drug blood levels are obtained. Recent neuroimaging study findings: Activation of central serotonin (5-HT)1A receptors, found in high density in brainstem raphe, hippocampus and temporal neocortex, exerts an anticonvulsant effect in various experimental seizure models. Previously, we found reduced 5-HT1A silent antagonist [18F]FCWAY binding in both lateral temporal neocortical and mesial temporal regions ipsilateral to temporal lobe epileptic foci. We now report magnetic resonance (MR)-based partial volume corrected (PVC) results. Two-hour dynamic PET images were acquired in 13 patients and 10 controls after bolus [18F]FCWAY injection. Image frames were registered to segmented T1-weighted MR volumes, and PVC applied using gray and white matter masks. Pixel data were corrected for intravascular radioactivity, labeled metabolite uptake, and spill-in of skull 18F-fluoride activity. Original and PVC pixel data were fitted to a 2-tissue compartment model using the metabolite-corrected input function to produce volume of distribution (V) images. ROIs drawn on MR volumes were applied to V images. Only gray matter voxels were sampled. Binding potential (BP) (mL/mL) was calculated as [V(ROI)-V(CEREBELLUM)]/ plasma free fraction. Both before and after PVC, patients had significantly lower BP, as well as greater side-side asymmetry than controls in ipsilateral hippocampus, parahippocampus and fusiform gyrus (p<.05, corrected for multiple comparisons). There were no contralateral differences before or after PVC. Hippocampal BP asymmetry before PVC was 64?b27% compared to 11?b9 % for FDG. After PVC, values were 51?b25% and 11?b12%, respectively. 5-HT1A BP is significantly reduced in TLE mesial temporal regions but not lateral neocortex. This reduction is not partial volume averaging artifact. These findings support our hypothesis of reduced 5-HT1A receptor binding in temporal lobe epileptic foci. Reduced 5-HT1A activity may contribute to regional hyperexcitability. Refractory partial epilepsy patients often exhibit regional hypometabolism. It is unknown whether the metabolic abnormalities are present at seizure onset or develop over time. In a previous study we found that abnormalities of glucose utilization are less common and profound in children with new onset partial seizures than in adults with chronic partial epilepsy. Studies using magnetic resonance imaging have shown that reduced hippocampal volume is associated with a history of febrile seizures, the duration of epilepsy, and the number of generalized tonic clonic seizures. It is uncertain whether these factors would have the same influence on functional as they do on structural measures of the integrity of the epileptogenic zone. We used positron emission tomography with fluorine 18 2 deoxy glucose to study 119 patients with temporal lobe seizure foci localized by ictal video-EEG. PET was performed in the awake interictal resting state with ears plugged and eyes patched. We recorded surface EEG during injection (5 millicuries) and the 30 minute uptake period. We used a standard template to analyze PET scans. Patients with a history of either any febrile seizures, or complex or prolonged febrile seizures, did not have greater hypometabolism ipsilateral to the epileptic focus than patients without hypometabolism. There was a non-significant trend for patients with more lifetime generalized tonic-clonic seizures to have greater relative hypometabolism ipsilateral to the epileptic focus. We found that there was a significant negative relation between the duration of epilepsy and hippocampal glucose metabolism ipsilateral to the epileptic focus (F=6.81; p <0.02). A functional measure of mesial temporal integrity in patients with temporal lobe epilepsy was affected by epilepsy duration, but not a history of febrile seizures. This suggests that hippocampal injury due to seizures may be a progressive process. In a parallel study in rats, we are in collaboration with ERS and the Laboratory of Functional and Molecular Imaging, we are using MRI to study the role of kainate and AMPA receptor activation in functional and structural alterations in animal models of epilepsy. We are attempting to model the limbic injury shown in human imaging, and investigate its physiologic basis. In our initial studies, we found that the specific gluR5 receptor agonist ATPA produces seizures when injected stereotactically into rat amygdala. These seizures were similar in terms of physiologic effects measured by MRI to those produced by injection of the less specific glutamate agonist kainic acid. This suggests that gluR5 receptor activation is sufficient by itself for seizure induction. In the next stage of the study, we will try to model chronic epileptogenesis via EEG and MRI after convulsant administration.
Showing the most recent 10 out of 68 publications
