Abstracts

Rationale: Accurate identification of the epileptic focus on preoperative tests is essential for achieving a favorable seizure outcome following epilepsy surgery. FDG-PET has been reported to be a useful preoperative test for localizing the epileptic focus in intractable epilepsy. In the present study, we created a normal database and used statistical analysis to determine 1) the rate of focus localization for temporal lobe epilepsy and 2) relationships between FDG-PET results and the condition of epilepsy (EEG, pathology).Methods: Subjects were 31 patients who underwent unilateral hippocampectomy after being diagnosed with temporal lobe epilepsy, and were divided based on MRI results into a group in which hippocampal sclerosis was present (n=14; five men and nine women; mean age, 30 years) and a group in which hippocampal sclerosis was not identified (n=17; 13 men and four women; mean age, 42 years) for comparison. Focus localization on FDG images was performed by indicating as percentages the ratios of Z-scores ?2 at each site in comparison to 1) visual assessment, 2) 3D-SSP, and 3) SEE method: normal database. Seizure type, correlation with EEG (EEG distribution), and pathological findings at the resection site were compared between the two groups.Results: In patients with hippocampal sclerosis, glucose hypometabolism simultaneously occurred in the uncus, parahippocampal gyrus, and lateral temporal lobe (T1-T3) on the side of the focus. In temporal lobe epilepsy unaccompanied by hippocampal sclerosis, glucose hypometabolism occurred only in the lateral region. In patients with hippocampal sclerosis on MRI, although lateralization can be determined without performing 3D-SSP analysis, statistical analysis enabled localization to the hippocampus and temporal cortex. In addition, 3D-SSP analysis may enable determination of lateralization in temporal lobe epilepsy unaccompanied by hippocampal sclerosis.Conclusions: Epilepsy diagnosis based on statistical analysis of 18F-FDG-PET using a normal database, which involves visualization of the temporal lobe (hippocampus) on the affected side, was considered useful as a complementary method in focus localization.