Abstracts

Rationale: Successful epilepsy surgery is largely dependent on accurate localization of the epileptogenic zone, and preoperative evaluation involves a series of assessments including detailed clinical history, interictal EEG, video-EEG monitoring, MRI, PET, SPECT, and neuropsychological assessment. In patients with medically refractory temporal lobe epilepsy, many patients have identifiable lesion or epileptic foci on MRI scans. However, there are many other patients whose MRI scans are normal despite the fact that semiology and history are suggestive of mesial temporal lobe seizures. We evaluated whether preoperative PET scan findings provide useful insight toward the surgical outcome in patients with normal MRI scans. Methods: We identified the total of 104 adult patients who underwent phase II monitoring with intracranial electrodes recordings (ICR) and PET studies for their intractable temporal lobe epilepsy from 2007 to 2009 at the Barrow Neurological Institute. We excluded patients who had abnormal MRI scans or who did not have sufficient number of seizures during the ICR. In order keep group homogeneity, we included patients who had standard 4 set of intracranial depth wires (orbitofrontal, lateral frontal, amygdale, and hippocampal depth wires on each hemisphere) with 8 contact electrodes per wire. We then reviewed the remaining patients ictal EEG localizations by ICR and preoperative PET scans retrospectively, and divided them into two study groups: Match Group (MG) and non-Match Group (nMG). When both PET scans and ICR showed same localization, patients were stratified into MG, and all others into nMG. The differences in seizure outcomes in 3, 6, and 12 months between two groups were evaluated postoperatively. Results: Out of 104 patients with ICR for their temporal lobe seizures, a total of 42 patients had normal MRI scan (3T) and sufficient seizures for ictal-focus localization. All 42 patients had FDG-PET scan of the brain prior to ICR. One year surgical outcome data was available in only 22 patients: 6 patients in MG and 16 in nMG. In nMG, PET was normal in 11/16 patients, showed bilateral hypometabolism in 4/16 patients, and showed extratemporal hypometabolism in 1/16 patient. Engel s Class I outcome in one year was observed in 5 out of 6 patients (83.3%) in MG compared to 13 out of 16 (81.3%) in nMG group. However, 3 patients (18.8%) had Class III or IV outcomes in nMG, compare to none in MG (5 Class I and 1 Class II).Conclusions: Our study suggests that the overall chance of seizure freedom is not affected by PET findings in MRI-negative patients with intractable temporal lobe epilepsy, and they may not provide further guidance in surgical decision making process. Even though the study population was small, this study also provides evidence that higher chance of worse surgical outcome may result if PET study and ICR results do not show concordant findings.