Abstracts

Patients with refractory temporal lobe epilepsy who demonstrate convergent data of non invasive presurgical evaluation are considered excellent surgical candidates. However, some presurgical candidates have mesial temporal sclerosis (MTS) or a temporal lesion with non convergent EEG data, including seizure semiology, interictal and ictal EEG recordings. Is positron emission tomography (PET) indicated in these patients and should F-18-fluorodeoxyglucose (FDG) PET or flumazenil PET be preferred in order to improve presurgical workup and possibly prevent intracranial EEG recordings ?
We studied ten patients with refractory epilepsy, eight had MTS and two demonstrated a low grade temporal lesion with non convergent EEG data. All ten patients were evaluated with dynamic FDG and flumazenil PET. One patient had intracranial EEG recordings before temporal lobe resection. Follow up after surgery varied from 1-4 years and outcome was scored according to the Engel classification.
Seven of ten patients had unilateral MTS, one asymmetric bilateral MTS. Two patients demonstrated a low grade temporal lesion, probably a dysembryoplastic neuroepithelial tumor (DNET). EEG studies of these patients were inconclusive and suggested bilateral epileptogenic zones in three and ipsilateral to the MRI abnormality a posterior temporal or extratemporal epileptogenic regions in six, whereas one patient had inconclusive EEG results. All ten patients demonstrated hypometabolic zones convergent with the lesional temporal lobe. Flumazenil PET was abnormal in nine patients and in two of these nine patients the area of decreased benzodiazepine receptor binding was less extensive compared to the area of hypometabolism in FDG PET. Nine patients were operated without and one after intracranial EEG recordings. Postoperative outcome was favourable (Engel classes 1 and 2) in seven and less favourable (Engel class 3) in three patients.
In ten patients with a temporal lesion (MTS or DNET) EEG studies were insufficiently localizing, lateralizing or even inconclusive and they were considered poor surgical candidates. In these patients FDG and flumazenil PET supported presurgical lateralization and localization of the epileptogenic temporal lobe in a way that made surgery possible without intracranial EEG recordings in nine of these ten patients. Only one patient needed confirmation with intracranial EEG recordings before surgery. Flumazenil PET was negative in one patient and therefore slightly less sensitive compared to FDG PET. This study supports the idea that either one of these PET techniques is sufficient to present the zone of functional deficit related to the epileptogenic zone in patients that otherwise would be rejected for surgery because of inconclusive results of non invasive EEG recordings or needed intracranial EEG recordings.