Prognostic factors and outcome after different types of resection for temporal lobe epilepsy

Object. It is unknown whether different resection strategies for temporal lobe epilepsy (TLE) produce alterations in seizure control or neuropsychological performance. Methods. A series of 321 patients who underwent surgery for TLE between 1989 and 1997 was submitted to a uniand multifactorial analysis of clinical, electrophysiological, neuroimaging, neuropsychological, and surgical factors to determine independent predictors of outcome. Until 1993, most patients with TLE underwent standard anterior temporal lobectomy (ATL); beginning in 1993, surgical procedures were increasingly restricted to lesions detected on magnetic resonance (MR) imaging and the presumed epileptogenic foci: for example, amygdalohippocampectomy (AH) or lesionectomy/corticectomy began to be used more often. The mean follow-up duration in this study was 38 months. Two hundred twenty-seven patients were classified as seizure free (70.7%), and 36 patients had rare and nondisabling seizures (11.2%); these groups were summarized as having good seizure control (81.9%). Twenty-four patients attained more than 75% improvement (7.5%), and no worthwhile improvement was seen in 34 cases (10.6%); these groups were summarized as having unsatisfactory seizure control (18.1%). On unifactorial analysis the following preoperative factors were associated with good seizure control (p < 0.05): single and concordant lateralizing focus on electroencephalography studies, low seizure frequency, absence of status epilepticus, concordant lateralizing memory deficit, clear abnormality on MR images, suspected ganglioglioma or dysembryoplastic neuroepithelial tumor (DNT), and absence of dysplasia on MR images. Stepwise logistic regression revealed a model containing five factors that were predictive for good seizure control (p < 0.1): 1) clear abnormality on MR images; 2) absence of status epilepticus; 3) MR imaging-confirmed ganglioglioma or DNT; 4) concordant lateralizing memory deficit; and 5) absence of dysplasia on MR images. Seizure outcome was mainly correlated with diagnosis and clinical factors. No significant differences were found regarding different resection types performed for comparable tumors. Neuropsychological testing revealed better postoperative results after limited resections compared with standard ATL, especially with regard to attention level, verbal memory, and calculated total neuropsychological performance. Conclusions. Different strategies for surgical approaches in TLE result in equally good outcomes. Seizure outcome is mainly dependent on the diagnosis and clinical factors, whereas the neuropsychological results are more beneficial after resections limited to an epileptogenic lesion and focus.