Multimodality image-guided surgery for the treatment of medically refractory epilepsy.

OBJECT: The aim of this study was to review seizure outcome, imaging modalities used, and complications following surgery in patients with epilepsy who had undergone multimodality image-guided surgery at our institution.
METHODS: Data from patients with epilepsy who had undergone surgery between April 1999 and October 2001 were reviewed. During this time period, 116 operations were performed in 109 patients with medically refractory epilepsy. Among these patients, 22 were selected to undergo multimodality image-guided surgery primarily on the basis of whether they had no lesion visible on conventional magnetic resonance (MR) imaging sequences, multiple lesions, or one very large lesion that could not be completely resected without the risk of significant postoperative morbidity. A fourth group of patients in whom there was a single lesion in the eloquent cortex, a location associated with a significant risk of postoperative morbidity, was also included in the analysis. This latter group was assessed with the aid of intracranial grid electrodes that were coregistered to the MR image and were used intraoperatively to minimize electrode position error. Other imaging modalities used included positron emission tomography (PET), fluid-attenuated inversion recovery (FLAIR) MR imaging, and subtracted ictal-interictal single-photon positron emission computerized tomography (SPECT) coregistered with MR imaging (SISCOM). After coregistration, images were then downloaded onto an image-guided surgical system and the epileptogenic area was then resected. The mean patient age was 33 years (range 17-46 years), and there was a mean follow up of 27 months (range 14-41 months). Multimodality coregistrations used were as follows: nine PET scans, seven subdural electrode grids, four SISCOM studies, one FLAIR MR image, and one combined PET/subdural grid. Seizure outcome was excellent in 17 patients (77%) and not excellent in five (23%), or favorable in 19 (86%) and unfavorable in three (14%). Six patients (27%) had a transient neurological deficit, one patient (5%) a permanent major deficit, and three patients (15%) a permanent minor deficit. Five patients (24%) had a transient psychiatric problem postoperatively.
CONCLUSIONS: Multimodality image-guided surgery offers a new perspective in surgery for epilepsy. Functional imaging modalities previously lateralized and often localized a seizure focus, but did not provide enough anatomical information to resect the epileptogenic zone confidently and safely. The coregistration of these modalities to a volumetric MR image and their incorporation into an image-guided system has allowed surgeons to offer surgery to patients who may not previously have been considered eligible, with outcomes comparable to those in patients with more straightforward lesional epilepsy.