Magnetic source imaging guidance of gamma knife radiosurgery for the treatment of epilepsy

Object. The purpose of this study was to determine if magnetic source (MS) imaging could provide useful information in the planning and performance of gamma knife radiosurgery (GKS) for epilepsy. Methods. Magnetic source imaging of interictal epileptiform dipoles was stu died in 53 epilepsy surgery candidates. All patients underwent volumetric m agnetic resonance (MR) imaging. Subsequently, magnetoencephalography (MEG) was performed using single or dual 37-channel units. The MR images and MEG recordings were then coregistered to produce the MS imaging data. Magnetic source imaging epileptiform data were reviewed in a blinded fashion and spa tial distributions were classified as focal, regional, multiple, scattered, or none. Postresection operative photographs were compared with MS image r esults to determine whether extensive or partial/no resection of the MS ima ge focus had been accomplished. Magnetoencephalography dipoles were identif ied in 47 patients (89%), in 46 of whom the lesions were resected. This inc luded 20 (80%) of 25 anterior temporal lobe (ATL) cases, and 26 (93%) of 28 extratemporal lobe (ETL) cases. Of the six patients who underwent extensiv e ATL resections, three (50%) were seizure free. Of 14 patients who underwe nt partial/no resection of the ATL, seven (50%) were seizure free. There wa s no clear relation between MS image spatial distribution and surgery-relat ed outcome. Of the seven ATL cases with hippocampal atrophy, five patients (71%) were seizure free. Of 12 ETL cases (three lesional), 10 patients (83% ) were seizure free. Of 14 patients who underwent partial/no ETL resections (three lesional), two (14%) were seizure free. Of five nonlesional ETL cas es with focal MS image dipoles, four patients (80%) were seizure Gee. Of fi ve nonlesional ETL cases with regional dipoles, three patients (60%) were s eizure free. Of eight ETL cases with multiple MS image dipoles, two patient s (25%) were seizure free. Spatial agreement of MS imaging and electrograph ic data had no apparent effect on outcome of either ATL or ETL cases. Conclusions. Nonlesional ETL cases with focal (and in some cases multiple o r regional) epileptiform MS image dipole distributions benefit significantl y from inclusion of the MS image epileptiform focus in the resections. Nonl esional ETL cases suitable for GKS may similarly benefit from including the MS image focus in the irradiated area.