Abstracts

Rationale: Interictal magnetoencephalography (MEG) is often done as one of multiple tests to localize the likely epileptogenic zone in patients with medically intractable epilepsy undergoing evaluation for resective treatment. We compared interictal MEG results with intracerebral depth EEG ictal onsets in the context of postoperative seizure control. Methods: The clinical courses of 45 consecutive patients with medically intractable epilepsy who underwent implantation of intracerebral depth electrodes at UCLA Medical Center between May 2000 and October 2005 were retrospectively reviewed. Of these 45 patients, 30 had interictal MEG. We evaluated the concordance of interictal MEG epileptiform dipoles with depth EEG ictal onsets. Concordance was classified as intralobar (group A), lobar (B), lateralized (C), bilateral homotopic (D1), bilateral homotopic interictal MEG with unilateral depth (D2), non-concordant unilateral MEG and bilateral depth (D3), non-concordant multifocal MEG and depth (D4), and contralateral non-concordant (E). The lack of depth lateralization or localization was classified as group (F) and the lack of MEG dipoles as (G). Results: Twenty-two of the 30 presurgical candidates who underwent both MEG and depth EEG studies subsequently had resective surgery; and 20 resected patients had postoperative seizure outcome follow-up with durations ranging from 10 to 96 months (mean 50). Seizure outcomes corresponded to our MEG-depth EEG concordance classification as follows. The concordant groups (A, B, C, D1, D2) comprised 22 of the 30 total patients; 17 of these patients underwent resection, with postoperative seizure outcomes of: Engel class I (11 patients) & class IV (6). The non-concordant groups (D3, D4, E) comprised 2 of the 30 total patients; one underwent resection, with Engel class IV outcome. The groups in which one of the two tests yielded no information (F, G) comprised 6 of the 30 total patients; 2 underwent resection, with seizure control outcomes of Engel class I for the one resected patient with no interictal MEG dipoles and Engel class II for the one resected patient with no lateralizing or localizing depth EEG ictal findings. Conclusions: Clinical use of MEG is becoming increasingly routine and has been supported by clinical research results and professional guidelines. However, standardized use has not developed. In our results, the majority of our presurgical candidates had some degree of concordance of interictal MEG epileptiform dipoles and depth EEG ictal onsets. Our results suggest that the significance of concordance of interictal MEG with depth EEG in presurgical evaluation for resective epilepsy surgery is complex and requires further analysis. A larger case series may be needed to address this question.