Neuropediatrics 2006; 37 - P1142
DOI: 10.1055/s-2006-953558

Usefulness of combined MEG and voxel-based morphometric MRI analysis in cryptogenic frontal lobe epilepsy

T Bast 1, A Schulze-Bonhage 2, S Fauser 2, J Zentner 3, C Kurth 4, BJ Steinhoff 4, H Brückner 5, JP Ernst 5, S Siebert 6, IEB Tuxhorn 7, D Rating 1, HJ Huppertz 8
  • 1Universitätsklinik für Kinder- und Jugendmedizin, Abteilung für Pädiatrische Neurologie, Heidelberg
  • 2Universitätsklinik Freiburg, Neurozentrum, Sektion Prächirurgische Epilepsiediagnostik, Freiburg
  • 3Universitätsklinik Freiburg, Abteilung Allgemeine Neurochirurgie, Freiburg
  • 4Epilepsiezentrum, Klinik für Erwachsene, Kehl-Kork
  • 5Epilepsiezentrum, Kinderklinik, Kehl-Kork
  • 6Abteilung für Neuroradiologie, Universitätsklinik, Heidelberg
  • 7Klinik Mara, Epilepsiezentrum, Bethel
  • 8Schweizerisches Epilepsiezentrum, Zürich, CH

Rational: Diagnostic work-up in cryptogenic frontal lobe epilepsy (FLE) is challenging because ictal and interictal EEG patterns are frequently diffuse. We report pathbreaking results from magnetoencephalography (MEG) and morphometric MRI analysis in such patients.

Methods: Seven patients suffering from FLE were investigated by 122-channel whole-head MEG and morphometric MRI analysis. Repetitive MRI investigations by tertiary epilepsy centers had not revealed clear epileptogenic lesions. MEG analysis of interictal spikes with BESA followed a standard protocol (Bast et al., Epilepsia 2004). A regional source was modeled within a spherical head model to localize the center of averaged interictal spikes. Ictal Beta patterns of two patients were filtered, averaged and modeled by a regional source. For morphometric MRI analysis, algorithms of SPM2 software and special filters were applied to generate maps highlighting typical features of focal cortical dysplasia (FCD) like abnormal gyration or blurring of the grey-white matter junction (Huppertz et al., Epilepsy Research 2005).

Results: Two patients have not been operated so far. In one of them, MEG was not analyzable because of technical problems, in the other MEG revealed a frontal focus, but MRI analysis was normal. Five patients (1 f, 4 m; age 8.5–50.5 years) underwent a frontal resection and became free from disabling seizures (Engel 1). Histopathology revealed FCD type 2 according to Palmini in all cases. Morphometric MRI analysis showed suspect regions with disturbed gyration or abnormal cortical to subcortical differentiation within the frontal lobe in all 5 patients. Interictal and ictal sources localized within or near these cortical areas in all cases. Invasive recordings (subdural/depth electrodes n=3, intraoperative electrocorticography n=2) confirmed non-invasive results.

Conclusion: Voxel-based morphometric MRI analysis can visualize cortical regions with discrete abnormalities but adds no information on function. MEG source analysis can model the center of epileptic activity, but the extent of an activated cortical area remains undefined. The combination of both methods is promising in cryptogenic (frontal lobe) epilepsies and may guide target-oriented invasive recordings