Abstracts

Rationale: Children with tuberous sclerosis complex (TSC) often present with medically intractable seizures very early in life. The seizures can be associated with developmental delay. Numerous studies have reported spatial correlation of the epileptogenic zone and the prominent tubers. Since TSC presents multifocally in both hemispheres and patients can have several types of seizures, finding the dominant focal seizure and the corresponding epileptogenic zone is crucial for surgical success. We studied TSC patients with MEG source localization of interictal discharges and and evaluated both the distance to nearest tuber and to the ictal onset zone (IOZ) in patients who underwent resective surgery. Methods: 24 patients (10 males and 14 females; mean age 8.8 years, range 2 to 26 years; seizure onset mean age; 0.9 years, range 4 weeks to 18 months) with medically intractable epilepsy secondary to TSC underwent presurgical evaluation including scalp video EEG, MEG, and MRI. We reviewed ictal semiology, scalp interictal/ictal EEG lateralization/localization, MRI lesions (presence of prominent tubers/calcification), MEG interictal/ictal source localization with synthetic aperture magnetometry (SAM), SAM kurtosis (SAM(g₂)) for all patients, and single rotating current dipole (SRCD), standardized low resolution brain electromagnetic tomography (sLORETA) and multiple signal classification (MUSIC) for all except 4 patients who had VNS implanted. For 14 of 24 patients who underwent resective surgery, ictal onsets of intracranial EEG (ICEEG) were also compared to the results of non-invasive studies. Results: 8 patients (33%) had clear ictal semiology lateralization. 5 of 8 (63%) patients had a prominent tuber in ipsilateral hemisphere but not always the same region as ictal onset. SAM virtual sensors, SRCD, the peak of sLORETA and MUSIC distribution showed adjacent locations to one of prominent or larger tubers in 16 cases (67%). For 14 patients who underwent resective surgery, MEG and ICEEG source localizations were compared and measured from the center of ictal onset from the visual inspection in ICEEG. 4 of 14 patients (29%) who had one area of limited corticectomy showed MEG source localization within 2cm maximum distance from the tuber and the center of the IOZ according to visual inspection. In 7 patients (50%) who had multi lobe corticectomy/lobectomy, MEG sources were part of the resected areas. For only one patient (7%) with epileptic spasms, MEG failed to localize the epileptogenic zone. For two (14%) had technical difficulty to localize MEG sources because of excessive diffuse beta activities. One patient who had seizure during MEG recording showed closer and better source localization in terms of the depth of source. Conclusions: MEG source localization with multiple algorithms was able to localize IOZ in patients who had both single and multiple epileptogenic zones. These methods would be helpful to reveal not only IOZ but also distributed epileptogenic zones in TSC.