Purpose: Patients with tuberous sclerosis complex (TSC) are potential surgical candidates if the epileptogenic region(s) can be accurately identified. This retrospective study determined whether FDG-PET/MRI coregistration and diffusion-tensor imaging (DTI) showed better accuracy in the localization of epileptogenic cortex than structural MRI in TSC patients.
Methods: FDG-PET/MRI coregistration and/or DTI for apparent diffusion coefficient (ADC) and fractional anisotropy (FA) were utilized in 15 TSC patients. Presurgery scalp EEG and postsurgery seizure control identified epileptogenic tubers (n = 27) and these were compared with nonepileptogenic tubers (n = 204) for MRI tuber volume, volume of FDG-PET hypometabolism on MRI coregistration, DTI, ADC, and FA values.
Results: Compared with nonepileptogenic tubers, epileptogenic regions had increased volume of FDG-PET hypometabolism (p < 0.0001), and increased ADC values in subtuber white matter (p < 0.0001). In contrast, the largest MRI identified tuber (p = 0.046) and decreased FA values (p = 0.58) were less accurate in identifying epileptogenic regions. Larger volumes of FDG-PET hypometabolism correlated positively with increased ADC values (p = 0.029), and localized to areas of cortical dysplasia adjacent to the tuber in four cases.
Conclusions: Larger volumes of FDG-PET hypometabolism relative to MRI tuber size and higher ADC values identified epileptogenic tubers and adjoining cortex containing cortical dysplasia in TSC patients with improved accuracy compared with largest tuber by MRI or lowest FA values. Used in conjunction with ictal scalp EEG and interictal magnetoencephalography, these newer neuroimaging techniques should improve the noninvasive evaluation of TSC patients with intractable epilepsy in distinguishing epileptogenic sites for surgical resection.