Abstracts

Patients without an unequivocal epileptogenic lesion identified on MRI remain a large challenge in presurgical epilepsy evaluations. FDG-PET and magnetoencephalography (MEG) are both established interictal noninvasive neuroimaging tools that can aid the goal of localization for surgery. However, the value of these imaging modalities in patients with non-localizing MRI (normal, questionable or ambiguous abnormalities) is unclear.
As part of a prospective multimodality imaging study in presurgical epilepsy evaluations we reviewed the localization results of FDG-PET and MEG exams acquired consecutively between September 2001 and January 2003 at UAB. Twenty-five patients had adequate imaging data sets and sufficient ultimate localizing information from video EEG and other tests to allow comparison studies. FDG-PET was acquired in the resting condition with no evidence of seizure activity during the scan. PET analysis was based on inspection by two independent blinded reviewers with consensus on identification of relative focal deficits of FDG uptake. MEG was recorded with a 148 channel whole-head system with simultaneous scalp EEG using the standard International System of electrode placement. Epileptiform spikes and sharp waves were analyzed with a single equivalent current dipole model.
Of the 25 patients, the mean age was 13.9 years (range 1-56, 12 female). Only nine patients had localized potential epileptogenic pathology on MRI. See the Table for distribution of patients based on lobar localization and MRI classification. Note, 5 of the 6 patients with [quot]lesions[quot] had only subtle or questionable abnormalities on MRI[ndash]3 suggestive of focal cortical dysplasia and 2 with slight increase in T2 in the amygdala. Twenty patients had localizing MEG studies and 18 had relative focal hypometabolism on FDG-PET (only one case with both negative MEG and PET studies). Concordant lobar localization was present between MEG and PET in 13 patients. Seven of the 8 patients localized with MEG and not PET had normal MRI scans and extratemporal lobe epilepsy. All three patients with localized PET and negative MEG had frontal lobe epilepsy with colocalized subtle MRI abnormalities (as noted above), pathologically confirmed as cortical dysplasia.
MEG and FDG-PET exams are useful in the presurgical evaluation of patients with non-localized MRI whether they have temporal or extratemporal lobe epilepsy. Further study (with larger numbers) should be designed to determine if patient subgroups can be identified who particularly benefit from one or the other modality, when the combination of the two is optimal, and when they are redundant.[table1]
[Supported by: NIH K-23 Award, RCK
EFA Clinical Epilepsy Fellowship Award, JB]