Abstracts

We used continuous EEG-fMRI to evaluate the regions activated as a result of generalized spike-wave (GSW) discharges in patients with idiopathic generalized epilepsy.
We selected 23 patients with active GSW. During EEG-fMRI study, EEG (21-channel) was recorded in a 1.5T MR. Artefacts from the imaging sequence were removed from the EEG ([italic]Hoffmann et al 2000 Magn Reson Med 44:791[/italic]). Eight patients were excluded due to lack of GSW during the study and one for technical reasons. Areas activated were found by statistical comparison to a model including five hemodynamic response functions: the standard response to auditory stimuli, peaking at 5.4 s ([italic]Glover 1999 NeuroImage 9:416[/italic]), and four gamma functions peaking at 3, 5, 7 and 9 s. Activations were considered significant if three contiguous voxels had a t-stat [ge] 3.0. The highest t-stat for each activation in different models was determined. Activations in the brainstem, basal ganglia and cerebellum were excluded.
Fourteen patients were evaluated (13 women; mean age, 36; age at seizure onset, 10). Two had no fMRI activation. Cortical activation was largely symmetrical in the 12 other patients. It was diffuse in three, and multiple regional in nine: mesial and lateral frontal in nine, parietal convexity in nine, mesial parietal in eight, mesial occipital in seven, occipital convexity in two, mesial central in six, central convexity in five, mesial or lateral temporal in six, and insula in 3. Activation was almost equally distributed over anterior and posterior head regions in five subjects, with anterior predominance in five and posterior in two.
We found negative and positive cortical activations: predominantly negative in seven patients (mean volume of negative activation 164.9 vs 20.4 cm3 for positive activation), and predominantly positive in five (mean volume of positive activation 233.9 vs 65.5 cm3). Maximum t-stat varied from 3.5 to 14.3 for positive and from -3.4 to -17.6 for negative activations.
All 12 patients showed thalamic activation, bilateral in 10: positive in eight, negative in three, and both positive and negative in one. Thalamic activations had positive maximum t-stat of 4.3 to 10.5 (volume 0.4 to 38.9 cm3), and negative max t-stat of -3.2 to -5.6 (volume 0.4 to 38.9 cm3).
GSW resulted in fMRI cortical responses that were largely symmetrical, involved primarily frontal and parietal regions, but also central and occipital regions. The thalamus was involved in all patients showing cortical activation. A surprising result was the presence of positive and negative activations in cortex and thalamus. Whereas positive activation results primarily from increased synaptic activity, excitatory or inhibitory, negative activation is thought to result from decreased neuronal input, which may also play a significant role in GSW.
[Supported by: Canadian Institute of Health Research]