Abstracts

Rationale: Patients with idiopathic generalized epilepsy (IGE) show a characteristic blood oxygenation level-dependent (BOLD) response to generalized spike-and-wave (GSW) and polyspike-and-wave (GPSW) activity. We investigated in various sub-syndromes of IGE whether the type of seizure in adulthood influences the BOLD pattern. Methods: We studied 25 patients (mean age 35.2 years, range 18-66, 13 males) and subdivided them in two groups according to the presence (Ab+) or absence (Ab-) of absence seizures. Patients underwent a continuous two-hours EEG-fMRI study using a 3 Tesla MRI scanner (Siemens) and 27 MRI compatible Ag/AgCl electrodes (10–20 system). Eleven studies were excluded: 8 patients had no GSW discharges during the fMRI acquisition, 2 had atypical abnormalities and one a failure in the EPI sequence. We studied the GSW- or GPSW-related BOLD pattern of the 14 remaining patients, 8 in group Ab+ and 6 in group Ab-. In group Ab+: one patient had eyelid myoclonia (eyelid myoclonia with absences); 4 had generalized tonic clonic seizures (GTCS) and myoclonia (epilepsy with myoclonic absences, n=1; juvenile myoclonic epilepsy [JME], n=3); and 3 had GTCS without myoclonia (juvenile absence epilepsy). The six Ab- patients never had absences throughout their whole life (JME, n=4; GTCS only, n=2). Positive (activation) and negative (deactivation) BOLD activity was studied in response to epileptic discharges, and assessed in 22 cortical and sub-cortical regions of interest (ROI).Results: EEG acquired in the scanner disclosed 3-3.5 Hz GSW or GPSW. Ab+ patients had a median of 13.5 (range 4-259) GSW (n=3), GPSW (n=3) or both (n=2) per study with a median duration of 1.5 sec (range 0.5-20), while Ab- patients had a median of 9 (range 4-38) GSW (n=1), GPSW (n=3) or both (n=2) with a median duration of 1.6 sec (range 0.3-2). One Ab+ patient had 259 GSW discharges lasting on average 3.9 sec (range 1-20). Thirteen patients had both activations and deactivations and one, only deactivations. All Ab+ versus two Ab- patients showed thalamic activations. Ab+ showed more frequently than Ab- patients activations in dorsolateral frontal (7/8 vs 2/6), SMA (6 vs 2), anterior cingulate (7 vs 1), and insular (4 vs 0) cortices. Also, cortical deactivations were more frequent in Ab+ patients: ventromedial frontal (6/8 vs 3/6), parietal (6 vs 3), posterior cingulate (6 vs 2) and temporo-occipital (5 vs 2). In the remaining ROIs, no differences in the BOLD responses were observed between the two groups. Conclusions: The typical pattern of thalamic activation and associative cortical deactivation is frequent in the epileptic discharges of IGE patients with absences and conversely rare in those of IGE patients without absences. Involvement of the thalamus and of the fronto-parietal and posterior cingulate network of the default mode of brain function may be the important factor that associates GSW and GPSW with clinical absence seizures.