Abstracts

Rationale: Even in cases of clinically diagnosed symptomatic generalized epilepsies (SGEs), asymmetry in blood perfusion or glucose metabolism is sometimes seen on functional neuroimaging such as single photon emission computed tomography (SPECT) and Positron Emission Tomography (PET). This suggests potential differences in epileptogenicity between hemispheres. The present study examined correlations between quantitatively assessed electroencephalogram (EEG) and SPECT findings to test the above hypothesis.Methods: Subjects comprised 11 children (5 boys, 6 girls; median age, 5.9 years; range, 1-14 years) diagnosed with SGE by long-term video-EEG monitoring as a part of presurgical evaluation for pharmacoresistant epilepsy. Underlying pathology was categorized as Lennox-Gastaut syndrome (n=3), infantile spasms (n=4) or other SGEs (n=4). One or more types of generalized seizure, such as spasms, atypical absence or tonic, atonic or myoclonic seizures were observed daily in all patients. Bilaterally synchronous spike and waves and multifocal spikes on EEG, and symmetrical features on MRI, including bilateral frontal polymicrogyria in 1 patient and lissencephaly in 1 patient, were observed in all. All EEG spikes of 16 selected electrodes were extracted from an arbitrarily selected 10- to 60-min period from artifact-free sleep records. Detected spikes from each channel were averaged for measurement of amplitude. Interictal SPECT data were analyzed using a 3-dimensional stereotactic region-of-interest template technique, and radioisotope (RI) counts of 14 cortical regions were semi-quantitatively assessed. Spike amplitudes of 8 unilateral electrodes and RI counts of 7 unilateral regions in each hemisphere were respectively averaged as hemispheric values. Laterality indices (LIs) of both parameters by region or hemisphere were calculated as (L-R)/(L+R), where L and R represent values of the measurement in question. Finally, correlations were statistically tested using regression analysis.Results: LIs of hemispheric and regional spike amplitudes ranged from -0.36 to 0.22, and from -0.77 to 0.3, respectively. Conversely, LIs of RI counts ranged from -0.005 to -0.001, and from -0.1 to -0.03, respectively, indicating that distributions of RI counts were less asymmetrical than spike amplitudes. However, LIs of spike amplitudes and RI counts displayed positive correlations to both hemispheric and regional evaluations (p=0.006; R2=0.59, p<0.001; R2=0.28, respectively). Conclusions: Particularly in localization-related epilepsies, hypo- or hyperperfusion in the epileptogenic focus is usually seen on interictal or ictal SPECT, respectively. However, in SGEs, RI counts on interictal SPECT are increased in the more susceptible hemisphere as postulated by EEG spike amplitudes. These results suggest that ictal-like perfusion images are often observed in SGEs due to frequent epileptiform discharges on EEGs.