Abstracts

Temporal lobe seizures cause diverse ictal behaviors including loss of consciousness, dystonic movements, and neuroendocrine changes, suggesting that widespread networks beyond the temporal lobes are involved. To elucidate these networks we used single photon emission computed tomography (SPECT) ictal-interictal imaging to compare complex partial seizures (CPS), with impaired consciousness, to simple partial seizures (SPS), with spared consciousness, in patients with mesial temporal epilepsy.
Patients with surgically confirmed mesial temporal lobe epilepsy were selected based on mesial temporal sclerosis on pathology, and lack of recurrent seizures with at least one year follow-up. CPS were studied during the late ictal and early postictal periods, respectively, with Tc99m-HMPAO injections 60-90s after seizure onset (N=10) and at [gt]90s after seizure onset (N=8). Patients with SPS were also studied (N=6). Group analyses were done in SPM99 using a paired t-test model. The extent threshold below which clusters were rejected was 125 voxels (1cc) and height threshold p=0.01 (z=2.33). To study network involvement, correlation analysis was done between brain regions.
Significant cerebral blood flow (CBF) increases in the CPS 60-90s group were found in the temporal lobe and the basal ganglia ispilateral to seizure onset. In the [gt]90s group, the temporal lobe was no longer activated, but significant hyperperfusion was present in midline subcortical structures with limbic connections, including the medial thalamus, hypothalamus, and upper brainstem. In both CPS groups, there was marked hypoperfusion of the bilateral fronto-parietal association cortex. In contrast, individual patients with SPS had CBF increases confined to the temporal lobe. The group data from SPS patients revealed no significant network involvement of midline subcortical structures or of the association cortex as was seen in CPS. Correlation analysis across all patients (N=24) revealed that hyperperfusion in the ipsilateral medial thalamus was strongly correlated with hypoperfusion in the ipsilateral frontal association cortex (r=0.67, p[lt]0.001).
Abnormal excitatory and inhibitory network mechanisms may lead to behavioral phenomenon associated with partial seizures. Spread of seizure activity from the medial temporal lobe to midline subcortical structures, especially during the late ictal and early postictal periods may be crucial for these long-range network effects. We propose that during complex partial seizures, abnormal inhibition of higher order association areas as well as abnormal excitation of medial diencephalon and upper brainstem may result in disruption of normal information flow, and thus lead to loss of consciousness.