Summary:Purpose: EEG recordings of spontaneous seizures should provide clues to epilepsy pathogenesis and pathology. We examined onset, propagation, and termination characteristics of spontaneous seizures recorded by intracranial EEG in temporal lobe (TL) epilepsy in relation to neuronal, glial, and synaptic changes in the same tissue.Methods: All of our patients with intracranial EEG recordings of spontaneous TL seizure onset, subsequent TL resection, and quantitative pathologic analysis of resected tissue were included. Seizure parameters were mean time to initial propagation, mean total electrical duration, uniformity of seizureonset distribution and location, and percentage of seizures with spiking onset per patient. Tissue was analyzed for glial and neuronal density in hippocampal fields and presence of sprouting. Outcome was classified as seizure free or not.Results: All seizures with onset in resected TL in 62 patients were analyzed. The percentage of each patient's seizures with spike onsets was significantly correlated with glial density in CA3 (p i 0.01). Initial propagation time was significantly and inversely correlated with neuronal density in CA4 (p < 0.02).Electrical seizure duration was significantly correlated with glial density in CA2 and CA3 (p < 0.02). Neuronal and glial density were significantly (and inversely) related to one another only in CAI (p < 0.001). Outcome was most significantly related to uniform hippocampal seizure onset. Presence or abhence of sprouting was not significantly related to outcome or any EEG measure.Conclusions: These results suggest both glia and neurons exert independent influences on the expression of ictal discharges in seizures of medial TL onset. Glial density influenced interictal-ictal transition, whereas neuronal density influenced seizure propagation. These findings may have implications for pathogenesis. Key Words: Neuronal changes-Glial changes-Temporal lobe.Mesial temporal lobe epilepsy is often medically refractory and may be treated successfully by surgical resection (1 -4). Quantifiable and predictable neuronal, glial, and synaptic changes accompany this epileptic process, which together define the entity called mesial temprovide clues to the meaning of those tissue changes and to the pathogenesis of mesial temporal lobe epilepsy.Some reports have described morphologic characteristics of ictal discharges as recorded from intracranial electrodes in mesial temporal lobe epilepsy (13-19). The poral sclerosis (MTS). Features that are consistently described include >50% neuronal loss in hippocampal fields CAI, CA3, and the dentate gyrus, accompanied by varying amounts of gliosis, and sprouting of fibers imrnunoreactive for neuropeptide Y, somatostatin, and dynorphin into the inner molecular layer of the dentate gyrus (5-12). The relation of these pathologic findings to the pathogenesis of mesial temporal lobe epilepsy remains uncertain. Correlations, therefore, of the pathologic alterations with specific EEG characteristics of the spontaneous seizures th...