50 Hz hippocampal stimulation in refractory epilepsy: Higher level of basal glutamate predicts greater release of glutamate

OBJECTIVE: The effect of electrical stimulation on brain glutamate release in humans is unknown. Glutamate is elevated at baseline in the epileptogenic hippocampus of patients with refractory epilepsy, and increases during spontaneous seizures. We examined the effect of 50 Hz stimulation on glutamate release and its relationship to interictal levels in the hippocampus of patients with epilepsy. In addition, we measured basal and stimulated glutamate levels in a subset of these patients where stimulation elicited a seizure. METHODS: Subjects (n = 10) were patients with medically refractory epilepsy who were undergoing intracranial electroencephalography (EEG) evaluation in an epilepsy monitoring unit. Electrical stimulation (50 Hz) was delivered through implanted hippocampal electrodes (n = 11), and microdialysate samples were collected every 2 min. Basal glutamate, changes in glutamate efflux with stimulation, and the relationships between peak stimulation-associated glutamate concentrations, basal zero-flow levels, and stimulated seizures were examined. RESULTS: Stimulation of epileptic hippocampi in patients with refractory epilepsy caused increases in glutamate efflux (p = 0.005, n = 10), and 4 of ten patients experienced brief stimulated seizures. Stimulation-induced increases in glutamate were not observed during the evoked seizures, but rather were related to the elevation in interictal basal glutamate (R(2) = 0.81, p = 0.001). The evoked-seizure group had lower basal glutamate levels than the no-seizure group (p = 0.04), with no stimulation-induced change in glutamate efflux (p = 0.47, n = 4). Conversely, increased glutamate was observed following stimulation in the no-seizure group (p = 0.005, n = 7). Subjects with an atrophic hippocampus had higher basal glutamate levels (p = 0.03, n = 7) and higher stimulation-induced glutamate efflux. SIGNIFICANCE: Electrical stimulation of the epileptic hippocampus either increased extracellular glutamate efflux or induced seizures. The magnitude of stimulated glutamate increase was related to elevation in basal interictal glutamate, suggesting a common mechanism, possibly impaired glutamate metabolism. Divergent mechanisms may exist for seizure induction and increased glutamate in patients with epilepsy. These data highlight the potential risk of 50 Hz stimulation in patients with epilepsy.