Microelectrode-Based Epilepsy Therapy: A Hybrid Neural Prosthesis Incorporating Seizure Prediction And Intervention With Biomimetic Maintenance Of Normal Hippocampal Function

Researchers and clinicians have become intrigued with coupling seizure prediction and local electrical stimulation or pharmacotherapy for treatment of the approximately one million individuals with persistent uncontrolled epilepsy in the USA. Critical for interventions aimed at halting seizure onset are (i) the ability to detect and characterize aberrant neuronal activity indicative of imminent seizure and (ii) the ability to maintain cognitive function normally supported by the brain tissue that must be treated to prevent seizures. A major goal of our research has been the development and refinement of an enzyme-based, mass-fabricated microelectrode array technology that can be implanted into the mammalian central nervous system and utilized for second-by-second assessment of localized glutamate neurotransmission. Based upon our experience, extracellular fluctuations in glutamate do not exhibit the spontaneity and complexity observed in the electrical characteristics of aberrantly firing neurons. We believe predictive modeling based upon chemical neuronal communication may present a very accurate seizure detection system with minimal false-positives and false-negatives currently complicating electroencephalography (EEG)-based seizure prediction. Though a prediction and local intervention therapeutic paradigm could offer a dramatic improvement in treatment of refractory seizures, we propose that achieving the ultimate goal of enhancing the daily functioning of these patients necessitates consideration of consequences experienced by the patient when hippocampal circuitry is 'shut-down' to thwart seizure onset. Therapeutic termination of seizure activity may produce memory-loss and other gross impairments of information processing in the brain; therefore, the second goal of our research stated above is the development of a hippocampal prosthesis to 'patch' therapeutically impaired hippocampal circuitry. Our prosthetic technology utilizes non-linear modeling to generate outputs characteristic of the normal functioning of the epileptic hippocampal circuit (i.e. when that circuit is not epileptiform) during the time when that circuit must be 'shut down' to suppress seizure. The outputs are computed from afferent neuronal input recorded from an indwelling microelectrode array 'upstream' from the impaired hippocampal tissue so as to bypass the malfunctioning portion of the hippocampus. The present chapter documents our current capabilities of measuring glutamate in rodents, primates and humans with in vivo amperometry and our progress with a hippocampal prosthesis, as we work toward development of a 'closed-loop' system for seizure prediction and termination with maintenance of normal hippocampal function.