Abstracts

Rationale: Several types of implantable medical devices are either approved or under investigation for the treatment of epilepsy. In order to better understand the mechanism of action of these therapies and to investigate next generation features, it is desirable to have a large animal model in which actual devices can be evaluated. The study reported here was conducted to assess the suitability of an ovine model of epilepsy (penicillin injection into the hippocampus) to address these needs. Specifically, the study assessed: 1) the ability to target and stereotactically implant deep brain stimulation (DBS) leads into the anterior nucleus of the thalamus (AN) and hippocampus (HC); 2) reciprocal stimulation and recording from AN and HC electrodes; 3) induction of ictal activity with penicillin injection and monitoring from HC and AN sites; 4) effects of AN DBS on pre-ictal and ictal HC neural activity. Methods: This study was conducted under an IACUC approved protocol. Animals were anesthetized with a standard regimen for surgery. Following anesthesia, 1.5T MRIs were collected and transferred to a surgical planning station. Trajectories for unilateral AN DBS leads (Medtronic 3387), HC depth electrodes (Ad-Tech), and HC catheters for penicillin delivery were calculated, and devices implanted using a frameless stereotactic system (NexFrame). Stimulation was delivered via a custom-designed system based upon implantable DBS hardware. ECoG recordings were collected using standard EEG amplifiers and stored for off-line analysis. Penicillin solution was injected directly into the HC region via the implanted catheter. Following the procedure the animals were euthanized and post-operative MRIs were collected to verify lead and catheter position. Results: Despite limited neuroanatomical atlas information for this species, it was possible to reliably target the AN and HC using MRI based techniques. Stimulation of the AN produced robust evoked potentials (EPs) in the HC which were electrode contact, stimulus intensity, and frequency dependent. Stimulation of HC electrodes generated shorter latency EPs recorded in the AN. Penicillin injection into the HC resulted in a slow progression of ictal activity, dominated by high amplitude spiking. AN DBS at therapeutic parameters resulted in distinct changes in both pre-ictal and ictal HC neural activity. Conclusions: These preliminary results suggest that the sheep model may be suitable for further investigation of implantable devices for epilepsy. The ability to assess the effects of different stimulation and recording configurations on ictal and inter-ictal activity may provide information regarding AN DBS mechanism of action, and potential future therapy advancements.