Suppressing limbic seizures by stimulating medial dorsal thalamic nucleus: factors for efficacy.

Abstract

The optimal sites and stimulation protocols for brain stimulation in epilepsy have not been found. Clinical trials, which have shown modest benefit in seizure reduction, have involved patients with poorly localized intractable focal epilepsy and stimulation sites without clear relations to specific underlying seizure circuits. The medial dorsal thalamic nucleus is a key node in limbic seizure circuits, and we wished to know what stimulation parameters might control seizures in a kindling model of limbic epilepsy. In urethane-anesthetized rats, we induced limbic seizures by stimulation of the piriform cortex or CA3 of the hippocampus while recording in the entorhinal cortex or CA1 of the contralateral hippocampus to determine the effect of specific stimulation parameters on seizure duration. Stimulation consistently suppressed seizure duration from baseline by over 80% (p < 0.001), frequently completely preventing the seizures. Position of the thalamic electrode, stimulus intensity and frequency had a significant influence, with higher stimulus intensities (40 V vs. 20 V) and frequencies (20 Hz vs. 7 Hz) significantly suppressing seizures. The most effective position was the lateral dorsal area of the medial dorsal nucleus (MD), which corresponded to the region of axon entry. Stimulation in the MD center was not effective. An anterior-posterior relationship of the stimulating electrode pair was effective, whereas a medial lateral orientation was not. Successful stimulation suppressed the evoked responses in the entorhinal cortex or CA1. Position and orientation of the stimulating electrode has to be precise, which suggests that the placement of the electrodes must be tailored to the individual's own seizure circuit. The data also indicate that successful deep brain stimulation induces a fundamental change in system physiology, which could be a marker to guide the development of stimulation parameters for each patient.