Abstracts

Rationale: Drug-resistant epilepsy patients may undergo surgical intervention to achieve seizure relief, either through tissue resection or stimulation therapy if the seizure onset zone cannot be resected. Responsive neurostimulation (RNS) therapy targets the seizure onset zone with a short burst of high frequency stimulation to interrupt synchronous electrical activity during a seizure. The efficacy of stimulation therapy is variable, with only 14% of patients experiencing periods of seizure freedom for at least a year. The intra-amygdala kainic acid mouse model mimics the features seen in human temporal lobe epilepsy and serves as an ideal model to evaluate human stimulation therapy in the mesial temporal lobe. The purpose of this study is to evaluate the influence of the standard RNS electrical pulse train on the synchronous electrophysiological activity in an in vitro preparation of the mouse mesial temporal lobe.

Methods: In a blinded fashion, kainic acid or saline was administered into the amygdala in 7 mice, and a recording electrode was implanted into the CA1 region in the hippocampus to monitor the development of epilepsy. Four weeks post injection, horizontal brain slices of the hippocampus and entorhinal cortex were collected for electrophysiology recordings. In vitro spontaneous, long-duration seizure-like events were generated using 100 µM 4-AP, a potassium channel blocker, and field potentials were measured from a recording electrode in the medial entorhinal cortex. A stimulation electrode in the angular bundle deployed stimulation during detected seizure-like events. Using the continuous wavelet transform, we measured the power across various frequency bands for pre- and post-stimulation intervals to quantify the reduction of synchronous oscillations in response to RNS.

Results: We analyzed the effectiveness of the classic RNS waveform in 3 kainic-acid treated mice (n=10 brain slices) and 4 vehicle-treated mice (n=14 brain slices). In the one second period following stimulation, broadband power significantly decreased in brain slices from both the kainic acid and saline-treated animals. In both groups, power in the gamma band (30-70 Hz), often associated with the development of spontaneous seizures, was significantly reduced. High gamma power (70-150 Hz) was only significantly reduced in non-epileptic animals.

Conclusions: These results show that clinically used stimulation waveforms have an acute effect in reducing highly synchronous seizure-like activity in the entorhinal cortex. Broadband activity and gamma power post-stimulation were reduced in both epileptic and naïve animals. Notably, high gamma activity, which may play a role in seizure initiation, was only significantly reduced in saline-treated animals. This study may provide evidence that seizure-like events generated in vitro from a temporal lobe epilepsy mouse may be partially refractory to desynchronization through stimulation compared to control animals. In future work, evaluating the efficacy of various waveform parameters on desynchronization may identify promising stimulation strategies that may improve RNS outcomes.

Funding: Please list any funding that was received in support of this abstract.: NIH NINDS F32, NS 114322 (Anderson), NIH NINDS K23, NS 114178 (Rolston).