Hippocampal High-Frequency Oscillation in-vivo and ex-vivo: A pilot study combining stereo-EEG and human organotypic slices

Introduction

Invasive video-EEG monitoring using depth electrodes (iEEG) is a reliable and safe method for determining the resection zone in the pre-surgical evaluation of non-lesional epilepsy. Since this examination can only be performed for relatively short period, safe lateralization of the resection zone based solely on electroclinical seizures may be difficult. Invasive recordings also facilitate the examination of high-frequency oscillations (HFOs), with frequencies >80Hz. The role of hippocampal HFOs in the ripple range (frequency 80-250Hz) is controversially discussed in the literature. Moreover, the underlying physiologic and pathophysiologic mechanisms of HFO emergence are not sufficiently understood. Ex-vivo analyses in resected epileptogenic tissue could help elicit this pertinent question.

Objectives

We aim to investigate the presence and frequency Ripple-HFOs in-vivo and ex-vivo. In iEEG-patients HFO frequency could provide diagnostic benefits for determining the lateralization of the resection zone in temporal lobe epilepsy, especially when comparing the number of HFOs on both sides. In postsurgical resection tissue, we investigate and strive to replicate our in-vivo findings of HFOs, using a multi-electrode array (MEA) system.

Material & Methods

We include subjects undergoing iEEG with bitemporal depth electrodes for presurgical evaluation. We analyze the presence and frequency of Ripple HFOs in hippocampal electrode contact for each hemisphere. Due to lower artifact overlay, evaluation is performed on nocturnal EEG segments from 2 different nights. HFO determination is semi-automated using the Matlab-based program Ripplelab, and the ratio of HFOs/minute between the seizure onset side and the contralateral side, respectively between both side in patients with bilateral seizure onset, is documented. After epilepsy surgery we retrieve tissue specimen from the putative epileptogenic zone and, if available, from access tissue to prepare organotypic slices. Network Activities and HFOs are then examined in a 256-channel MEA system. HFO extraction is performed using Ripplelab. Slice cultures are then exposed to different temperatures and medical agents to induce or inhibit the occurrence of epileptic network activity.

Results

So far, 5 subjects have been included in the ripple ratio analysis. 3 of them had a unilateral temporal lobe epilepsy and showed a significantly reduced occurrence of interictal Ripple HFOs in the side-by-side comparison across all mesiotemporal depth electrode contacts on the side of the electroclinical seizure onset (Ratio 0.19, 0.21, 0.11). In the 2 patients with bilateral temporal lobe epilepsy the ratio was > 0.25, so we propose a cut-off-ratio of 0.25 to distinguish between these two groups. In organotypic slice cultures, we were able to show the occurrence of HFO in hippocampus tissue, suggesting that the HFO emerge in this area. We could show that treatment with norepinephrine increased the frequency, whereas AMPA-inhibitors decreased HFO-occurrence. Furthermore, HFO frequency increased with elevated temperature.

Conclusions

Our data suggests that the Ripple HFO ratio is helpful as an additional diagnostic tool for identifying focal lateralization in temporal lobe epilepsy. We could further show that HFOs are also present in human hippocampal organotypic slices and responsive to temperature and medical treatment. This opens new perspectives to better understand HFO pathophysiology.