Abstracts

Rationale: Sudden Unexpected Death in Epilepsy (SUDEP) is the leading cause of non-accidental death in patients with refractory epilepsy, with an annual incidence of 1 in 1000 adults. Prolonged PGES is a biomarker of decreased arousability leading to SUDEP; and spontaneous cessation of non-terminal PGES provides a unique window into understanding the underlying network mechanisms. Currently, the role of the thalamus is widely known to be crucial for cortical arousal. Our first-in-human study, however, shows involvements of extra-thalamic pathways in the ascending arousal network that are activated preceding the first motor response, which can be harnessed toward neuromodulatory rescue therapy.

Methods: A total of 12 adult patients with temporal lobe epilepsy were implanted with SEEG electrodes to record local field potentials (LFP) from one of the four different thalamic nuclei (ANT, MD, CM, and CL) as well as - Amygdala, Insula, Temporal lobe, Frontal lobe, Cingulate gyrus, Orbitofrontal Cortex, Anterior, and Posterior Hippocampus. Band-specific power was estimated using Thomson’s Multitaper power spectral density estimation for a frequency range of 1 - 150 Hz. The Continuous Morse Wavelet Transform was used to perform time-frequency decomposition of the post-ictal state, which were normalized with respect to the means and standard deviations of stage I isoelectric suppression. Pairwise normalized cross-correlation histograms (NCCH) were computed for two second windows centered around the first motor response, and the temporal order of activation was estimated. The significance of the difference in mean band-specific power and high-frequency activity was tested using paired t-test and one-way ANOVA with Bonferroni corrections, respectively.

Results: Three distinct states were identified during scalp EEG confirmed PGES state. Stage I consisted of absent arousal response and suppression of all recorded LFPs in the lower-frequencies (1 - 70 Hz) compared to the pre-ictal segments (Figure 1A). Concurrently, the high-frequency activity (70 - 150 Hz) was significantly lower in the post-ictal period, when compared with selected awake and sleep segments, which indicated reduced ensemble neural activity within the thalamocortical network (Figure 1B). Stage II was characterized by the emergence of narrow spectral bands in lower frequencies in extra-thalamic regions (insula, orbitofrontal, or posterior hippocampus regions) or CM and CL, but not ANT thalami. Concurrently, the patients were in a state of low vigilance. Stage III consisted of a sharp increase in LFP broadband power throughout the channels which was followed by the first motor response (Figure 2A-C). It was observed that the thalamus was not the first region to activate (Figure 2D-E), which suggests an involvement of extra-thalamic pathways during motor arousal.

Conclusions: This first-in-human study confirms the absence of arousal response during PGES. The return to consciousness is characterized by arousal response in the extra-thalamic and thalamic nodes. Our study departs from the “thalamocentric” focus on arousal and suggests incorporating extra-thalamic structures as neuromodulatory targets.

Funding: The authors received no specific funding for this work.