Abstracts

Rationale: Neuromodulation significantly reduces the seizure burden for many individuals with epilepsy, however, seizure freedom is rare, latency to effect is long, and some individuals do not respond. Epilepsy is heterogenous, with subject-specific seizure network dynamics. To efficiently screen and optimize neuromodulation, short-latency biomarkers are needed. Single pulse electrical stimulation and the resulting evoked responses provide a measure of network effective connectivity (directed causal influence between neural populations ). Here, individuals undergoing stereo-EEG including a thalamus electrode completed a trial of high-frequency (HF) SANTE-like (145 Hz) thalamic stimulation, with single pulse thalamocortical evoked potentials measured at baseline and following HF stimulation to assess changes in network connectivity and excitability. Modulation of interictal epileptiform discharges was assessed.


Methods: Ten patients undergoing clinical sEEG including a thalamus lead completed a trial of HF (145 Hz) thalamic duty-cycle stimulation. Single electrical pulses were delivered to the thalamus (~0.2 Hz, 200 µs pulse width, 6mA) at baseline and following HF stimulation. Thalamocortical evoked potential root-mean-square power assessed network effective connectivity between the thalamic stimulation site and each recording contact. The rate of interictal epileptiform spikes was assessed in the hours preceding the stimulation trial, and during the stimulation trial, using a validated classifier (Janca R, Brain Topogr. 2015). Modulation of network effective connectivity was assessed by Cohen’s D effect size. Statistical significance was assessed by paired-sample t-test, or Mann-Whitney U test, with 0.05 significance level. The Mayo Clinic IRB approved the study and all patients provided informed consent.


Results: HF stimulation delivered for >1.5 hours (duration of active phase only of duty-cycle) was associated with a significant reduction in network excitability (effective connectivity) (n=6 subjects), and the degree of modulation was proportional to the strength of baseline effective connectivity. Suppression of network excitability was not seen for individuals who received < 1.5 hours of active-phase HF stimulation. Suppression of interictal epileptiform discharges during active-phase HF stimulation was immediate and was more pronounced for seizure networks with greater baseline effective connectivity (Cohen’s D > 1.0) to the thalamic stimulation site (P< 0.001).


Conclusions: Single pulse and repetitive stimulation delivered to the thalamus during sEEG provide new insights into large scale brain dynamics. The immediate suppression of interictal epileptiform spikes by HF stimulation may reflect direct electrical effects, while delayed modulation of network effective connectivity may reflect changes associated with synaptic plasticity. This work also provides direct electrophysiological evidence for the network theory of DBS. Short latency biomarkers of seizure network excitability/irritability could enable new approaches to efficiently optimize neuromodulation and improve patient care.


Funding: National Institute of Mental Health Award R01MH122258