Abstracts

Rationale: Interictal discharges (IIDs) are believed to result from desynchronization of a large region of cortex. We hypothesize that stimulating at the time of IIDs will result on a different evoked response to stimulation compared to random stimulation. Methods: We have developed a closed-loop stimulation system for electrical stimulation in humans (CLoSES) to detect IIDs and in real-time send a single-pulse direct electrical stimulation (SPES) to a nearby electrode pair. An IID was detected if the smooth absolute power of the intracranial EEG was above threshold for 12.5ms-25ms. Random interleaved stimulation was used as comparison. Three patients with pharmaco-resistant complex partial seizures implanted for clinical reasons voluntarily participated after fully informed consent (1 or 2 stimulation sites each). SPES pulses were 4mA, biphasic width=90us, inter-pulse interval = 53us. Minimum inter-SPES interval was 2s. For each stimulation site (N=5), we analyzed the cortico-cortical evoke potentials (CCEP) in the channel where IIDs were detected and in another channel also adjacent to the stimulation pair (OTHER channel). K-means clustering separated sharp waves from spike and wave and distinct detected phases within a channel (total 11 clusters). We computed maximum absolute amplitude (10-100ms following stimulation) and latency of the z-normalized CCEP with respect to baseline (500ms duration, 100 ms before stimulation). ANOVA, followed by multiple comparison, compared detected and stimulated IIDs (DetSTIM), detected IIDs but not stimulated (DetNoStim), and random stimulation (RandSTIM). Results: Our system successfully detected IIDs and delivered SPES in real-time (latency <10ms). On detected channels, DetSTIM was similar to DetNoSTIM. When separating IIDs in clusters, DetSTIM was significantly different than RandSTIM in 7 out of 11 clusters and DetSTIM was not significantly different than DetNoSTIM in all clusters. On OTHER channels, DetSTIM was similar to RandSTIM in all channels but one in which there were large IIDs (simultaneous to the detected IIDs). For clusters, DetSTIM was not significantly different than RandSTIM in 9 out of 11 clusters. Importantly, RandSTIM was similar in Detected and OTHER channels (pooled t-test: p=0.3). Conclusions: Brain’s response to stimulation was modified when stimulation occurred during an IID. This could be related to the IIDs depleting the region of activation. If confirmed in larger cohorts, this could have implications for neuromodulatory therapies. Funding: This work was supported by the Defense Advanced Research Projects Agency (DARPA) under Cooperative Agreement Number W911NF-14-2-0045, issued by the Army Research Office contracting office in support of DARPA'S SUBNETS program. The views, opinions, and/or findings expressed are those of the authors and should not be interpreted as representing the official views or policies of the Department of Defense or the U.S. Government.