Authors :
Xiaoxuan Liu, MS – Washington University Schoold of Medicine in St. Louis
Thomas Foutz, MD, PhD – Washington University School of Medicine in St. Louis
Markus Adamek, MS – Washington University School of Medicine in St. Louis
Kara Donovan, PhD – Washington University School of Medicine in St. Louis
Phillip Demarest, BS – Washington University Schoold of Medicine in St. Louis
Ziwei Li, MS – Washington University School of Medicine in St. Louis
Gansheng Tan, MEng – Washington University School of Medicine in St. Louis
Xie Tao, PhD – Washington University School of Medicine in St. Louis
Enrico Opri, PhD – University of Michigan
Sean McEvoy, MD – Washington University School of Medicine in St. Louis
Jarod Roland, MD – Washington University School of Medicine in St. Louis
Peter Brunner, PhD – Washington University School of Medicine in St. Louis
Presenting Author: Jon Willie, MD, PhD – Washington University Schoold of Medicine in St. Louis
Rationale:
Deep brain stimulation (DBS) of the centromedian (CM) nucleus of the thalamus has been demonstrated to be an effective therapy in controlling generalized epilepsy. However, targeting the CM, a wide range of stimulation parameters, and the prolonged timescale in assessing seizure outcomes pose challenges for optimizing such therapy. Therefore, we identified an accessible electrophysiological biomarker to inform electrode placement in CM-DBS surgery and subsequent programming using intraoperative scalp EEG responses to single-pulse electrical stimulation of the CM.
Methods:
This study included eight patients who had been diagnosed with generalized epilepsy. During lead placement, we delivered biphasic stimulation (1 Hz, six mA, 200 μs) via pairs of adjacent DBS contacts and captured the evoked potentials (EPs) on scalp EEG. Each stimulation condition lasted for two minutes. We evaluated the magnitude of the EPs via peak amplitude and root mean square (RMS) values. We compared these metrics between CM and non-CM stimulation across EEG regions (anterior, central, and posterior). To assess whether the EPs may predict the clinical outcome, we categorized patients into responder and non-responder groups based on their seizure reduction rate, which was greater than or less than 50%. We then determined whether patients who are responsive to CM-DBS exhibit a stronger EP magnitude across EEG regions compared to non-responders. We first used two-way ANOVA tests to determine which factors (i.e., on- or off-target CM stimulation, EEG region, and seizure outcome) had dominant effects on EP magnitudes. We used two-sample t-tests for pairwise comparisons. The statistical threshold was adjusted with Bonferroni’s method.
Results:
At the individual patient level, we observed differential effects on EPs from on- and off-target CM stimulation. Both EEG regions and stimulation locations had statistically significant effects on peak amplitude and RMS values (p < 0.05, two-way ANOVA test). Stimulation of on-target contacts within the CM produced significantly greater posterior-region peak amplitudes than off-target stimulation outside the CM (p< 0.05, two-sample t-test). Patients responding to CM-DBS with over 50% seizure reduction exhibited larger RMS values but smaller peak amplitudes in the posterior region; however, these differences were not statistically significant.