Authors :
Presenting Author: Vishal Shah, MD – University of Kansas Medical Center
Arun Sebastian, Clinical Engineer – Neuropace
Ryan Lay, MD – University of Kansas Medical Center
Mark Farrenburg, MD – St Francis Medical Center
Margaret Logan, MD – University of Kansas Medical Center
Murtaza Khan, MD – University of Kansas Medical Center
Nancy Hammond, MD – University of Kansas Medical Center
Carol Ulloa, MD – The University of Kansas Medical Center
Patrick Landazuri, MD – Department of Neurology, University of Kansas Medical Center
Utku Uysal, MD – The University of Kansas Medical Center
Rationale:
Responsive neurostimulation stimulation (RNS) is considered a beneficial adjunctive treatment in drug resistant epilepsy. RNS may be employed after intracranial EEG study (iEEG). iEEG seizure onset patterns (SOP) have been well established. Our study compares patients’ own iEEG SOPs to their RNS SOPs, characterizes electrophysiological properties of the RNS SOPs, and assesses the impact of SOP patterns on efficacy.
Methods:
A retrospective chart review was performed for patients who were followed at the University of Kansas Medical Center and underwent RNS placement before 06/01/2024. Patients with >12 months follow-up duration were included in the data analysis.. SOPs recorded in the habitual clinical seizures captured during iEEG and those recorded from RNS were compared. Visual analysis of the SOP (Table 1) was performed based on the Marseilles group classification system into the following eight onset patterns: low voltage fast activity (LVFA), preictal spiking followed by LVFA, burst of polyspikes followed by LVFA, slow wave/DC shift followed by LVFA, sharp theta/alpha waves, beta sharp waves, rhythmic spikes/spike-waves, and delta-brushes. Frequency based analysis of the SOP was performed and stratified to the delta, theta, alpha, beta and gamma bands. Efficacy was gauged on patient reported percentage seizure frequency reduction classified as no change, 1-24%, 25 – 50%, 50 -75%, 75 – 90%,90 – 99%, seizure free and worse. Further analysis of SOP type from location of RNS implantation was performed to assess value of each SOP type from individual location and its effect on efficacy.
Results:
Forty five patients met inclusion criteria. Concordance between iEEG and RNS SOP was found in 34/45 patients (75%) ( Fig 1) . The most common RNS SOP pattern was LVFA (20), sharp theta / alpha waves (9), rhythmic spikes / spike waves (8) and preictal spiking followed by LVFA (6). No RNS SOP described as burst of polyspikes followed by LVFA, slow wave / DC shift LVFA or delta brushes were identified. iEEG captured SOP pattern described as burst of polyspikes followed by LVFA but this SOP was not captured in any RNS SOP. Fifteen (75%) of patients with LVFA RNS SOP reported > 50% improvement in seizure frequency, increasing to 100% for insula and neocortical extra temporal RNS. Only 50% of patients with LVFA SOP from mesial or extra temporal placemen reported > 50% improvement in seizure frequency. Frequency analysis of RNS SOP revealed the most common SOP frequency being gamma frequency (14) followed by beta and delta (12) and theta activity (7). RNS SOP in gamma frequency showed > 50% reduction rate in 78% of patients which increased to 100% (4) when gamma activity is obtained from an insular RNS electrode often showing 90-99% seizure reduction.
Conclusions:
RNS SOP with LVFA activity / gamma frequency activity is associated with > 50% seizure reduction. Subgroup analysis identified LVFA / gamma frequency activity from insular or neocortical extratemporal lobe may have more favorable outcomes. Additionally, there is wider SOP variability in the remaining lobes with an equally variable seizure reduction.
Funding: None