Authors :
Presenting Author: Sameera Vedantam, BA – Baylor College of Medicine
Zulfi Haneef, M.B.B.S, M.D. – Baylor College of Medicine; Alexandra Vallera, BS – Baylor College of Medicine; Stephen Cleboski, M.Eng. – Neuropace; Jay Gavvala, M.D., M.Sci. – Baylor College of Medicine
Rationale:
The Responsive Neurostimulation (RNS) System improves seizures by detecting programmed electrocorticogram (ECoG) patterns and delivering responsive electrical stimulation. It is unknown whether seizure onset patterns (SOPs) detected by RNS are similar to those described in prior literature of intracranial electroencephalography (IC-EEG) (Epilepsia 2018; 60(1):85–95). This study attempts to describe SOPs commonly seen in RNS long episode ECoGs and evaluate the correlation between seizure outcomes and particular SOPs.
Methods:
Among 58 patients with RNS implantations screened, 40 were included after excluding thalamic implants (n=7) and those with insufficient data (n=11). Five ECoGs typical of each seizure pattern were reviewed and various SOPs identified following the seven patterns characterized by the Montreal group (Epilepsia 2018; 60(1):85–95). Each seizure type was assigned an SOP if a majority of the five SOPs were found to be identical. Correlation of seizure outcomes was evaluated with two categories of SOPs – low-voltage fast activity (LVFA) vs. non-LVFA – with ≥ 50% improvement defined as “best outcome."
Results:
Among the 40 patients included, hippocampal-onset seizures were seen in 24 (60%). For each patient, the SOP that occurred most frequently (mode) was identified as the main SOP type. Representative images of the various SOPs identified with associated spectrogram images are shown in Figure 1. The distribution of the various SOPs identified in our study, along with a comparison distribution reported in IC-EEG by Lagarde et al (Epilepsia. 2018; 60(1):85–95) is shown in Figure 2. The most common SOP observed was LVFA, followed by rhythmic slow spikes. There was no correlation found between the presence of LVFA and best outcome (Fisher exact test, p = 0.715, Odds Ratio = 1.63). Inter-rater agreement at the patient level was moderate (Cohen’s Kappa 0.63) for the main SOP pattern.
Conclusions:
This is an initial description of seizure onset patterns commonly seen in RNS ECoG recordings according to the established schema for IC-EEG patterns. Our findings will guide further research of RNS SOPs that could lead to determination of the impact of SOPs on RNS outcomes and may help define specific stimulation parameters that may help with different SOPs.
Funding: None