Neural Fragility of the Intracranial EEG Network Decreases Intraoperatively After Surgical Resection of the Epileptogenic Zone in Children with Epilepsy
Abstract number :
3.103
Submission category :
2. Translational Research / 2C. Biomarkers
Year :
2021
Submission ID :
1825561
Source :
www.aesnet.org
Presentation date :
12/6/2021 12:00:00 PM
Published date :
Nov 22, 2021, 06:44 AM
Authors :
Adam Li, BS, MS, PhD - Johns Hopkins University; Patrick Myers, BS, MS - Johns Hopkins University; Chester Huynh, BS, MS - Johns Hopkins University; Nebras Warsi, MD - The Hospital for Sick Children; Kristin Gunnarsdottir, PhD - Johns Hopkins University; Soo Kyung Kim, BS, MPH - Stonybrook University; Viktor Jirsa, PhD - Aix-Marseille University; Sridevi Sarma, PhD - Johns Hopkins University; George Ibrahim, MD - The Hospital for Sick Children
Rationale: For selected patients with drug-resistant epilepsy (DRE), the most effective treatment remains surgical resection of the epileptogenic zone (EZ). Successful outcomes from resective surgery, critically depend on accurate identification and resection of the EZ. In six pediatric patients with DRE, intraoperative electrocorticography (iEEG) was performed using the same electrode arrays that were implanted for extraoperative monitoring. We applied a neural fragility algorithm, based on dynamical systems theory, to the iEEG data and compared pre and post resection intraoperative recordings. Neural fragility is a nodal metric that captures how susceptible a region is to small perturbations generating unstable seizure phenomena. We hypothesized that neural fragility of the entire brain network will decrease when the EZ is successfully resected.
Methods: Extraoperative and intraoperative recordings data from the same chronically implanted electrodes were obtained from 6 patients with DRE who underwent invasive monitoring between January 2017 and December 2019 from The Hospital for Sick Kids (SickKids). We applied the neural fragility algorithm which transforms iEEG data over a recording session into a spatiotemporal heatmap. We computed effect sizes (Cohen’s D) and p-values on differences between the pre and post resection session neural fragility heatmaps. To estimate a valid 95% confidence interval, we used contiguous bootstrap samples (n=100) over the time-axis. To compute p-values comparing neural fragility heatmaps, we used a nonparametric K-Sample MANOVA test, and to compare HFO rates, we used a Wilcoxon rank-sum test. To compute HFOs, we used an RMS detector.
Results: Of the six SickKids patients, one had Engel class III outcome (patient E1), while the rest had Engel I outcome. A fragility heatmap for each patient was generated that shows the fragility of EEG contacts over time. In the patient with Engel III outcome, the network increases in fragility after resection, whereas in all patients with Engel I outcome, we see a decrease in fragility after resection. Moreover, we observe that in terms of Cohen’s D effect size and pvalue generated using a nonparametric bootstrap procedure, subject E1 had a significant increase in fragility after surgery, whereas the rest had a significant decrease. Modulation after resection is not observed in HFOs in the same manner.
Conclusions: The novel network fragility algorithm accurately separates the Engel I and Engel II-IV surgical outcomes. The patient with high fragility after resection continued to have seizures, while all patients with low fragility after resection remained seizure free for at least 12 months. Importantly, we did not observe this trend when we analyzed HFOs or other spectral metrics. Neural fragility as a tool may enable clinicians to identify the seizure focus in DRE patients and guide intraoperative decision-making surrounding extent of resection.
Funding: Please list any funding that was received in support of this abstract.: NIH T32 EB003383, NSF GRFP (DGE-1746891), ARCS Scholarship, Whitaker Fellowship, Chateaubriand Fellowship, NIH R21 NS103113, Human Brain Project SGA3.
Translational Research