Sevoflurane-based Enhancement of Phase-amplitude Coupling and Localization of the Epileptogenic Zone
Abstract number :
3.085
Submission category :
2. Translational Research / 2C. Biomarkers
Year :
2023
Submission ID :
557
Source :
www.aesnet.org
Presentation date :
12/4/2023 12:00:00 AM
Published date :
Authors :
Presenting Author: Masaki Sonoda, MD, PhD – Yokohama City University
Keiko Wada, MD, PhD – Department of Anesthesiology – National Center Hospital, National Center of Neurology and Psychiatry; Ethan Firestone, PhD – Children’s Hospital of Michigan, Detroit Medical Center, Wayne State University; Kazuki Sakakura, MD, PhD – Wayne state university; Naoto Kuroda, MD, PhD – Wayne state university; Yutaro Takayama, MD, PhD – Yokohama City University; Keiya Iijima, MD, PhD – National Center of Neurology and Psychiatry; Masaki Iwasaki, MD, PhD – National Center of Neurology and Psychiatry; Takahiro Mihara, MD, PhD – Yokohama City University; Takahisa Goto, MD, PhD – Yokohama City University; Tetsuya Yamamoto, MD, PhD – Yokohama City University; Eishi Asano, MD, PhD – Wayne state university; Tomoyuki Miyazaki, MD, PhD – Yokohama City University
Rationale:
Phase-amplitude coupling between high-frequency (≥150 Hz) and delta (3-4 Hz) oscillations - modulation index (MI) - is a promising, objective biomarker of epileptogenicity. Sevoflurane has been reported to increase the occurrence rate of interictal spike discharges and high-frequency events, in a dose-dependent manner. On the other hand, too high a concentration of sevoflurane could induce sharp transients and high-frequency discharges unrelated to epileptogenicity. First, we examined whether MI measured under sevoflurane anesthesia preferentially increases at epileptogenic sites and whether it is useful in localizing that area. Next, we identified which anesthetic stage optimizes the accuracy of MI-based classification of epileptogenic and non-epileptogenic sites.
Methods:
This is an observational study of intraoperative electrocorticography data from 621 electrodes in eight patients with drug-resistant, focal epilepsy. All patients were anesthetized with sevoflurane during resective surgery, which subsequently resulted in seizure control. We classified 'removed' and 'retained' brain sites as epileptogenic and non-epileptogenic, respectively. Mixed model analysis and ROC analysis determined what anesthetic stage optimized MI-based classification of the epileptogenic sites.
Results:
The mixed model analysis demonstrated that more advanced sevoflurane anesthetic stages were associated with increased MI (fixed effect 95% confidence interval [CI]: +0.012 to +0.015; p-value: < 0.001; degree of freedom [DF]: 3040; Figure 1A). Furthermore, mixed model analysis incorporating the interaction between the anesthetic stage and baseline MI values suggested that more advanced sevoflurane stages increased MI preferentially at electrode sites with higher baseline MI values (fixed effect 95% CI: +0.48 to +0.62; p-value: < 0.001; DF: 3038). The logistic mixed model analysis indicated that the 'last 1/3' anesthetic stage optimized the MI-based classification of epileptogenic and non-epileptogenic sites. For each increase of 0.2 MI during the 'last 1/3' anesthetic stage, the odds of a given site being classified as epileptogenic increased by 2.44 times. The fixed effect estimates (odds ratio; p-value) of the local MI values were -2.04 (0.13; 0.44) during ‘first 1/3’, +2.90 (18.2; 0.10) during ‘middle 1/3’, +4.46 (86.6; 0.003) during ‘last 1/3’, and +1.38 (3.97; 0.11) during ‘2.0 MAC maintenance’ stage (Figure 2). The ROC analysis during the ‘last 1/3’ anesthetic stage indicated that MI classified the epileptogenic site with an area under the curve (AUC) of 0.71 (bootstrap 95% confidence interval: +0.65 to +0.77, n=5,000). Thereby, electrode sites with a model probability above 0.45 were classified as epileptogenic with a specificity of 0.95 and a sensitivity of 0.23 (Figure 1B).
Translational Research