Phase Lag Analyses on Ictal Scalp Electroencephalography May Predict Outcomes of Corpus Callosotomy for Epileptic Spasms.

Masayoshi Oguri,Ayataka Fujimoto,Yoshihiro Maegaki,Sotaro Kanai,Kaoru Ogo,Hideo Enoki,Tohru Okanishi,Kazuo Okanari,Shimpei Baba,Mitsuyo Nishimura,Takashi Himoto

doi:10.3389/fneur.2020.576087

Abstract

Objective: We aimed to clarify the patterns of ictal power and phase lag among bilateral hemispheres on scalp electroencephalography (EEG) recorded pre-operatively during epileptic spasms (ESs) and the correlation with the outcomes following corpus callosotomy.Methods: We enrolled 17 patients who underwent corpus callosotomy for ESs before 20 years of age. After corpus callosotomy, seven patients did not experience further ESs (favorable outcome group), and the remaining 10 patients had ongoing ESs (unfavorable outcome group). We used pre-operative scalp EEG data from monopolar montages using the average reference. The relative power spectrum (PS), ictal power laterality (IPL) among the hemispheres, and phase lag, calculated by the cross-power spectrum (CPS) among symmetrical electrodes (i.e., F3 and F4), were analyzed in the EEG data of ESs from 143 pre-operative scalp video-EEG records. Analyses were conducted separately in each frequency band from the delta, theta, alpha, beta, and gamma range. We compared the means of those data in each patient between favorable and unfavorable outcome groups.Results: Among all frequency bands, no significant differences were seen in the individual mean relative PSs in the favorable and unfavorable outcome group. Although the mean IPLs in each patient tended to be high in the unfavorable outcome group, no significant differences were found. The mean CPSs in the delta, theta, and gamma frequency bands were significantly higher in the unfavorable than in the favorable outcome group. Using the Youden index, the optimal cutoff points of those mean CPS values for unfavorable outcomes were 64.00 in the delta band (sensitivity: 100%, specificity: 80%), 74.20 in the theta band (100, 80%), and 82.05 in the gamma band (100, 80%). Subanalyses indicated that those CPS differences originated from pairs of symmetrical electrodes in the bilateral frontal and temporal areas.Significance: Ictal power and laterality of the ictal power in each frequency band were not associated with the outcomes of CC; however, the phase lags seen in the delta, theta, and gamma frequency bands were larger in the unfavorable than in the favorable outcome group. The phase lags may predict outcomes of CC for ESs on pre-surgical scalp-ictal EEGs.

Highlights

Epileptic spasms (ESs) are seizures leading to muscular contraction, typically involving the axial muscles and proximal limb segments [1]; they appear mainly in patients with West syndrome
The large phase lags of delta, theta, and gamma activities among bilateral hemispheres during ES correlated with unfavorable outcomes following CC
Written informed consent to participate in this study was provided by the participants’ legal guardian/ of kin

Summary

Introduction

Epileptic spasms (ESs) are seizures leading to muscular contraction, typically involving the axial muscles and proximal limb segments [1]; they appear mainly in patients with West syndrome. The patterns of electroencephalography (EEG) and electromyography (EMG) activity observed in these spasms are similar [2]. Ictal EEG findings in these patients comprise three contiguous phases: [1] 15- to 20-Hz spindle-like fast activity in posterior areas, [2] diffuse polyphasic delta/theta waves, and [3] electrodecremental activity [1, 3]. Diffuse polyphasic delta/theta waves occur in 100%, and electrodecremental activity occurs in 70% of patients with ESs [4]. Ictal EMG findings in these patients have rhombus or diamond shapes. When electroencephalographic/video monitoring first came into clinical use, several studies investigated ictal EEG patterns of ESs, including isolated spindle-like activity, high-amplitude slow wave, the spindle-like activity followed by the slow wave, and decremental activity, which follows the slow wave [1]. The scalp EEG data of ESs showed components of wide frequencies from delta to high gamma bands, and the coupling of high gamma and slow wave EEG components associated with the response to medical treatment [5]

Methods

Results

Conclusion