Automated Quantitative EEG Analysis for Neonates with Hypoxic Ischemic Encephalopathy
Abstract number :
1.102
Submission category :
2. Translational Research / 2C. Biomarkers
Year :
2021
Submission ID :
1826497
Source :
www.aesnet.org
Presentation date :
12/4/2021 12:00:00 PM
Published date :
Nov 22, 2021, 06:54 AM
Authors :
Eva Catenaccio, MD - The Children's Hospital of Philadelphia; Rachel Smith, PhD - Postdoctoral research fellow, Department of Biomedical Engineering, Johns Hopkins University; Raul Chavez-Valdez, MD - Pediatrics/Neonatology - Johns Hopkins University School of Medicine; Vera Joanna Burton, MD, PhD - Neurology/Pediatric Neurology - Kennedy Krieger Institute, Johns Hopkins University School of Medicine; Ernest Graham, MD - Obstetrics and Gynecology - Johns Hopkins University School of Medicine; Charlamaine Parkinson, BSN, MS - Pediatrics/Neonatology - Johns Hopkins University School of Medicine; Dhananjay Vaidya, MBBS, MPH, PhD - General Internal Medicine - Johns Hopkins University School of Medicine; Aylin Tekes, MD - Radiology/Pediatric Radiology - Johns Hopkins University School of Medicine; Frances Northington, MD - Pediatrics/Neonatology - Johns Hopkins University School of Medicine; Allen Everett, MD - Pediatrics/Pediatric Cardiology - Johns Hopkins University School of Medicine; Carl Stafstrom, MD, PhD - Neurology/Pediatric Neurology - Johns Hopkins University School of Medicine; Eva Ritzl, MD, MBA - Neurology - Johns Hopkins University School of Medicine
Rationale: Neonatal hypoxic ischemic encephalopathy (HIE) causes significant neurologic morbidity, which may be partially ameliorated by therapeutic hypothermia (TH). Early biomarkers are needed to predict outcomes. Neonates undergoing TH are often monitored with continuous electroencephalography (cEEG). Preliminary quantitative analyses of EEG data (qEEG) in neonates have prognostic utility but have been limited by labor-intensive customized methodology. In this study we piloted a fully automated method for extracting qEEG measures from clinical cEEG and investigated the performance of qEEG against established markers of severity in HIE.
Methods: Neonates at Johns Hopkins Hospital who underwent TH for HIE from 4/2018-11/2019 were screened for study eligibility criteria and availability of EEG and clinical data. QEEG measures from the start of recording through 6 hours post rewarming were calculated using Persyst (Prescott, AZ), which provides automated (1) spectral power analysis calculated using fast Fourier transformations, and (2) calculation of the suppression ratio (SR; the fraction of the EEG in suppression, defined as ≥0.5 seconds at ≤3 microvolts, averaged over a 60 second epoch), and averaged by phase of TH. Serial plasma levels of central nervous system necrosis (Tau) and inflammation (interleukin [IL]-6,) were measured from remnant clinical laboratory samples.
Non-parametric tests for trends across ordered groups were conducted to determine if qEEG at each phase of TH differed by Sarnat score (mild, moderate, or severe). Kruskall-Wallis tests were conducted to determine if qEEG at each phase of TH differed between neonates with normal versus abnormal imaging (brain MRI). Spearman’s correlation coefficients were calculated between serum biomarkers and qEEG.
Results: In 30 neonates meeting inclusion criteria, SR increased and delta power decreased with higher Sarnat score at every phase of TH (p-values < 0.02). In neonates with abnormal MRI, delta power was lower during TH and during and after rewarming (p-values < 0.05). SR trended higher, but differences were not statistically significant. SR positively correlated with serum IL-6 and Tau at the start of TH and with Tau during rewarming (Table). Delta power was negatively correlated with serum IL-6 and Tau at the start of TH.
Conclusions: Automated quantitative processing of EEG in neonates with HIE correlates with other markers of injury severity including degree of encephalopathy, presence of abnormal imaging findings, and serum biomarkers. These relationships were stronger at the start of TH and during and after rewarming, likely due to the impact of TH.
Funding: Please list any funding that was received in support of this abstract.: This work was supported by the American Academy of Pediatrics Resident Research Grant and by NIH Grant Number NIH/NICHD R01HD086058.
Translational Research