Scalp High-frequency Oscillations Differentiate Neonates with Seizures from Healthy Neonates and Indicate Postneonatal Epilepsy Risk
Abstract number :
3.214
Submission category :
2. Translational Research / 2C. Biomarkers
Year :
2024
Submission ID :
1026
Source :
www.aesnet.org
Presentation date :
12/9/2024 12:00:00 AM
Published date :
Authors :
Presenting Author: Panagiota Karatza, MSc – University Children’s Hospital Zurich
Dorottya Cserpan, PhD. – University Children’s Hospital Zürich
Lo Biundo Santo Pierto, R.EEG T. – University Children’s Hospital Zürich
Andrea Rüegger, Dr. med. – University Children’s Hospital Zurich
Francesco Pisani, Prof. med. – Child Neuropsychiatry Unit, Neuroscience Department, University of Parma
Johannes Sarnthein, Prof. Dr. – University Hospital Zürich
Georgia Ramantani, Prof. Dr. med – University Children’s Hospital Zürich
Rationale: This study aimed to investigate (1) whether an automated detector capture scalp high-frequency oscillations (HFO) in neonates, (2) whether scalp HFO rates differentiate neonates with seizures from healthy neonates, (3) whether HFO rates differ according to seizure etiology, and (4) whether scalp HFO rates differentiate neonates with postneonatal epilepsy from neonates with normal development.
Methods: The study included 45 neonates with EEG-confirmed seizures, 43 with 1-year follow-up, and 6 healthy neonates. A previously validated automated HFO detector was employed to determine scalp HFO rates during quiet sleep.
Results: Neonatal seizure etiology included hypoxic-ischemic encephalopathy (HIE, n=19), structural vascular lesions (n=14), genetic (n=10), metabolic (n=1), and infectious causes (n=1). Significantly higher HFO rates were observed in neonates with seizures (0.17 ± 0.16 HFO/min/ch) compared to healthy neonates (0.065 ± 0.029 HFO/min/ch; p=0.001). An HFO rate of 0.17 HFO/min/ch yielded the highest Youden index for discriminating neonates with seizures from healthy neonates. Furthermore, HFO rates varied according to etiology (p=0.005): neonates with HIE had significantly higher HFO rates (0.24 ± 0.26 HFO/min/ch) compared to those with structural vascular lesions (0.067 ± 0.049 HFO/min/ch; p = 0.008). In neonates with HIE, therapeutic hypothermia did not affect the HFO rates (p=0.27). EEG background activity did not correlate with HFO rates in neonates with seizures. At follow-up, 8 of 34 surviving neonates with seizures presented postneonatal epilepsy, and 9 developmental delay. Neonates with postneonatal epilepsy had higher HFO rates (0.28 ± 0.24 HFO/min) compared to those without epilepsy or development delay (0.15 ± 0.16 HFO/min/ch). An HFO rate of 0.15 HFO/min/ch yielded the highest Youden index for discriminating neonates with postneonatal epilepsy from neonates without epilepsy or development delay, with a specificity of 0.7, a sensitivity of 0.75, and an AUC of 0.7.
Conclusions: Neonatal scalp HFO can be detected using automated methods, differentiate neonates with seizures from their healthy counterparts, and indicate those neonates at higher risk of postneonatal epilepsy. These observations hold significant implications for neuromonitoring in neonatal care, representing an initial step toward establishing neonatal HFO as a potential biomarker for seizure propensity and epileptogenesis.
Funding: This research was supported by the Swiss National Science (SNSF: 208184) Foundation, and the Vontobel Foundation.
Translational Research