Authors :
Nigel Colenbier, PhD – Clouds of Care NV, Ghent, Belgium
Caroline Neura, M.D. – Centre for Cognitive Neuroscience, Salzburg, Austria
Emiel Vereycken, M.Sc. – Clouds of Care NV, Ghent, Belgium
Gert Vanhollebeke, Ph.D. – Clouds of Care NV, Ghent, Belgium
Velislava Zoteva, Ph.D., M.Sc. – Clouds of Care NV, Ghent, Belgium
Barry Ticho, M.D., Ph.D. – Stoke Therapeutics
Kimberly A. Parkerson, M.D., Ph.D. – Stoke Therapeutics
Presenting Author: Pieter van Mierlo, PhD – Clouds of Care, NV, Ghent, Belgium
Rationale:
Dravet syndrome (DS) is a severe developmental and epileptic encephalopathy typically caused by voltage-gated sodium channel α subunit 1 gene (SCN1A) haploinsufficiency. This haploinsufficiency leads to reduced firing of fast-spiking parvalbumin-positive inhibitory interneurons that form a large part of the GABAergic inhibitory neuron system [1]. Studies investigating how to reliably distinguish DS from age-matched healthy individuals using electroencephalograms (EEGs) are based on small sample sizes [2], and comprehensive characterization of a large cohort is lacking. Here, we used spectral EEG analysis to characterize the differences between children with DS and neurotypical control participants (NCPs) in a large cohort and to evaluate spectral power as a potential biomarker to assess disease modification.Methods:
Spectral power from clinical routine EEG recordings of children with DS (age: 2–18 years; n=74) and age-matched NCPs (n=50) were compared. DS data were collected from baseline recordings of the MONARCH/ADMIRAL Phase 1/2a open-label, multi-center studies of zorevunersen (NCT04442295 [USA]/2020-006016-24 [UK]). NCP data were obtained from epilepsy-unrelated EEGs at Saint-Luc University Hospital (Brussels, Belgium). All data were acquired in a clinical setting using an international 10-20 system. After pre-processing and quality control, 15 minutes of cleaned EEG data per recording (total N=124) were subjected to spectral analysis. Power spectra (1–32 Hz, log-scaled) were computed using Morlet wavelet decomposition, and absolute power values were scaled and log-transformed to express differences in decibels. Spectral differences were quantified by deriving frequency-specific power estimates averaged across electrodes.Results:
Broadly elevated spectral power (1–32 Hz) was observed in children with DS compared with NCPs, with the largest difference observed in the δ-power frequency range (1–4 Hz; Fig. 1). Quantification of spectral differences showed significantly elevated δ-power in patients with DS compared with NCPs (5.02 dB, P < 0.001; Fig. 2A). Further examination of the developing trajectories of δ-power across age demonstrated significant decline of δ-power with age in both groups (NCP: −2.33 dB, P < 0.001; DS: −2.34 dB, P < 0.001; Fig. 2B), without significant difference in slopes between the groups (difference: −0.01 dB, P = 0.983). Conclusions:
Our findings demonstrate that EEG δ-power is useful in distinguishing between children with DS and neurotypical individuals. Although children with DS showed decreasing δ-power with age, similarly to neurotypical individuals [3], they demonstrated a stable elevation across development (ages 2–18 years). This persistent δ-band abnormality may reflect DS-related pathophysiology, warranting its further exploration as a potential biomarker to assess disease modification induced by novel therapies such as zorevunersen.
References:
1. Lopez-Santiago L et al. Epilepsy Curr 2019; 19 (1): 51–53.
2. Hall JC et al. Ann Child Neurol Soc 2024; 2 (2): 92–105.
3. Gasser T et al. Electroencephalogr Clin Neurophysiol 1998; 69 (2): 91–99.
Funding:
Stoke Therapeutics