Authors :
Presenting Author: Caroline Neuray, MD – Epilog, Clouds of care NV
Susan Boronat, MD, PhD – Hospital de la Santa Creu i Sant Pau, Barcelona, Spain; Brecht Nerinckx, MSc – Epilog, Clouds of care NV; Pieter Van Mierlo, PhD – Epilog, Clouds of care NV; Ekatherina Garzon, MSc – Epilog, Clouds of care NV; Andreas Brunklaus, MD – School of Health & Wellbeing, University of Glasgow, Glasgow, Scotland; Matt Lallas, MD – Nicklaus Children’s Hospital, Miami, Florida USA; Kelly Knupp, MD – Children’s Hospital Colorado, Aurora, Colorado USA; Scott Perry, MD – Cook Children’s Medical Center, Fort Worth, Texas USA; Joseph Sullivan, MD – University of California at San Francisco, San Francisco, California USA; Salvador Rico, Md, PhD – Encoded Therapeutics, South San Francisco, California USA; Jacqueline Gofshteyn, MD, PhD – Encoded Therapeutics, South San Francisco, California USA
Rationale:
Dravet syndrome (DS) is a severe developmental and epileptic encephalopathy that has gained increased focus for development of novel therapeutics that address both seizure and non-seizure manifestations. Correspondingly, there is considerable interest in exploring potential electrophysiologic diagnostic, prognostic, and predictive biomarkers. Traditional EEG methods have been instrumental in the diagnosis and management of DS but their utility as a biomarker has been limited due to high inter-individual variability. This is particularly relevant in the very young in whom interictal epileptic discharges (IED) may yet not be present. We compared routine EEGs collected from young children participating in the ENVISION natural history study of SCN1A+ DS with age-matched healthy controls (HCs) provided by Epilog and applied advanced analytics to understand potential distinctions. Methods:
ENVISION was an international, multicenter, longitudinal, prospective study of children with SCN1A+ DS, aged six months to five years. EEGs were analyzed using a semi-automated approach. Background activity was centrally assessed by an expert pediatric epileptologist. IEDs were detected, automatically clustered, and verified by the same expert. We extracted spectral features using Fast Fourier Transformation and calculated functional connectivity between the atlas-based regions using weighted phase-lag index. Graph analysis, degree, and efficiency were computed to define the characteristics of the functional network. Neuropsychological assessments were administered every six months throughout the study to assess developmental trajectory and potential correlation with EEG findings.
Results:
The study included 45 EEGs: 21 from DS individuals (five recorded during sleep) and 24 from HCs (10 recorded during sleep), each with a 30-minute epoch. Mean age was 23 months (range four months to five years and three months) for DS children and 14 months (range two months to three years and two months) for HCs. Background frequency during wakefulness was age-appropriate in 89% (11/14) of HCs and in 69% (11/16) of DS individuals. Only 38% (8/21) of EEGs from DS individuals exhibited IEDs. These occurred during wakefulness and more frequently in younger individuals aged < 1.5 years. When present, IED rates were significantly higher in individuals with DS compared with HCs. We observed significantly lower mean and peak power of the gamma frequency band in DS vs HCs across all channels. Finally, connectivity was globally decreased in DS except within the frontal regions, where some areas of increased connectivity during wakefulness were observed.