Authors :
Presenting Author: Joni Saby, PhD – Children's Hospital of Philadlephia
Tim Benke, MD, PhD – Children's Hospital of Denver; Jacinta Soldaris, PhD – Telethon Kids Institute; Scott Demerest, MD – Childrens's Hospital of Denver; Jenny Downs, PhD – Telethon Kids Institute; Peter Jacoby, PhD – Telethon Kids Institute; Helen Leonard, MD – Telethon Kids Institute; Gina Vanderveen, MPh – Children's Hospital of Denver; Eric Marsh, MD, PhD – Children's Hospital of Philadlephia
Rationale:
As clinical trials begin for CDKL5 Deficiency Disorder (CDD) and related developmental encephalopathies, there is a need for objective biomarkers to quantify treatment response. Our previous work provided initial evidence that quantitative EEG (qEEG) measures have utility in representing cortical function and disease severity in CDD (Saby et al., 2022, Brain Communications). This study aims to further validate qEEG as a biomarker of brain function in a predominantly new cohort of individuals with CDD using new CDD-specific clinical assessments developed as part of the International CDKL5 Clinical Research Network.
Methods:
Data from 41 participants with CDD (mean age=9 years, range = 10 months–30.5 years) were analyzed. EEG data were acquired across three sites (Denver, Philadelphia, Boston children’s hospitals; COMIRB 19-2756). Resting EEG was acquired for 15 minutes during which time participants were permitted to watch a silent movie. The EEG files were analyzed using custom MATLAB scripts. Raw files were re-referenced to a Laplacian montage, filtered 1–70 Hz (60 Hz notch), and split into 0.5 s segments. Segments exceeding amplitude and line length thresholds were automatically rejected. Non-overlapping segments of non-rejected EEG of four second duration were then identified and used for calculations of qEEG amplitude and power features. All participants were assessed by a child neurologist and assigned a clinical severity score based on the CDKL5 Clinical Severity Assessment-Clinician (CCSA). A subset of participants also completed video-based assessments of hand (
n = 22) and gross motor function (
n = 19). Hand function was rated on an eight point scale. Gross motor function was scored from seventeen tasks enabling a total score of 51. Higher scores indicated higher function. Linear regressions tested for associations between the qEEG features and the clinical variables.
Results:
No qEEG amplitude features, but several qEEG power features, were significantly associated with severity on the CCSA including 1/f slope and
alpha/delta, beta/delta, alpha/theta, and beta/theta ratios (
ps< .01). The parameters with the strongest linear relationships with CCSA were alpha/delta and beta/delta ratios (Fig. 1). The same qEEG power features that were associated with the CCSA were also associated with hand function and gross motor video scores, specifically, 1/f, alpha/delta, beta/delta, alpha/theta, and beta/theta ratios (