Properties of beta oscillations in Dup15q syndrome

• Main Authors: Vidya Saravanapandian, Joel Frohlich, Joerg F. Hipp, Carly Hyde, Aaron W. Scheffler, Peyman Golshani, Edwin H. Cook, Lawrence T. Reiter, Damla Senturk, Shafali S. Jeste
• Format: Article
• Language: English
• Published: BMC 2020-08-01
• Series: Journal of Neurodevelopmental Disorders
• Subjects: Dup15q syndrome; Autism; Biomarkers; EEG; GABA; UBE3A
• Online Access: http://link.springer.com/article/10.1186/s11689-020-09326-1

Abstract Background Duplications of 15q11.2-q13.1 (Dup15q syndrome) are highly penetrant for autism, intellectual disability, hypotonia, and epilepsy. The 15q region harbors genes critical for brain development, particularly UBE3A and a cluster of gamma-aminobutyric acid type A receptor (GABAAR) genes. We recently described an electrophysiological biomarker of the syndrome, characterized by excessive beta oscillations (12–30 Hz), resembling electroencephalogram (EEG) changes induced by allosteric modulation of GABAARs. In this follow-up study, we tested a larger cohort of children with Dup15q syndrome to comprehensively examine properties of this EEG biomarker that would inform its use in future clinical trials, specifically, its (1) relation to basic clinical features, such as age, duplication type, and epilepsy; (2) relation to behavioral characteristics, such as cognition and adaptive function; (3) stability over time; and (4) reproducibility of the signal in clinical EEG recordings. Methods We computed EEG power and beta peak frequency (BPF) in a cohort of children with Dup15q syndrome (N = 41, age range 9–189 months). To relate EEG parameters to clinical (study 1) and behavioral features (study 2), we examined age, duplication type, epilepsy, cognition, and daily living skills (DLS) as predictors of beta power and BPF. To evaluate stability over time (study 3), we derived the intraclass correlation coefficients (ICC) from beta power and BPF computed from children with multiple EEG recordings (N = 10, age range 18–161 months). To evaluate reproducibility in a clinical setting (study 4), we derived ICCs from beta power computed from children (N = 8, age range 19–96 months), who had undergone both research EEG and clinical EEG. Results The most promising relationships between EEG and clinical traits were found using BPF. BPF was predicted both by epilepsy status (R 2 = 0.11, p = 0.038) and the DLS component of the Vineland Adaptive Behavior Scale (R 2 = 0.17, p = 0.01). Beta power and peak frequency showed high stability across repeated visits (beta power ICC = 0.93, BPF ICC = 0.92). A reproducibility analysis revealed that beta power estimates are comparable between research and clinical EEG (ICC = 0.94). Conclusions In this era of precision health, with pharmacological and neuromodulatory therapies being developed and tested for specific genetic etiologies of neurodevelopmental disorders, quantification and examination of mechanistic biomarkers can greatly improve clinical trials. To this end, the robust beta oscillations evident in Dup15q syndrome are clinically reproducible and stable over time. With future preclinical and computational studies that will help disentangle the underlying mechanism, it is possible that this biomarker could serve as a robust measure of drug target engagement or a proximal outcome measure in future disease modifying intervention trials.