Electrophysiological Markers of ASD in Infants with Tuberous Sclerosis Complex: A Genetics-First Approach to the Search for Predictive Biomarkers

Thursday, May 11, 2017: 5:30 PM-7:00 PM
Golden Gate Ballroom (Marriott Marquis Hotel)
K. J. Varcin1, A. Dickinson2, J. Frohlich2, D. Senturk3, S. Huberty3, L. M. Baczewski4, C. A. Nelson5 and S. S. Jeste6, (1)Telethon Kids Institute, Perth, Australia, (2)University of California, Los Angeles, Los Angeles, CA, (3)University of California Los Angeles, Los Angeles, CA, (4)Boston Children's Hospital Labs of Cognitive Neuroscience, Cambridge, MA, (5)Boston Children's Hospital, Boston, MA, (6)UCLA, Los Angeles, CA
Background: Tuberous sclerosis complex (TSC) is one of the most commonly-occurring single-gene disorders associated with ASD. Approximately 50% of children with TSC will meet criteria for ASD. Recent evidence has demonstrated striking phenotypic homology in the profile of social communication impairment amongst toddlers with TSC/ASD and toddlers with non-syndromic ASD (Jeste, 2016). In addition, infants with TSC/ASD can be distinguished from infants with TSC no-ASD as early as 12-months of age on the basis of non-verbal cognitive ability. However, there is currently limited understanding regarding the neurobiological mechanisms underlying the development of ASD in TSC. Conversely, animal and human model research has identified characteristic neurobiological aberrations in the context of a TSC mutation, namely in atypical white matter development (including hypomyelination). There is preliminary evidence from older children that these alterations may be most severe in children with TSC/ASD.

Objectives: We aimed to identify potential predictive biomarkers of ASD in TSC in order to inform our understanding of mechanistic pathways to ASD in TSC. Specifically, we examined whether developmental trajectories of electroencephalographic (EEG) power and peak alpha band frequency distinguish infants with TSC/ASD from infants with TSC no-ASD. We examined these specific electrophysiological markers as they have been linked to alterations in white matter development, and hence, may be sensitive to atypical development in TSC in early infancy.

Methods: Data reported here form part of a larger, multisite, prospective study of infants with TSC across the first three years of life. We collected high spatial density resting-state EEG recordings from infants with TSC (n = 40) and typically developing infants (TD; n = 32) at 12, 18, 24 and 36 months. ASD diagnosis at 24 and 36 months was determined using the Autism Diagnostic Observation Schedule and clinical best estimate. We calculated whole-brain relative power in delta (1-4Hz), theta (4-6Hz), low-alpha (6-8Hz), high-alpha (8-11Hz), beta (11-35Hz) and gamma (35-50Hz) bands, as well as peak alpha frequency (between 6-11Hz) for each infant at each age.

Results: Mixed-effects models revealed reduced low-alpha power across early development in infants with TSC/ASD compared to infants with TSC no-ASD at all ages (p < .01). As a group, infants with TSC exhibited lower relative low-alpha (p < .001) and gamma (p < .001) power from 12-36 months and reduced peak alpha frequency from 18-36 months compared to TD infants (ps < .05). Gamma power at 12 months was positively correlated with non-verbal abilities in infants with TSC (r = 0.34, p = .01).

Conclusions: Our results suggest that reduced whole-brain low-alpha power across early development may be a neurophysiological signature associated with ASD in TSC. Reductions in peak alpha frequency and relative whole-brain gamma power are associated with TSC as a whole, rather than ASD. Alterations in EEG power and alpha peak frequency are consistent with the characteristic disruptions to white matter development in TSC. Follow-up analyses will examine patterns of neural connectivity and brain-behavior associations to further inform our understanding of cortical alterations associated with ASD in the context of TSC.