Topic: Resting State Activity

Clinical Trial Readiness of MEG Biomarkers in Children Across the Autism Spectrum

PROJECT SUMMARY Biological and phenotypic heterogeneity of autism spectrum disorder (ASD) poses a major challenge for clinically focused research and interventions. Brain electrophysiological phenotyping holds promise for parsing this heterogeneity. Using magnetoencephalography (MEG), findings of diminished and delayed auditory evoked responses (e.g. the ~50ms component, M50 and, specifically, its latency: M50L) have reproducibly been shown in ASD, with correlation to behavior. Additionally, abnormal resting state activity and network functional connectivity has been identified as an electrophysiological hallmark. Such passively-acquired signatures may serve as objective biomarkers in subtyping autistic individuals, including stratifying patients for inclusion in clinical trials according to biology, rather than behavior alone. However, despite their abundant promise, these measures are not yet permeating clinical trial design, nor being utilized in clinical practice, in part because of their lack of standardized implementation and analysis. This proposal seeks to remedy this by using rigorous and standardized, scalable and sharable methods with two leading MEG measures to determine their measurement- reliability as well as their sensitivity to inter-individual differences in clinically-relevant aspects of autism features, general cognitive ability and language and communication. Specifically adopting a 12-week repeated scanning design, mimicking the duration of a typical pharmaceutical trial or behavioral intervention, we will acquire each of these two MEG metrics at baseline and 12-week follow-up to assess interval change. Additionally, we will evaluate test-retest variability with an intermediate measurement point 4-weeks after baseline. As such we will characterize both intra-subject variability (measurement precision) and inter-subject variability which will be correlated with dimension axes of autism features, general cognitive ability and language skills, as well as major co-occurring condition confounds. These studies will recruit a broad range of 240 autistic children, paralleling the CDC’s prevalence data on intellectual ability and encompassing the group considered as having “profound autism”. This is enabled by our adoption of MEG-PLAN, a strategy developed over the last decade in our group and demonstrated to enhance inclusive participation in MEG scanning studies, even in non-verbal participants. Data will be compared to a control group of age-matched typically-developing peers. The two MEG measures will also be assessed for their ability to identify clusters of less heterogeneous neurophysiological phenotype as a novel basis for stratification or subtyping of the heterogeneous autism population. In culmination, this study addresses key “clinical readiness” aspects of utilization of MEG biomarkers for ASD including profound autism, for both stratification (inclusion/trial selection) and monitoring of response to intervention, and will, ultimately, pave the way for the adoption of such biomarkers as adjunctive tests in increasingly-routine clinical practice.