Rare Copy Number Variation As Predictors for Treatment Response to Social Skills Training in Children with Autism Spectrum Disorder

Background: Social skills group training (SSGT) is one the most commonly used interventions for individuals with autism spectrum disorder (ASD). However, the response to SSGT can vary. Limited information exists of predictors of the SSGT treatment outcome. Chromosomal microarray (CMA), to analyze the presence of rare copy number variation (CNVs), is currently the recommended first-tier molecular diagnostic test for children with ASD. However, studies investigating the treatment outcomes in the individuals with CNVs are lacking.

Objectives: Analyze the association between rare CNVs and response to SSGT in children and adolescent with ASD

Methods: Saliva-derived DNA was used to genotype 207 consenting participants from the KONTAKT SSGT study (Choque-Olsson et al. J Am Acad Child Adolesc Psychiatry 2017; 56(7):585–592) using the Affymetrix Cytoscan HD microarray. CNVs were called using two different algorithms followed by annotation and selection of rare CNVs using population-based dataset. The rare CNVs were further stratified by size and gene context. The response to SSGT was measured using changes in the parental reported Social Responsiveness Scale (SRS-2) from baseline to post-treatment and 6-months follow-up. Secondary outcome measures included parental reported Adaptive Behavior Assessment System II (ABAS-II) and trainer reported Developmental Disabilities modification of the Children’s Global Assessment Scale (DD-CGAS). Linear mixed models were used to test associations between rare CNVs and change in SRS-2 or other outcome measures. The outcome was considered first using data from all time points, then at the posttreatment or follow-up assessments specifically. We additionally investigated differences in the baseline phenotypic measures, such as full-scale IQ and ADOS-2 total scores, between individuals with and without rare CNVs

Results: A total of 101 rare CNVs or chromosomal abnormalities were identified in the participants. These included known microdeletion and duplication syndromes such as 7q11.23dup but also rare CNVs affecting newly identified ASD genes such as CHD8. Our analysis reveal that participants in the active KONTAKT SSGT group that were carriers of exonic CNVs had overall worse outcomes at all times point (β=12.1, p=0.018). This association was mainly driven by CNVs larger than 500kb, for which significant association was shown for both posttreatment (β=15.3, p=0.017) and follow-up (β=14.2, p=0.028). These results indicate a significant increase in the symptom severity after SSGT. Similar results were obtained for ABAS-II. However, no significant association was found for the trainer reported DD-CGAS outcome and rare CNVs. The individuals with rare CNVs had significantly lower IQ than non-carriers, but they did not differ in other baseline measures tested.

Conclusions: This is the first study to report that carriers with rare exonic CNVs had worse and even adverse outcome after completing SSGT. Our results suggest that currently used molecular diagnostic test CMA can potentially be used to pinpoint subgroup of individuals that do not benefit from SSGT. In the future, genomic testing will lead to recommendations on how to better tailor the limited resources in treatments for affected individuals thus enhancing the personalized medicine approach in ASD.