There is a need for objective biomarkers to assist clinicians with the diagnosis of childhood neurodevelopmental disorders. A number of investigators have reported changes in blood gene expression associated with autism spectrum disorders; a recent paper by Voineagu [1] reviews this literature.
Objectives:
The aim of this study was to assess whether blood gene expression could provide a biomarker to distinguish children on the autism spectrum from children with other conditions that might present in the same clinical setting.
Methods:
We used Affymetrix U133plus2.0 gene expression microarrays to profile blood samples from 235 subjects from the CHARGE (CHildhood Autism Risks from Genetics and the Environment, [2]) Study. 103 of these subjects were diagnosed as being on the autism spectrum (ASD) based on ADI-R and ADOS scores. Of the remaining subjects, 83 had been referred for evaluation for developmental delay and found not to be on the autism spectrum (NAR, for non-autism referred); this group included 16 found to be typically developing on evaluation. An additional group of 49 subjects were recruited as typically developing controls (TD). Individual genes were evaluated for differential expression by t-test between the different pairs of groups, and between each group and the other two.
Results:
Although initial evaluations showed no significant differences between the groups, after adjustment for sample and process variables, including batch effects, and removal of the largest principal components to account for unknown masking variables, a significant excess of differentially expressed genes compared to random class label permutations was observed. The largest differences were observed between the ASD and NAR groups. 5-fold cross-validated machine learning using a radial basis support vector machine classifier yielded an AUC of .67 for distinguishing between these two groups.
Conclusions:
These results provide support for further research and larger studies to determine whether gene expression differences can be informative for differentiating autism spectrum disorders from other forms of developmental delay. Planned followup studies include increasing the sample sizes and use of RNASeq for expression profiling.
[1] Voineagu, I., Gene expression studies in autism: Moving from the genome to the transcriptome and beyond, Neurobiol. Dis. (2011), doi:10.1016/j.nbd.2011.07.017
[2] Hertz-Picciotto I, Croen L, Hansen R, Jones C, Pessah IN (2006). The CHARGE Study: An epidemiologic investigation of genetic and environmental factors contributing to autism. Environ Health Persp 114(7):1119-1125. http://www.ehponline.org/members/2006/8483/8483.pdf