Subphenotype-Dependent Disease Markers for Diagnosis and Personalized Treatment of Autism Spectrum Disorders

Abstract

Abstract. Autism spectrum disorders (ASD) are a collection of neurodevelopmental disorders that are currently diagnosed solely on the basis of abnormal reciprocal language and social development as well as stereotyped behaviors. Without genetic or molecular markers for screening, individuals with ASD are typically not diagnosed before the age of 2, with milder cases diagnosed much later. Because early diagnosis is tantamount to early behavioral intervention which has been shown to improve individual outcomes, an objective biomarker test that can diagnose at-risk children perinatally is a medical imperative. The rapidly increasing prevalence of ASD in the United States (now estimated at 1 in 88 individuals) also makes early diagnosis and intervention a public health imperative. This article reviews recent genome-wide (genomic) approaches to the identification of disease markers that may be used not only for diagnosis of ASD, but also for the informed development of novel drugs that target specific core symptoms of ASD. Because of the heterogeneity of clinical manifestations associated with the ASD population, this review also addresses the importance of dividing individuals with ASD into clinically relevant subphenotypes in the quest to identify appropriate biomarkers.Keywords: Autism, clinical phenotypes, genomics, gene expression, genetics, epigenetics, gene-environment interactions Diagnosis of autism spectrum disordersAutism spectrum disorders (ASD) are a group of neurodevelopmental disorders that are characterized behaviorally on the basis of difficulties in initiating and maintaining reciprocal social interactions, delayed or abnormal language development and usage, and stereotyped repetitive behaviors, often with restricted interests [1]. Although recently reported to affect 1 in 88 individuals in the United States with a male-to-female ratio exceeding 4:1 [2], there are still no genetic or molecular biomarkers that can be used to unequivocally diagnose idiopathic, or nonsyndromic, ASD. This is in contrast to the genetically defined syndromic disorders, such as Fragile X, Rett, and Smith-Lemli-Opitz Syndromes, which are associated with a relatively high risk for ASD (reviewed in [3]). The difficulty in identifying genes and other biological markers for ASD arises at least in part from the phenotypic diversity associated with this broad spectrum disorder, which undoubtedly is the result of multiple etiologies. Thus, a key strategy in the identification of potential biomarkers for ASD is the stratification of the population into more homogeneous clinical subgroups, each of which may reflect a shared biological phenotype. Also implicit in the design of studies to identify disease markers for ASD is the need to use non-neuronal surrogate tissues, such as blood, that can be easily accessed for diagnostic screening in the clinic. In this article, I will provide a brief overview of various methods that have been used to reduce the heterogeneity of ASD for genetic/genomic analyses, and then describe some of our re...

Phenotypic subgrouping and multi-omics analyses reveal reduced diazepam-binding inhibitor (DBI) protein levels in autism spectrum disorder with severe language impairment

“…In the present study, we therefore sought to investigate the proteome profiles of patients with ASD by reducing the heterogeneity of the ASD population using a phenotypic subgrouping strategy that we employed in recent studies [10,16,17,39–41], followed by transcriptome and proteome profiling analyses. First, a cluster analysis of clinical phenotypes obtained from the Autism Diagnostic Interview-Revised (ADI-R) scores from individuals with ASD was performed to identify subgroups/clusters of ASD based on clinical phenotypes.…”

Trends in documented co-occurring conditions in children with autism spectrum disorder, 2002–2010

“…Investigating CoCs may provide clues into shared etiologies (Bradley & Isaacs, 2006; Hu, 2012; Ibrahim, Voigt, Katusic, Weaver, & Barbaresi, 2009; Zhang, Xu, Liu, Li, & Xu, 2012), as the neurological mechanisms of CoCs may provide insight in to the mechanisms behind ASD (Maski, Jeste, & Spence, 2011; Sinzig, Walter, & Doepfner, 2009). These conditions complicate ASD presentation and need to be considered when creating interventions specific to a child’s needs.…”

Identification of Subgroups of Children in the Australian Autism Biobank using Latent Class Analysis

“…It is encouraging to see that over time, the number of studies that have focused on characterizing potential ASD subgroups has increased and that emphasis has shifted from theoretically derived classi cations of subtype to data-driven approaches [7]. A range of con rmatory and exploratory statistical approaches have been utilised for this purpose, such as different types of cluster analysis [5,8], and latent class or pro le analysis for cross-sectional and latent transition pro le analysis for longitudinal data [9,10]. These approaches all seek to identify similarities in patterns of observed data between individuals, and are therefore dependent upon the data variables selected for inclusion in the analysis [7].…”