Comprehensive immunoprofiling of neurodevelopmental disorders suggests three distinct classes based on increased...

Neurodevelopmental disorders (NDDs) are a spectrum of conditions with commonalities as well as differences in terms of phenome, symptomatome, neuropathology, risk factors and underlying mechanisms. Immune dysregulation has surfaced as a major pathway in NDDs. However, it is not known if neurodevelopmental disorders share a common immunopathogenetic mechanism. In this study, we explored the possibility of a shared immune etiology in three early-onset NDDs, namely Autism Spectrum Disorder (ASD), Attention Deficit Hyperactivity Disorder (ADHD) and Intellectual Disability Disorder (IDD). A panel of 48 immune pathway-related markers was assayed in 135 children with NDDs, represented by 45 children with ASD, ADHD and IDD in each group, along with 35 typically developing children. The plasma levels of 48 immune markers were analyzed on the Multiplex Suspension Assay platform using Pro Human cytokine 48-plex kits. Based on the cytokine/chemokine/growth factor levels, different immune profiles were computed. The primary characteristics of NDDs are depletion of the compensatory immune-regulatory system (CIRS) (z composite of IL-4, IL-10, sIL-1RA, and sIL-2R), increased interleukin (IL)-1 signaling associated with elevated IL-1α and decreased IL-1-receptor antagonist levels, increased neurogenesis, M1/M2 macrophage polarization and increased IL-4 as well as C-C Motif Chemokine Ligand 2 (CCL2) levels. With a cross-validated sensitivity of 81.8% and specificity of 94.4%, these aberrations seem specific for NDDs. Many immunological abnormalities are shared by ASD, ADHD and IDD, which are distinguished by minor differences in IL-9, IL-17 and CCL12. In contrast, machine learning reveals that NDD group consists of three immunologically distinct clusters, with enhanced neurogenesis, Th-1 polarization, or IL-1 signaling as the defining features. NDD is characterized by immune abnormalities that have functional implications for neurogenesis, neurotoxicity, and neurodevelopment. Using machine learning, NDD patients could be classified into subgroups with qualitatively distinct immune disorders that may serve as novel drug targets for the treatment of NDDs.