Common and unique multimodal covarying patterns in autism spectrum disorder subtypes

Area of Science:

• Neuroscience
• Psychiatry
• Medical Imaging

Background:

• Autism spectrum disorder (ASD) is characterized by significant heterogeneity, complicating diagnosis and personalized treatment approaches.
• Observed variations in ASD include biological etiologies, genetics, neural systems, neurocognitive profiles, and clinical phenotypes.

Purpose of the Study:

• To investigate the heterogeneity within autism spectrum disorder (ASD) using a multimodal brain imaging approach.
• To identify distinct neuroimaging patterns associated with different ASD subtypes, including Asperger's disorder, PDD-NOS, and Autistic disorder.

Main Methods:

• Utilized functional and structural MRI data fused using the Autism Diagnostic Observation Schedule (ADOS) as a reference.
• Employed a discovery cohort from the ABIDE II dataset (Asperger's: 79, PDD-NOS: 58, Autistic disorder: 92) and a replication cohort from ABIDE I (400 participants).
• Analyzed fractional amplitude of low-frequency fluctuations (fALFF) in subcortical brain areas to identify subtype-specific functional differences.

Main Results:

• Common functional-structural covarying cortical areas across ASD subtypes include the dorsolateral prefrontal cortex and superior/middle temporal cortex.
• Distinct subcortical functional differences were identified: negative putamen-parahippocampus fALFF in Asperger's, negative anterior cingulate cortex fALFF in PDD-NOS, and negative thalamus-amygdala-caudate fALFF in Autistic disorder.
• Subtype-specific brain patterns correlated with different ADOS subdomains, primarily social interaction, and predicted subtype-specific symptoms.

Conclusions:

• While ASD shares a common neural basis related to social interaction deficits, each subtype exhibits unique brain system associations and symptom profiles.
• Multimodal neuroimaging provides valuable insights into the underlying mechanisms driving ASD heterogeneity.
• The identified subtype-specific neural patterns offer a potential avenue for more targeted diagnostic and therapeutic strategies.