Amygdala and Hippocampal Morphology in Youth with High Functioning Autism Spectrum Disorders

Background: The amygdala and hippocampus are two medial temporal lobe regions implicated in emotion processing and psychopathology, and are hypothesized to play a role in the neuropathology of autism spectrum disorder. Controversy exists, however, regarding the nature of these abnormalities in ASD. For example, some data indicate enlarged amygdala volumes in childhood (Schumann et al., 2004), whereas other data indicate that amygdala volumes normalize by late childhood (Barnea-Goraly et al., 2013). Hippocampal data are also inconsistent and show enlarged (Sparks et al., 2002), reduced (Aylward et al., 1999) or no change (Piven et al., 1998; Haznedar et al., 2000) in volumes. These discrepant findings may be attributed to sample heterogeneity as well as limitations in imaging methodologies used to examine amygdala and hippocampus volumes. A fine grain analysis of these brain structures is therefore warranted in order to better delineate neuropathology. The current study uses large deformation diffeomorphic metric mapping (LDDMM), a powerful brain mapping tool, to examine shape abnormalities in the amygdala and hippocampus in youth with ASD.

Objectives: 1) To examine differences in amygdala and hippocampus morphology (shape and volume) between children with high functioning autism (HFA) and typically developing children; 2) To determine if localized differences in amygdala and hippocampal morphology (shape and volume)  in the HFA group are associated with clinical symptoms.

Methods: High resolution T1 scans from 82 subjects with high functioning autism or Asperger’s Disorder (mean age = 10.5 years, 85 % male, IQ > 80), along with 190 typically developing children (mean age =10.3, 69% male, IQ > 80) were examined. All children underwent a comprehensive assessment that included administration of the Diagnostic Interview for Children and Adolescents (DICA), Autism Diagnostic Observation Schedule (ADOS), Autism Diagnostic Interview (ADI-R), dimensional measures of psychopathology, and an IQ test. Amygdala and hippocampal volumes will be manually delineated on a subset of the sample in order to create the multiple atlas set, which will then be used to segment the two structures of interest. For each structure, LDDMM-surface mapping will be used to transfer a common template surface onto a triangulated surface contouring the amygdala and hippocampus for each subject. Deformation maps will then be computed and used for examining group differences in local shape measurements as well as correlations between shape and clinical measures.

Results: Investigators of this study have extensive experience in conducting morphological analyses of brain structures (e.g., Qui et al., 2010). Delineation of amygdala and hippocampus shapes and volumes are currently underway. Data analyses will examine the effect of diagnoses on morphological variables. Correlations will be performed to examine for associations between brain morphology and behavioral measures, including ASD specific symptom domains, an ADOS based measure of autism symptom severity, and measures of anxiety.

Conclusions: This is one of the first, and potentially the largest, structural MRI study examining amygdala and hippocampal morphology in a relatively homogeneous group of children with ASD. Findings may provide insights into abnormal neuronal organization within these structures and their relation to the clinical phenotype of ASD.