Investigating Transverse Relaxation Time Abnormalities of White Matter in Autism

Background: Autism is a developmental disorder characterized by social deficits, impaired communication, and restricted and repetitive patterns of behavior. Although there is strong evidence that autism is associated with abnormal brain development, the anatomical extent and timing of these neurobiological differences are unknown. One method to examine tissue abnormalities in vivo is quantitative transverse relaxation time (T2) imaging. T2 is influenced by the molecular environment and tissue properties. We have reported an increase in overall white matter T2 in children and adolescents with autism, with a follow-up study finding disproportionally localized increases in the frontal and parietal lobes.  This pattern of increased T2 parallels that of increased white matter volume of some volumetric studies in autism.  A recent volumetric study has attempted to further localize volume increases in white matter by radially parcellating it into radiate and inner zones and appropriate subcompartments.
Objectives: The purpose of this study was to investigate T2 differences in patients with autism with a similar parcellation scheme to determine if the pattern of T2 abnormalities remained similar to that of volumetric studies.

Methods: Twenty-one male patients with autism between the ages of 6 and 16 and 20 male controls in the same age range participated in this study. The diagnosis was made according to DSM-IV-TR criteria using the Autism Diagnostic Interview-Revised, the Autism Diagnostic Observation Schedule. All patients had non-verbal IQ greater than 70. Control subjects were drawn from the local community and were assessed to rule out any psychiatric disorders.  The groups did not differ significantly in age, sex, race, full-scale IQ, or non-verbal intelligence.  Ten patients were medication-naïve at the time of their scan, while 3 others had discontinued their medication prior to the scan. 16 patients required sedation with oral midazolam in order to complete the scan.  T2 data were acquired on a 3T magnetic resonance scanner using a Gradient Echo Sampling of the Free Induction Decay and Echo (GESFIDE) sequence. T2 maps were reconstructed for each subject.  Spatial normalization was performed using SPM5 using an adult T2 template as a target.  Masks for each region of interest (superficial white matter, radiate white matter and deep/bridging white matter) in standard space were generated using the Pickatlas toolbox for Matlab and the DTI-81 white matter atlas.  Mean T2 values for each region of interest were then calculated by multiplying the normalized T2 maps by these binary masks.
Group differences in white matter T2 were investigated using a Repeated-Measures Analysis of Covariance.  Given the changes in T2 described in childhood, we covaried the statistical analysis of T2 for age. 

Results: In a preliminary analysis, a repeated measures ANCOVA revealed no significant main effect or interactions involving diagnosis, though the main effect for diagnosis approached significance level (p<0.08).

Conclusions: Patients in this study did not show an increase in superficial or radiate white matter T2 paralleling that of volumetric studies.  Further work will focus on optimizing and validating image registration and parcellation techniques used in the analysis.