Quantification of the Gray/White Matter Boundary in ASD

Background:

Magnetic resonance imaging (MRI) measures of cortical thickness in autism spectrum disorders (ASD) have revealed conflicting findings, reporting both increased cortical thickness and no differences in cortical thickness between ASD subjects and typical subjects. Because cortical thickness measurements in MRI often depend on thresholding to establish a boundary between the cortex and underlying white matter, any differences between ASD subjects and typicals may be attributable to neuroanatomical differences at this boundary. Previous neuroanatomical studies have suggested this boundary may be less distinct in ASD (Hutsler et al, 2007).

Objectives:

The aim of the present study was to objectively quantify the gray/white matter boundary in ASD and age-matched control subjects. If the ASD brain shows a more indistinct boundary, this could contribute to findings of cortical thickening using MRI techniques.

Methods:

Postmortem tissue was acquired from 16 male subjects (8 ASD and 8 age-matched controls) ranging in age from 10-51 years old. Tissue blocks were taken from the superior temporal gyrus (BA 21), dorsolateral frontal lobes (BA 9), and dorsal parietal lobes (BA 7). The tissue samples were sectioned perpendicular to the gyral axis and cell bodies were labeled with a Nissl stain. Approximately 300 digital photomicrographs of the gray/white matter boundary were acquired from these tissue samples. The images were then converted to grayscale and subjected to a noise removal filter. A sigmoid curve was fitted to the transition zone between layer VI and white matter. The slope of this curve indicated the spatial extent of the transition zone. A ‘shallow’ sigmoid curve represents an indistinct boundary, while a ‘steeper’ sigmoid curve represents a distinct boundary.

Results:

The results from the sigmoid curve analyses indicate that ASD subjects showed an indistinct boundary between cortical layer VI and white matter. For all three cortical regions examined, ASD subjects showed ‘shallower’ sigmoid curves compared to typicals, (F[1,14] = 6.45, p = .02).

Conclusions:

These results support previously documented neuroanatomical abnormalities in the lower boundary of cortical layer VI. An indistinct boundary suggests that during neuronal migration a portion of neuroblasts are not reaching their proper position within the developing cortical plate. Furthermore, these results raise question about the use of cortical thickness measurement techniques that rely on gray/white matter thresholding; an indistinct transition zone could lead to a misplaced cortical boundary resulting in an overestimation of cortical thickness. We would like to thank the Autism Tissue Program for their assistance in acquiring the brain tissue used in this study.