Diffusion MRI of the developing cerebral cortical gray matter can be used to detect abnormalities in tissue microstructure associated with fetal ethanol exposure

Abstract

Fetal alcohol spectrum disorders (FASDs) comprise a wide range of neurological deficits that result from fetal exposure to ethanol (EtOH), and are the leading cause of environmentally related birth defects and mental retardation in the western world. One aspect of diagnostic and therapeutic intervention strategies that could substantially improve our ability to combat this significant problem would be to facilitate earlier detection of the disorders within individuals. Light microscopy-based investigations performed by several laboratories have previously shown that morphological development of neurons within the early-developing cerebral cortex is abnormal within the brains of animals exposed to EtOH during fetal development. We and others have recently demonstrated that diffusion MRI can be of utility for detecting abnormal cellular morphological development in the developing cerebral cortex. We therefore assessed whether diffusion tensor imaging (DTI) could be used to distinguish the developing cerebral cortices of ex vivo rat pup brains born from dams treated with EtOH (EtOH; 4.5 g/kg, 25%) or calorie-matched quantities of maltose/dextrin (M/D) throughout gestation. Water diffusion and tissue microstructure were investigated using DTI (fractional anisotropy, FA) and histology (anisotropy index, AI), respectively. Both FA and AI decreased with age, and were higher in the EtOH than the M/D group at postnatal ages (P)0, P3, and P6. Additionally, there was a significant correlation between FA and AI measurements. These findings provide evidence that disruptions in cerebral cortical development induced by EtOH exposure can be revealed by water diffusion anisotropy patterns, and that these disruptions are directly related to cerebral cortical differentiation.

Figure 1

The two middle columns of images are laterally-facing mid-cortical surface models of one rat postnatal day (P)0, P3 and P6 right hemisphere for each treatment group (ethanol (EtOH) and maltose/dextrin (M/D)), on which cortical FA at each mid-cortical surface node is projected. The outer columns represent mid-coronal FA maps for the right hemisphere of the same subjects depicted in the middle columns. Cortical FA decreased significantly with age. Additionally, cortical FA was largest, and isocortical volume smallest, in the EtOH group compared to the M/D group. This group difference is most visible in the outer layers of the cortex. Error bars represent one standard error. Scale bar is 4 mm. D = dorsal, V= ventral, M = medial, L= Lateral, Cd = Caudal, R = Rostral.

Figure 2

A, B. Brain volume and isocortical volume in the rat (n = 3 – 4/age/group) increased significantly with age in both groups, and were both significantly lower in the EtOH than in the M/D group at all three ages [postnatal day (P)0, P3, P6]. C. Isocortical FA in both treatment groups decreased significantly with age. Additionally, across age groups, the maltose/dextrin (M/D) had the lowest mean isocortical FA, while the ethanol (EtOH) group had the highest FA. D. The anisotropy index (AI) for a subset of subjects (1–3/age/group), in both treatment groups decreased significantly with age. Additionally, across age groups, the maltose/dextrin (M/D) group had the lowest mean isocortical AI, while the ethanol (EtOH) group had the highest AI. Numbers of subjects in each group at each age are represented on the x-axes above the age group labels. Please note that for the brain volume, isocortical volume and FA analyses, numbers represent numbers of subjects. For the AI analysis numbers represent the numbers of histological slices included in the analysis. Statistical results for effects of age and group on each dependent variable are represented in each graph in the form of p-values.

Figure 3

An FA map for one P6 M/D subject is shown on the left corresponding to the directional FA map (center). The directional FA map is shown for the same section with only the in-plane orientation for direct comparison with the directional AI map (far right), which is necessarily constrained in two dimensions. Green color indicates tissues oriented in a dorsal-ventral direction, while red indicates a medial-lateral direction. Please note that while blue coloring is typically used to represent water diffusion anisotropy oriented in the dorsal – ventral plane (e.g. Huang et al. 2008), green coloring was substituted here to increase contrast for visualization purposes. There is a close correspondence between the directional FA and AI maps.

Figure 4

A regression analysis investigating the relationship between AI and FA resulted in a significant direct overall association between AI and FA (p < 0.001). This relationship was also significant in both treatment groups separately (EtOH p < 0.0001, M/D p < 0.0001). This suggests that the morphological characteristics of neurons, represented by AI, underly FA meausurements in the developing cerebral cortex, and that this is true in both normal as well as FASD development.