Myelin-water imaging and multi-shell diffusion-weighted imaging in adults with adrenoleukodystrophy

Abstract

The pathophysiology of X-linked adrenoleukodystrophy (ALD) is not well-understood. New quantitative MRI (qMRI) sequences, such as myelin-water imaging (MWI) and multi-shell diffusion-weighted imaging (DWI), are non-invasive techniques that can investigate microstructural abnormalities in the brain. Using these techniques, this study investigated abnormal white matter of cerebral ALD (cALD) lesions and normal-appearing white matter (NAWM) in ALD and in controls. Adult participants were scanned on 3T MRI scanners. Fractional anisotropy (FA), axial diffusivity (AD) and radial diffusivity (RD) were measured with diffusion tensor imaging (DTI). The neurite density index (NDI) was established using neurite orientation dispersion and density imaging (NODDI). Myelin water fractions (MWF) were evaluated with multi-compartment relaxometry diffusion-informed MWI (MCR-DIMWI) and multi-echo T2-relaxation imaging with compressed sensing (METRICS). Measures were evaluated in 10 regions of interest (ROIs) in participants with and without cALD, and in controls. A total of 81 ALD participants (69% males, of whom 17 had cALD), and 21 controls were included. FA, NDI, MWF_METRICS and MWF_MCR-DIMWI were significantly lower in lesions in the corpus callosum of cALD participants than in controls. FA, NDI and MWF_METRICS also significantly differentiated the NAWM of male, but not female, ALD participants from controls. FA and NDI had the largest effect sizes for males. In the NAWM, RD was increased, whereas AD was mostly unaffected. In a few NAWM regions, qMRI measures also differed between males with and without cALD. The NAWM of males with ALD contains microstructural abnormalities. Myelin and myelinated axons are impacted by ALD pathophysiology. These techniques have potential for clinical applications.

Keywords: biomarker; imaging; leukodystrophy; radiology; translational research.

Graphical Abstract

Figure 1

Anatomical and lesion segmentation. (A) Sagittal, coronal (posterior) and axial (centrum semiovale) images of a male with cALD. (B) Automatic lesion segmentation shown with yellow overlays. (C) Segmentation of the white matter of the cerebral lobes and of the cerebellum. Purple: frontal lobe; dark-blue: parietal lobe; yellow: occipital lobe; red: temporal lobe; light-blue: cerebellar white matter. (D) Yellow: genu of CC; purple: body of CC; dark-blue: splenium of CC; light-blue: internal capsule. Probabilistic tract segmentation of the corticospinal tracts is shown in red-yellow colours where yellow indicates a more certain localization. cALD, cerebral ALD; CC, corpus callosum.

Figure 2

Participant inclusion and exclusion. DWI, diffusion-weighted imaging; cALD, cerebral ALD; MGRE, multi-echo gradient echo; METRICS, multi-echo T2 relaxation imaging with compressed sensing. Twelve individuals were excluded from the analyses for technical reasons, the presence of brain abnormalities not considered to be related to ALD or because they did not match the population inclusion criteria.

Figure 3

FA, NDI and MWF in the NAWM of the corpus callosum. From left to right: male controls (light blue, n = 10), males without cALD (medium blue, n = 39), males with cALD (dark blue, n = 17), female controls (beige, n = 11), females without cALD (brown, n = 25). Data points represent individual participants. Boxplots represent means and standard deviations. Red crosses in the cALD bars indicate participants with lesions in the corpus callosum (n = 9). For these participants, qMRI measures were determined in NAWM, which did not include lesion voxels. FA, NDI, and MWF are unitless fractions between 0 and 1. GLMs showed that FA (A), NDI (B) and MWF_METRICS (D) are significantly decreased in the NAWM of the corpus callosum in males with ALD compared to male controls. No significant differences were observed for MWF_MCR-DIMWI (C). In females, no significant differences were observed in this region. NAWM, normal-appearing white matter; FA, fractional anisotropy; NDI, neurite density index; MWF_MCR-DIMWI, myelin-water fraction with diffusion-informed myelin-water imaging for the multi-echo gradient echo sequence; MWF_METRICS, myelin-water fraction for the multi-echo T2 relaxation imaging with compressed sensing sequence. *Significant (P < 0.05) after Bonferroni correction for multiple testing corrected for age and for scanner.

Figure 4

AD and RD in the NAWM of the corpus callosum of males. Boxplots represent mean and standard deviation of AD and RD in the NAWM of the corpus callosum. (A) GLMs showed that AD is not significantly different between the NAWM of males with cALD (n = 17), males without cALD (n = 39) and male controls (n = 10). (B) RD is significantly increased within the NAWM of both males with cALD and males without cALD. Crosses in the cALD group indicate participants with a lesion in the corpus callosum (n = 9). qMRI measures were determined in NAWM, which did not include lesion voxels. NAWM, normal-appearing white matter; AD, axial diffusivity; RD, radial diffusivity—both in units of 10⁻⁵ mm²/s. *Significant (P < 0.05) difference with control values after Bonferroni correction for multiple testing corrected for age and scanner.