Distinct signatures of white matter damage in cerebral small vessel disease and Alzheimer’s disease: a fixel‐based analysis study

Abstract

AbstractBackgroundAlzheimer’s disease (AD) and cerebral small vessel disease (SVD) are the two main causes of age‐related cognitive decline and frequently co‐exist in memory clinic patients. White matter diffusion alterations have been described as key features of both diseases with different spatial patterns in AD and SVD. Biomarkers that capture and disentangle the effects of each disease on white matter integrity are of great need but are still lacking. Advanced diffusion MRI‐based techniques on the fixel level (fiber population within a voxel) yield promise, since one can derive measures of both fiber density and fiber bundle cross‐section (Dhollander et al., Neuroimage, 2021). In this multimodal imaging study, we assessed the ability of these fixel metrics to detect disease‐specific signatures of white matter fiber tract damage yielding the path towards personalized medicine.MethodWe included genetically defined SVD (n=73) and the full continuum of biomarker‐characterized AD from ADNI (n=34 Aβ–T–; n=19 Aβ+T–, n=18 Aβ+T+). Using multi‐shell diffusion MRI, we assessed the fiber density and fiber bundle cross‐section of six well‐established white matter fiber tracts previously shown to be damaged or related to cognitive decline in AD and SVD (see Figure 1). Using simple linear regression analyses and while controlling for intracranial volume, we determined associations between tract‐specific fixel metrics and imaging markers of SVD (white matter hyperintensity volume, lacune and microbleed count) and AD (global amyloid‐ and tau‐PET SUVR) as well as age and brain volume.ResultFiber density of all fiber tracts was strongly associated with white matter hyperintensity volume in SVD (e.g. inferior fronto‐occipital fasciculus, R2adj=73%). In contrast, fiber bundle cross‐section was highly associated with cerebral atrophy in the AD sample (e.g. corpus callosum, R2adj=36%). Of note, both fiber density and fiber bundle cross‐section were only weakly associated with amyloid‐ and tau‐PET (e.g. fiber density of uncinate fasciculus ∼ amyloid‐PET, R2adj=11%; fiber bundle cross‐section of inferior longitudinal fasciculus ∼ tau‐PET SUVR, R2adj =5%).ConclusionFiber density captures the effects of SVD on white matter integrity whereas fiber bundle cross‐section indicates tract‐specific neurodegeneration in AD. Fixel‐based analysis allows to assess disease‐specific effects of SVD and AD on white matter fiber tracts.