Neurology: Association of data-driven high-intensity spatial features of white matter with different cerebral small vessel disease etiologies

White matter damage or hyperintensity (WMH) is a very common radiological phenotype that is increasingly recognized as a marker of poor brain health, characterized by hyperintensity on T2-weighted magnetic resonance imaging
.
WMH is a major manifestation of cerebral small vessel disease (CSVD) and a major component
of vascular contributions to cognitive impairment and dementia (VCID) and Alzheimer's disease (AD).
WMH reflects a range of different underlying etiologies that mask its homogeneous large-scale appearance
.
This heterogeneity poses a major challenge
to unraveling the underlying pathogenesis of WMH.
Therefore, efforts are needed to identify WMH features or patterns that distinguish between different disease etiologies, such as AD, vascular-associated CSVD (arteriosclerosis), and amyloid-associated CSVD (cerebral amyloid angiopathy (CAA)).

However, due to erroneous assumptions of voxel independence and a lack of statistical power for high-dimensional and multi-comparison problems, previous studies using whole-brain, voxel-based comparisons have failed to identify different patterns
.
A study published in Neurology hypothesized that the topographic distribution of white matter hyperintensity (WMH) varies
depending on cerebrovascular risk factors.
Use unbiased pattern discovery methods to identify different spatial patterns of WMH and investigate their relationship
with different causes of WMH.

A cross-sectional study of participants in the Alzheimer's Disease Neuroimaging Program (adNI) was conducted to identify spatially distinct WMH distribution patterns using SVD analysis of voxel-based aligned WMH probability
plots.
All participants from the ADNI Big Chance/ADNI 2 study were included and baseline 2D-FLAIR MRI scans were provided with no prior stroke history or radiographic infarction
.
The association of these WMH spatial patterns with vascular risk factors, amyloid-β PET, and imaging biomarkers of cerebral amyloid angiopathy (CAA) was evaluated, characterizing different forms of cerebral small vessel disease (CSVD)
using multivariate regression.
Linear regression models are also used to investigate whether the spatial distribution of WMH affects cognitive impairment
.

Data-driven spatial patterns of WMH reflect different underlying causes, including arteriosclerosis, CAA, AD, and normal aging
.
Global approaches to measuring WMH volumes may overlook important spatial differences
.
Spatial features of WMH can serve as etiology-specific imaging markers to address heterogeneity of WMH, identify major underlying pathological processes, and improve prediction of clinically relevant trajectories affecting cognitive decline
.

Source: Phuah CL, Chen Y, Strain JF, et al.
Association of Data-Driven White Matter Hyperintensity Spatial Signatures With Distinct Cerebral Small Vessel Disease Etiologies [published online ahead of print, 2022 Sep 19].
Neurology.
2022; 10.
1212/WNL.
0000000000201186.
doi:10.
1212/WNL.
0000000000201186