Clin Anat. 2026 Jun 27. doi: 10.1002/ca.70155. Online ahead of print.
ABSTRACT
Cerebral small vessel disease (CSVD) is a leading cause of vascular cognitive impairment, yet its earliest stages remain clinically silent and poorly detected. Although white matter hyperintensities (WMHs) are widely used neuroimaging markers, conventional cardiovascular risk scores and cognitive testing lack sensitivity to subclinical cerebrovascular injury. This study investigated whether circulating microparticles (MPs) and fractal analysis of the Circle of Willis (CoW), reflecting cerebrovascular network-hemodynamic complexity, could provide an early, mechanistically informative marker of silent CSVD. Sixty asymptomatic adults with low-to-moderate cardio-cerebrovascular risk (QRISK3) underwent 3 T MRI, cognitive testing, and circulating MPs profiling. Cerebrovascular fractal dimension of the CoW (Df [W]) was computed from 3D time-of-flight magnetic resonance angiography. Multivariable regression, mediation analysis (10,000 bootstraps), and ROC analyses were performed. Reduced Df (W) was strongly associated with greater WMHs burden (p < 0.001) and significantly outperformed QRISK3 and MPs in discriminating WMHs (AUC = 0.928 vs. ~0.75). Leukocyte-derived (CD62L+) and platelet-derived (CD62P+) MPs were elevated in participants with WMHs and correlated with both WMHs burden and Df (W), but lost independent significance after adjustment for WMHs, indicating upstream systemic vascular injury. Mediation analysis confirmed that MPs influenced WMHs' burden primarily through their effect on cerebrovascular Df (W). A combined biological-imaging model integrating MPs and Df (W) achieved near-perfect diagnostic accuracy (AUC = 0.952). Despite marked vascular and microstructural abnormalities, neurocognitive performance was preserved, with only a weak association between processing speed and Df (W), consistent with network reserve in early CSVD. Thus, cerebrovascular fractal complexity may capture the structural imprint of cumulative vascular injury and enable biologically grounded detection of preclinical CSVD.
PMID:42365442 | DOI:10.1002/ca.70155