Quantitative assessments of white matter hyperintensities and plasma biomarkers can predict cognitive impairment and cerebral microbleeds in cerebral small vessel disease patients

The objective of this study is to examine the efficacy of magnetic resonance imaging (MRI) features and peripheral blood biomarkers in assessing cognitive function in patients with cerebral small vessel disease (CSVD). A total of 58 CSVD patients were recruited. Six features of white matter hyperintensities (WMHs) were derived from MRI scans. Additionally, five neurodegenerative biomarkers (Aβ40, Aβ42, t-tau, p-tau181, NfL) and 13 serum inflammatory cytokines (VILIP-1, CCL2, IL-6, IL-18, TNF-α, CX3CL, sTREM-1/2, VEGF, s-RAGE, BNDF, TGF-β1, β-NGF) were quantified. Cognitive assessments were conducted using standardized neuropsychological scales. Spearman analysis revealed that the volumetric characteristics (absolute area, upper area, bottom area, absolute area percentage, upper percentage, and bottom percentage) of WMHs were negatively correlated with performance on all cognitive scale measures except the verbal fluency test (VFT) (r < -0.3, p > 0.05), while they were positively correlated with plasma neurofilament light (NFL) levels (r > 0.4, p < 0.05). Additionally, serum tumor necrosis factor-α (TNF-α) and soluble receptor for advanced glycation end-products (s-RAGE) showed significant correlations with scales of speech function. An integrated model incorporating WMHs features, neurodegenerative biomarkers, and neuroinflammatory markers was developed, demonstrating high predictive accuracy for cognitive impairment with an area under the curve (AUC) of 0.95 (accuracy 0.88, sensitivity 0.87, specificity 0.89). Another integrated model that includes features of WMHs and inflammatory cytokines for predicting cerebral microbleeds (CMBs) achieved an AUC of 0.95 (accuracy 0.88, sensitivity 0.82, specificity 0.92). Our findings suggest that these markers have the potential to be used for the early detection of cognitive decline and CMBs in patients with CSVD.