Distinct cerebrovascular pathways underlying Alzheimer's disease-related neurodegeneration.

The etiology of cerebrovascular pathology is heterogeneous. Independent or synergistic role of this pathology relative to Alzheimer's disease (AD) pathology is necessary to clarify distinct neurodegenerative pathways. We evaluated the interplay of various cerebrovascular markers postmortem and their in vivo neuroimaging, clinical and neuropathologic correlates using data from the Alzheimer's Disease Neuroimaging Initiative (ADNI). In 109 individuals, postmortem cerebrovascular pathology (atherosclerosis of the circle of Willis, cerebral amyloid angiopathy [CAA], arteriolosclerosis, white matter rarefaction, old infarcts, microinfarcts, hemorrhages, other ischemic/vascular changes) was characterized. Additionally, we assessed in vivo neuroimaging (cortical thickness, subcortical volume, white matter lesion burden, glucose standardized uptake value ratio, fractional anisotropy of white matter tracts, cerebral blood flow), cognitive, and neuropathologic measures (atrophy, AD pathology and copathologies including Lewy body, TDP-43, hippocampal sclerosis). The study sample had mean (standard deviation) age of 82.9 (7.2) years and included 29 women (27%) and 84 (77%) with intermediate/high AD neuropathologic change. Arteriolosclerosis and CAA emerged as dominant cerebrovascular markers using multiple correspondence analysis. More severe arteriolosclerosis was explained by higher white matter lesion burden and greater postmortem hippocampal atrophy (β = 143.2, 95% CI 63.9 to 230.1, p = 0.0003), but not AD pathology. More severe CAA was explained by fractional anisotropy (β = - 20, 95% CI - 41.5 to -3.1, p = 0.02) adjusted for AD pathology and reduced integrity of superior cerebellar peduncle, posterior thalamic radiation, and sagittal stratum tracts (rho < - 0.6, false discovery rate corrected p < 0.05). More severe CAA was also explained by cortical atrophy and AD pathology (β = 0.6, 95% CI 0.2 to 1.2, p = 0.007), and associated with poorer memory (β = - 0.2, 95% CI - 0.3 to -0.09, p = 0.0009). Results demonstrate two dominant cerebrovascular pathways. An arteriolosclerosis-driven pathway is unspecific to AD pathology, whereas a CAA-driven pathway is specific to AD pathology. Cerebrovascular pathology is associated with AD pathology in an etiology-dependent manner which may influence eligibility for treatment or treatment-emergent adverse events in disease-modifying therapies for AD.