Gray matter hypoperfusion is a late pathological event in the course of Alzheimer's disease

Abstract

Several studies have shown decreased cerebral blood flow (CBF) in Alzheimer's disease (AD). However, the role of hypoperfusion in the disease pathogenesis remains unclear. Combining arterial spin labeling MRI, PET, and CSF biomarkers, we investigated the associations between gray matter (GM)-CBF and the key mechanisms in AD including amyloid-β (Aβ) and tau pathology, synaptic and axonal degeneration. Further, we applied a disease progression modeling to characterize the temporal sequence of different AD biomarkers. Lower perfusion was observed in temporo-occipito-parietal cortex in the Aβ-positive cognitively impaired compared to both Aβ-negative and Aβ-positive cognitively unimpaired individuals. In participants along the AD spectrum, GM-CBF was associated with tau, synaptic and axonal dysfunction, but not Aβ in similar cortical regions. Axonal degeneration was further associated with hypoperfusion in cognitively unimpaired individuals. Disease progression modeling revealed that GM-CBF disruption Followed the abnormality of biomarkers of Aβ, tau and brain atrophy. These findings indicate that tau tangles and neurodegeneration are more closely connected with GM-CBF changes than Aβ pathology. Although subjected to the sensitivity of the employed neuroimaging techniques and the modeling approach, these findings suggest that hypoperfusion might not be an early event associated with the build-up of Aβ in preclinical phase of AD.

Keywords: Alzheimer’s disease; arterial spin labeling; cerebral blood flow; neurodegeneration; tau.

Figure 1.

Inclusion/exclusion pathway. Flowchart shows the initial participant population with ASL scans and those excluded due to non-AD neurodegeneration, missing CSF sampling, participant’s withdrawal, vascular or other sources of artefacts e.g., excessive motion or poor background suppression, missing tau-PET data and young age. The non-AD group comprised the following diagnoses: behavioral variant fronto-temporal dementia (N = 10), Parkinson’s disease with dementia (N = 4), vascular dementia (N = 10), dementia with Lewy bodies (N = 24), Parkinson’s disease (N = 33), progressive supranuclear palsy (N = 17), multiple system atrophy (N = 7), corticobasal syndrome (N = 2), semantic variant primary progressive aphasia and primary non-fluent aphasia (N = 5 and 1, respectively), Aβ-negative mild cognitive impairment (N = 60) and unspecified dementia (N = 19). The diagnoses were made by neurologists following clinical and neuropsychological evaluation fulfilling the respective criteria based on Diagnostic and Statistical Manual of Mental Disorders (DSM-5) or the corresponding above-mentioned references. Note that of 45 individuals excluded due to the young age, one was positive for Aβ pathology and therefore was not included in the reference group for disease progression modeling.

Figure 2.

A priori-defined CBF-ROIs projected on the cortical surface. The left and right panels represent the lateral and medial views, respectively. The ROIs are displayed with distinct colors (lateral temporal = cyan, lateral parietal = green, superior lateral occipital = red, middle frontal = violet, medial parietal = yellow, and medial temporal = green).

Figure 3.

Voxel-wise group differences between (a) Aβ-negative CU vs. Aβ-positive CI and (b) Aβ-positive CU vs. Aβ-positive CI individuals adjusted for age, sex and ASL sequence version. Clusters with significantly decreased GM-CBF in Aβ-positive CI patients are overlaid on the mean population T1-weighted volume in radiological space (p < 0.05, FWE corrected). The top and bottom panels display different representative slices.

Figure 4.

Voxel-wise associations between GM-CBF and tau uptake in (a) early, (b) temporal, (c) late ROIs and (d) CSF concentrations of P-tau217 in the AD spectrum adjusted for age, sex and ASL sequence version. CBF reduction was primarily observed in lateral temporal, lateral/medial parietal and superior lateral occipital regions although its spatial extent was moderately attenuated for P-tau217 compared to tau-PET (p < 0.05, FWE corrected). Conventions as for Figure 3.

Figure 5.

Voxel-wise associations between CBF and CSF levels of NfL and NPTX2/T-tau in CU individuals and those along the AD spectrum adjusted for age, sex and ASL sequence version. Increased levels of NfL were associated with lower CBF in both (a) CU subjects and (b) AD spectrum. (c) Decreasing levels of NPTX/T-tau were accompanied by CBF reduction mainly in temporo-occipito-parietal regions (p < 0.05, FWE corrected). Conventions as for Figure 3.

Figure 6.

Progression pattern of AD predicted by the linear event progression model with four different estimations of the GM-CBF. In panel (a), the weighted mean GM-CBF was obtained in temporo-occipito-parietal regions demonstrating significant voxel-wise associations with tau-PET burden in the early ROI in the AD spectrum (see Figure 4(a)). The same mask of significant voxel-wise associations was used to extract CBF values in Aβ-negative controls. In panel (b), GM-CBF was derived as the mean of all predefined ROIs (see Figure 2). Panel (c) shows the temporal ordering of the weighted mean GM-CBF in precuneus and posterior cingulate along with other biomarkers and in panel (d) the total GM-CBF value was fed into the model. While the earliest disease stages comprised abnormalities in CSF markers of Aβ and tau, GM-CBF variations occurred during the latest stages of the disease progression. There is a small variation in the staging of the biomarkers between the four diagrams, particularly for the volumetric measure. This is due to the interdependence between variables. In panels a, c, and d the estimated GM-CBF values are highly collinear resulting in analogous positional variance diagrams while in panel b, the CBF values are slightly less collinear which in turn affects the staging of all the variables. At each stage, the accumulative probability for each biomarker going from one z-score to another is indexed by distinct colors i.e., red, magenta, and blue representing z-scores of 1, 2 and 3, respectively. Higher opacity indicates more confidence in the ordering.