Validation of a novel prospective circulatory biomarker for Alzheimer's Disease using the ADNI dataset.

点击翻译按钮获取中文摘要

Summary This supplement will evaluate the utility of hemodynamic parameter maps extracted from resting state fMRI data, as early markers of cerebrovascular dysfunction in Alzheimer’s Disease (AD). We have developed a method to derive maps of cerebrovascular function and dysfunction from resting state data, including through retrospective analysis of existing public datasets. These metrics are sensitive to both macrovascular and microvascular changes. In healthy tissue, vasodilation modulates cerebral perfusion in response to changing demands for oxygen and nutrients; this ability is reduced or absent in many forms of cerebrovascular pathology. Cerebrovascular reactivity (CVR), is a measure of brain blood vessels’ capacity for vasodilation, which offer useful information on the health of local vasculature. Traditional analysis methods of hypercarbic CVR data underestimate CVR magnitude in regions where the response is delayed with respect to the gas administration schedule, and give no information at all about blood flow delay, a crucial feature of vascular dysfunction. We have validated a method to quantitatively map delays in blood flow arrival, and other hemodynamic parameters, even in the absence of external manipulations, allowing us to differentiate the circulatory and metabolic components of cerebrovascular compromise from resting state data alone. We have already demonstrated the utility of this approach in primarily circulatory disorders, such as stroke and moyamoya disease; this supplement seeks to establish the utility of this approach to probe early circulatory dysfunction in AD. Specific patterns of local alteration in circulation may precede (or even lead to) neuronal degeneration, and may serve as an effective biomarker for following disease progression. This supplement would be used to test this hypothesis in two ways. First, we will add a focused cohort of patients with mild cognitive impairment or probable mild Alzheimer’s Disease (16 subjects) to our active protocol studying subjects at risk for stroke. We will perform an abbreviated version of our extensive circulatory evaluation protocol to determine circulatory markers of early AD relative to our comparison group. Second, we will perform a broad retrospective analysis on the resting state fMRI data in the ADNI (Alzheimer’s Disease Neuroimaging Initiative) database to measure circulatory parameters in a very large group that includes a range of diagnoses, and longitudinal data from a subset of subjects, to determine how regional bloodflow changes with disease progress. Finally, we will design and test machine learning classifiers to evaluate the ability of these circulatory markers to detect early AD/PRAD. These complementary efforts will test the hypothesis that specific, regional patterns of circulatory alteration progress with Alzheimer’s Disease, and may in fact precede other symptoms, allowing early, objective and quantitative detection of AD, and tracking of disease progression. Furthermore, this will serve as excellent preliminary data for designing more comprehensive efforts to develop this technique as an effective clinical tool.