An MRI measure of degenerative and cerebrovascular pathology in Alzheimer disease

Abstract

Objective: To develop, replicate, and validate an MRI-based quantitative measure of both cerebrovascular and neurodegeneration in Alzheimer disease (AD) for clinical and potentially research purposes.

Methods: We used data from a cross-sectional and longitudinal community-based study of Medicare-eligible residents in northern Manhattan followed every 18-24 months (n = 1,175, mean age 78 years). White matter hyperintensities, infarcts, hippocampal volumes, and cortical thicknesses were quantified from MRI and combined to generate an MRI measure associated with episodic memory. The combined MRI measure was replicated and validated using autopsy data, clinical diagnoses, and CSF biomarkers and amyloid PET from the Alzheimer's Disease Neuroimaging Initiative.

Results: The quantitative MRI measure was developed in a group of community participants (n = 690) and replicated in a similar second group (n = 485). Compared with healthy controls, the quantitative MRI measure was lower in patients with mild cognitive impairment and lower still in those with clinically diagnosed AD. The quantitative MRI measure correlated with neurofibrillary tangles, neuronal loss, atrophy, and infarcts at postmortem in an autopsy subset and was also associated with PET amyloid imaging and CSF levels of total tau, phosphorylated tau, and β-amyloid 42. The MRI measure predicted conversion to MCI and clinical AD among healthy controls.

Conclusion: We developed, replicated, and validated an MRI measure of cerebrovascular and neurodegenerative pathologies that are associated with clinical and neuropathologic diagnosis of AD and related to established biomarkers.

Figure 1. Examples of primary MRI measures used in the study

(A) White matter hyperintensities (WMHs). The first column shows a single axial slice. The second column displays labeled WMHs in in-house-developed software. The third column illustrates 3D rendering of WMH burden. The top row displays a single participant with relatively mild WMH burden, whereas the bottom row shows a participant with relatively severe WMHs. (B) Examples of observed brain infarcts in a single participant. Top row shows T1-weighted and T2-weighted (fluid-attenuated inversion recovery image) with hyperintense ring around the lesion. The bottom row shows magnification of infarcted areas. (C) 3D rendering of bilateral hippocampus (in yellow) with FreeSurfer in a single participant. FreeSurfer-derived cortical thickness measurements in a single participant demonstrate regions that comprise the Alzheimer disease cortical signature.

Figure 2. Relation between quantitative MRI measure and brain pathology

Each patient's specific brain pathology was rated on a scale from 0 to 5 as described in the Methods. The quantitative MRI measure for each patient is represented by a black dot within the rating of the neuropathologic findings. The red dot represents the mean quantitative MRI measure within each category.

Figure 3. Scatterplot representing the distribution of the quantitative MRI measure in relation to the PET–Pittsburgh compound B (PiB) average uptake in the Alzheimer's Disease Neuroimaging Initiative–1 cohort