COMBINATORIAL SENSORY MODALITY ASSESSMENT IN PRODROMAL ALZHEIMER’S DISEASE: RELATION TO MRI AND AMYLOID AND TAU PET

Abstract

Measures of sensory function, including vision and smell, have been shown to be impaired in Alzheimer's disease (AD), even in prodromal stages. Our goal was to evaluate whether visual function and smell identification combined would provide a better prediction of AD pathology (i.e., cortical amyloid deposition, medial temporal lobe (MTL) tau deposition, MTL atrophy) than either measure alone in a cohort of cognitively normal older adults (CN), older adults with subjective cognitive decline (SCD), and mild cognitive impairment patients (MCI). 34 individuals (14 CN, 11 SCD, 9 MCI) underwent assessments of visual contrast sensitivity (via frequency doubling technology), olfactory identification (the UPSIT), amyloid PET with either [18F]Florbetapir (FBP) or [18F]Florbetaben (FBB), tau PET with [18F]Flortaucipir (FTP), and MRI. Scans were collected using ADNI-2/3 protocols for amyloid PET, tau PET, and MPRAGE (http://www.adni-info.org). Scans were processed using standard techniques to generate smoothed, normalized static images from 50-70min (FBP), 90-110min (FBB), or 80-100min (FTP). Standardized uptake value ratio images (SUVR) were generated using the appropriate reference region (whole cerebellum for FBP, cerebellar grey matter for FBB, cerebellar crus for FTP). SUVR values from the cerebral cortex (for amyloid measures) and the MTL (includes entorhinal cortex, fusiform, and parahippocampal gyri; for FTP) were extracted. Entorhinal cortex (EC) thickness was determined using Freesurfer v5.1. Linear regression models were completed to determine whether contrast sensitivity and UPSIT total score together predict amyloid (z-scored relative to ADNI CN), tau, and atrophy better than either measure alone, covaried for age as needed. In combination, the visual and olfactory measures predicted amyloid, tau, and EC thickness, with poorer performance associated with more amyloid and tau and thinner EC (p<0.05; Table 1, Figure 1A, 2A, 3A). Lower visual contrast sensitivity alone predicted higher tau (p<0.001; Figure 2B), while lower UPSIT score alone predicted a thinner EC (p=0.021, Figure 3C). Relationship of Sensory Measures with Amyloid Deposition Relationship of Sensory Measures with Tau Deposition Relationship of Sensory Measures and Entorhinal Cortex Thickness In general, the combination of both a visual and an olfactory measure showed a better prediction of amyloid, tau, and neurodegeneration. These findings suggest that multi-modal vs single modality sensory assessment may provide a better estimation of AD-related pathology during preclinical and prodromal disease stages.