LOCUS COERULEUS INTENSITY IS ASSOCIATED WITH EARLY AMYLOID AND TAU PATHOLOGY IN PRECLINICAL AUTOSOMAL DOMINANT ALZHEIMER’S DISEASE

Autopsy studies reported the locus coeruleus (LC) as the initial site of tau accumulation in younger adults. Neuronal loss in the LC coincides with mounting Braak stage pathology. Carriers of AD-causing mutations have elevated amyloid (Aβ) several years before tau accumulation. We hypothesized that LC MRI signal intensity, a measure of neuromelanin cell density, would be associated with both Aβ and tau and consequently memory performance in carriers but not in non-carriers. 14 cognitively unimpaired carriers and 13 age-matched non-carriers from the Colombian-kindred (age range=28-51 years, Table1) underwent PiB-PET, Flortaucipir-PET, 3T-MRI and memory testing. Brainstem regions of interest on the MRI (LC and pontine tegmentum as reference) were registered to each subject using high-dimensional diffeomorphic normalization transformations. Interslice variability was removed by normalizing the within-slice intensity to the reference region and LC intensity was thresholded by selecting iteratively the highest five connecting voxels. Global Aβ (cut-off: 1.32 DVR) and entorhinal tau were calculated and partial volume corrected. Associations between LC intensity, age, Aβ, tau and memory were evaluated with Spearman correlations in both groups.Groups were compared with the Mann-Whitney-U test. The two groups did not differ in LC intensity (p=0.26), but carriers showed a negative association with age (r=−0.55, p=0.043), non-carriers did not (r=−0.25, p=0.40). There was a marginal negative association with Aβ in the carriers (r=−0.49, p=0.08), and a significant association in the non-carriers (r=−0.57, p=0.045) and overall group (r=−0.51, p=0.007; controlling for group: r=−0.55, p=0.004). LC intensity correlated negatively with entorhinal tau in the overall group (r=−0.47, p=0.014), in the carriers (r=−0.59, p=0.03), but not in non-carriers (r=−0.12, p=0.71, Fig2). Finally, LC intensity correlated positively with word list learning at trend-level in carriers (r=0.52, p=0.055), but not in non-carriers (r=0.008, p=0.98, Fig3).