Longitudinal Locus Coeruleus Imaging across the Alzheimer’s Spectrum

Abstract

AbstractBackgroundNoradrenergic neurons of the locus coeruleus (LC) may play a critical role in the pathophysiology of Alzheimer’s Disease (AD). LC imaging with MRI is a promising in vivo method to study the role of this system in AD. Here we investigate longitudinal change in LC integrity across the AD spectrum.MethodThis study included 291 participants from the TRIAD cohort at the McGill Research Center for Studies in Aging (cognitively normal: n = 166, mild cognitive impairment: n = 79, AD: n = 46). Follow‐up data was available for 123 participants. All participants underwent MR imaging including a turbo spin echo (TSE) sequence for LC imaging, tau positron emission tomography (using [18F]MK6240), and clinical and cognitive assessment including the Mini Mental State Exam (MMSE). LC signal contrast was assessed relative to a pontine reference region. Linear regression analyses related LC signal to clinical measures.ResultConsistent with prior work in an earlier version of this sample, at baseline, LC signal was reduced in tau positive participants (t(285) = ‐4.56, p<0.0001) and negatively correlated to AD severity (Braak stage: t(253) = ‐4.79, p<0.0001; MMSE score: t(254) = 4.30, p<0.0001). There was no significant change in LC signal over time (t(122) = 1.36, p = 0.18) and no significant impact of tau status on LC change (t(119) = 0.39, p = 0.70). Counterintuitively, there was evidence of increasing LC signal in the most severe participants, as those with low MMSE score (<27) showed a significant signal increase over time (t(16) = 2.16, p = 0.046). Moreover, increasing LC signal in tau positive participants was correlated to reduction in MMSE performance over time (t(35) = ‐2.13, p = 0.040). In contrast, tau negative participants showed the opposite trend (t(52) = 1.97, p = 0.054).ConclusionWhile cross‐sectional analyses found the expected negative correlation between LC signal and AD severity, longitudinal analyses obtained the surprising result that increasing signal correlated to worsening AD severity. This finding could suggest that the LC imaging sequence used (TSE) may imperfectly capture LC degeneration and signal increases could be driven by other factors occuring simultaneously with LC degeneration (perhaps elevated neuromelanin or tau deposition). Future work could determine if different types of LC sequence (TSE versus magnetization transfer) differ in their ability to capture degeneration and provide complementary information.