Lower locus coeruleus integrity in older COVID‐19 survivors: initial findings from an international 7T MRI consortium

Abstract

AbstractBackgroundThe SARS‐CoV‐2 coronavirus has been associated with structural brain changes, consistent with its neurological manifestations. Recent studies showed a specific predilection for brainstem glial activation and hypometabolism, possibly indicating involvement of the locus coeruleus. The locus coeruleus (LC) modulates many cognitive functions and behaviors and its norepinephrine projections regulate both immune responses and vascular reactivity. We aimed to examine differences in LC integrity between COVID‐19 survivors and controls.MethodParticipants are enrolled across 3 US and 1 UK sites using harmonized cognitive and 7T MR‐imaging protocols. Here, we analyzed data from 18 participants enrolled at Houston Methodist (12 COVID‐19 survivors, 6 controls; Figure 1). COVID‐19 survivors were required to have had a positive antigen test and an illness syndrome consistent with COVID‐19. Healthy controls were required to have no significant pre‐existing medical, neurologic, or psychiatric illness and no illness requiring hospitalization in the last 2 years. LC imaging was performed using a dedicated 7T MT‐TFL sequence (0.4×0.4×0.5mm). A site‐specific normalized template was constructed using ANTs/FSL. The entire average LC integrity as well as voxel‐wise integrity values were compared between COVID‐19 survivors and controls using a robust linear regression (age‐controlled and threshold free cluster enhancement corrected). LC integrity was correlated with age, sex, ethnicity and cognition using Spearman’s rank correlation.ResultAverage LC integrity was not correlated with age, sex, or Hispanic ethnicity (p>0.3). COVID‐19 survivors did not differ from Controls when examining the entire LC (p = 0.54). Voxel‐wise analyses revealed a small cluster (19 voxels) in the middle portion of the left LC where COVID‐19 survivors exhibited lower LC integrity than controls (p = 0.005; Figure 2). Integrity of this cluster was not related to age or Hispanic ethnicity (p = 0.9). LC integrity did not correlate with cognitive performance within the COVID‐19 survivors (Trail Making Test B: p = 0.43; Craft Story delayed recall p = 0.47; MoCA p = 0.84).ConclusionConsistent with previous animal and human studies, our initial findings provide evidence for neuroinvasive potential of SARS‐CoV‐2 localized in the middle LC. In the future, we aim to expand our sample and link these observations to the neurocognitive sequelae of COVID‐19.