Disrupted connectivity of the locus coeruleus in healthy adults with parental history of Alzheimer?s disease

DEL CERRO, INES; VILLAREAL, MIRTA F.; ABULAFIA, CAROLINA; DUARTE ABRITTA, BARBARA; SANCHEZ, STELLA M; CASTRO, MARIANA N; BOCACCIO, HERNAN; FERRER, ISIDRO; MENCHON, JOSE M; SEVLEVER, GUSTAVO; SORIANO-MAS, CARLES; GUINJOAN, SALVADOR M

Niza

Workshop; ECNP Workshop for Early Career Scientists in Europe; 2019

European College of Neuropsychopharmacology

Background: Neurodegenerative disorders are the mostincapacitating and prevalent diseases in the elderly, andthe first cause of dementia [1]. Previous studies have associated initial pathological changes with abnormalities inthe Locus Coeruleus (LC) [2], a small pontine noradrenergicnucleus involved in attentional function [3], targetingwidespread cortical and subcortical brain regions [4]. Structural alterations in the LC have previously been reported inneurodegenerative processes [5]. However, no studies, todate, have evaluated its functional alterations in individualsat a high risk of developing Alzheimer?s disease (AD).Aim of the study: Based on previous evidence concerningalterations of the LC in early AD stages, we used functionalmagnetic resonance imaging (fMRI) to assess the functionalconnectivity (FC) of the LC in a large sample of asymptomatic middle-aged adults with a parental history of ADin comparison with an equivalent group of subjects withoutparental history of AD.Methods: Thirty-one offspring of late-onset AD (O-LOAD)(mean age ± SD=50.36 ± 8.32, 22 females) and 28 healthycontrols (HC) (mean age ± SD=53.90 ± 8.44, 20 females)were assessed. All participants underwent a completeneuropsychological evaluation and a fMRI assessment. Neurocognitive measurements were analyzed with SPSS v25,while images were preprocessed, denoised and analyzedwith the MATLAB-based CONN toolbox v17. Specifically, weperformed a global connectivity voxel-to-voxel (V2V) analyses to detect regions of altered FC within the pons, whichwas followed by a seed-based FC analysis using the V2V result as a seed. All analyses were family-wise error (FWE) corrected for multiple comparisons by means of the thresholdfree cluster enhancement (TFCE) non-parametric test.Results: HC and O-LOAD showed significant differencesin the Mini-Mental State Examination (MMSE) (t= 2.921;p=0.005); the Rey Auditory-Verbal Learning Test (RAVLT)Recognition measure (t=2.547; p=0.014), and the RAVLTDelayed Recall score (t=2.699, p=0.009).The O-LOAD group, in comparison with HC, showed adecreased global connectivity degree in a cluster encompassing the left locus coeruleus (peak at x= - 4, y= -34,z= -32 (kE= 8, pFWE= 0.047, TFCE= 168.36)). Moreover,global connectivity of the left LC correlated positively withMMSE total score (r= 0.397, p= 0.002), RAVLT total score(r= 0.364, p= 0.005) and RAVLT delayed recall (r= 0.411,p= 0.003).In seed-to-voxel analysis O-LOAD individuals showed adecreased connectivity between left LC and the bilateralcerebellar Crus I (left; pFWE= 0.018; TFCE= 418.86, right;pFWE= 0.033; TFCE= 354.19), the left lobule VI of thecerebellum (pFWE= 0.025; TFCE= 383.04), the left cerebellar IX lobule (pFWE= 0.026; TFCE= 380.41), and theright cerebellar lobule X (pFWE= 0.047; TFCE= 310.13).Finally, MMSE and RAVLT Delayed Recall scores correlatedpositively with LC ? right cerebellum connectivity patternin O-LOAD patients.Conclusions: Decreased FC between the LC and the cerebellum during resting state seems to characterize the OLOAD group. Moreover, the positive correlations betweenLC-cerebellum FC patterns and neuropsychological scorescould be suggesting the existence of a continuum in LCfunction in this pre-symptomatic population, since O-LOADparticipants performing worse in cognitive measurementsshowed larger alterations in LC-cerebellum FC patterns.