Influence of amyloid deposition and neuroinflammation on brain connectivity and cognition in Alzheimer's disease spectrum: a multimodal imaging study

Abstract

Background: Neuroinflammation has been recognised as an important factor in the pathogenesis of Alzheimer’s disease (AD). In neurodegeneration, preclinical evidence has suggested that activated microglia, the main cells in neuroinflammation, could damage neuronal circuit structures and disrupt neuronal activity. β-amyloid (Aβ) deposition is one of the hallmarks of AD and is associated with neuroinflammation. However, there is limited clinical evidence on whether neuroinflammation and amyloid deposition influence brain connectivity and cognition. Therefore, this project interrogated the relationship between neuroinflammation, Aβ deposition and brain connectivity/cognition in Alzheimer’s spectrum, which covers preclinical, mild cognitive impairment and dementia stage of the AD. Methods: Cross-sectional multimodal imaging data at baseline and neuropsychological assessments at both baseline and follow-up visits were obtained from the ‘Amyloid and Neuroinflammation’ study, including 18 healthy volunteers, 54 MCI and 21 AD patients. 11C-PBR28, 18F-Flutemetamol and 18F-AV1451 PET were used to quantify neuroinflammation, amyloid and tau pathology, respectively; T1-weighted MRI, diffusion tensor imaging, resting state functional MRI, and arterial spin labelling MRI measured the grey matter structure, structural connectivity, functional connectivity and cerebral blood prefusion status. Linear models, including principal component analysis, linear regression, and sparse canonical correlation analysis were then applied to interrogate the association between PET markers of Alzheimer’s pathology, MRI markers of brain connectivity and cognitive impairment. Results: 11C-PBR28 uptake was negatively associated with fractional anisotropy of white matter tracts in Aβ positive cohort (standardised β=-0.51, p=0.014, in multiple regression), independent of 18F-Flutemetamol uptake and cortical thickness. 11C-PBR28 uptake was associated with increased amplitude of low frequency fluctuation of BOLD signal in multiple regions including medial prefrontal cortex (MPFC), while 11C-PBR28 in MPFC was associated with decreased MPFC connectivity with posterior cingulate and precuneus (suggesting impaired communication within default mode network). MPFC 18F-Flutemetamol uptake was associated with local connectivity alterations mainly within frontal lobe. Global measure of 11C-PBR28 uptake was associated with decreased local efficiency of brain’s functional organisation (standardised β=-0.38, p=0.001), while 18F-Flutemetamol was not. 11C-PBR28 uptake was also negatively associated with cerebral blood perfusion in multiple regions including precuneus and posterior cingulate. MRI measures of neuronal damage performed better in predicting baseline cognitive performance, while PET markers of amyloid, tau pathologies and neuroinflammation outperformed MRI measures in predicting future cognitive decline in patients. Conclusions: (1) neuroinflammation is associated with both structural and functional connectivity disruption in Alzheimer’s spectrum, which is independent of the amount of amyloid deposition; (2) neuroinflammation, amyloid deposition and tau pathology markers together have been found to predict longitudinal cognitive decline in MCI and AD patients, underlining their independent contributions; (3) neuroinflammation is associated with abnormal neuronal hyperactivity and impaired neurovascular coupling in AD.