Resiliency to Alzheimer’s disease neuropathology can be distinguished from dementia using cortical astrogliosis imaging
Stephanie Barsoum,Caitlin S. Latimer,Amber L. Nolan,Alexander Barrett,Koping Chang,Juan Troncoso,C. Dirk Keene,Dan Benjamini
DOI: https://doi.org/10.1101/2024.05.06.592719
2024-05-08
Abstract:Despite the presence of significant Alzheimer’s disease (AD) pathology, characterized by amyloid β (Aβ) plaques and phosphorylated tau (pTau) tangles, some cognitively normal elderly individuals do not inevitably develop dementia. These findings give rise to the notion of cognitive ‘resilience’, suggesting maintained cognitive function despite the presence of AD neuropathology, highlighting the influence of factors beyond classical pathology. Cortical astroglial inflammation, a ubiquitous feature of symptomatic AD, shows a strong correlation with cognitive impairment severity, potentially contributing to the diversity of clinical presentations. However, noninvasively imaging neuroinflammation, particularly astrogliosis, using MRI remains a significant challenge. Here we sought to address this challenge and to leverage multidimensional (MD) MRI, a powerful approach that combines relaxation with diffusion MR contrasts, to map cortical astrogliosis in the human brain by accessing sub-voxel information. Our goal was to test whether MD-MRI can map astroglial pathology in the cerebral cortex, and if so, whether it can distinguish cognitive resiliency from dementia in the presence of hallmark AD neuropathological changes. We adopted a multimodal approach by integrating histological and MRI analyses using human postmortem brain samples. cerebral cortical tissue specimens derived from three groups comprised of non-demented individuals with significant AD pathology postmortem, individuals with both AD pathology and dementia, and non-demented individuals with minimal AD pathology postmortem as controls, underwent MRI at 7 T. We acquired and processed MD-MRI, diffusion tensor, and quantitative T and T MRI data, followed by histopathological processing on slices from the same tissue. By carefully co-registering MRI and microscopy data, we performed quantitative multimodal analyses, leveraging targeted immunostaining to assess MD-MRI sensitivity and specificity towards Aβ, pTau, and glial fibrillary acidic protein (GFAP), a marker for astrogliosis. Our findings reveal a distinct MD-MRI signature of cortical astrogliosis, enabling the creation of predictive maps for cognitive resilience amid AD neuropathological changes. Multiple linear regression linked histological values to MRI changes, revealing that the MD-MRI cortical astrogliosis biomarker was significantly associated with GFAP burden (standardized β=0.658, pFDR<0.0001), but not with Aβ (standardized β=0.009, =0.913) or pTau (standardized β=-0.196, =0.051). Conversely, none of the conventional MRI parameters showed significant associations with GFAP burden in the cortex. While the extent to which pathological glial activation contributes to neuronal damage and cognitive impairment in AD is uncertain, developing a noninvasive imaging method to see its affects holds promise from a mechanistic perspective and as a potential predictor of cognitive outcomes.
Neuroscience