Heterogeneity in Alzheimer's disease: a multi-modal perspective

Key findings: I. Does clinical heterogeneity among individuals with AD and FTD relate to differences in the underlying neurobiological manifestation of the disease? We observed dissociable atrophy patterns among individuals with an atypical AD variant (i.e., PCA), and these are associated with distinct visual processing deficits, e.g., ventral atrophy is associated to object perception. Cognitively-defined subgroups of individuals within the spectrum of typical late-onset AD dementia exhibit differential patterns of regional atrophy, e.g., left-lateralized temporal atrophy in the language subgroup. Cognitively-defined subgroups of individuals with AD dementia display distinct hypometabolic patterns and longitudinal trajectories of brain metabolism, e.g., faster parietal decline in metabolism for the visuospatial subgroup. Individuals with FTD and prominent right-sided anterior temporal atrophy (i.e., rtvFTD) display a clinical phenotype that is most notably characterized by prosopagnosia, and which is distinct from AD and other FTD variants. II. What are the associations between genetics and the clinical/neuroanatomical manifestation of AD? Cognitive subgroup-specific atrophy patterns are related to distinct regional gene expression profiles. By examining biological functions related to these gene expression profiles, we found that some biological pathways are related to atrophy in multiple subgroups (e.g., mitochondrial dysfunction), while other pathways were subgroup-specific (e.g., cell-cycle in the memory subgroup). APOEe2 carriership among individuals with AD dementia is related to a clinical phenotype with more non-amnestic (i.e., language, attention and executive) impairments, and a radiological phenotype with greater vascular involvement (i.e., more microbleeds and white matter hyperintensities) compared to APOEe3 homozygotes and APOEe4 carriers. The prevalence of APOEe4 among amyloid-beta positive individuals along the AD clinical spectrum is 51% for preclinical AD, 64% for prodromal AD and 66% for AD dementia. These estimates are higher than reported in non-biomarker confirmed populations, which indicates that APOEe4 carriership is even more closely related to AD (i.e., amyloid-positivity) than previously thought. III. Are factors related to reserve and resilience associated with clinical heterogeneity and severity in AD and other dementias? Cognitive and brain reserve have independent and differential protective effects on cognition in individuals with AD. Across the clinical spectrum of AD, cognitive and brain reserve are associated with attenuated rates of cognitive decline. Furthermore, cognitive reserve specifically accelerates cognitive decline in the dementia stage and both cognitive and brain reserve are moderately associated with attenuated mortality risk. Our meta-analysis shows that, after adjustment for the level of pathology present, the intracranial volume is positively related to cognition. This indicates that the intracranial volume is a valid, easily obtained, although rather crude, measure of brain reserve. Our meta-analysis of randomized controlled trials reveals that physical activity interventions offered to individuals diagnosed with dementia are effective in slowing down clinical progression. Suggesting that cognitive reserve is still a viable target for treatment in the dementia stage. IV. What is the role of amyloid and tau pathology in individuals who are not diagnosed with AD dementia? Mesial temporal tau is associated with worse cognitive performance and higher neocortical tau among amyloid-negative individuals performing within the normal limits of cognitive functioning. [11C]PiB-PET does not detect amyloid pathology associated with intracerebral amyloidoma. Histological examination of brain biopsy tissue revealed that amyloid is present within the tissue but not amyloid-beta. These findings indicate that amyloid, often only associated with AD, may be present in other atypical forms that are not detectable by PiB. Overall, amyloid-positivity is related to more depressive symptoms in VCD, and, in major VCD, to better attention and executive functioning. Furthermore, in mild-VCD, amyloid-positivity is related to fewer white matter hyperintensities, while the opposite was true in major-VCD. Hence, co-morbid amyloid pathology affects the manifestation of VCD in a disease stage specific manner.