Deciphering Neuropathological Heterogeneity in Alzheimer's Disease: Beyond Plaques and Tangles

Alzheimer’s disease (AD) is especially clinically recognized as heterogeneous. Research from the past decade shows that the disease is also radiologically and pathologically more diverse than previously assumed. This thesis first aimed to gain more insight into the pathological diversity of different AD subtypes. The second aim was to translate pathological heterogeneity to radiology by studying the correlation between in-vivo and post-mortem measurements. The key findings Aim 1 • Although neuroinflammation is observed in and around plaque-like structures, its distribution measured by activated microglia and complement factors parallels the occurrence of pTau in typical and atypical AD. • Atypical AD cases with an early disease-onset were associated with a relative increase and aberrant distribution of neuroinflammation compared to typical AD with a late disease-onset. • Innate immune cells (e.g., microglia and astrocytes) correlate regionally differently with AD pathological hallmarks in various clinical subtypes. • The distribution of reactive astrocytes differentiates PCA and typical AD better than pTau distribution does. • Atypical AD cases showed increased presence of a fibrillar plaque-type referred to as the ‘coarse-grained’ plaque. • The coarse-grained plaque is especially observed in early-onset homozygous APOE ε4 AD cases, who are affected by cerebral amyloid angiopathy (CAA). The plaque is a unique fibrillar deposit located at the parenchymal border of the blood-brain barrier that shares similarities and differences with both parenchymal and vascular Aβ deposits. • Fibrillization of α-helical Aβ into a β-sheet confirmation increases along the hypothesized ascending plaque-maturation stages. During plaque progression, this protein conformation gradually increases and localizes within the plaque’s center. • The Aβ β-sheet fibril confirmation is different in dense-cored compared to non-cored fibrillar plaques. Carotenoids, associated with inflammation, are predominantly found in the dense-cored deposits. Aim 2 • Aβ cerebrospinal fluid (CSF) and amyloid positron emission tomography (PET) biomarker status do not per se reflect Aβ pathology, as inflammation may induce a positive amyloid signal. A positive Aβ biomarker status does not invariably equal a primary diagnosis of AD at autopsy since Aβ can be concomitant to other pathologies. • In AD, Aβ, pTau, and activated microglia are differently associated with post-mortem MRI-derived cortical thickness. Aβ is associated with a globally reduced cortical thinning, whereas pTau and activated microglia are regionally associated with increased cortical atrophy. • Post-mortem in-situ MRI-derived cortical thickness but not volumetry measurements can be used as a proxy for ante-mortem MRI, provided that the obtained measurements are corrected for within-subject. In this thesis, I show that various AD subtypes not only have different regional distribution patterns of AD pathological hallmarks, I also show that certain Aβ pathology is predominantly seen in a particular subgroup and that the underlying neuroinflammatory process may be a crucial factor in linking tau to Aβ. It seems that disease mechanisms, in particular neurodegeneration and neuroinflammation may be differently involved in various subtypes. Therefore, I propose a paradigm change in AD research: instead of focusing on AD as a single pathological entity, future research should acknowledge the individual differences in immunology related to AD pathology. To grow a better etiological understanding, we should start to disentangle those individual neuro-immunological differences at the immunohistochemical and genetical level. To measure the status and progression of the innate immunity in-vivo, we also need to continue our search for derivates of this innate immune system in the CSF and blood, or by imaging. Not only will this approach help us in understanding the disease, but also in 1) categorizing patients based on their neuro-immunologic fingerprint and 2) developing personalized treatment strategies that are tailored to both anti-Aβ and pTau, as well as the innate immune system.