Different aspects of Alzheimer's disease-related amyloid β-peptide pathology and their relationship to amyloid positron emission tomography imaging and dementia

Abstract

Alzheimer's disease (AD)-related amyloid β-peptide (Aβ) pathology in the form of amyloid plaques and cerebral amyloid angiopathy (CAA) spreads in its topographical distribution, increases in quantity, and undergoes qualitative changes in its composition of modified Aβ species throughout the pathogenesis of AD. It is not clear which of these aspects of Aβ pathology contribute to AD progression and to what extent amyloid positron emission tomography (PET) reflects each of these aspects. To address these questions three cohorts of human autopsy cases (in total n = 271) were neuropathologically and biochemically examined for the topographical distribution of Aβ pathology (plaques and CAA), its quantity and its composition. These parameters were compared with neurofibrillary tangle (NFT) and neuritic plaque pathology, the degree of dementia and the results from [18F]flutemetamol amyloid PET imaging in cohort 3. All three aspects of Aβ pathology correlated with one another, the estimation of Aβ pathology by [18F]flutemetamol PET, AD-related NFT pathology, neuritic plaques, and with the degree of dementia. These results show that one aspect of Aβ pathology can be used to predict the other two, and correlates well with the development of dementia, advancing NFT and neuritic plaque pathology. Moreover, amyloid PET estimates all three aspects of Aβ pathology in-vivo. Accordingly, amyloid PET-based estimates for staging of amyloid pathology indicate the progression status of amyloid pathology in general and, in doing so, also of AD pathology. Only 7.75% of our cases deviated from this general association.

Trial registration: ClinicalTrials.gov NCT01165554 NCT02090855.

Keywords: Alzheimer’s disease; Amyloid PET; Amyloid load; Amyloid maturation; Amyloid β peptide; Insoluble amyloid; Soluble amyloid; Staging; [18F]flutemetamol.

Conflict of interest statement

DRT received consultancies from GE-Healthcare (UK) and Covance Laboratories (UK), speaker honorarium from Novartis Pharma Basel (Switzerland), travel reimbursement from GE-Healthcare (UK), and UCB (Belgium), and collaborated with GE-Healthcare (UK), Novartis Pharma Basel (Switzerland), Probiodrug (Germany), and Janssen Pharmaceutical Companies (Belgium). TGB received a consultancy from GE-Healthcare (UK). GF, CB, APLS are employees of GE-Healthcare (UK, USA). JL and KH were employees of GE-Healthcare (Sweden). JL is currently employee of Hermes Medical Solutions (Sweden). AC and AI received personal fees from GE-Healthcare via the University of Leeds. CAFvA received honoraria from serving on the scientific advisory board of Nutricia GmbH (2014), Roche (2018) and Hongkong University Research council (2014) and has received funding for travel and speaker honoraria from Nutricia GmbH (2014–2015), Lilly Deutschland GmbH (2013–2016), Desitin Arzneimittel GmbH (2014), Biogen (2016–2018), Roche (2017–2018) and Dr. Willmar Schwabe GmbH &Co. KG (2014–2015). MO received honoraria from serving on the advisory board of Axon and Fujirebio.

Figures

Fig. 1

Comparison of the relationship of the Aβ phases (a), AβMTL phases (b), the A-scores (c), CAA stages (d) and CAA severities (e) with Aβ plaque loads for Aβ, AβN3pE, and AβpSer8-containig plaques in cohort 1, depicted by boxplot diagrams. a: The plaque-loads for Aβ (Spearman correlation analysis: r = 0.888) and AβN3pE (Spearman correlation analysis: r = 0.882) correlated better with the Aβ phases than that for AβpSer8-positive plaques (Spearman correlation analysis: r = 0.810). b-e: Likewise, the AβMTL phases (b), the A-scores (c), the CAA stage (d) and the CAA severity (e) correlated with the Aβ, AβN3pE, and AβpSer8 loads (r = 0.582–0.899; p < 0.001; for detailed statistical analysis see Additional file 1: Table S5a)

Fig. 2

Boxplot diagrams representing the distribution of soluble (Sol.), dispersible (Disp.), membrane-associated (Memb. ass.), and plaque-associated (Plaq. Ass.) Aβ (a, d, g, j, m), AβN3pE (b, e, h, k, n), and AβpSer8 levels (c, f, i, l, o) in relation to the phase of Aβ plaque distribution (Aβ phase; a-c), the AβMTL phase (d-f), the A-score (g-i), the CAA stages (j-l) and the CAA severity (m-o) in cohort 1. The correlation for these three different forms of Aβ was best for Aβ detected with antibodies against non-modified forms of Aβ (Spearman correlation analysis: r = 0.603–0.809) followed by AβN3pE (Spearman correlation analysis: r = 0.572–0.756) whereas AβpSer8 was not detectable in soluble Aβ aggregates. Dispersible, membrane-associated and plaque-associated AβpSer8 showed a week correlation with the Aβ phases (Spearman correlation analysis: r = 0.324–0.556) due to the fact that it was seen only in Aβ phases 4 and 5 but not earlier, except for single cases. Increased levels of Aβ and AβN3pE were found already in cases without CAA (m, n). Cases with CAA showed high levels of soluble, dispersible, membrane-associated, and plaque-associated Aβ and AβN3pE in all stage and severity degrees of CAA. Only the presence of AβpSer8 was restricted to cases with CAA. The detailed correlation analysis is provided in Additional file 1: Table S5a

Fig. 3

Boxplot diagram comparing the relationship of the Aβ phases (a), AβMTL phases (b), A-scores (c), CAA stages (d) and CAA severities (e) with the biochemical stages of Aβ aggregate maturation reflecting the hierarchical occurrence of Aβ, AβN3pE, and AβpSer8 in Aβ aggregates in brain homogenates (B-Aβ stages) or in Aβ plaques (B-Aβ plaque stage) in cohort 1. a: The B-Aβ stages and the B-Aβ plaque stages increased with increasing Aβ phase until Aβ phase 4 when the maximum B-Aβ and B-Aβ plaques stages were reached. b, c: AβMTL phases and A-scores also correlated with the B-Aβ stages and B-Aβ plaque stages but exhibiting no ceiling effect. d, e: In the presence of CAA regardless of the stage (d) or severity (e), end-stage B-Aβ stage and B-Aβ plaque stage were observed, probably indicative for a ceiling effect. (Spearman correlation analysis: r = 0.694–0.906) (for detailed statistical analysis see Additional file 1: Table S5a)

Fig. 4

Boxplot diagrams representing the distribution of Aβ phase, AβMTL phase, A-score, CAA stage, CAA severity (a) and Aβ loads (b) in relation to the PET-Aβ phase estimate in cohort 3. All parameters correlated with the PET-Aβ phase estimates (r = 0.610–0.835; p < 0.001; for detailed statistical analysis see Additional file 1: Table S6). Note that the main increase in Aβ load occurs after Aβ became detectable in PET-Aβ phase estimate 1. The mean value of the Aβ load in PET-Aβ phase estimate 1 was 1.396% (median 1.42%) indicative for the threshold of Aβ detectability in the brain by PET

Fig. 5

Boxplot and scatter diagrams depicting the correlation of the Braak NFT stages, CERAD-scores for neuritic plaque pathology, NIA-AA scores of AD pathology, and the clinical dementia scores (CDR for cohorts 1 and 2 and MMSE for cohort 3) with the topographical Aβ parameters Aβ phase (a-d) and CAA stage (e-h), the quantitative measure of the Aβ load (i-l), and the qualitative aspect provided by the B-Aβ plaque stages (m). The boxplots are depicted separately for cohorts 1 (a, e, i, k, m), 2 (b, f), and 3 (c, d, g, h, j, l). The Braak NFT stages, CERAD scores, NIA-AA degrees of AD pathology, and CDR scores correlated with all parameters depicted here (r = 0.287–0.920, p < 0.001). Likewise, the MMSE scores showed a negative correlation with the Aβ phase and the CAA stages in cohort 3 (r = − 0.514/− 0.315, p ≤ 0.012) except for the Aβ load (p = 0.051) which showed only a trend (for detailed statistical analysis see Additional file 1: Table S7)