Welcome to the AD/PD™ 2024 Interactive Program
The conference will officially run on Western European Standard Time (Lisbon, UTC+0)
To convert the conference times to your local time Click Here
LONGITUDINAL MARKERS OF CEREBRAL AMYLOID ANGIOPAHTY AND RELATED INFLAMMATION IN RTG-DI RATS
Abstract
Aims
Cerebral amyloid angiopathy (CAA) is a frequent small vessel disease of the elderly and a common comorbidity of Alzheimer’s disease leading to inflammation, cerebral microbleeds, intracerebral hemorrhages, and vascular cognitive impairment and dementia (VCID). Despite the clinical prevalence of this disease there currently exists no validated biomarkers for CAA and related inflammation (CAAri). Here, we used our rTg-DI rat model of CAA-ri to evaluate transcriptomic and proteomic changes that occur during progression from emergent to late stages of disease.
Methods
Cohorts of rTg-DI and wild-type rats were aged to emergent stages of disease (≈4 months) and to advanced stages of disease (12 months). Comparative analyses were performed to evaluate longitudinal cerebral vascular amyloid deposition and CAA-ri. Further, we utilized data-independent acquisition (DIA) protein mass spectroscopy to identify longitudinally differentially expressed proteins (DEPs) in the rTg-DI rat model of capillary CAA and CAA-ri.
Results
Capillary CAA emerges in younger rTg-DI rats at ≈4 months and progressed to severe stages of disease at 12 months. Increased reactive astrocytes and activated microglia similarly appeared at emergent stages of CAA but were more numerous at stages of disease. Brain regional proteomic analysis of rats in emergent and late disease stages revealed longitudinal differentially expressed proteins (DEPs) including ANXA3, HTRA1, APOE, CST3, CLU, and HSPB1. Pathway analysis of DEPs indicated activation of TGF-β1, and immunolabeling confirmed upregulation of TGF-β1. Further analysis revealed numerous astrocyte and microglia DEPs possibly indicating CAA-ri specific disease associate astrocyte (DAA) and disease associated microglia (DAM) populations. Potential CAA-ri DAAs include MLC1 and HSPB1, while potential CAA-ri DAMs include ANXA3, TGF-β1, and SPARC.
Conclusions
We have identified potential new early-stage biomarkers for CAA and CAA-ri, and proteins that may contribute to perivascular neuroinflammatory responses.
BLOOD-BRAIN BARRIER AND CHOROID PLEXUS DYSFUNCTION ARE ASSOCIATED WITH DISTINCT LIPID PROFILES IN CSF OF ALZHEIMER’S DISEASE PATIENTS
Abstract
Aims
Most lipid transport takes place across the blood-brain and blood-CSF barriers, which is disrupted in Alzheimer’s disease (AD). We recently discovered a distinct blood-brain barrier (BBB) dysfunction and a choroid plexus dysfunction subtype in AD based on CSF proteomics (Tijms, 2023). Here we studied if these AD subtypes were associated with specific lipid profiles in CSF.
Methods
We analysed CSF lipidomics data in individuals from the Amsterdam Dementia Cohort that were previously labelled as having a BBB dysfunction AD subtype (n=56) or choroid plexus dysfunction AD subtype (n=76) and 186 controls (intact cognition and normal AD biomarkers; table). Untargeted complex lipidomics was measured with CSH-QTOF mass spectrometry. Of 3532 lipids detected, 270 could be mapped to known classes. We compared AD barrier subtypes to controls on CSF lipid levels, controlling for sex and age with general linear models, and report results with p<0.05 to capture patterns of co-dysregulated lipids.
Results
Compared to controls, the BBB AD subtype had mostly increased levels of 302 lipids (148 with a known class; figure), whereas the choroid plexus subtype had mostly decreased levels of 314 lipids (163 known class) with an overlap of 150 lipids. These lipids included mostly glycerophospholipids, ceramides, and sphingomyelins. The BBB AD subtype further had increased levels of 14 tryglicerides, which were unaltered in the choroid plexus subtype. No specific alterations of lipids with a known class were observed in the choroid plexus subtype.
Conclusions
Our main finding is that two AD subtypes with different types of brain barrier dysfunction are related to CSF alterations of lipids from specific classes in opposite directions. This implies that these barriers have a distinct role in lipid alterations in AD, which may require specific treatment.
HIGH GLYCOSYLATION OF APOE3 IS ASSOCIATED WITH BRAIN AMYLOID BURDEN, INFLAMMATION, AND AD COGNITIVE STATUS
Abstract
Aims
Glycosylation of apolipoprotein E (APOE) affects the metabolization of APOE-containing particles in the brain, which impacts brain lipid metabolism. Here, we investigated the effects of APOE glycosylation on Aβ, tau, and inflammatory biomarkers in individuals in the early clinical stages of Alzheimer’s disease (AD).
Methods
We assessed 188 individuals from the Alzheimer’s Disease Neuroimaging Initiative [ADNI; 91 CU, 97 MCI] with APOE3 and APOE4 glycosylation levels in the cerebrospinal fluid (CSF). We divided individuals into APOE glycosylation level groups using tertile analysis. We tested the association between biomarkers with Pearson correlations while ANCOVA compared groups.
Results
Individuals Aβ-positive exhibited higher levels of APOE3 glycosylation than Aβ-negative (p = 0.002, Fig. 1A). Interestingly, MCI Aβ-positive showed the highest number of individuals presenting abnormally high levels of APOE3 glycosylation (the highest tertile, 46% of individuals) in comparison to MCI Aβ-negative (26%) and CU individuals (Aβ negative: 26%; Aβ-positive: 36%) (Fig. 1C, D). No significant differences were observed in APOE4 glycosylation levels according to Aβ status (Fig. 1B). CSF APOE3, but not APOE4, glycosylation was negatively correlated with sTREM2 levels (r = - 0.308 p = 0.003, Fig. 1E; r = -0.187, p = 0.202, Fig. 1F). No associations between APOE3 or APOE4 glycosylation and CSF p-tau181 were observed (r = - 0.020 p = 0.827, Fig. 1G; r = - 0.243, p = 0.070, Fig. 1H).
Conclusions
We have shown that glycosylation of APOE3, but not APOE4, was associated with increased Aβ burden. In addition, APOE3 glycosylation is also associated with brain inflammation, but not with tau phosphorylation. Together, these results point to alterations in brain lipid metabolism as an early event associated with brain amyloidosis and inflammation.
PLASMA AMYLOID-Β BIOMARKERS DISTINGUISH CO-MORBID ALZHEIMER’S DISEASE PATHOLOGY IN PATIENTS WITH LEWY BODY DISEASE
Abstract
Aims
Although comorbidity with Alzheimer’s disease (AD) is not uncommon in Lewy body disease (LBD), no blood biomarkers which can identify AD pathology in LBD have ever been established. To investigate if plasma amyloid-β (Aβ) biomarkers can distinguish AD comorbidity in LBD patients.
Methods
We examined plasma Aβ40, and Aβ42 in 15 patients with Aβ negative LBD, 10 with Aβ positive LBD, and 12 with AD. In addition, patients underwent amyloid-positron emission tomography imaging (11C Pittsburgh compound-B) or cerebrospinal fluid Aβ42 test.
Results
The plasma Aβ42/Aβ40 ratio was significantly decreased in patients with Aβ+ LBD compared with those with Aβ- LBD (p = 0.02) and could accurately classify the two diseases (area under the curve, 0.727).
Conclusions
Plasma Aβ42/Aβ40 ratio might be a useful marker for comorbid AD pathology in LBD.
EFFECT OF KIDNEY FUNCTION ON PLASMA PTAU217 CONCENTRATIONS USING DIFFERENT PTAU217 ASSAYS.
Abstract
Aims
Chronic kidney disease (CKD) has been associated with increased plasma phosphorylated Tau217 (pTau217) concentrations, potentially confounding its utility in the evaluation of Alzheimer's disease (AD). We assessed the association of estimated glomerular filtration rate (eGFR) with plasma pTau217 concentrations measured by various assays.
Methods
We included 100 participants from the Mayo Clinic Study of Aging or the Alzheimer’s Disease Research Center with diagnoses of cognitively unimpaired (CU, n=58), mild cognitive impairment (MCI, n=33), and dementia (n=9). Amyloid-PET was performed with Pittsburgh 11C-Compound B; abnormal amyloid (A+) was defined as SUVR ≥1.52 (Centiloid ≥25). eGFR was determined based on serum creatinine concentrations. pTau217 was quantified using ALZpath Simoa pTau217 immunoassay, Fujirebio Lumipulse G pTau217(N-Terminal) immunoassay, and a pTau217 mass spectrometry (MS) assay (C2N Diagnostics). Additionally, a pTau217 ratio (pTau217 to non-phosphorylated [npTau217]) was calculated using the MS assay. Linear regression was used to assess associations of continuous eGFR with plasma pTau217 and amyloid PET.
Results
Median pTau217 concentrations were higher in individuals with eGFR <60 mL/min/1.73 m2 compared to those with eGFR >60 across different clinical diagnoses and amyloid groups (Figure). A 10-unit worsening of eGFR was associated with higher pTau217 mean concentrations as follows: 18% (95% CI, 2–36%) for C2N pTau217, 13% (3–24%) for ALZpath pTau217, and 14% (4–26%) for Lumipulse pTau217. eGFR was not significantly associated with the pTau217/npTau217 ratio or amyloid-PET.
Conclusions
CKD was associated with increased pTau217 concentrations when measuring pTau217 but not when using the pTau217 ratio. Additional research is needed to better understand the confounding effects of CKD in the clinical interpretation of plasma pTau217 assays.
ALZPATH PTAU217 PERFORMANCE IDENTIFYING ALZHEIMER’S PATHOLOGY IN THE BRAIN AND BODY DONATION PROGRAM (BBDP) AT BANNER SUN HEALTH RESEARCH INSTITUTE
Abstract
Aims
Blood tests enable accessible screening and monitoring for Alzheimer’s disease (AD) disease-modifying therapeutics (DMT). The phosphorylated tau (pTau) in plasma considered to have the highest accuracy in identifying AD pathology is pTau217. We evaluated the relationship between plasma pTau217 and amyloid and tau load obtained via pathological exam in Banner BBDP (n=300).
Methods
The ALZpath pTau217 assay reported here is an ultra-sensitive blood-based assay developed on the semi-automated single-molecule array Simoa® platform. Banner BBDP is characterized by a variety of clinically meaningful comorbidities (e.g., cerebral amyloid angiopathy), mid-to-late stage AD pathology (Amyloid load mean = 8.53; Tau load mean = 8.06; Braak Score mean = 4.18), and a high co-occurrence of amyloid (A) and tau (T), A+T+ = 117, A+T- = 35, A-T- = 1, A-T+ = 0.
Results
Amyloid and tau load were correlated (r=0.43, p= <0.001). A Kendall rank analysis compared the relationship between plasma pTau217 and total A and T determined from postmortem pathology, as well as A and T within the temporal lobe (Fig 1). A machine learning model evaluated predictive capabilities of pTau217 for AD pathology accounting for sex, comorbidities, time between blood draw and death, and MMSE.
Conclusions
pTau217 is a high performing biomarker of CNS amyloid and tau burden. The ALZpath pTau217 performance capabilities can support timely AD diagnosis and intervention and facilitate scalable DMT implementation.
OPTIMAL COMBINATION OF CSF PHOSPHORYLATED-TAU FORMS TO PREDICT ALZHEIMER’S DISEASE PATHOLOGICAL BURDEN AS MEASURED BY PET
Abstract
Aims
Objectives: Evaluate the potential of cerebrospinal fluid (CSF) phosphorylated-tau (pTau) ratio’s derived from mass spectrometry (MS) to track Alzheimer’s disease progression.
Methods
Methods: We included 443 participants from the BioFINDER-2 cohort with available CSF-IP/MS quantification (Washington University), CSF amyloid-β42/40 (Aβ) and [18F]RO948 tau-PET. [18F]flutemetamol-PET was available for a subset (57.6%) and missing data was imputed based on Aβ42/40 (Table-1). PET burden was determined in 3 previously defined Aβ-stages1 and Braak stages (I-II, III-IV, V-VI)2. Using a ‘pseudotime’ approach3, we placed each individual on a two-dimensional space based on their (Aβ/tau) staging burden, and inferred the trajectory using SCORPIUS, creating a continuous measure of AD pathological-PET burden. The ratio of tau phosphorylated to non-phosphorylated peptides was determined for 9 forms4, and the microtubule-binding region (MTBR-tau243)5 log-transformed, and z-scored based on Aβ-negative participants. We determined the parsimonious model of CSF biomarkers (R-package MuMIn) to predict pathological-PET burden.
Results
Results: All %pTau forms, except 199 and 202, showed an early increase in abnormality with increasing AD pathological-PET burden (driven by Aβ) and an attenuation effect with higher levels (driven by tau), where only %pTau205, %pTau217, and MTBR-tau243 continued to show an increase in abnormality, with the latter showing the steepest slope (Figure-1). While %pTau217 increasing slowed down with higher AD pathological-PET burden, continued change of pTau217 concentration was observed, suggesting that overall phosphorylation of tau persists (Figure-2). AD pathological-PET burden was optimally predicted by combining Aβ42/40 ratio, %pTau205, %pTau217, MTBR-tau243, and age (R2=0.87).
Conclusions
Conclusions: Concurrent assessment of several pTau forms and MTBR-tau243 can accurately track PET-based AD pathological burden. This can support staging efforts at the individual level using CSF and in the future plasma, which is an increasing need for clinical trial design.
COMPARISON OF [18F]MK6240 AND [18F]AV1451 FOR PET-BASED BRAAK STAGING: THE HEAD STUDY
Abstract
Aims
Tau-PET tracers allow for in vivo Braak staging of individuals in the Alzheimer’s disease (AD) continuum. The impact of the properties of different tracers for Braak staging using tau-PET remains unclear. Therefore, we performed a head-to-head comparison of Braak staging using first- and second-generation tau-PET tracers.
Methods
We assessed 39 cognitively impaired and 34 cognitively unimpaired (CU) older adults with [18F]MK6240 and [18F]AV1451 tau-PET scanned at McGill University as part of the HEAD Study. We calculated standardized uptake value ratios (SUVR) in Braak-like regions of interest (ROI) for each ligand and investigated their association using Spearman’s correlation. Thresholds defined the presence of tau abnormality in each Braak ROI, which was used to assign a Braak stage to each participant. Finally, we assessed the agreement between the Braak staging provided by the two tracers.
Results
In all Braak ROIs, we found positive correlations between the SUVR values of the two tracers (Fig. 1). The strongest correlation was observed in Braak I (r=0.89), and the weakest in VI (r=0.54). The mismatch regarding tau positivity was higher in early Braak ROIs, mainly driven by cases positive solely for [18F]AV1451 (Fig. 2). Agreement regarding the latest stage of abnormality was substantial for both the traditional (kappa=0.651) and the simplified (kappa=0.734) frameworks (Fig. 3). [18F]AV1451 had higher probabilities of overestimating the classification given by [18F]MK6240 for both the traditional (22%vs15%) and simplified (14%vs11%) frameworks. [18F]AV1451 (15%) showed more Braak non-conformant cases than [18F]MK6240 (4%).
Conclusions
Off-target binding of [18F]AV1451 to the choroid plexus might explain the disagreement at early stages. [18F]MK6240 seems more promising to identify tau accumulation in late Braak stages. Overall, substantial agreement was observed between [18F]MK6240 and [18F]AV1451 for in vivo staging of AD severity.